Draft Ecological Screening Assessment Report Long-Chain (C9–C20) Perfluorocarboxylic Acids, their Salts and their Precursors

Ecological Exposure Assessment

Atmosphere

Loewen et al. (2008) studied atmospheric concentrations and lake water concentrations of FTOHs over an altitudinal transect in western Canada. Lake water samples were collected at Cedar Lake (a small lake near Golden, British Columbia), at Bow Lake in Banff National Park (Banff, Alberta) and at another unnamed small lake in Banff National Park (Banff, Alberta). Passive air samplers were deployed on altitudinal transects (800–2740 above sea level) from Golden, British Columbia, to Banff National Park. Loewen et al. (2008) noted that the amount of 8:2 and 10:2 FTOHs (<2.0 ng/sampler) increased with increasing altitude.C10 PFCA lake water concentrations were below 0.2 ng/L.

Stock et al. (2007) took atmospheric particle/gas phase samples on Cornwallis Island, Nunavut where mean values of total concentrations of FTOHs ranged from 2.8 (10:2 FTOH) and 14 pg/m ³ (8:2 FTOH). The 8:2 and 10:2 FTUCAs were measured at mean concentrations of 0.06-0.07 pg/m³. C9 and C10 PFCAs were measured at mean concentrations of 0.4 pg/m³ while C11, C13 and C14 PFCAs were measured at mean concentrations ranging from 0.02 to 0.06 pg/m³. Shoeib et al. (2006) took twenty high-volume air samples during a crossing of the North Atlantic and Canadian Archipelago in July 2005 (Gothenburg, Sweden to Barrow, Alaska via the North Atlantic and Canadian Archipelago). The highest concentrations (sum of gas- and particle-phases) of FTOHs were for 8:2 FTOH at 5.8 -26 pg/m³, followed by 10:2 FTOH at 1.9-17 pg/m³ and 6:2 FTOH at below detection to 6.0 pg/m³. For comparison purposes, Shoeib et al. (2006) also collected air samples at a semi-urban site in Toronto in March 2006 where the mean 8:2 FTOH concentration in Toronto was 41 pg/m³. Studies from Toronto measured levels of 4:2, 6:2, 8:2 and 10:2 FTOHs from non-detect (ND) to 650 pg/m³ over a two-year period, with 8:2 FTOH dominating in the first half of the period and 10:2 FTOH dominating in the later half (Stock et al. 2005). Dreyer et al. (2009) conducted high volume air sampling in the Atlantic Ocean, the Southern Ocean and the Baltic Sea. C9 PFCA, C10 PFCA, C11 PFCA, C12 PFCA and C13 PFCA were all detected in the particle fraction (< 1 pg/m³). 6:2 FTOH and 8:2 FTOH were dominant in the gas-phase fraction. The concentrations of 8:2 FTOH were between 1.8 and 130 pg/m³. The sum of all the FTOHs (4:2 FTOH, 6:2 FTOH, 8:2 FTOH, 10:2 FTOH and 12:2 FTOH) ranged between 0.3–47 pg/m³.

As previously discussed, FTOHs may degrade to long-chain (C9-C20) PFCAs; therefore, levels of FTOHs in the atmosphere may contribute to levels of long-chain (C9-C20) PFCAs in the environment, including the Canadian Arctic. Atmospheric degradation of FTOH is expected to produce relatively equal quantities of adjacent odd-and even-chained perfluorinated PFCAs (Ellis et al. 2004b), whereas aerobic biodegradation of FTOH tends to yield predominantly even-chain-length PFCAs (Dinglasan et al.2004).

Water

Long-chain (C9-C20) PFCAs have been measured in precipitation from 1998 to 1999 in Canadian remote areas in Saturna Island ( British Columbia), Algoma (Ontario) and Kejimkujik (Nova Scotia) (Scott et al. 2006b). Scott et al. (2006b) measured C9 PFCA concentrations ranging from < 1 to 7.6 ng/L, with Algoma, Ontario, having the highest concentration, at 7.6 ng/L. In urban areas (Egbert and north Toronto, Ontario), C9 PFCA concentrations ranged from 0.4 to 9.7 ng/L. C9 to C12 PFCAs were detected in urban areas only, at concentrations ranging from < 0.07 to 5.2 ng/L. The authors also detected 8:2 and 10:2 FTCAs and FTUCAs at two urban sites in Canada (Egbert and north Toronto, Ontario) at concentrations ranging from < 0.07 to 8.6 ng/L. Loewen et al. (2005) also measured C10 and C12 FTCAs and FTUCAs in rainwater samples in Winnipeg, Manitoba.

Simcik and Dorweiler (2005) measured levels of long-chain (C9-C20) PFCAs in remote areas (Tettegouche and Nipisiquit) located along the north shore of Lake Superior and Voyageurs National Park along the U.S.-Canada border (Loiten and Little Trout lakes). These four remote lakes are all hike-in lakes (no roads) and have no surface inflow. Only C9 PFCA was found at a concentration greater than 1 ng/L, in Nipisiquit. Scott et al. (2006a, 2003) measured C9 PFCAs in the Great Lakes (lakes Ontario, Erie, Superior and Huron) at concentrations ranging from <0.5 to 0.11ng/L. Three connected urban lakes in Minneapolis, Minnesota (Lake of the Isles, Lake Calhoun and Lake Harriet), and the Minnesota River, a major tributary to the Mississippi River, were studied (Simcik and Dorweiler, 2005). C9 PFCA was found in Lake Calhoun (< 1 ng/L) and Minnesota River (< 10 ng/L). C10 PFCA was found in Lake Calhoun (< 1 ng/L).

Stock et al. (2007) measured concentrations of C9, C10, C11 and C12 PFCAs in lake water samples from three remote Arctic Lakes (Resolute, Char, and Amituk) on Cornwallis Island, Nunavut in 2003. C9 PFCA mean concentrations ranged from 0.3 – 4.1 ng/L. C10 PFCA mean concentrations ranged from 1.1 – 19 ng/L and C11 PFCA concentrations ranged from 0.3 – 4.9 ng/L. C12 PFCA mean concentrations ranged from 0.2 – 0.8 ng/L. 8:2 FTUCA and 10:2 FTUCA were measured in all three lakes at concentrations ranging from 1.1 to 6.4 ng/L.

Ahrens et al. (2009) analyzed surface water samples from the Atlantic Ocean along the longitudinal gradient from Las Palmas (Spain) to St. John’s (Canada) and along the latitudinal gradient from the Bay of Biscay to the South Atlantic Ocean, in 2007. No long-chain (C9-C20) PFCAs were detected in the particulate phase and in the two deep water samples at 200 m and 3800 m. C9 PFCA was detected in concentrations ranging from < 0.0051 to 0.107 ng/L, and C12 PFCA was not detected generally above 0.0014 ng/L. Del Vento et al. (2009) measured up to 0.13 ng/L C9 PFCA in seawater in eastern part of the Beaufort Sea near Banks Island. Del Vento et al. (2009) also measured snow concentrations in the Amundsen Gulf for C9 – C11 PFCAs ranging from 0.035 – 1.3 ng/L. C9 PFCA was measured in Asian seawaters at ng/L levels, with values reported three times higher and with higher variability in coastal regions than in open oceans (So et al. 2004; Yamashita et al. 2004, 2005).

Sediments

Stock et al. (2007) measured concentrations of C9, C10, C11, and C12 PFCAs in sediment core samples taken from three remote Arctic lakes--Resolute, Char, and Amituk Lakes on Cornwallis Island, Nunavut in 2003. Concentrations of the PFCAs decreased with depth and sediment age. C9 –C12 PFCAs were present in Amituk Lake. C9 -C11 PFCAs were measured in Char Lake at 0.6 – 3.3 ng/g. In Resolute Lake, C9 PFCA was measured up to 3.2 ng/g while C10, C11 and C12 PFCAs were at the limit of detection of 0.5 ng/g. 10:2 FTUCA was measured in Amituk and Char Lakes at concentrations ranging from 0.5 to 2 ng/g.

Biota

Levels of long-chain (C9–C20) PFCAs have been reported in a number of freshwater and marine animals in North America and Greenland, including polar bear (De Silva and Mabury 2004; Kannan et al. 2005; Smithwick et al. 2005a,b; Smithwick et al. 2006; Dietz et al. 2008), ringed seal (Bossi et al. 2005; Butt et al. 2008), bottlenose dolphins (Houde et al. 2005), animals from a Lake Ontario food web (Martin et al. 2004b), lake trout (Furdui et al. 2007, 2008), northern fulmar and thick-billed murre (Butt et al. 2007a).

Martin et al. (2004b) measured C9–C15 in whole homogenate samples from a variety of fish from Lake Ontario. Concentrations ranged from < 0.5 ng/g-ww (detection limit of 0.5 ng/g) to 39 µg/kg-ww (sculpin; C11). In an accidental spill situation at Etobicoke Creek (Ontario) of aqueous fire-fighting foam, Moody et al. (2002) showed the presence of C10, C11, and C13 PFCAs in the livers of common shiner fish (Notropus cornutus). It should be noted that the level of biota concentrations may not necessarily be all associated with the spill. Concentrations ranged from 8.8 to 390 ng/g-ww, with the highest concentration observed for the C10 PFCA.

Martin et al. (2004a) have shown the presence of C9 through C15 PFCAs in the liver of a variety of species including seals, foxes, fish, polar bears and birds sampled between 1993 and 2002 in the Canadian Arctic. Liver concentrations in all species ranged from non-detect to 180 ng/g-ww (detection limit = 0.5 ng/g). Liver concentrations were greatest for polar bears (Ursus maritimus; maximum 180 ng/g-ww, C9) and decreased as the chain length increased. Stern (2009) measured C9 – C11 PFCAs were measured in burbot liver from the Mackenzie River at Fort Good Hope, Northwest Territories. Thirty-seven burbot were sampled with mean concentrations ranging from 0. 89 – 7.97 ng/g-ww for C9, 1.2 – 36.85 ng/g-ww for C10, and 7 – 2.25 ng/g-ww for C11. Butt et al. (2007a) noted a predominance of C11 to C15 PFCAs in Arctic seabirds, although the PFCA detected in most wildlife, e.g., ringed seals, is often C8 to C11 PFCAs. Powley et al. (2008) detected C9 to C12 PFCAs in a variety of organisms from Banks Island (eastern edge of the Beaufort Sea in the Northwest Territories). Concentrations in zooplankton (Calanis hyperboreus, Themisto libellula, Chaetognatha) ranged from non-detect to 1.1 ng/g-ww. In Arctic cod (Boreogadus saida), concentrations ranged from non-detect to

0.6 ng/g-ww. In ringed seal (Phoca hispida), concentrations in blubber ranged from non-detect to 0.2 ng/g-ww, concentrations in blood ranged from 1 to 2.5 ng/g, and concentrations in liver ranged from 1 to 6.9 ng/g-ww. Concentrations were not detected in bearded seal (Eriganthus barbatus) blubber, blood or liver. It should be noted that the sample sizes were small, ranging from 1 to 5. Tomy et al. (2009b) measured C8-C12 PFCAs in the liver of various top trophic level mammals in an eastern Arctic marine food web from Cumberland Sound, Nunavut in 2007. Liver-based concentrations in beluga ranged from 38.07 – 47.6 ng/g-ww;narwhal liver-based concentrations ranged from 11.71 – 50.78 ng/g-ww; harp seal liver-based concentrations ranged from 2.93- 12.78 ng/g-ww; ringed seal liver-based concentrations ranged from 2.18 – 23.4 ng/g-ww; and Greenland shark liver-based concentrations ranged from 17.76 - 110.79 ng/g-ww. In general, C11 PFCA concentrations were dominant except in the Greenland shark where C12 PFCA dominated.

Houde et al (2006c) assessed the concentrations of C9 to C12 PFCAs, and C14 PFCAs in plasma, milk and urine of bottlenose dolphins residing in and around Sarasota Bay , Florida, USA. During the past 35 years, the year-round resident population (approx. 160 animals) has been the subject of a long-term monitoring project. Houde et al (2006c) investigated the relationships between PFCA concentrations and known biological parameters (age, gender, reproductive history and morphometrics). The dominant PFCA detected was C11. C9 PFCA concentrations in plasma ranged from 11.7 to 24.5 ng/g-ww, the concentration in milk was 2.2 ng/g-ww, and the concentration in urine was below detection limit. C10 PFCA concentrations in plasma ranged from 15.8 to 35.7 ng/g-ww, the concentration in milk was 2.4 ng/g, and the concentration in urine was below detection limit. C11 PFCA plasma concentrations ranged from 31.4 to 64.7 ng/g-ww , 3 ng/g ww in milk, and 0.06 ng/g in urine. C12 PFCA concentrations in plasma ranged from 2.7 to 8.2 ng/g-ww, the concentration in milk was 2.9 ng/g, and the concentration in urine was below detection limit. C14 PFCA plasma concentrations ranged from 1.1 to 3.4 ng/g-ww, and no analyses were made for milk or urine. No significant differences were found between dolphins inhabiting the northern end of Sarasota Bay and those frequenting the southern part. Sarasota Bay is a semi-enclosed environment surrounded by a highly residential urban area, which may explain the relatively high concentrations detected in resident dolphins. Temporal trend analyses showed that PFC concentrations in plasma were not significantly greater in dolphins captured in the summer of 2003 and winter 2004 compared to other sampling seasons. Results showed significant negative correlations between C9–C12 PFCAs and age of the dolphins. No significant relations were found for gender. Concentrations were found to decrease with age for both male and female dolphins.

Worldwide, a number of studies have reported levels of long-chain (C9-C20) PFCAs in biota, including porpoises (van der Vijver et al. 2004, 2007), harbour seals (van der Vijver et al. 2005) and glaucous gulls (Verreault et al. 2005). van de Vijver et al. (2007) collected liver samples from harbour porpoises (Phocena phocena relicta) along the Ukranian coast of the Black Sea. Concentratons ranged from 1.4 to 19 ng/g-ww with C10 PFCA having the highest concentration. van de Vijver et al. (2003) also has shown the presence of C9-C11 PFCAs in the livers of several mammals taken from the North Sea coast, including a sperm whale, a white-sided dolphin and white-beaked dolphins. Concentrations ranged from non-detection to 480 ng/g-ww for all four species (detection limit 30–90 ng/g). Concentrations were highest in the white-beaked dolphin (Lagenorhynchus albirostris). Leonel et al. (2008) measured C9-C12 PFCAs in Franciscana dolphin (Pontoporia blainvillei) collected from southern Brazil. Liver concentrations ranged from < 0.1 to 0.46 ng/g-ww with C11 PFCA having the highest concentration. Leonel et al. (2008) also measured C9-C12 PFCAs in Subantarctic fur seal (Arctocephalus tropicalis) also collected from southern Brazil. Liver concentrations ranged from <0.1 to 0.74 ng/g-ww – again C11 PFCA had the highest concentration. C9-C12 PFCAs were measured in liver of the Indo-Pacific humpback dolphins (Sousa chinensis) and finless porpoises (Neophocaena phocaenoides) in Hong Kong (Yeung et al. 2009c). C9-C12 PFCA concentrations in the humpback dolphins ranged from 0.243 to 120 ng/g-ww. C9-C12 PFCAs were detected in finless porpoises at concentrations ranging 0.522 to 34.3 ng/g-ww.

Tseng et al (2006) found C10 PFCA in oysters (Crassostrea gigas), in tilapia (Oreochromis sp.) and Japanese seaperch (Lateolabrax japonicus) in Taiwan. C10 PFCA concentrations in the oysters ranged from 140 – 320 ng/g-ww. Liver and fish muscle concentrations of C10 PFCA in tilapia was 390 and 250 ng/g-ww, respectively. C10 PFCA concentration in the Japanese seaperch was 480 ng/g-ww.

Concentrations of C9-C12 PFCAs were measured in egg yolks of three species of birds--the little egret (Egretta garzetta), little ringed plover (Charadrius dubius) and vinous-throated parrotbill (Paradoxornis webbiana)--collected around Lake Shihwa, Korea (Yoo et al. 2008). C9-C12 PFCAs concentrations ranged from 5.7 to 675 ng/g-ww. The highest concentration was found in the little ringed plover with a C11 PFCA concentration of 675 ng/g ww. C9 PFCA was not detected in the liver of the northern fulmar (Fulmarus glacialis) along the coast of Svalbard and BjØrnØya in the Barents Sea (Norwegian Arctic) (Knudsen et al. 2007). However, Holmstrom and Berger (2008) measured C9 – C16 PFCAs in common guillemot (Uria aalge). C16 PFCA was below the detection limit. However C9 – C15 PFCAs were measured in concentrations ranging from 0.17 to 32 ng/g-ww. Wang et al. (2008) measured concentrations of C9 –C12 PFCAs in waterbird eggs in South China. C9-C12 PFCA concentrations in egg samples from black-crowned night herons (Nycticorax nycticorax) ranged from 0.072 to 41.3 ng/g-ww, concentrations in great egrets (Ardea alba) ranged from 0.225 to 5.79 ng/g-ww and concentrations in little egrets (Egretta garzetta) ranged from 0.77 to 39.4 ng/g-ww.

C9 – C12 PFCA was detected in beaver liver collected from Poland at concentrations ranging from < 0.04 to 4.46 ng/g-ww with C9 PFCA having the highest concentration (Taniyasu et al. 2005). Male wild rats (Rattus norvegicus) were collected from eight sites in Japan (i.e., a WWTP, a port, two industrial areas, a seafood market, a marketplace, two landfill sites and a seafood port)(Yeung et al. 2009b). Whole-blood samples were analyzed for C9 to C12 PFCA with mean concentrations ranging from 0.792 to 7.3 ng/mL. C9 and C10 PFCAs were measured in serum of the Chinese Amur tiger (Panthera tigris altaica) (Li et al. 2008) found in northeastern China, far eastern Russia and North Korea. C9 PFCA at concentrations of 0.13–0.89 ng/mL was found to be one of the most prevalent perfluorinated compounds in the serum of the Amur tigers. C10 PFCAs was found at a mean concentrations of 0.1 – 0.15 ng/mL. Gender differences were found for C9 and C10 PFCA accumulation where concentrations were slightly higher in females than in males.

Temporal and Spatial Trends

A temporal study over a 22-year period between 1980 and 2002 examined suspended sediment in Niagara River discharge where reported PFCA levels generally increased over time (Lucaciu et al. 2004). Myers et al. (2009) examined spatial distribution and temporal trends of perfluorinated compounds in Great Lakes sediment and surface waters. It was found that spatial distributions of PFCAs (C7 – C12 PFCAs) indicated that urban and industrial activities influenced concentrations in Great Lakes sediment and water. In Lake Ontario surface water, tributary samples showed the highest C7-C12 PFCA concentrations relative to near-shore and open lake samples; however, for sediment, open lake samples showed the highest concentrations. Myers et al. (2009) also noted that C7-C12 PFCA sediment concentrations are increasing in Lake Ontario. However, an observed increase and levelling –off was observed in Lake Superior sediments which may reflect atmospheric transport in C7-C12 PFCAs.

Furdui et al. (2007) determined the spatial trends of long chain PFCAs in lake trout (Salvelinus namaycush, age class equal to 4 years) collected from the Great Lakes in 2001. C9 to C15 PFCAs were detected in concentrations ranging from 0.37 to 4.9 ng/g ww. The highest concentrations were observed in Lake Erie, followed by Lake Huron, Lake Ontario, Lake Michigan and Lake Superior. No significant correlation was determined between concentrations and fish weight. The temporal trends of long chain PFCAs were determined in lake trout collected between 1979 and 2004 from Lake Ontario (Furdui et al. 2008). PFCA concentrations were generally low (non-detect to 3 ng/g) with C11 PFCA, C12 PFCA and C13 PFCA having the greatest concentrations. Most PFCA concentrations in 1988 and/or 1993 were generally higher than in 1979 followed by a levelling or decrease in concentrations. Regression analyses for individual PFCAs were not of sufficient significance to indicate declines in recent years since the peaks in 1998 and/or 1993.

An investigation of the temporal trends in liver of northern fulmar (Fulmarus glacialis) and thick-billed murre (Uria lomvia) from the Canadian Arctic reported overall increases in PFCAs over time for both species (Butt et al. 2007a). The geometric mean concentrations for the SPFCAs in thick-billed murres and northern fulmars were 23.9 ng/g and 12.4 ng/g, respectively. C13 PFCA was the predominant compound followed by C11 and C14 PFCA. Thick-billed murres showed increasing concentrations of PFCAs through 1975-2004 with doubling times ranging from 2.3 years for C15 PFCA to 9.9 years for C12. Doubling times for northern fulmars ranged from 2.5 years for C15 PFCA and 11.7 years for C12 PFCA. However, in the case of the northern fulmars, most PFCAs showed maximum concentrations in 1993 or statistically similar concentrations between 1987, 1993, and 2003. This may be the result of differing migratory patterns in bird species (Butt et al 2007a). Gebbink et al. (2009a) determined the spatial distribution, trends and sources of C9 – C15 PFCAs in 16 colonies of gull species sampled from eastern (Nova Scotia, New Brunswick, Newfoundland), central (Quebec, Ontario, Manitoba) to western Canada (Alberta, British Columbia). The four species are glaucous-winged gull (Larus glaucescens), California gull (Larus californicus), ring-billed gull (Larus delawarensis), and herring gull (Larus argentatus). The authors noted that the SPFCAs were greatest in the herring gull eggs collected in southern Ontario and western Quebec colonies close to urban sources. C11 and C13 PFCAs were generally dominant for most of the colonies although differences were observed among colonies. Overall, the spatial distribution of PFCAs in gull eggs across Canada was considered to be primarily influenced by location and proximity to local sources as opposed to diet. Gebbink et al (2009b) collected herring gull (Larus argentatus) eggs from 15 colonies located at Canadian and some American sites across the Laurentian Great Lakes. PFCAs ranging from C9 to C15 were detected with C11 and C13 PFCAs as the most dominant. C9 PFCA was more abundant in Lake Superior and Lake Michigan colonies and C11 PFCA was more abundant in the Lake Erie and Lake Ontario colonies. The highest SPFCAs were found in Lake Huron at 113 ng/g-ww followed by colonies in Lake Erie and Ontario. 8:2 and 10:2 FTOHs were not detected in any herring gull eggs.

Verreault et al. (2007) showed temporal trends for whole eggs of herring gulls (Larus argentatus) from two geographically isolated colonies (HornØya and RØst) in northern Norway. These colonies encompassed the southern and northern distribution range of herring gulls breeding in northern Norway. The dominant long chain PFCA in the herring gull eggs was C11 PFCA (4.2 ng/g ww, HornØya), followed by C13 (2.8 ng/g-ww, RØst). C9 to C13 PFCAs for both colonies showed significant increases between 1983–1993 folowed either by an increase post-1993 (i.e., C9, C10 and C11 PFCAs) or a leveling off (i.e., C12 and C13 PFCAs). Spatial trends between the two colonies were not different with the exception of C9 PFCA which was highest in the RØst colony. The eggs from the RØst colony collected in 1993 also had higher proportions of the C14 and C15 PFCAs compared to the HornØya colony and other sampling years. Verreault et al. (2007) suggested that the direct and indirect local-sources and/or remote-sources of long chain PFCAs may have changed over the last two decades in northern Norway. Alternately, there might have been shifts in the dietary preferences for adult herring gulls in northern Norway – these gulls have a limited annual feeding range and are primarily fish-feeders although they may also feed on crustaceans, seabird chicks, eggs, and other terrestrial food sources (human refuse). Löfstrand et al. (2008) determined spatial trends in guillemot (Uria aalge) eggs collected from Iceland, Sweden, the Faroe Islands, and Norway (Sklinna and HjelmsØya). C9 PFCA was detected only in Sweden at 48 ng/g-ww. C10 PFCA was detected only in Iceland and Norway at 38 – 42 ng/g-ww. C11 PFCA appears to be the most dominant compound with concentrations ranged from 9 – 140 ng/g-ww followed by C12 PFCA with concentrations ranging from ND to 81 ng/g-ww. The SPFCAs were highest in Sweden (150 ng/g-ww) followed by Iceland (96 ng/g ww), and the Faroe Islands (76 ng/g-ww). Lofstrand et al. (2008) suggested that the spatial patterns differ likely due to the differing feeding habits of the guillemot across the Atlantic and that the Swedish eggs were sampled in locations nearest industrial areas and heavily populated areas.

Levels of PFCAs in ringed seal livers from eastern and western Greenland measured from 1980 and 2000 increased 3.3 and 6.8% per year for C10 and C11, respectively (Bossi et al. 2005). Butt et al. (2007b) examined temporal trends in liver samples from two ringed seal (Phoca hispida) populations in the Canadian Arctic: Arviat (1992, 1998 and 2005) and Resolute Bay (1972, 1993, 2000 and 2005). C9–C15 PFCAs showed concentration increases from1992 to 1998 but later sampling points (1998, 2003, and 2005) were not statistically different. Concentrations increased by 117% for C14 PFCA to 310 % for C9 PFCA. No significant differences between sex were identified for any PFCA with any of the two populations. Overall, concentrations of the PFCAs increased from 1993-2005; however, the increases in the most recent years were not statistically significant. Doubling times ranged from 10.0 (C9 PFCA) to 19.4 (C12 PFCA). Butt et al. (2008) detected long-chain PFCAs in liver samples from eleven ringed seal populations in the Canadian Arctic from 2002 to 2005. Concentrations of C9–C11 PFCAs ranged from 1 to 10 ng/g-ww, whereas those of C12–C15 PFCAs were less than 1 ng/g-ww. In addition, 8:2 and 10:2 FTUCAs were detected in all populations; however, concentrations were less than the detection limit. Quantifiable levels were measured in two ringed seal populations, from 4 to 6 ng/g ww. Some statistically significant spatial trends were observed between individual populations; however, it was concluded that variations were largely attributable to elevated levels in two populations and lower levels in another population (Butt et al. 2008).

Temporal trends were investigated in the Baikal seal (Pusa sibirica) from Lake Baikal, eastern Siberia, Russia (Ishibashi et al. 2008b). C9 to C12 PFCAs were measured in the liver and serum of the Baikal seal. The Baikal seal is an endemic species and is a high trophic level predator in the food web of Lake Baikal. In male and female Baikal seal liver, the concentration of C9-C12 PFCAs ranged from <0.56 to 72 ng/g-ww. In male and female Baikal seal serum, C9-C12 PFCAs concentrations ranged from <0.33 -4 ng/g-ww (Ishibashi et al. 2008b). The mean concentrations of C9 and C10 PFCA in livers of seals collected in 2005 were 1.2 and 1.7-fold greater than in seals collected in 1992. C10 PFCA concentrations were significantly higher in 2005 than in 1992. For C9 PFCA, although there was a trend of increasing concentration from 1992 to 2005, no significant differences were observed.

In the studies by De Silva and Mabury (2004) and Smithwick et al. (2005a), levels of longer-chain PFCAs in polar bear liver were found to be generally higher in the east (Greenland), with evidence of C9 and C10 higher in the west. Examination of the tropospheric airflow patterns indicates that the central eastern regions would tend to receive air from eastern North America and those in the east (Greenland) would receive airflow from North America and Europe (De Silva and Mabury 2004). The higher C9 and C10 in western Arctic polar bears may be due to higher emissions of these congeners from Asia (Smithwick et al. 2005a). A similar west-to-east trend was observed with PFCA levels in ringed seal liver samples, although the highest levels were reported in the southern Hudson Bay region (Butt et al. 2008). Dietz et al (2008) examined a subsample of 128 subadult (3–5 years old) polar bears from 1984 to 2006 from Ittiqqortoormiit (Scoresby Sound) in central Greenland. Linear regression analysis of logarithmic-transformed median concentrations showed annual increases for C9 PFCA (6.1%), C10 PFCA (4.3%), C11 PFCA (5.9%), C12 PFCA (52%), and C13 PFCA (8.5%). Mean doubling times for concentrations in polar bear livers from North American Arctic regions ranged from 5.8 years in the east to 9.1 years in the west for C9, C10 and C11 from 1972 to 2002 (Smithwick et al. 2006).

Tomy et al. (2009a) measured the temporal trends of SPFCAs (C8-C12) in male beluga whales from Pangnirtung, Nunavut which showed an annual increase of 1.8 +/- 0.5 ng/g-ww between liver-based concentrations for the years 1980 to 2010. The SPFCAs (C8-C12) concentrations in the male beluga whales from Pangnirtung ranged from 2.4 to 171.05 ng/g-ww. However, male beluga whales from Hendrickson Island showed a decline of 7.41 +/- 0.71 ng/g-ww between liver-based concentrations for the years 1980 to 2009 (Tomy et al. 2009a). Concentrations for the SPFCAs (C8-C12) in the Hendrickson Island male beluga whales ranged from 4.87 to 313.10 ng/g-ww.

Ecological Effects Assessment

Aquatic Organisms

Boudreau et al. (2002) examined the toxicity of C10 on the aquatic macrophyte Lemna gibba, anddetermined that the IC₅₀ (inhibitory concentration) value (based on growth) was 99 mg/L. The IC₅₀ values (based on growth) of C10 on the freshwater algae Selenastrum capricornutum and Chlorella vulgaris were 218 mg/Land 198 mg/L respectively, indicating little difference in sensitivity between the two species (Boudreau et al. 2002). The acute and chronic toxicity of C10 on two species of water fleas, Daphnia magna and Daphnia pulicaria, were investigated. The acute LC₅₀ (based on mortality) and chronic EC₅₀ values (based on immobilization) were 259 and 130 mg/L, respectively for Daphnia magna and 285 and 150 mg/L, respectively for Daphnia pulicaria (Boudreau et al. 2002). These values indicate that there may be few sensitivity differences between the two species. Hoke et al. (2009) determined acute toxicity values for C10 PFCA: a 96hr rainbow trout (Oncorhynchus mykiss) LC₅₀ value of 32 mg/L, a 48hr EC₅₀ for Daphnia magna of >100 mg/L, and a 72hr EC₅₀ for the green algae, Pseudokirchneriella subcapitata, of 10.6 mg/L.

Chronic toxicity studies of C9 on the two species of water fleas indicated increased sensitivity of Daphnia pulicaria versus Daphnia magna (Boudreau et al. 2002). The 21-day LC₅₀ values (based on mortality) were 8.8 mg/L and 39 mg/L, respectively. Chronic toxicity studies also indicated that, for Daphnia magna, the number of young produced was a more sensitive endpoint than mean time to first brood. The 21-day no-observed-effect concentration and lowest-observed-effect concentration values (both based on number of young produced) were 13 and 25 mg/L, respectively for Daphnia magna and 6 and 13 mg/L, respectively for Daphnia pulicaria.

Terrestrial Organisms

The 48-hour EC₅₀ (based on acute lethality) for C9 for the soil-dwelling nematode Caenorhabditis elegans was 0.66 mM (306.29 mg/L) from surface contact exposure in a 1.7% agar nematode growth medium (Tominaga et al. 2004). Multi-generation effects following nematode exposure to C9 were found at concentrations as low as approximately 1 nM (0.000464 mg/L), which induced a 70% decline in fecundity by the fourth generation (Tominaga et al2004). Generation-response relationships and concentration-response relationships were not observed, although the results suggest that C9 could have longer-term multi-generational effects at relatively low exposure concentrations.

Yeung et al. (2009a) exposed one-day old male chickens (Gallus gallus) to C10 PFCA at 0.1 and 1.0 mg/kg body weight three times a week for three weeks. No adverse effects on body weight, organ indexes, blood clinical parameters or organ histopathology were observed at the 0.1 or the 1.0 mg/kg body weight of C10 PFCA. However, the half-life for PFDA at 0.1 and 1.0 mg/kg body weight was 11 and 16 days, respectively indicating the bioaccumulative properties of C10 PFCA in chickens.

Other Effects

Stevenson et al. (2006a) examined the toxicity of C9 and C10 with respect to the multi-xenobiotic resistance mechanism in the marine mussel Mytilus californianus. This mechanism acts as a cellular first line of defense against broad classes of xenobiotics exporting moderately hydrophobic chemicals from the cell via adenosine triphosphate (ATP)-dependent, transmembrane transport proteins. The most studied transporter is the p-glycoprotein which, in humans, is active in the kidney, adrenal gland, liver, blood-testes barrier and blood-brain barrier. This defense mechanism is fragile and can be compromised by some xenobiotics. This increased sensitivity, referred to as chemosensitization, arises from the ability of the p-glycoprotein to recognize and bind to multiple xenobiotic substrates, resulting in the saturation of the binding capacity. Even non-toxic substances can be chemosensitizers and cause adverse effects on organisms by allowing normally excluded toxic substances to accumulate in the cell. Stevenson et al. (2006a) found that C9 and C10 had average IC₅₀s (based on p-glycoprotein inhibition) of 4.8 µM (2.23 mg/L) and 7.1 µM (3.65 mg/L), respectively, which significantly inhibited the p-glycoprotein in Mytilus californianus. This result indicates that C9 and C10 are chemo sensitizers for this organism. C9 inhibits the p-glycoprotein by an indirect mechanism, and this inhibition is reversible. C9 also induces expression of the p-glycoprotein transporter after a 2-hour exposure--a stress response that may result in a metabolic cost to the organism.

Liu et al. (2008) investigated the effects of C12 and C14 PFCA on the membrane systems of the freshwater alga species, Scenedesmus obliquus. C12 and C14 PFCA inhibited algal growth rate in a concentration-dependent manner (i.e inhibition increased with increasing exposure concentration). The IC₁₀, IC₅₀, and IC₉₀ for cell density calculated were 90 uM (46.27 mg/L), 183 uM (94.08 mg/L), and 367 uM (188.67 mg/L) for C10 PFCA and 41 uM (29.28 mg/L), 134 uM (95.69 mg/L), and 292 uM (208.52 mg/L) for C14 PFCA. For C10 PFCA, an enhancement of the mitochondrial membrane potential was observed between 30 and 100 uM (15.42 – 51.40 mg/L). For C14 PFCA, an enhancement of the mitochondrial membrane potential was observed between 50 and 100 uM (35.70 – 71.41 mg/L). The increase in mitochondrial membrane potential indicates damage to the mitochondrial function – mitochondria are multi-tasking organelle involved in oxidative energy metabolism as well as apoptosis by integrating death signals. In addition, C12 and C14 PFCA caused an increase in cell membrane permeability at 20 – 100 uM (12.28 - 61.41 mg/L) for C12 PFCA and 50 – 100 uM (35.70 – 71.41 mg/L) for C14 PFCA. Effects on the permeability status of the cell membrane could play a role in mediating the adverse effects of other contaminants. Liu et al. (2008) used flow cytometric measurements to investigate the effects of C12 and C14 PFCA on the membrane systems of the freshwater alga Scenedesmus obliquus. The IC₅₀ values (cell density) for C12 and C14 were 183 µM (112.38 mg/L) and 134 µM (95.69 mg/L), respectively. Liu et al (2008) noted that the impairment of energy production by the interruption of the mitochondrial function may account for the inhibition of cell division caused by C12 and C14 observed as a reduction in growth rate (i.e., cell density).

Benninghoff et al. (2007) found that C9–C12 PFCA significantly induced vitellogenin, a biomarker of estrogen exposure, in rainbow trout. C9–C12 PFCAs demonstrated weak affinity to the rainbow trout hepatic estrogen receptor. In juvenile rainbow trout, C10 increased vitellogenin in a dose-dependent manner in vivo and significantly increased plasma vitellogenin at moderately low (0.0256–2000 µg/g diet) concentrations. The authors note that the trout hepatic estrogen receptor has greater affinity to more xenoestrogens than other mammalian estrogen receptors, including humans.

Nakayama et al. (2008) studied the common cormorant (Phalacrocorax carbo), a fish-eating bird that is the top predator in the ecosystem of Lake Biwa in Japan. C9 concentrations were measured in the liver of wild common cormorants (male and female) and related to gene expression. C9 concentrations for females ranged from < 0.005 to 0.0088 µg/g-ww, and for males, concentrations ranged from < 0.005 to 0.043 µg/g-ww. Significant sex differences were not detected. Gene expression analysis showed significant positive relationships between C9 and glutathione peroxidase 1 (enzyme in the antioxidant system) and heterogenous nuclear ribonucleoprotein U (RNA processing). The induction of the antioxidant enzymes may be an adaptive response to oxidative stress caused by C9.

Ishibashi et al. (2008a) showed that C9-C11 PFCAs induced peroxisome proliferator-activated receptor a (PPARa) in Baikal seal (Pusa sibirica) livers at lowest observed effect concentrations of 125 uM (58.00 mg/L)(C9 PFCA), 125 uM (64.26 mg/L) (C10 PFCA), and 62.5 uM (35.25 mg/L)(C11 PFCA). PPARa plays a critical physiological role as a lipid sensor and a regulator in lipid metabolism. Expression levels of PPARa mRNA showed a positive correlation with C9 PFCA and the expression of the hepatic CYP4A-like protein was correlated with the hepatic concentrations of C9 and C10 PFCAs suggesting modulation of the PPARa-CYP4A signalling pathway in wild Baikal seals.

The potential impact from exposure to perfluorinated compounds was investigated on liver lesions in east Greenland polar bears (Sonne et al. 2007). Parameters included mononuclear cell infiltrations, lipid granulomas, steatosis, Ito cells and bile duct hyperplasia/portal fibrosis. The population size consisted of 28 females and 29 males sampled by local hunters from 1999 to 2002. Liver samples were analyzed for several perfluorinated compound including C9, C10, C11, C12 and C13 PFCAs. Sixty-five percent of the polar bears had ∑PFA concentrations above 1000 ng/g-ww. In female bears, the ∑PFA ranged from 256 to 2770 ng/g-ww, and in male bears the ∑PFA ranged from 114 to 3052 ng/g-ww. Given that all PFA compounds in the analysis were summed, a direct cause-effect correlation with a particular perfluorinated compound, such as the long-chain PFCAs, is not possible to deduce. In addition, east Greenland polar bears are also contaminated with other substances such as organochlorines (PCBs, DDTs) and mercury which may be confounding synergistic co-factors in the development of the lesions. The authors concluded that the statistical analysis did not provide evidence as to whether chronic exposure to perfluorinated compounds is associated with liver lesions in polar bears; however, these lesions were similar to those produced by perfluorinated compounds under laboratory conditions (Sonne et al. 2007).

Characterization of Ecological Risk

The approach taken in this ecological screening assessment was to examine various supporting information and develop conclusions based on multiple lines of evidence such as persistence, exposure, trends, ecological risk, inherent toxicity, bioaccumulation and widespread occurrence in the environment.

In traditional toxicity studies, long-chain (C9-C20) PFCAs were found to be low to moderately toxic, with acute toxicity values ranging from 8.8 to 285 mg/L. There are two studies on the toxicity of long-chain (C9-C20) PFCAs in terrestrial species. In one study, no adverse effects were observed up to 1.0 mg/kg body weight for male chickens dosed with C10 PFCA. In another study, a soil-dwelling nematode showed acute lethality at 306 mg/L and multi-generation effects (decreased fecundity) at 0.000464 mg/L when exposed to C9 PFCA.

There is the potential for PFAs, including long chain (C9-C20) PFCAs to cause hepatotoxicity, i.e., liver lesions in polar bears at 114 – 3052 ng/g-ww total PFAs, and the activation of the PPARa in Baikal seal livers at 35.25 – 64.26 mg/L C9-C11 PFCAs. There is also the potential for long chain (C9-C20) PFCAs to cause affect endocrine function such as vitellogenesis in rainbow trout at 0.0256- 2000 ug/g diet C10 PFCA. C9-C10 PFCAs are also chemical sensitizers for the marine mussel, Mytilus californianus, by allowing normally excluded toxic substances to accumulate in the marine mussel. C12 and C14 PFCAs increased the mitochondrial membrane potential in the freshwater alga, Scenedesmus obliquus, indicating damage to the mitochondrial function.

Long-chain (C9-C20) PFCAs have been measured in the Canadian aquatic environment in concentrations ranging from <0.5 ng/L to 19 ng/L. C9-C12 PFCAs were measured in sediment from the Canadian Arctic ranging in concentration from 0.5 – 3.3 ng/g. C9 to C15 PFCAs were measured in the liver of seals, foxes, fish, polar bears, Greenland shark, narwhals, beluga whales and birds either in the Canadian Arctic or the Great Lakes region. Concentrations ranged from below detection levels to 180 ng/g liver-ww with concentrations greatest for polar bears followed by Greenland shark, narwhals and beluga whales. Worldwide, levels of C9 to C15 have been reported in ringed, fur and harbour seals, dolphins (i.e. white-sided, bottlenose, white-beaked, Franciscana, humpback), finless porpoises, glaucous gulls, sperm whale, beavers, Amur tigers, wild rats and several species of birds (little egret, little ringed plover, parrotbills, black-crowned herons). Concentrations ranged from below detection levels to 480 ng/g-ww, with concentrations highest in the white-beaked dolphin.

For C11 (2700<BCF<11000), C12 (18000<BCF<40000), and C14 (23000<BCF<30000) PFCAs, there is a high potential for bioconcentration in fish and potential for biomagnification in fish and marine mammals. For the other long-chain PFCAs, there is evidence of biomagnification and trophic magnification in the environment. There are no experimental or predicted bioaccumulation data available for long-chain PFCAs greater than C14. Nevertheless, there is the potential that long chain PFCAs could bioaccumulate or biomagnify in marine and/or terrestrial species based on chemical conformations. In addition, C14 and C15 PFCAs have been found in fish, invertebrates and polar bears.

Increasing trends of long-chain (C9-C20) PFCAs have been shown in polar bears, ringed seals and birds. From 1980 to 2000, C10 and C11 PFCAs in ringed seal livers from Greenland increased 3.3 and 6.8% per year, respectively. From 1992 to 2005, the mean concentrations of C9 and C10 PFCA in the livers of Baikal seals were 1.2 to 1.7-fold higher. From 1972 to 2002, mean doubling times for concentrations in polar bear livers from the Arctic ranged from 5.8 to 9.1 years for C9 to C11. From 1993 to 2004, concentrations in ringed seal liver samples increased, with a doubling time of 4 to 10 years for C9 to C12. In northern fulmar liver samples, C9 to C15 levels increased from 1987 to 1993 and remained steady from 1993 to 2003. Thick-billed murre liver samples showed an increase in C9 to C15 concentrations from 1975 to 2004. Concentrations of C9 to C13 increased significantly in whole eggs of herring gulls in Norway from 1983 to 1993. Temporal increases of C9-C12 were observed in male beluga whales from Nunavut at 1.8 ng/g-ww liver from 1980-2010.

Uncertainties in Evaluation of Ecological Risk

Certain data gaps and uncertainties exist such as limited physical and chemical properties, experimental persistence data, and toxicity data. There is, nonetheless, a substantial body of information on long-chain (C9-C20) PFCAs and their precursors. For example, while the mechanisms of transport of long chain (C9-C20) PFCAs and their precursors to the Arctic are not clear, they appear to be mobile in some form, as both the long-chain PFCAs (C9-C15) and their precursors have been measured in biota throughout the Canadian Arctic, far from known sources.

Environmental pathways of long-chain (C9-C20) PFCAs to biota are not well understood, as there are relatively few monitoring data on concentrations of various precursors in air, water, effluents and sediments in Canada. While mechanisms of toxic action of long-chain (C9-C20) PFCAs are not well understood, a range of toxicological effects, including vitellogenin induction and hepatotoxicity have been reported in a variety of species. In addition, analytical results from individual laboratories may not be directly comparable, according to studies by van Leeuwen et al. (2006), indicating variability in analytical results between individual laboratories.

There is also limited information on the toxicology of long-chain PFCA precursors, their relative contribution from different sources (e.g., significance of precursors from the degradation of fluorotelomer-based substances), and their potential for combined or synergistic effects with other perfluorinated compounds.