Persistence and Bioaccumulation Potential

The predicted half-life for atmospheric degradation of HBCD due to reaction with the hydroxyl radical is 2.13 days (AOPWIN 2000).

HBCD is not expected to undergo hydrolysis in the environment, due to a lack of hydrolyzable functional groups and low water solubility (Harris 1990; ACC 2002). Velsicol Chemical Corporation (1979) conducted a hydrolysis experiment using the commercial product, Firemaster 100. No significant hydrolysis occurred over the 39-day test period.

MITI (1992) observed only 1% biodegradation over 28 days in a ready biodegradation test for HBCD. The results indicate that the ultimate degradation half-life in water is likely to be much longer than 182 days (more than 5 years assuming first-order degradation kinetics) and that the substance is therefore likely to persist in this environmental compartment. Similarly no biodegradation was reported in 28-day ready biodegradation testing conducted using a composite sample of HBCD (purity 93.6%) comprised of 6.0% α-isomer, 8.5% ß-isomer and 79.1% γ-isomer (CMABFRIP 1996; ACC 2002).

Although experimental data on the biodegradation of HBCD in water are available, model estimates derived from quantitative structure-activity relationships (QSARs) were also considered (Environment Canada 2007; see Table 3). BIOWIN (2000) sub-model 4 predicts that HBCD is amenable to primary degradation (estimated half-life of = 182 days). However, with respect to ultimate degradation, sub-model 3 predicts that HBCD biodegrades slowly. Both BIOWIN (2000) sub-models 5 and 6 (both ultimate biodegradation models) also predict a low probability of rapid biodegradation. CPOPs (2008), which predicts ultimate biodegradation, estimates a biochemical oxygen demand (BOD) of only 0.1%, which further suggests very slow biodegradation. When results of the empirical ready biodegradation tests are considered together with the model data, it appears likely that HBCD will undergo some primary biodegradation in water but that the time to ultimate biodegradation may exceed 182 days, making the substance persistent in this medium. As well?as noted below?there is evidence for the formation of a stable and potentially persistent transformation product, 1,5,9-cyclododecatriene.

ACCBFRIP (2003b) and Davis et al. (2005) examined the degradability of HBCD using aerobic and anaerobic water/sediment microcosms and soils. Disappearance half-lives were 11 and 32 days in the aerobic microcosms,1.1 and 1.5 days in the anaerobic microcosms and 6.9 days for anaerobic soil. No degradation products were detected in the sediment, overlying water or headspace of the microcosms. In their analysis of the study, EU RAR (2008) noted that recoveries of HBCD in the test vessels varied from 33 to 125%, with most recoveries below 70%. An interfering chromatographic peak with characteristics identical to that of γ-HBCD was also present in one of the two river sediment samples, indicating possible contamination of the sample with HBCD. In addition, the very low initial HBCD concentration resulted in levels of the α- and ß-diastereomers being below detection limits by the completion of the test. For this reason, quantification was only possible for the γ-isomer, and no information is available on the fate of α- and ß-HBCD. This is particularly significant given the evidence for a predominance of the α-isomer in biota, suggesting that this isomer may have greater environmental stability (see Bioaccumulation section below). As no degradation products, including carbon dioxide, were identified in the study, biotic processes could not be conclusively linked to the observed rapid disappearance of HBCD, and the results are therefore presented in terms of disappearance times rather than biodegradation (EU RAR 2008).

In a high-quality study, EBFRIP (2004b) and Davis et al. (2006) investigated biodegradation of HBCD in activated and digester sludge, river sediment, and surface soil. The study objectives emphasized identification of degradation pathways and products, and transformation half-lives were not reported for the various test media. Substantial transformation occurred in the anaerobic digester sludge and in freshwater aerobic and anaerobic sediment microcosms. Degradation rates were slower in the activated sludge samples, and no degradation of HBCD was observed in the aerobic soil microcosms. Tetrabromocyclododecene, dibromocyclododecadiene and 1,5,9-cyclododecatriene were identified as primary biotransformation products, providing evidence that degradation of HBCD in the environment may occur through a process of sequential debromination.

Gerecke et al. (2006) reported a degradation half-life of 0.66 days for technical HBCD incubated with digested sewage sludge under anaerobic conditions. Beta- and γ-HBCD degraded more rapidly than α-HBCD, leading the researchers to propose that differential degradation rates may contribute to the relative enrichment of α-HBCD observed in biota samples. Findings from the study contrasted with those of EBFRIP (2004b), which determined there were no differences in the transformation behaviour of the three isomers.

No information could be found on the degradation properties and toxicities of tetrabromocyclododecene and dibromocyclododecadiene; however, some limited data are available for 1,5,9-cyclododecatriene, the final debromination product. The substance is classified as not readily biodegradable, with only 1% biodegradation observed in standard 28-day ready biodegradation testing (du Pont 2003). Bridié et al. (1979a, 1979b) measured a BOD of 0.02 g/g and a 24-h LC₅₀ (median lethal dose) for goldfish (Carassius auratus) of 4 mg/L, suggesting that 1,5,9-cyclododecatriene is resistant to microbial oxidation processes and is potentially toxic to aquatic species. Other measured and estimated data support the finding that the substance presents high hazard to aquatic organisms. For instance, NITE (2002) reports a 48-hour LC₅₀ of 0.166 mg/L for rice fish (Oryzias latipes), and ECOSAR (2009) predicts acute toxicity to aquatic organisms below 1 mg/L (i.e., fish 96-hour LC₅₀ = 0.104 mg/L; daphnid 48-hour LC₅₀=0.098 mg/L; and green algae 96-hour EC₅₀ = 0.214 mg/L, Appendix A). Data from NITE (2002) further indicate that the substance has a high bioconcentration potential, with measured BCFs for carp of 2360 to 12 500 and 1920 to 14 800, resulting from 10-week exposures to 0.01 and 0.001 mg/L, respectively. Using the Arnot and Gobas (2003) bioaccumulation model, calculated BCF values for 1,5,9-cyclododecatriene range from 9813 (corrected for metabolic transformation) to 18 620 L/kg (no metabolism), and BAF values range from 66 360 (corrected for metabolism) to 177 828 (no metabolism) (Appendix A). Enhanced aerobic ready biodegradation testing conducted using the isomer trans, trans, trans-1,5,9-cyclododecatriene determined that although the substance is not readily biodegradable, it will undergo primary biodegradation following a lag phase of approximately 14 days (EBFRIP 2006). Conclusive results with respect to complete mineralization were not possible from the study. A subsequent study conducted under similar conditions and using lower test concentrations (Davis 2006) documented the formation of carbon dioxide over the course of the 77-day test period, indicating that mineralization of the substance was occurring under the conditions of the study. While this study provides evidence that 1,5,9-cyclododecatriene will biodegrade under the conditions of enhanced aerobic ready biodegradation testing, information is needed on the potential for biodegradation under low oxygen conditions, as these are most likely to prevail in subsurface layers of the soil and sediment compartments to which HBCD preferentially partitions. Additionally, complete mineralization of HBCD has not yet been demonstrated, an indication that degradation products such as 1,5,9-cyclododecatriene remain stable under some study conditions. Based on the available information, 1,5,9-cyclododecatriene is considered to be potentially persistent in the environment.

Sediment core studies in Europe and Japan have reported HBCD concentrations in sediment layers that date back to the 1960s and 1970s (Remberger et al. 2004; Minh et al. 2007; Kohler et al. 2008; Tanabe 2008). For example, Rembergeret al. (2004) measured concentrations of HBCD in sediment layers approximately 30 and 40 years old in cores from the Stockholm archipelago; these concentrations were 25 - 33% of HBCD concentrations found in the top layer of the cores. Such studies suggest that degradation half-lives under field conditions are not as fast as simulation degradation studies (e.g., ACCBFRIP 2003b) might indicate (EU RAR 2008).

In summary for sediment, data for HBCD suggest that the substance is persistent in sediment. Primary degradation half-lives are relatively long, but likely less than 365 days. However, ultimate degradation half-lives are likely much longer than 365 days based on an extrapolation ratio of 1:4 for a water:sediment biodegradation half-life (Boethling et al. 1995). Furthermore, sediment core measurements suggest that degradation in the environment may be on the order of years to decades. Information gathered to date on the HBCD degradation products suggests that these products are expected to be bioaccumluative and toxic, like HBCD itself.

ACCBFRIP (2003c) also investigated the degradation of HBCD in aerobic and anaerobic soil microcosms. An average HBCD decrease of 75% was observed in the aerobic soil microcosms over the 119-day test period. In the anaerobic test system, HBCD decreased by 92% over 21 days in the test microcosms. Based on the results of the study, disappearance half-lives of 63 and 6.9 days were determined in the aerobic and anaerobic soils, respectively. No degradation products were detected in the soil or headspace of the microcosms. EU RAR (2008) noted that, as with the water/sediment microcosm study described above, only the γ-isomer was quantified and therefore this study provides no information on the fate of α- and ß-HBCD in soil. As well, only one soil type was tested, making it difficult to evaluate the representativeness of the determined half-lives to conditions in the environment. Finally, in the absence of identified transformation products, the mechanism behind the observed disappearance of HBCD remains unclear and may in part be due to adsorption to soil, given the large differences observed between measured and nominal HBCD concentrations in the soil at test initiation (EU RAR 2008).

The absence of observable degradation in the aerobic soil microcosms of EBFRIP (2004b) contrasted markedly with results obtained by ACCBFRIP (2003c), which reported a disappearance half-life of 63 days in aerobic soils. The test substances used in the two studies were comparable in composition, although the dosing was higher in EBFRIP (2004b) and the test substance contained a higher proportion of γ-isomer, making it closer in composition to the current commercial product. The test soils were collected at different times of year (April for ACCBFRIP 2003c and November for EBFRIP 2004b) from the same site in North Dakota (EBRIP 2004b), and exposure periods were of comparable duration (119 vs. 112 days). The longer pre-stabilization period of 35 days used in the ACCBFRIP (2003c) study may have produced a more stable microbial population at test initiation; however, the 15-day period employed by EBFRIP (2004b) was well within the OECD Guideline’s recommended range of 2 days to 4 weeks (OECD 2002). A key difference was the addition of activated sludge to the microcosms of ACCBFRIP (2003c), a procedure designed to investigate possible degradation outcomes following the addition of biosolids containing HBCD to surface soils during land treatment. While ACCBFRIP (2003f) reported an almost 30% inhibition of activated sludge micro-organisms following treatment with HBCD, it is likely that the presence of these organisms in the soil microcosms of ACCBFRIP (2003c) significantly enhanced degradation rates relative to those of EBFRIP (2004b).

In summary for soils, existing data for HBCD suggest that the substance is persistent in soil. The ultimate degradation half-life in soil is likely much longer than 182 days, based on an extrapolation ratio of 1:1 for a water:soil biodegradation half-life (Boethling et al. 1995). Primary degradation rates appear to be variable, but may also be longer than 182 days (EBFRIP 2004b).

Based on empirical and modelled data, HBCD meets the persistence criteria in air, water, soil and sediment (half-life in air = 2 days, half-lives in soil and water = 182 days, and half-life in sediment = 365 days) as set out in the Persistence and Bioaccumulation Regulations (Canada 2000).

Wania (2003) used a modelling approach to evaluate the potential for long-range atmospheric transport of HBCD and concluded that, based on physical and chemical properties, the substance should have low potential to reach remote areas. In a subsequent study, Brown and Wania (2008) identified HBCD as a potential Arctic contaminant based on an atmospheric oxidation half-life of greater than two days and structural similarities to known Arctic contaminants. The low volatility of HBCD likely results in significant sorption to atmospheric particulates and for this reason, the long-range transport potential of HBCD may depend upon the transport behaviour of the atmospheric particulates to which it sorbs. HBCD has been measured in air, sediment and biota samples collected from remote sites such as the Arctic (e.g., Remberger et al. 2004; Verreault et al. 2005, 2007a, 2007b; Muir et al. 2006; Evenset et al. 2007; Svendsen et al. 2007; Tomy et al. 2008). As there is no evidence for the natural production of HBCD, these data are indicative of contamination from anthropogenic sources. While this contamination may be local in origin, it is also possible that the findings represent evidence that under some circumstances HBCD may be capable of atmospheric transport over long distances and to remote locations. Based on the available information, it is considered that HBCD meets the persistence criterion of being subject to atmospheric transport from its source to a remote area, as specified in CEPA 1999 (see Table 4).

Additional evidence for the persistence of HBCD is its potential for biomagnification (see section below: studies by Morris et al. 2004; Tomy et al. 2004a; and Law et al. 2006a). The occurance of biomagnification is also indicative of environmental persistence and/or a lack of significant metabolism, for in order to biomagnify significantly, a substance must persist long enough to be transferred successively from lower to higher trophic levels and/or not be subject to metabolic transformation.

Veith et al. (1979) measured a bioconcentration factor (BCF) of 18 100 in fathead minnow, Pimephales promelas, exposed to 0.0062 mg/L HBCD for 32 days, while CMABFRIP (2000) calculated bioconcentration factor values ranging from 4650 to 12 866 in rainbow trout, Oncorhynchus mykiss, exposed for 35 days to 0.0034 mg/L HBCD.

Law et al. (2006b) and Law (2006) measured biomagnification factors (BMFs) of 9.2, 4.3 and 7.2 for α-, ß- and γ-HBCD, respectively, by exposing juvenile rainbow trout, Oncorhynchus mykiss, to single isomer concentrations ranging from 12 ng/g to 29 ng/g lipid weight in the diet. Bioaccumulation of γ-HBCD was linear, while that of α- and ß-HBCD increased exponentially with respective doubling times of 8.2 and 17.1 days. Both ß- and γ-HBCD followed first-order depuration kinetics, with depuration rate constants (k_d) of 0.44 x 10^-2 and 0.48 x 10^-2 d^-1 and calculated half-lives of 157 (±71) and 144 (±60) days, respectively. A k_d value and half-life could not be calculated for α-HBCD, since depuration out of the muscle tissue did not obey a first-order rate process. Assimilation efficiencies, calculated by comparing concentrations measured in the fish with those in the food, were determined to be 31.1, 41.4 and 46.3% for α-, ß- and γ-HBCD, respectively. Bioisomerization of HBCD was also reported in the study, with statistically significant amounts of α-HBCD measured in the muscle tissue of trout exposed exclusively to the γ-isomer. Similarly, both α- and γ-HBCD were present in statistically significant quantities in fish exposed only to ß-HBCD. The results suggested that juvenile rainbow trout were able to bioisomerize the ß- and γ-isomers of HBCD, with preferential formation of the α-isomer. The α-isomer appeared recalcitrant to bioisomerization in this fish species. Selective bioisomerization of HBCD has the potential to contribute appreciably to determining isomer distributions within organisms.

Tomy et al. (2004a) reported a strong positive linear correlation between tissue concentrations of HBCD and trophic level in a Lake Ontario pelagic food web, evidence that bioaccumulation and biomagnification was occurring within the web. Species examined in the study included a top predator?lake trout (Salvelinus namaycush)?and prey species such as alewife (Alosa pseudoharengus), rainbow smelt (Osmerus mordax), slimy sculpin (Cottus cognatus), mysid (Mysis relicta), amphipod (Diporeia hoyi) and zooplankton, such as copepods and cladocerans. Lipid-normalized BMFs exceeded 1 for most feeding relationships, and ranged from 0.4 to 10.8 for the α-isomer and 0.2 to 9.9 for γ-HBCD. A BMF for the ß-isomer was not determined from the study. A trophic magnification factor was calculated for HBCD in the food web by comparing HBCD concentrations with those of the stable nitrogen 15 isotope (d15N). Trophic magnification factors of around 0 suggest that a chemical moves through the food web without being biomagnified, while those exceeding 1 indicate that biomagnification is occurring (Broman et al. 1992; Fisk et al. 2001). A trophic magnification factor of 6.3 was calculated for HBCD, comparable to that of known biomagnifying substances, such as the persistent organochlorines p,p'-DDE (6.1) and polychlorinated biphenyls (PCBs) (5.7).

Law et al. (2006a) calculated trophic magnification factor values for a Lake Winnipeg pelagic food web, using zooplankton, mussels (Lampsilis radiate), walleye (Stizostedion vitreum), whitefish (Coregonus commersoni), emerald shiner (Notropis atherinoides), burbot (Lota lota), white sucker (Catostomus commersoni) and goldeye (Hiodon alosoides). The trophic magnifcation factors were 2.3, 2.3 and 4.8 for α-, ß- and γ-HBCD, respectively, while that for total HBCD was 3.1. The highest individual biomagnifaction factors were associated with the predator/prey pairs of goldeye/mussel (8.2), burbot/emerald shiner (6.3), walleye/whitefish (5.3), burbot/mussel (5.0) and emerald shiner/plankton (5.0). The results indicated that biomagnification was occurring, but at a lesser rate than it was taking place in a comparable Lake Ontario food web (Tomy et al. 2004a).

Biomagnification of HBCD in a North Sea food web was evaluated by comparing concentrations in species from various trophic levels (Morris et al. 2004). Amounts in top predators, such as harbour porpoise (Phocoena phocoena) and harbour seal (Phoca vitulina), were several orders of magnitude higher than those measured in aquatic macro-invertebrates such as sea star (Asterias rubens) and common whelk (Buccinium undatum) collected from the same area. Similarly, high concentrations were detected in liver samples from cormorant (Phalacrocorax carbo), a top predator bird, and in eggs of the common tern (Sterna hirundo). Intermediate amounts were found in cod (Gadus morhua) and yellow eel (Anguilla anguilla). Results from the study were considered to indicate bioaccumulation and biomagnification up the aquatic food chain.

Velsicol Chemical Corporation (1980) reported rapid metabolism of HBCD in the blood, muscle, liver and kidneys of rats given a single oral dose of radiolabelled substance. Elimination occurred primarily via the feces (70%) and urine (16%), with 86% of the radiocarbon removed over the three days following dosing. The test substance distributed throughout the body, with the highest amounts in the fatty tissue, followed by the liver, kidney, lung and gonads. HBCD remained mostly unchanged in fatty tissue. The study concluded that HBCD was capable of accumulating in the fatty tissue of rats following repeated exposure.

CMABFRIP (2001) examined the presence of individual diastereomers in adipose tissue of rats dosed with 1000 mg/kg body weight per day for up to 90 days. Concentrations of the α-isomer exceeded those of ß- and γ-HBCD, accounting for 65% to 70% of the total HBCD present. Gamma-HBCD accounted for 14% to 20% of the total, while the ß-isomer was present at from 9% to 15%. This contrasted markedly with proportions present in the test substance, which contained 84.5% γ-isomer, 8.9% α-isomer and 6.6% ß - isomer. The highest tissue concentrations were measured on study day 89, and the amounts were consistently higher in female rats as compared with males.

Although empirical bioaccumulation data are available for HBCD, QSARs were also applied (Environment Canada 2007) using the predictive models shown in Table 5. Model estimates range from approximately 275 400 to 6 457 000 for the BAF and from 20 400 to 24 000 for the BCF.

Based on empirical and modelled data, HBCD meets the criteria for bioaccumulation (bioaccumulation and bioconcentration factors of 5000 or more) as set out in the Persistence and Bioaccumulation Regulations (Canada 2000).

While Canadian and North American exposure data are limited, HBCD has been detected in all environmental media in many parts of the world, with highest levels occurring near urban and industrial areas (see Tables 6 and 7).

Air

Concentrations of up to 0.011 ng/m³ were measured in the particle phase of air samples collected in 2002 and 2003 at five sites from Lake Michigan through the U.S. Midwest to the Gulf of Mexico (Hoh and Hites 2005). Based on similarities in spatial concentration patterns of HBCD and the brominated diphenyl ether flame retardant PBDE-209 (decabromodiphenyl ether), the researchers speculated that the brominated flame retardant market may be shifting from diphenyl ether products to HBCD (Hites and Hoh 2005).

Precipitation samples collected from the Great Lakes basin contained up to 35 ng/L (Backus et al. 2005). All three major diastereomers were detected, with an average distribution of 77%, 15% and 8% for α-, ß- and γ-HBCD, respectively.

European concentrations tend to be higher than those measured in North America. Remberger et al. (2004) analyzed HBCD in air and rainfall samples collected in 2000 and 2001 from various locations in Sweden . Air concentrations near potential sources (e.g., an extruded polystyrene manufacturing facility, landfill for construction and demolition waste, textile industry facility) ranged from 0.013 ng/m³ to 1070 ng/m³ while those at urban stations in Stockholm were 0.076 ng/m³ to 0.61 ng/m³. The highest concentration, 1070 ng/m³, was recorded close to the exhaust of an air ventilation system at an extruded polystyrene manufacturing facility.

Surface Waters

Law et al. (2006a) reported a mean dissolved phase concentration of 0.011 ng/L for α-HBCD in surface water samples collected from the south basin of Lake Winnipeg in 2004. Beta- and γ-HBCD were not detected (detection limit: 0.003 ng/L). The researchers commented that detection of only α-HBCD in the samples was consistent with its much greater aqueous solubility (4.88 x 10⁴ ng/L; see Table 2) relative to that of the ß- (1.47 x 10⁴ ng/L) and γ- (2.08 x 10³ ng/L) isomers. Surficial sediment grab samples from the same region contained a mean concentration of 0.05 ng/g dry weight of γ-HBCD. Alpha- and ß-HBCD were not detected in the samples (detection limit: 0.04 for ß- and γ-HBCD to 0.08 ng/g dry weight for α-HBCD). The results were consistent with the γ-isomer being the most hydrophobic of the three isomers.

In a draft study, filtered surface water and suspended solids samples were collected upstream of a sewage treatment plant in the United Kingdom (U.K.). Filtered water samples contained 57 ng/L to 1520 ng/L; HBCD was not detected (detection limit: 50 ng/L) in a single sample taken approximately one kilometre downstream of the plant (Deuchar 2002). Concentrations in the suspended solids of the upstream samples were up to 1310 ng/L, while the single downstream sample contained 215 ng/L. Two U.K. locations considered remote from industrial activity contained from less than 50 g/L to 210 ng/L.

Sediment

Marvin et al. (2004, 2006) measured HBCD in suspended sediments collected along the Detroit River from Lake St. Clair to the outflow to Lake Erie, and determined that occurrence of the substance was strongly associated with urban and industrial activities. Annual mean concentrations ranged from 0.012 ng/g to 1.14 ng/g dry weight, with the highest levels being found downstream of the urban region surrounding the city of Detroit. About two thirds of the samples had isomeric profiles similar to those found in commercial technical mixtures, with a predominance of the γ-isomer, while the remaining samples were dominated by the α-isomer. The ß-isomer was present at substantially lower levels, consistent with its lower prevalence in commercial mixtures. The researchers concluded that distribution of HBCD in the Detroit River appeared to be heavily influenced by HBCD associated with shoreline-based urban and industrial activities. In addition, the widespread occurrence of relatively low concentrations suggested that large urban areas may act as diffuse sources of HBCD.

Four surficial sediment grab samples collected in 2003 from four sites in the south basin of Lake Winnipeg contained a mean concentration of 0.05 ng/g dry weight γ-HBCD (Law et al. 2006a). Alpha- and ß-HBCD were not detected in the samples (detection limit: 0.04 ng/g for ß- and γ-HBCD to 0.08 ng/g dry weight for α-HBCD). The researchers commented that the results were consistent with the γ-isomer being the most hydrophobic of the three isomers.

Concentrations of less than 1.7 ng/g to 1680 ng/g dry weight were measured in river and estuarine sediments collected from 2000 to 2002 at various locations throughout the U.K. (Morris et al. 2004). The highest concentration occurred close to a brominated fire retardant manufacturing plant in northeast England that closed in 2003 and was demolished in 2004 (EU RAR 2008). The same study examined sediments from the region surrounding the Western Scheldt (the Netherlands ) and Scheldt Basin (Belgium). Concentrations of up to 950 ng/g dry weight were measured in the samples, with highest levels occurring near areas of industrial activity. Most samples contained isomeric patterns closely resembling that of the commercial formulations, with a predominance of γ-HBCD. In some instances, however, sediments contained higher percentages of α- and ß-HBCD. Thermal rearrangement of HBCD isomers at temperatures greater than 160°C has been documented, resulting in the conversion of γ-HBCD into the α-isomer (Peled et al. 1995). As these temperatures are commonly employed in processes to incorporate HBCD into a polymer matrix, the presence of higher proportions of α- and ß-isomers in the sediment samples was considered to indicate use of HBCD in processing operations such as polymer and textile applications (Morris et al. 2004).

Soil

The existing literature contains few references to soil concentrations of HBCD. Four shallow soil samples (actual depth not provided) taken from the vicinity of a U.K. flame retardant coating manufacturing facility in 1999 contained 18 700 to 89 600 ng/g dry weight HBCD (mean concentration 62 800 ng/g dry weight) (Dames and Moore 2000a). Remberger et al. (2004) analyzed soil samples collected in 2000 at distances of 300 m, 500 m and 700 m from a Swedish facility known to manufacture extruded polystyrene with HBCD. Concentrations of HBCD in the samples ranged from 140 ng/g to 1300 ng/g dry weight, and decreased with increasing distance from the plant.

Waste Effluent and By-products

No North American data on concentrations in waste treatment products were found in the literature.

Morris et al. (2004) sampled landfill leachates in 2002 from sites in southeast England , Ireland and the Netherlands . HBCD was not detected in the U.K. samples (detection limits: 15 ng/L for the dissolved phase and 3.9 ng/g dry weight for the particulate phase; de Boer et al. 2002). However, concentrations of 2.5 ng/g to 36 000 ng/g dry weight (mean 5906 ng/g dry weight) were measured in the samples collected in the Netherlands . The substance occurred only in the particulate phase, and the ?-isomer predominated in the samples.

Concentrations of 3 ng/L and 9 ng/L were measured in two leachate samples collected in 2000 at a landfill site for construction and demolition waste near Stockholm (Remberger et al. 2004). Sediment from the leachate sedimentation basin contained less than the detection limit of 0.1 ng/g dry weight.

Concentrations of up to 29.4 ng/g dry weight (particulates) and 24 ng/L (dissolved phase) were measured in influent samples collected in 2002 from five sewage treatment plants in southeast England (Morris et al. 2004). The substance was not detected (detection limit: 3.9 ng/g dry weight) in the effluents, but was present at 531 ng/g to 2683 ng/g dry weight (mean 1401 ng/g dry weight) in sludge samples taken from the sites. The γ-isomer predominated in the samples, with α- and ß-HBCD present in smaller and almost equal quantities. The researchers proposed that release of HBCD from contaminated dust, such as office dust containing brominated flame retardants, may account, at least in part, for the presence of the substance in sewage treatment plant influents and sludge.

Sludge sampled in 2000 from 50 sewage treatment plants throughout Sweden contained from 3.8 ng/g to 650 ng/g dry weight (mean 45 ng/g dry weight; Law et al. 2006c). Higher concentrations occurred in samples collected near known or suspected sources, such as textile industries, producers of extruded polystyrene and a company that upholstered cars.

HBCD was present in all of 19 samples collected from 16 Swiss wastewater treatment plants from May to July 2003 and in January 2005 (Kupper et al. 2008). Concentrations in the samples ranged from 39 ng/g to 597 ng/g dry weight, with a mean value of 149 ng/g dry weight and a median of 123 ng/g dry weight.

Zennegg et al. (2005) reported concentrations of 19 to 170 ng/g dry weight (mean 85 ng/g dry weight) in urban compost collected from six composting facilities in Switzerland . The study also evaluated levels of several other brominated flame retardants, including polybrominated diphenyl ethers (PBDE congeners 28, 47, 99, 100, 153, 154, 183 and 209) and tetrabromobisphenol A. HBCD was the most prominent brominated flame retardant in the samples.

Biota

HBCD has been detected in North American organisms, as well as organisms from other parts of the world.

Archived samples of Lake Ontario lake trout, Salvelinusnamaycush, contained from 16 ng/g to 33 ng/g lipid weight (2 ng/g to 4 ng/g wet weight) total HBCD, with the amounts decreasing significantly between 1979 and 2004 (Ismail et al. 2009). The α-isomer predominated in the samples (15 ng/g to 27 ng/g lipid weight; 1.7 ng/g to 3.4 ng/g wet weight), with lower levels of ß- (0.16 ng/g to 0.94 ng/g lipid weight; 0.03 ng/g to 0.11 ng/g wet weight) and γ-HBCD (1.4 ng/g to 6.5 ng/g lipid weight; 0.23 ng/g to 0.77 ng/g wet weight). The researchers proposed that alterations to food web processes in the lake, such as changes to the lake trout diet and/or changes at the base of the food web, as well as possible temporal variations in contaminant loadings and voluntary emission-limiting measures undertaken by industry, may be factors in the downward trend in concentration. However, the need for further research was emphasized, given the conflicting evidence of increasing temporal trends reported in other studies (see below).

Mean concentrations ranging from 3 ng/g to 65 ng/g lipid weight were measured in fish, mussels and zooplankton collected from the south basin of Lake Winnipeg between 2000 and 2002 (Law et al. 2006a). The ß-isomer was consistently detected at much lower levels than were the α- and γ-isomers, while the proportions of α- and γ-HBCD varied between species.

Tomy et al. (2004a) examined bioaccumulation and biomagnification of HBCD in a Lake Ontario pelagic food web by measuring concentrations in lake trout (Salvelinus namaycush, a top predator) and several of its major prey. Alpha- and γ-HBCD were detected at all trophic levels, with the highest concentrations present in lake trout (mean total HBCD 1.68 ng/g wet weight). Concentrations of α-HBCD were consistently higher than those of γ-HBCD, while the ß-isomer was below the method detection limit (estimated at 0.03 ng/g wet weight) in all the species tested.

Pooled homogenates of herring gull (Larus argentatus) eggs collected from six colonies around the Great Lakes contained from 2.1 ng/g to 20 ng/g wet weight α-HBCD (Gauthier et al. 2007). Highest levels were measured at Gull Island on northern Lake Michigan, likely a result of this lake being the most urbanized and industrialized of the Great Lakes (Norstrom et al. 2002). Beta-HBCD was not detected in the samples; however, low levels of γ-HBCD were present in two of the six. It should be noted, however, that the southern portions of the lake are more heavily industrialized as compared to the areas from which the samples were taken. The findings confirm the presence of HBCD in the aquatic food web associated with herring gulls in the Great Lakes, with mother gulls exposed via their diet and subsequent in vivo transfer to the eggs (Gauthier et al. 2007).

HBCD was not detected (detection limit: 0.01 ng/g wet weight) in 29 blood samples collected from 2001 to 2003 from nestling bald eagles (Haliaeetus leucocephalus) in British Columbia and southern California (McKinney et al. 2006). Sampling was conducted at four locations in southwestern British Columbia (Barkley Sound, Nanaimo/Crofton, Delta/Richmond, Abbotsford/Chilliwack), one location in northern B.C. (Fort St. James) and one southern California site (Santa Catalina Island).

Blubber and liver samples collected from Atlantic white-sided dolphin (Lagenorhynchus acutus) stranded on the east coast of the United States between 1993 and 2004 contained from 14 ng/g to 280 ng/g wet weight (19 ng/g to 380 ng/g lipid weight) and 0.051 ng/g to 3.6 ng/g wet weight (2.9 ng/g to 140 ng/g lipid weight), respectively (Peck et al. 2008). The α-isomer was present in all samples, while ß- and γ-HBCD were not detected (detection limit: 0.4 ng/g wet weight for both isomers). No significant trend in concentration over time was evident in the samples.

Almost all (50 out of 52) fish samples collected in 2003 from Chesapeake Bay of the northeastern United States contained at least one stereoisomer of HBCD (Larsen et al. 2005). Total HBCD concentrations ranged from 1.0 ng/g lipid weight (white perch) to 73.9 ng/g lipid weight (channel catfish), with the highest levels measured in samples collected from historically contaminated areas. Isomer distributions differed significantly between benthic fish (e.g., catfish, eel), which had a predominance of α-HBCD, and pelagic species (e.g., striped bass), in which the γ-isomer dominated.

Johnson-Restrepo et al. (2008) measured concentrations in the blubber of bottlenose dolphin (Tursiops truncates) and the muscle tissue of bull shark (Carcharhinus leucas) and Atlantic sharpnose shark (Rhizoprionodon terraenovae) collected from the coastal waters of Florida from 1991 to 2004. HBCD was present in all samples at concentrations ranging from 0.460 ng/g to 72.6 ng/g lipid weight in bottlenose dolphin, 9.15 ng/g to 413 ng/g lipid weight in bull shark, and 1.83 ng/g to 156 ng/g lipid weight in Atlantic sharpnose shark. The α-isomer predominated in the samples, although most also contained smaller amounts of both ß- and γ-HBCD.

Concentrations in European biota tend to be higher than those measured in North America, likely reflecting the substantially higher market demand for HBCD in Europe and possibly the higher human population density.

Allchin and Morris (2003) reported concentrations of 39.9 - 75 ng/g wet weight in yellow eel (Anguilla anguilla) and < 1.2 - 6758 ng/g wet weight in brown trout (Salmo trutta) collected from eight locations along the rivers Skerne and Tees in the United Kingdom.

Morris et al. (2004) examined biomagnification in the North Sea food web by comparing concentrations present in species from various trophic levels from 1998 to 2001. The highest levels were found in top predator species, such as harbour porpoise (Phocoena phocoena; 440 - 6800 ng/g lipid weight), harbour seal (Phoca vitulina; 63 - 2055 ng/g lipid weight) and cormorant (Phalacrocorax carbo; 138 - 1320 ng/g lipid weight) and in the eggs of the common tern (Sterna hirundo; 330 - 7100 ng/g lipid weight). HBCD was also present in cod (Gadus morhua; maximum 50 ng/g lipid weight), yellow eel (Anguilla anguilla; maximum 690 ng/g lipid weight), sea star (Asterias rubens; maximum 84 ng/g lipid weight) and common whelk (Buccinium undatum; maximum 47 ng/g lipid weight). The α-isomer strongly dominated the diastereomeric profile, particularly in top predator species such as fish.

HBCD was detected in all of 85 samples of harbour porpoise blubber collected from 1994 to 2003 from animals stranded or caught in waters off the U.K. coast (Law et al. 2006d). The α-isomer predominated in the samples, with concentrations ranging from 10 ng/g to 19 200 ng/g wet weight. Concentrations in the blubber increased sharply from about 2001 onward, suggesting changing patterns in the use of HBCD. The researchers postulated that limitations on production and use of two commercial polybrominated diphenyl ether (PBDE) formulations (i.e., commercial pentaBDE and octaBDE) may have been driving the increase, since HBCD may be being used as a substitute for these formulations in some applications.

In a subsequent study, analyses were conducted of an additional 138 samples collected from the same region from 2003 to 2006 (Law et al. 2008). Concentrations of total HBCD in the samples ranged from less than 10 ng/g to 11 500 ng/g wet weight (up to 12 800 ng/g lipid weight), with the maximum value determined for an animal stranded or caught in 2003. A statistically significant decrease in levels was seen between 2003 and 2004, with the downward trend continuing between 2004 and 2006. The researchers attributed this to possibly being the result of the closure in 2003 of an HBCD manufacturing plant in northeastern England and two voluntary schemes to reduce emissions to the environment that took effect in 2006.

Lindberg et al. (2004) analyzed peregrine falcon (Falco peregrinus) eggs collected from 1991 to 1999 from wild and captive breeding populations in Sweden . Eggs from a northern wild breeding population contained 34 - 590 ng/g lipid weight, while those from the south contained 79 - 2400 ng/g lipid weight. HBCD was not detected in eggs collected from the captive breeding population (detection limits: 4 - 8 ng/g lipid weight). Dietary differences were considered primarily responsible for the observed range in HBCD levels. Birds from the northern wild population prey mainly on aquatic species, such as waders and ducks, while those in the south feed on birds in the terrestrial food web (Lindberg and Odsjö 1983). The captive breeding population received a controlled diet of domestic chickens. These samples were later re-examined alongside eggs collected from the same regions from 1987 to 1999. These tests confirmed higher concentrations of HBCD in the two wild populations compared with the concentrations in the captive population (Johansson et al. 2009).

Studies from Asia indicate that HBCD is widely distributed among aquatic species in the Asia-Pacific region. Ueno et al. (2006) reported a maximum concentration of 45 ng/g lipid weight in muscle samples of skipjack tuna (Katsuwonus pelamis) collected from 1997 to 2001 in offshore waters near Japan, Taiwan, the Philippines, Indonesia, the Seychelles and Brazil, as well as various locations in the Japan Sea, East and South China seas, Indian Ocean and North Pacific Ocean. The presence of HBCD in all but three of the 65 samples, including those taken from remote regions in the mid-Pacific Ocean, was considered evidence of widespread contamination in the global marine environment. Similar concentrations were observed in tuna collected from remote regions of the North Pacific Ocean (up to 29 ng/g lipid weight) and those from coastal Asian areas (28-45 ng/g lipid weight in samples from off the coast of Japan and East China Sea). This was considered indicative of an unknown local pollution source in the North Pacific or evidence of long-range atmospheric transport of HBCD with subsequent deposition in cold-water regions through the process of global distillation, or both. Other recent studies report the presence of HBCD in aquatic invertebrates (Ramu et al. 2007), fish (Xian et al. 2008) and marine mammals (Isobe et al. 2008) collected from coastal areas of Korea and China, as well as terrestrial vertebrates in Japan (Kunisue et al. 2008).

Presence in Remote Regions

HBCD has been measured in air, sediment and biota collected in regions considered to be remote from potential sources, including the Arctic.

Remberger et al. (2004) reported concentrations of up to 0.28 ng/m³ in air samples collected at remote sampling locations in Sweden and the Arctic areas of Finland.

Concentrations of 0.43 ng/g dry weight (α-isomer) and 3.88 ng/g dry weight (γ-isomer) were measured in sediment collected from Lake EllasjØen on Bjornoya (Bear Island) in the Norwegian Arctic (Evenset et al. 2007). The ß-isomer was not detected in the samples (detection limit: 0.06 ng/g dry weight).

Yolk of newly hatched European shag (Phalacrocorax aristotelis), a fish-eating top predator related to the cormorant, contained a mean concentration of 417 ng/g lipid weight of HBCD (Murvoll et al. 2006a). The samples were collected in 2002 from a Norwegian island considered remote and free from pollution. HBCD was present in all of 30 samples. The samples were also analyzed for several of the more persistent and bioaccumulative PBDE congeners. The mean concentration of HBCD in the yolk samples exceeded that of any PBDE congener measured, including PBDE-47 (mean concentration of 5.59 ng/g wet weight), PBDE-99 (1.56 ng/g wet weight) and PBDE-100 (6.16 ng/g wet weight), as well as total PBDEs (17.2 ng/g wet weight; sum of seven tri- to hexaBDE congeners).

A similar study was conducted on North Atlantic kittiwake (Rissa tridactyla) collected from an island off Norway and at Svalbard in the Norwegian Arctic (Murvoll et al. 2006b). Yolk sacs collected from newly hatched chicks contained mean concentrations of 260 ng/g lipid weight (island location) and 118 ng/g lipid weight (Arctic location). The presence of HBCD in Arctic kittiwake hatchlings provides further evidence of possible transport of the substance to regions remote from its source.

Muir et al. (2006) reported total HBCD in adipose tissue of polar bears (Ursus maritimus) from Alaska, Eastern Greenland and Svalbard in the Norwegian Arctic. Concentrations of up to 35.1 ng/g lipid weight were measured in two of eight female bears collected from 1994 to 2002 in the Bering-Chukchi Sea of Alaska. Male bears in the region contained no detectable HBCD (detection limit: 0.01 ng/g lipid weight). HBCD was present in all 11 samples collected from 1999 to 2001 from female polar bears in Eastern Greenland. Concentrations ranged from 32.4 ng/g to 58.6 ng/g lipid weight in the samples. HBCD was also present in all 15 samples collected in 2002 from female bears in the Svalbard area, with concentrations of 18.2 - 109 ng/g lipid weight.

Concentrations of 0.07 - 1.24 ng/g wet weight were measured in the blood plasma of adult glaucous gulls (Larus hyperboreus) collected in the Norwegian Arctic during May and June 2004 (Verreault et al. 2005). Plasma collected from female polar bears (Ursus maritimus) living in the same region contained up to 0.85 ng/g wet weight. While HBCD was present in all 27 gull samples, only 2 of the 15 polar bear plasma samples contained levels above the detection limit (0.03 ng/g wet weight). The researchers hypothesized that the lower occurrence in the bears may indicate a superior ability to detoxify and eliminate HBCD. Alternatively, the lower levels may reflect differences in diet and feeding rate between the two species. Plasma levels averaged 1.73 - 2.07 ng/g wet weight in gulls collected from the same region in May and June of 2006 (Verreault et al. 2007a). HBCD was found in around 60% of the 49 plasma samples; however, the substance was present in all 31 gull eggs sampled in the study, with an average concentration in the yolk of 19.8 ng/g wet weight and a maximum measured value of 63.9 ng/g wet weight. The results provide evidence of potential maternal transfer of HBCD to the eggs of glaucous gulls.

An earlier study by Verreault et al. (2007b) measured average concentrations of 3.29 ng/g and 75.6 ng/g wet weight in blood and liver, respectively, collected from Norwegian Arctic glaucous gulls in early July 2002. Whole body concentrations ranged from 52.6 ng/g to 270 ng/g wet weight (mean of 117 ng/g wet weight) with feathers, and from 38.4 ng/g to 194 ng/g wet weight (mean 91.0 ng/g wet weight) when content in the feathers was not included.

SØrmo et al. (2006) analyzed representative species from various trophic levels of the polar bear food chain, using samples collected from 2002 to 2003 at Svalbard in the Norwegian Arctic. HBCD was below detection limits (minimum 0.012 ng/g lipid weight) in the amphipod, Gammarus wilkitzkii. Concentrations increased from polar cod (Boreogadus saida; 1.38 ng/g to 2.87 ng/g lipid weight) to ringed seal (Phoca hispida; 14.6 ng/g to 34.5 ng/g lipid weight), but decreased in the top predator, polar bear (Ursus maritimus, 5.31 ng/g to 16.51 ng/g lipid weight). The results suggested that substantial biomagnification was occurring from polar cod to ringed seal but none from ringed seal to polar bear. The lower levels in the polar bear samples were considered to indicate possible enhanced metabolic capability in the bears.

Gebbink et al. (2008) measured a mean concentration of 41 ng/g wet weight in adipose tissue collected from 10 adult male and 10 adult female polar bears in central East Greenland between 1999 and 2001. The substance was not detected in blood, brain and liver samples from the bears (detection limit not specified). Morris et al. (2007) reported a concentration of 0.38 ng/g lipid weight in the blubber of ringed seal (Phoca hispida) from the Barrow Strait, Nunavut.

Tomy et al. (2008) investigated isomer-specific accumulation of HBCD at several trophic levels of an eastern Canadian Arctic marine food web. Alpha- and γ-HBCD were present in all species examined (beluga whale, Delphinapterus leucas; walrus, Odobenus rosmarus; narwhal, Monodon monoceros; arctic cod, Boreogadus saida; deepwater redfish, Sebastes mentella; shrimp, Pandalus borealis and Hymenodora glacialis; clam, Mya truncata and Serripes groenlandica; and mixed zooplankton) with total HBCD concentrations ranging from 0.6 ng/g (geometric mean) to 3.9 ng/g lipid weight. The ß-isomer was below detection limits (0.0004 - 0.0059 ng/g lipid weight) in all samples. No clear trend was evident in the diastereomeric profile of the animals; however α-HBCD contributed greater than 70% of the total HBCD burden in shrimp, redfish, arctic cod, narwhal and beluga, while zooplankton, clams and walrus contained more than 60% γ-HBCD. The observed differences in diastereoisomer predominance were attributed, at least in part, to the differing environmental fates and behaviours of the isomers, with the least water-soluble γ-isomer more likely to diffuse passively from the water into zooplankton, which have proportionately high lipid content. Similarly, as benthic filter feeders, clams may be more likely to absorb a large proportion of the γ-isomer from the surrounding sediment, where this isomeric form has been shown to predominate. The presence of large proportions of α-HBCD, such as in the beluga and narwhal, may indicate enhanced metabolic capability based on evidence of stereoisomer-specific biotransformation of the γ-isomer into the α- form (see, for example, Zegers et al. 2005; Law et al. 2006b). The researchers reported a significant positive relationship of α-HBCD with trophic level, indicative of biomagnification throughout the food web, while a significant negative relationship was observed between concentrations of γ-HBCD and trophic level (i.e., trophic dilution).

Temporal Trends

Remberger et al. (2004) reported concentrations of 0.8 - 1.5 ng/g dry weight in surface sediments (2 - 4 cm in depth) collected in 1996 and 1997 from three locations in Stockholm. Deeper core samples (20 - 32 cm in depth) from the same sites contained 0.2-0.5 ng/g dry weight. Higher concentrations in the surface sediments were considered to indicate increasing deposition with time. Based on radioactive dating, the surface sediments were estimated to originate in the mid 1990s, while those in the deeper layers represented deposition from the 1950s and 1960s.

Kohler et al. (2008) reported a rapid and linear increase in HBCD levels present in successive layers of a sediment core collected in 2003 from the deepest point of a shallow suburban lake in Switzerland . HBCD first appeared in a sediment layer corresponding to approximately the mid 1970s and reached a maximum concentration of 2.5 ng/g dry weight at the surface layer of the core, estimated to be from approximately 2001. A similar trend was evident in a sediment core collected from a deep pre-alpine Swiss lake, with levels of less than 0.1 ng/g dry weight in samples from prior to 1980 and increasing rapidly to a maximum concentration of around 0.7 ng/g dry weight in the surface layer, corresponding to the early 2000s (Kohler et al. 2007).

HBCD was present in all three sediment cores and six surface sediment samples collected in 2002 from Tokyo Bay (Minh et al. 2007). Concentrations ranged from 0.056 ng/g to 2.3 ng/g dry weight, with the highest levels found near densely populated and industrialized areas. HBCD first appeared in the sediment cores at depths of 20 - 25 cm, estimated to date from the late 1960s and early 1970s, with the concentration increasing steadily to the highest levels at the surface. Based on the data, Tanabe (2008) estimated concentration doubling times of 7.1 - 12 years for HBCD in the sediment.

A number of studies examine HBCD concentrations in biota over time as a means of identifying possible trends in contamination levels. Braune et al. (2007) reported mean concentrations of 2.1 - 3.8 ng/g lipid weight in pooled samples of eggs of the ivory gull (Pagophila eburnea) collected from the Canadian Arctic from 1976 to 2004. Concentrations decreased from a highest value of 3.8 ng/g lipid weight in 1976 to 3.0 ng/g lipid weight in 1987 and 2.1 ng/g lipid weight in 2004.

Stapleton et al. (2006) measured 0.71 - 11.85 ng/g wet weight in blubber samples collected from male California sea lions (Zalopus californianus) stranded along the California coast between 1993 and 2003. HBCD was present in 80% of the samples analyzed, with the α-isomer predominant in all samples. Levels increased almost exponentially over the 10-year study period and, while the researchers cautioned that the sample size of 26 might have been too limited to allow accurate estimation of accumulation rates, the doubling time in the sea lion blubber over the study period was approximately two years, if the increase is assumed to be exponential (Stapleton et al. 2006).

Sellström et al. (2003) observed a steady and significant (p < 0.001) increase in concentrations present in the eggs of guillemot (Uria algae) collected from the Baltic Sea from 1969 to 2001. The observed increase was attributed to increasing use of HBCD, although this was difficult to substantiate due to a lack of industrial production and use information. The presence of HBCD in the eggs was considered to indicate possible biomagnification of the substance (Kierkegaard et al. 1999).

A marked increase was evident in blubber concentrations of juvenile male grey seals (Halicoerus grypus) collected in the Baltic Sea from 1980 to 2000 (Roos et al. 2001). Concentrations ranged from 16 ng/g to 177 ng/g lipid weight, with lowest levels in seals collected during the early 1980s.

Atlantic cod (Gadus morhua) collected in 2003 from the southern industrialized region of Norway, near Oslo, contained up to 16.9 ng/g wet weight (56.9 ng/g lipid weight), while those collected from the same region in 1998 contained up to 2.70 ng/g wet weight (22.67 ng/g lipid weight; Bytingsvik et al. 2004). This represents a more than six-fold increase when considered on a wet weight basis (a more than 2.5-times increase in terms of lipid weight).

Diastereomeric Differences

Studies providing a breakdown of the individual diastereomers commonly report a predominance of α-HBCD in biota samples, with the γ- and ß- isomers present at lower levels or below detection limits. This congener profile contrasts markedly with that seen in commercial formulations and sediment samples, in which the γ-isomer most often dominates. The isomeric pattern observed in biota may reflect differences in exposure potential, uptake, metabolism or depuration of the three isomers. There is evidence that conversion of γ-HBCD to α-HBCD occurs at temperatures above 160°C (Peled et al. 1995), suggesting that finished products subjected to high temperatures during processing may carry a much higher proportion of a-isomer than that present in the original technical formulation. This may increase the potential for organism exposure to α-HBCD during product use and disposal. As well, α-HBCD has higher water solubility (see Table 2), suggesting that it may more readily enter organisms through preferential transfer from particles through water (Morris et al. 2004). Janá k et al. (2005) reported consistently higher levels of the α-isomer compared with those of γ-HBCD in the livers of several fish species, and considered this a possible indication that the γ-isomer was more easily metabolized. Further evidence for differential rates of biotransformation was provided by in vitro assays in which ß- and γ-HBCD were significantly metabolized by rat and harbour seal liver microsomes, while α-levels remained mostly unchanged (Zegers et al. 2005). The net result was accumulation of the α-isomer relative to that of the other two isomers.

Research by Law et al. (2006b) demonstrated that bioformation or bioisomerization of HBCD appeared to occur in some species. Statistically significant amounts of α-HBCD were measured in the muscle tissue of rainbow trout (Oncorhynchus mykiss) exposed exclusively to γ-HBCD via the diet. Similarly, both α- and γ-HBCD were present in statistically significant quantities in fish exposed only to ß-HBCD. The results suggested that selective bioisomerization of HBCD, with preferential formation of the α-isomer, may contribute appreciably to determining isomer distributions in the environment. The α-isomer appeared recalcitrant to bioisomerization in the fish, a factor that may also contribute to its proportionately higher tissue levels in biota samples.

Ecological Effects Assessment

The ecotoxicity database for HBCD includes endpoint values from several pelagic trophic levels (i.e., fish, invertebrates, algae), as well as data for benthic and terrestrial species. Most data were derived using standard methods and species, although results from novel studies are also reported in the literature. Acute or chronic (partial life cycle) toxicity testing results (or both) are available for rainbow trout (Oncorhynchus mykiss), bluegill sunfish (Lepomis macrochirus), water flea (Daphnia magna), green algae (Selenastrum capricornutum, Chlorella sp.) and diatoms (Skeletonema costatum, Thalassiosira pseudonana). Toxicity data are also available for benthic organisms (Lumbriculus variegates, Hyalella azteca), earthworm (Eisenia fetida) and six terrestrial plant species. While most studies failed to determine a numerical endpoint value, indicating only that minimum effect levels can be expected to exceed that of the highest concentration tested, the quantity and quality of the available studies make HBCD a rich source of data compared to most brominated flame retardants.

It should be noted that toxicity studies generally utilize the commercial HBCD mixture; thus, organisms would be exposed to various amounts of each diasteriomer found in the commercial product. Inferences about which diastereoisomer is responsible for the observed effects are not possible, since organisms would be exposed to varying HBCD diasteriomers concurrently.

No information was found on a possible mode of toxic action for HBCD. ECOSAR (2004) classifies the substance as a neutral organic, based on its chemical structure. As a neutral organic, HBCD is expected to exhibit effects through nonpolar narcosis (i.e., through non-specific disruption of cellular membrane integrity or function, or both).

HBCD has demonstrated toxicity in both aquatic and terrestrial organisms, with significant adverse effects on survival, reproduction and development reported in algae, aquatic invertebrates, fish and terrestrial annelid worms. In aquatic species, a 21-day no-observed-effect concentration (NOEC) and lowest-observed-effect concentration (LOEC) of 3.1 µg/L and 5.6 µg/L, respectively, were determined for the water flea, Daphnia magna, based on significantly reduced growth (CMABFRIP 1998). Daphnids exposed to the highest test concentration of 11 µg/L exhibited statistically significant reductions in length, dry weight and number of young.

Walsh et al. (1987) examined the effect of HBCD on population density in two unicellular marine algae, Skeletonema costatum and Thalassiosira pseudonana, using six nutrient media. Depending on the nutrient medium used, the 72-hour median effective concentration (EC₅₀) values based on reduced population density ranged from 9.3 µg/L to 12.0 µg/L in S. costatum and from 50 µg/L to 370 µg/L in T. pseudonana.

Ronisz et al. (2004) injected juvenile rainbow trout, Oncorhynchus mykiss, with HBCD dissolved in peanut oil and observed the effects on several biomarkers relating to liver enzyme function and hormonal activity. Ethoxyresorufin-O-deethylase activity was significantly inhibited in fish receiving approximately 5 × 10⁵µg/kg body weight (kg-bw) for a period of 28 days, while fish dosed at 5 10⁴ and 5 × 10⁵ µg/kg-bw for 5 days displayed significantly increased catalase activity. Significant increases in the liver somatic index (LSI; liver weight as a percentage of whole body weight) were evident in high-dose fish following an exposure period of 28 days. The induction of catalase at 5 days, together with increased LSI in exposed fish after 28 days, suggested that HBCD may be a peroxisome proliferator, a negative hormonal response. Further investigation into this possibility by the researchers yielded inconclusive results. Peroxisome proliferators are considered to be tumor promoters through a non-genotoxic mechanism (Waxman 1999; Vanden Heuvel 1999) and have been associated with hepatocarcinogenesis (Ackers et al. 2000).

Altered thyroid status, including changes to circulating plasma thyroid hormone levels and hepatic metabolic enzyme activity, were reported in juvenile rainbow trout fed lipid-corrected concentrations of 29.14 µg/kg, 11.84 µg/kg and 22.84 µg/kg of α-, ß- or γ-HBCD, respectively (approximately 10 µg/kg to 30 µg/kg-bw) for 56 days followed by a clearance period of 112 days (Palace et al. 2008). The results provided evidence that HBCD exposure can affect the thyroid system in fish, with effects increasing at higher concentrations.

Atlantic salmon, Salmo salar L., exposed to low levels of HBCD (0.011 µg/L) in freshwater for 30 days over the peak natural smoltification period, and then transferred to clean seawater for 20 days, exhibited significant alterations in the levels and patterns of circulating thyroid hormones (Lower and Moore 2007). These hormones play a key role in smoltification and are critical to the imprinting of olfactory memory, which allows the fish to return to their natal river for spawning. Thyroid hormone (T4, T3) levels were significantly higher in control fish following transfer to seawater, peaking at the time of transfer. In contrast, the levels in HBCD-exposed fish did not show this increase at transfer, peaking earlier, at the end of the freshwater exposure period. Olfactory sensitivity was also significantly decreased in the HBCD-exposed fish. The researchers concluded that while all fish appeared to complete the parr-smolt transformation successfully and were able to survive and osmoregulate in saline conditions for a period of 20 days, the HBCD-exposed fish displayed evidence of disruption to thyroid hormone homeostasis during development, which may ultimately affect imprinting and other behaviour in the adult fish.

Increased microsomal enzyme activity and oxidative stress were observed in mature (4-6 months) Chinese rare minnow (Gobiocypris rarus) exposed to water concentrations of up to 500 µg/L of HBCD for duration periods of 28 and 42 days (Zhang et al. 2008). The researchers concluded that increasing the duration of HBCD exposure induced microsomal enzymes such as ethoxyresorufin-O-deethylase and pentaoxyresorufin-O-depentylase, and caused the formation of excess reactive oxygen species, finally resulting in oxidative damage to lipids, proteins and DNA, and decreased antioxidant capacities in the fish.

Kuiper et al. (2007) reported that immature European flounder, Platichthys flesus, exposed for 78 days to a wide range of concentrations in sediment and food (up to 800 µg/g total organic carbon (TOC) and 3000 µg/g lipid in sediment/food test systems and 8000 µg/kg TOC in sediment-only systems) exhibited no signs of hepatic microsomal enzyme induction, no alterations to thyroid gland activity or thyroid hormone levels, and no indications of endocrine effects as measured through production of the yolk precursor protein vitellogenin.

Sediment testing with the freshwater oligochaete, Lumbriculus variegates, yielded 28-day NOEC and LOEC values of 3.25 × 10³ and 2.93 × 10⁴ µg/kg dry weight of sediment, respectively, based on significant reductions in total worm numbers (Oetken et al. 2001). The researchers concluded that the sediment-bound fraction of HBCD is bioavailable and causes effects. ACCBFRIP (2003d, 2003e) conducted 28-day tests using the same species, as well as the amphipod, Hyalella azteca, and chironomid, Chironomus riparius, but found no dose-responsive, statistically significant effects in any of the three species up to concentrations of 1 × 10⁶ µg/kg dry weight of sediment.

The effects of HBCD on terrestrial plant seedling emergence and growth were evaluated in a 21-day study using corn (Zea mays), onion (Allium cepa), ryegrass (Lolium perenne), cucumber (Cucumis sativa), soybean (Glycine max) and tomato (Lycopersicon esculentum) (ACCBFRIP 2002). No apparent adverse treatment-related effects were observed on seedling emergence, survival or growth for any of the six species tested, and the 21-day NOEC for the study was equal to or greater than the highest test concentration of 5 × 10⁶ µg/kg dry weight of soil.

A toxicity study using the earthworm, Eisenia fetida, determined a 56-day NOEC and LOEC of 1.28 × 10⁵and 2.35 × 10⁵ µg/kg dry weight of soil, respectively, based on significantly reduced reproduction (ACCBFRIP 2003a). The 56-day EC₁₀ (10% inhibition) and EC₅₀ (50% inhibition) for reproduction were 2.16 × 10⁴ and 7.71 × 10⁵ µg/kg dry weight of soil, respectively. As the calculated EC₁₀ value was less than the lowest concentration tested, it was considered an estimate only. There was no significant effect on adult worm survival, and the 28-day NOEC for survival was equal to or greater than the highest test concentration of 4.19 × 10⁶ µg/kg dry weight of soil.

There are no published reports describing potential effects to wildlife species; however, a number of studies have examined toxicity in rodents. Theses studies are summarized in the Human Health portion of this assessment.

Crump et al. (2008) reported significant up-regulation of enzymes involved with the metabolism of xenobiotics (CYP enzymes and uridine 5'-diphospho-glucuronosyltransferase) in cultured chicken, Gallus domesticus, hepatocytes following 24- and 36-hour exposures to concentrations of 1 µM to 30 µM α-HBCD or technical HBCD. Significant down-regulation of proteins associated with the thyroid hormone pathway and lipid regulation also occurred in this concentration range.

Summaries of key toxicity studies used in the effects assessment of HBCD are provided in Table 16.

The approach taken in this ecological screening assessment was to examine various pieces of supporting information and to develop conclusions based on a weight-of-evidence approach, as required under section 76.1 of CEPA 1999. The screening assessment is a conservative assessment, intended to represent reasonable worst-case conditions. It integrates known or potential exposure to the target substance with known or potential effects on the environment.

The potential for HBCD to persist in the environment and accumulate within organisms formed primary lines of evidence in support of a decision relating to ecological harm. Evidence that a substance is persistent and bioaccumulative, together with evidence of commercial activity provides a significant indication of its potential to enter the environment under conditions that may have harmful long-term ecological effects (Environment Canada 2006). Substances that are persistent remain in the environment for a long time after being released, increasing the potential magnitude and duration of exposure. Substances that have long half-lives in mobile media (air and water) and that will exist within these media have the potential to cause widespread contamination. Releases of small amounts of bioaccumulative substances may lead to high internal concentrations in exposed organisms. Highly bioaccumulative and persistent substances are of special concern, since they may biomagnify in food webs, resulting in very high internal exposures, especially for top predators. Evidence that a substance is both persistent and bioaccumulative, when taken together with other information (such as evidence of toxicity at relatively low concentrations, and evidence of uses and releases) may therefore be sufficient to indicate that the substance has the potential to cause ecological harm.

HBCD has been detected in all environmental media, and there is evidence that the substance meets CEPA 1999 persistence criteria (half-life in air of 2 days or more, half-lives in soil and water of 182 days or more, and half-life in sediment of 365 days or more; see Table 3). In addition, the substance is present in samples collected from regions considered remote from potential sources, including the Arctic, indicating that it is sufficiently stable in the environment to allow long-range transport in air or water, or both. Atmospheric transport of a substance to an area remote from its source is a criterion for persistence in air, as defined by the Persistence and Bioaccumulation Regulations. under CEPA 1999.

Measured bioconcentration factors of up to 18 100 are reported in the published literature. Based on these data, HBCD meets CEPA 1999 bioaccumulation criteria (bioaccumulation and bioconcentration factors of 5000 or more; see Table 4).

HBCD has demonstrated toxicity in both aquatic and terrestrial species (21-day LOEC of 5.6 µg/L for reduced growth in Daphnia magna, for example; CMABFRIP 1998), with significant adverse effects on survival, reproduction and development reported in algae, daphnids and annelid worms. Recent studies indicate a potential link to altered hormonal status in fish, with reported impacts on the activity and normal functioning of liver enzymes (Ronisz et al. 2004; Zhang et al. 2008) and thyroid hormones (Lower and Moore 2007; Palace et al. 2008). The α-isomer has displayed a greater capacity to disrupt hormonal function in vitro and this apparent higher potency is of concern, given the higher prevalence of this isomer, compared to the other two, in biota samples.

As mentioned previously, combustion of HBCD under certain conditions may lead to the formation of brominated dibenzo-p-dioxins and dibenzofurans, brominated analogues of the Toxic Substances Management Policy Track 1 polychlorinated dibenzo-p-dioxins and dibenzofurans. Trace levels of these compounds and their precursors have been measured during combustion of flame-retarded polystyrene materials containing HBCD.

North American and global demand for HBCD may be on the rise. Higher concentrations are reported in surficial layers of sediment cores as compared with those in deeper layers, an indication of increasing deposition with time (Remberger et al. 2004; Minh et al. 2007; Kohler et al. 2008). As well, time-trend analyses conducted using birds (Sellström et al. 2003) and marine mammals (Roos et al. 2001; Stapleton et al. 2006; Law et al. 2006d) document nearly exponential increases in biota levels beginning in the early 1990s. While HBCD was first commercially introduced to the brominated flame retardant market in the 1960s, its application in extruded polystyrene did not commence until the 1980s (2007 email from Dow Chemicals Canada Inc. to Environment Canada; unreferenced). There is also evidence that HBCD may be replacing PBDE flame retardants, some of which are no longer in production. Spatial concentration patterns of HBCD in U.S. air samples were similar to those of PBDE-209, possibly signalling a shift from polybrominated diphenyl ether (PBDE) products to HBCD (Hoh and Hites 2005). This is further supported by comparison studies that report levels approaching or exceeding those of PBDEs in compost (Zennegg et al. 2005) and bird yolk (Murvoll et al. 2006a, 2006b).

The available information on the persistence, bioaccumulation potential, toxicity and use and release of HBCD in Canada therefore suggests that this substance has the potential to cause ecological harm in Canada .

Quantitative risk estimation methods are also used to evaluate potential to cause ecological harm. A summary of data used in the risk quotient analysis of HBCD is presented in Table 17. Exposure data used in the determination of predicted exposure concentrations can be found in Tables 6 and 7. Due to the general paucity of HBCD surface water and sediment concentrations in Canada, a fugacity modelling approach, based on principles described by Cahill et al. (2003) and, more generally, Mackay (1991), was applied for estimating exposure and determining predicted exposure concentrations (PECs) in water and sediments (see Appendix B). The database of soil HBCD concentrations was also considered inadequate, and so the soil predicted exposure concentration was derived using a simple calculation procedure involving the application of sewage sludge to agricultural soil and pastureland. Toxicity data used to determine critical toxicity values and predicted no effect concentrations are summarized in Table 16.

For pelagic organisms, risk quotients exceeded 1, indicating a current potential for risk, in surface water scenarios associated with handling raw materials and compounding HBCD. Application of secondary treatment processes greatly reduced the potential for risk; however, predicted exposure values still exceeded minimum effect levels for scenarios associated with large production quantities (e.g., 100 000 kg per year) or use of only primary wastewater treatment, or both. Similar trends were observed in the benthic compartment, in which predicted bulk sediment concentrations of HBCD exceeded minimum effect levels for facilities handling large volumes of raw materials (e.g., 100 000 kg per year) and for smaller volume facilities (e.g., 10 000 kg per year) using only primary wastewater treatment. Predicted bulk sediment concentrations were less than 1 for scenarios associated with compounding facilities, suggesting that current estimated HBCD exposure concentrations derived from compounding activities in Canada are unlikely to exceed minimum effects levels in organisms.

Risk quotients for the soil compartment were determined using exposure values calculated from concentrations measured in sewage sludge. This approach was used because the application of sewage sludge to agricultural soils and pasturelands is considered to represent a direct pathway for HBCD into soil. Since no Canadian or North American sewage sludge data were available, a European value was selected to represent possible levels in populated regions of Canada , such as southern Ontario. The risk quotient results suggested that current estimated exposure concentrations in Canadian soils are unlikely to exceed those leading to adverse effects in organisms.

The risk quotient derived for wildlife species highlights the potential for intake arising from the uptake of HBCD in food. In this analysis, the critical toxicity value is based on significant reductions in the levels of circulating thyroid hormones in rats receiving oral doses of 1 × 10⁵ µg/kg to 1 × 10⁶ µg/kg-bw per day over a 90-day period (CMABFRIP 2001). It should be noted that this level represents the lowest effect level and not the lowest adverse effect level, since no adverse effects were apparent in the affected animals. However, the endpoint is considered relevant to potential impacts in wildlife populations,since disruptions in thyroid hormone homeostasis may alter critical metabolic processes such as development of the central nervous system and cell metabolic rates (Dorland 2006). Allometric scaling was used to extrapolate data obtained from laboratory feeding studies with rats to a surrogate wildlife species, American mink. The results indicated that current HBCD concentrations in Canadian biota are unlikely to exceed minimum effects levels.

The analysis of risk quotients determined that HBCD concentrations in the Canadian environment have the potential to cause adverse effects in populations of pelagic and benthic organisms, but are unlikely to result in direct adverse effects to soil organisms and wildlife. However, it must be considered that the presence of even small amounts of HBCD in the environment warrants concern in light of strong evidence that the substance may be environmentally persistent and bioaccumulative.

There is some uncertainty regarding physical and chemical properties of the individual HBCD diastereomers and how these relate to persistence, bioavailability, bioaccumulation potential and toxicity of HBCD in the environment.

The assessment finds that HBCD may biodegrade based on laboratory studies. While there may be some lack of understanding respecting diastereoisomeric transformations in the environment (including biota), when modelled and monitoring data are considered together, the data on HBCD indicate a significant level of persistence in the environment as well as transportability to remote locations. HBCD is highly bioaccumulative in aquatic biota; however, there is some uncertainty respecting the potential to bioaccumulate in sediment and soil life, as well as biomagnification in terrestrial wildlife.

The role of partitioning to atmospheric particulates and the potential for long-range atmospheric transport of particle-bound HBCD warrants further consideration.

There is a general lack of data on HBCD concentrations in the Canadian environment, particularly in sediments, soils, sewage sludge and biota.

Clarification of toxicity to sediment and soil organisms is also required. Markedly divergent outcomes were reported in 28-day Lumbriculus testing (i.e., NOECs of 5 and = 1000 mg/kg sediment dry weight), suggesting that effects in soil and sediment tests may be significantly influenced by procedures used to incorporate the test substance, such as the use of a carrier substance. Uncertainties are also associated with toxicity to wildlife, including possible metabolic pathways and products, and effects on pelagic, benthic, soil and wildlife species resulting from prolonged (e.g., lifetime and multigenerational) exposure.