Perfluorooctane sulfonate in the environment: chapter 5
The spatial distribution of PFOS in air (gas and particle phase), sediment, fish and birds across Canada generally correlates to human activity (Figure 2). In many cases, elevated PFOSconcentrations were observed near cities, especially in southern Ontario. Other sources may include airports, where use of aqueous film-forming foam is permitted until May 2013. However, it is difficult to identify the dominant sources for each site with the information that is currently available. PFOS was also detected in rural and remote locations, albeit at concentrations lower than in urban centres. The PFOS in source regions is transported to background sites through atmospheric transport of its precursor compounds and/or transport of PFOS through river and ocean currents. PFOSconcentrations in media where environmental quality guidelines exist (i.e., water, fish tissue and bird eggs) were below the draft FEQGs. However, PFOSconcentrations in fish and bird eggs exceeded draft FEQGs for the protection of mammals and birds that eat the fish and bird eggs, suggesting that this compound could represent a current risk to wildlife predators.
Figure 2. PFOS concentrations in air, sediment, water, fish, and wildlife (European Starling and Herring Gull eggs) across Canada in 2006-2011. For air, PFOS concentrations are either for passive or high-volume samplers (denoted as HV), collected in 2009. Sediment samples were collected in 2008. Water samples were collected between 2007 and 2010 and fish samples were collected between 2006 and 2010. Starling eggs where collected as pooled samples in 2009, the sites denoted with a W were collected from waste sites (i.e., landfills). Gulls, eggs were collected as pooled samples in 2009-2011. The spatial trends for individual gull samples collected in 2008 are similar and not shown. The green circles represent sites where PFOS concentrations were not detected. Where there was more than one data point available for a given media/location, the average (geometric mean) value was plotted on the maps.
Air: Monitoring PFOS in air across Canada provides information on PFOS levels within the country as well as quantities entering Canada from international sources. Air measurements have been obtained using two methods: high-volume air sampling1,2 and passive air sampling.3 High-volume air samplers measure a larger volume of air and are better for detecting the low PFOSconcentrations often found in the environment. However, passive air samplers are able to gather information on long-term exposure, and can be advantageous under many circumstances because of their simplicity, ease of transport to remote sites and non-reliance on a power source.
Sampling using high-volume samplers was conducted at three locations in Canada in 2009, and it was observed that PFOSconcentrations (geometric mean) were more than three times higher in Toronto (1.5 pg/m3) compared with Lake Superior air (0.43 pg/m3). PFOS was below the detection limit of 0.2 pg/m3 at the Canadian High Arctic station of Alert, Nun.; however, its precursors were detected up to several pg/m3.
Sampling using passive samplers was conducted at eight locations across Canada over a three-month period in 2009. PFOS concentrations were detected in Toronto, Ont. (8 pg/m3), an agricultural site in Saskatchewan (5 pg/m3), Whistler, B.C. (4 pg/m3), and Alert, Nun. (2 pg/m3). One site in northern Ontario had elevated PFOS concentration of 18 pg/m3. However, these data points are based on only one sample. PFOSwas not detected at the other Canadian sites. The PFOS levels measured in Canada using passive samplers were substantially lower than in Paris, France (150 pg/m3), but comparable to Sydney, Florida (3.4 pg/m3), Tudor Hill, Bermuda (6.1 pg/m3), Malin Head, Ireland (3.3 pg/m3), and Hilo, Hawaii (6.6 pg/m3).3
In general, the results showed that PFOS air concentrations in urban locations (e.g. Toronto) were on the same order of magnitude as more remote sites (e.g. Lake Superior), demonstrating the widespread distribution of PFOS in the Canadian atmosphere. FEQGs do not exist for PFOSin air.
Sediment: In 2008, 65 surface sediment samples were collected at 18 sites across Canada. The highest PFOSconcentration in sediments was found in Lake Ontario (geometric mean = 10 ng/g dry weight). PFOSwas also detected in the open-water sediment of the other Great Lakes at concentrations of 0.89, 2.2 and 1.4 ng/g dry weight in Lake Erie, Lake Huron and Lake Superior, respectively. Sediment PFOSconcentrations at Hamilton and Toronto harbours (in Lake Ontario) and near Thunder Bay (in Lake Superior) were 0.64, 1.9 and 0.54 ng/g dry weight. The PFOSconcentration in Lake Simcoe, Ont., sediment (geometric mean = 0.76 ng/g dry weight) was comparable to the Great Lakes sites, except Lake Ontario.23 Downstream of the Great Lakes watershed, the PFOSconcentration in Lake Saint-Pierre, Que., was 0.16 ng/g dry weight. In the Atlantic provinces, PFOS was detected in the Nappan River, an industrial site in New Brunswick (2 ng/g dry weight), Kejimkujik Lake, N.S. (0.28 ng/g dry weight), and at Little Sackville, N.S. (0.19 ng/g dry weight). In western Canada, PFOS was found only in Osoyoos Lake, B.C. (0.36 ng/g dry weight). PFOS was non-detectable at the other sites monitored.
Overall, although the highest PFOS concentration in sediment was observed in the urbanized Lake Ontario, levels were not always associated with human population. For example, relatively low PFOS concentrations were found in the heavily developed Lake Erie, as well as the Hamilton and Toronto harbours, which were comparable to more remote sites (e.g. Lake Superior). FEQGs for PFOS do not exist for sediment.
Water: Water was sampled for PFOS in the Great Lakes, as well as in streams and rivers across the country4,5 from 2007 to 2010. PFOS was highest at a site in Mill Creek, located in Kelowna, B.C. (geometric mean = 10 ng/L). This section of Mill Creek is urbanized and influenced by urban stormwater. In addition, Wascana Creek, in Regina, Sask., had a relatively elevated PFOS concentration (geometric mean = 7.8 ng/L). This site in Wascana Creek is located 8.5 km downstream of a wastewater treatment plant (WWTP) outfall and is the water collection station most impacted by WWTP inputs. Detectable values (>2 ng/L) were also observed in southern Ontario (Hamilton Harbour, Niagara River at Fort Erie, Lake Ontario at Wolfe Island, Grand River and Thames River), St. Lawrence River at Quebec City, Vancouver, British-Columbia (Still Creek and Serpentine River), Abbotsford, British-Columbia (Fishtrap Creek), and at the three Atlantic sites (Napan River, N.B., Sackville River, N.S., and Waterford River, N.L.).
Many of the rivers and streams containing relatively elevated levels of PFOSwere located in cities, and thus are influenced by urban activities. In comparison, PFOSconcentrations were mostly not detected in water bodies from non-urban regions and background sites. The draft FEQG for PFOS in water is 6000 ng/L, which is 200 times greater than the highest measured water concentration in Canada (31 ng/L in Mill Creek, Kelowna, B.C.). Therefore, current PFOS levels in water at the selected Canadian sampling sites are of low risk to aquatic life.*
Fish: Top predator fish (e.g., lake trout and walleye) were collected at 21 sites across Canada from 2006 to 2010. PFOSlevels varied considerably, with the highest concentrations observed in lake trout from Lake Erie (geometric mean = 90 ng/g wet weight) and Lake Ontario (geometric mean = 62 ng/g wet weight). Relatively elevated PFOSconcentrations (geometric mean) were also found in walleye from the St. Lawrence River (30 ng/g wet weight), Codette Reservoir, Sask. (24 ng/g wet weight) and Lake Diefenbaker, Sask. (23 ng/g wet weight), and in lake trout from Peninsula Harbour, Ont. (24 ng/g wet weight) and Lake Champlain (17 ng/g wet weight). Levels in fish were mostly low (<3 ng/g wet weight) in the water bodies located in northern Canada, Pacific and Atlantic regions, and Lake Superior.
In general, the results for fish indicated relatively elevated PFOSconcentrations in urban areas, particularly in southern Ontario, compared to the more remote lakes sampled. The draft FEQG for fish tissue is 8,300 ng/g wet weight, which is 12 times greater than the highest measured fish concentration in Canada (189 ng/g wet weight in Lake Erie). This suggests a low probability of adverse effects to fish related to PFOS exposure. In contrast, the PFOSlevels could represent a potential risk to the fishes’ wildlife predators in some of the systems (draft FEQG = 4.6 ng/g wet weight and 8.2 ng/g wet weight for mammalian and avian predators, respectively).
Wildlife: PFOS is currently monitored in the eggs of two types of birds: gulls and European Starlings. Gulls often eat fish and provide an indication of contamination in the aquatic environment. Although gull sampling focussed on Herring Gulls, when they were not available, other related species (i.e., California Gull, Ring-billed Gull and Glaucous-winged Gull) were monitored. In 2008, gull eggs were measured individually, whereas between 2009 and 2011, the gull eggs were measured as pooled samples.8The results between individual and pooled eggs often differ, and they are therefore reported separately here. For the individual gull eggs, relatively elevated PFOSconcentrations (geometric mean) were found at urbanized areas of the Great Lakes and the St. Lawrence River, with levels greater than 260 ng/g wet weight. Concentrations were lower (geometric means ranged from 7-115 ng/g wet weight) in non-urban areas as well as at marine colonies on both the Atlantic and Pacific coasts. A similar result of elevated concentrations in the Great Lakes were also found in the pooled gull eggs collected between 2009 and 2011, with the highest levels observed in Lake Erie (676 ng/g wet weight). In contrast to gulls, European Starling eggs provide information on terrestrial systems, and European Starlings feed lower on the food web. European Starlings were collected from waste (i.e., landfills) and non-waste sites in 2009 and were measured as pooled samples. The highest geometric mean PFOS concentration (703 ng/g wet weight) was found in Starling eggs collected at the Brantford, Ont., landfill, located in a highly urbanized region in southern Ontario. Relatively elevated PFOS concentrations (geometric mean = 148 ng/g wet weight) was also found in starling eggs collected at the Calgary landfill. However, other than these two landfill sites, PFOS concentrations were not higher at the waste sites compared to non-waste sites. For example, PFOSconcentrations (geometric mean) were higher in the starlings collected from urbanized communities of Indus, Alta. (199 ng/g wet weight), Delta, B.C. (75 ng/g wet weight) and Hamilton, Ont. (41 ng/g wet weight) compared with landfill sites located in Langley, B.C. (5.6 ng/g wet weight), Halton, Ont. (29 ng/g wet weight), Stoney Creek, Ont. (28 ng/g wet weight) and Otter Lake, N.S. (18 ng/g wet weight).
As shown above, the spatial trends of PFOS concentrations in gull eggs were related to human population. In comparison, for starling eggs, although the highest PFOS concentration was observed in the heavily industrialized landfill site of Brantford, Ont., levels were not always higher in eggs from urbanized/industrialized locations compared to areas that were more remote. The draft FEQG for bird eggs (1900 ng/g wet weight) was 2.3 times greater than the highest PFOS concentration observed in gulls (811 ng/g wet weight in Lake Erie). When the Brantford, Ont., landfill site was excluded, the draft FEQG for bird eggs was 7.5 times greater than the highest PFOS concentration in starlings (254 ng/g wet weight at Indus, Alta.). The draft FEQG for bird eggs was 1.6 times greater than the PFOS concentrations in starling eggs at the Brantford, Ont., landfill site (maximum = 1184 ng/g wet weight). These results suggest that PFOS represents little risk to the birds themselves. However, guideline exceedances were common for their wildlife consumers (draft FEQG = 4.6 ng/g wet weight for mammalian predators and 8.2 ng/g wet weight for avian predators).
* Draft FEQG for water is not designed to protect against bioaccumulation - see corresponding dietary and tissue residue FEQGs.
Report a problem or mistake on this page
- Date modified: