POSSIBLE HCB EMISSION from a MAGNESIUM EXTRACTION PLANT using ASBESTOS tailings.

Introduction

The Comité de Citoyens du Project Magnola have also asked our research group to investigate any potential impacts of Magnola's emissions of hexachlorobenzene (HCB) on ecosystems and wildlife at both local and regional scales.
To answer this question we combined the results from the company's environmental impact assessment with a chemical fate model in order to estimate long term concentrations in the air, water, soil and sediment 'compartments' of the surrounding environment. The environmental impact assessment determined how the HCB emissions disperse from the factory and onto the surrounding landscape, calculating ground air concentration values. The chemical fate model determines how this concentration of HCB will partition itself into the 'compartments' of the environment. Then using information gathered from an extensive literature review, we estimated what the resulting impact will be on the plants, invertebrates, fish, birds and mammals of both aquatic and terrestrial ecosystems.
The limitations of our research include the accuracy and the choice of the chemical fate model and the model used in Magnola's Environmental Impact Assessment. The accuracy of Magnola's reported emissions is a concern. We are also limited by the shortcomings of the literature. There are no studies of impacts on wildlife populations, only impacts on individuals. The data we used come from different ecosystems, as there is insufficient data for southern Quebec. We will also have to generalize from one species to another, and even if closely related, their responses to HCB can differ greatly.

Project Issue in a Broader Context / Literature Review

HCB is considered to be a "non-threshold toxicant", a substance for which there is believed to be some chance of adverse health effects at any level of exposure (CEPA, 1993). Because of this, the ministers of the Environment and of National Health and Welfare have concluded that the concentrations of HCB present in the Canadian Environment may be a danger to the environment and to human life and health. Therefore, HCB is classified as "toxic" as interpreted under section 11 of the Canadian Environmental Protection Act, which states:

"...a substance is toxic if it is entering or may enter the environment in a quantity of concentration or under conditions
a) having or that may have an immediate or long-term harmful effect on the environment;
b) constituting or that may constitute a danger to the environment on which human life depends; or
c) constituting or that may constitute a danger in Canada to human life or health."

HCB has not been used commercially in Canada since 1972 due to concerns about adverse effects on the environment and human health (CEPA, 1993). Sedimentary core studies in the St. Lawrence River have shown a decline in HCB contamination (by a factor of 5-10) since its production was discontinued in the 70's (Carignan et al., 1994). However, it is still released into the environment as a by-product of industrial processes such as Magnola's.

The Fate of HCB in the Environment
The atmosphere plays an important role in the transport, distribution and cycling of chemicals that are volatile and semi-volatile in nature, such as HCB (Hart et al., 1993). Long-range atmosphereic transport has been shown to be a significant source and means of redistribution of HCB throughout the world (van Pul et al., 1998). In the atmosphere, chemicals attach to aerosols and either settle back down on the soil (dry deposition) or are dissolved by atmospheric moisture and returned to the ground with precipitation (wet deposition) (Ahrens, 1994). HCB is removed from the atmosphere primarily through wet deposition to aquatic and terrestrial systems (Eisenreich and Strachan, 1992). Wet deposition, combined with the low levels of HCB currently in the atmosphere, indicates that HCB is not likely to trap significant quantities of thermal radiation from the Earth's surface, nor is it expected to reach upper layers of the atmosphere. Therefore, HCB is not likely to be associated with global warming (CEPA, 1993).
Because of its mobility in the environment and its resistance to degradation, HCB is widely distributed throughout the Canadian environment. HCB is removed from the air phase via atmospheric deposition to water and soil, or undergoes very slow photolytic degradation (half-life: t ½ = 80 days) (Mill and Haag, 1986). Volatilization is the major removal process for soil at the surface while slow aerobic (t ½ = 2.7 - 5.7 years) and anaerobic biodegradation (t ½ = 10.6 - 22.9 years) are the major removal processes at lower depths (Howard et al., 1991). Due to its low solubility, HCB volatilizes easily from water to air and adsorbs to suspended particles eventually collecting in bottom sediments. Once in sediment, HCB will tend to accumulate and be trapped by overlying sediments (Oliver and Nicol, 1982). Chemical and biological degradation are not considered to be important removal processes of HCB from water or sediments (Oliver and Carey, 1986).
Organisms accumulate HCB from contaminated air, water and food, although benthic organisms may accumulate HCB directly from contaminated sediment (Oliver, 1984b). As with most organochlorides, HCB has a tendency to biomagnify up the food chain. Studies have shown that high trophic level organisms in natural aquatic ecosystems accumulate HCB to levels much greater than those at lower trophic levels (Oliver, 1987; Oliver and Niimi, 1988). For example, Allen-Gil et al. (1997) have shown for four Alaskan Lakes, that HCB concentrations increase from sediment (mean = 0.17 ng/g dry weight) to snails (mean = 0.15 ng/g wet weight) to greylings (mean = 0.64 ng/g wet weight of liver) and finally to trout (mean = 1.15 ng/g wet weight of liver). Allen-Gil et al. (1997) provide an integrated assessment of HCB distribution within freshwater systems by combining results from sediment, snails and freshwater fish. When viewed as a system, the largest contaminant sink was lake trout, the top predatory fish species. This indicates that, through bioaccumulation, the biological component of ecosystems may function as an important sink for HCB.

HCB concentrations in various media
Due to global long-rang transport of HCB, measurements of ambient air concentrations taken in the mid 1980's in southern and central Ontario (both downtown and rural regions) and in the Canadian high arctic were all similar (0.15 ng/m3) (CEPA, 1993). The Magnola standard is set at a maximum air concentration of 2.2 ng/m3 (Commité de Citoyens, pers. comm.). Sedimentary concentrations of HCB have been found to range from below the limit of detection (1.0 ng/g dry weight) to 351 ng/g (dry weight) in the St. Lawrence River (Kaiser et al., 1990); and from below the limit of detection (0.2 ng/g dry weight) to 10 ng/g (dry weight) in the Atlantic Provinces (Leger, 1991 in CEPA, 1993).
HCB has been detected in invertebrates, fish, reptiles, birds and mammals across Canada since the 1960's, when monitoring of organochlorines began. Appendix C1 summarizes many of the results of these studies.

Toxicity
Studies have been done on both the acute and chronic toxicity of HCB for various organisms. We concentrated on chronic toxicity studies because exposure to HCB emissions from the Magnola facility is continuous, not short-term. Data on the chronic toxicity of HCB are available for species from a number of trophic levels, including freshwater alga, protozoa, invertebrates and fish. For the terrestrial environment, toxicity data are available only for birds and mammals. The literature on chronic toxicity to various biota is summarized in Appendix C2. The significance of individual toxic responses to effects at the population level is unknown (CEPA, 1993).
Fish eating mammals and predatorial birds tend to be the most susceptible to HCB in the environment as they are at the end of the food chain and get relatively larger doses of HCB than lower trophic levels (because of HCBs tendency to bioaccumulate).

The local environment
Understanding the composition of the environment surrounding the factory is important when predicting potential impacts of HCB emissions. Magnola is situated at the tail end of the Appalachian range, characterized by mountains of low elevation (150-300 meters). The plant is located in part of a long train of serpentine outcrops beginning in Alabama and ending at the tip of the Ungava Peninsula in Northwestern Quebec. The outcrop Magnola is situated on is about 100 km long, 25 km wide and stretches from Richmond, to East Broughton in the Eastern Townships of Quebec. Historically, this area has yielded 40% of the world's asbestos production (Brooks, 1987).
The serpentine belt is characterized by highly infertile soils, which are largely composed of the minerals antigorite and chrysotile. Thus, the plants that evolved here are adapted to be tolerant of low nutrient levels and heavy metals in the soils. The principal types of forests in the region are the ash maple grove and the fir forest. The two most common species of conifer are white cedar (Thuja occidentalis) and balsam fir (Abies Balsamea). The terrestrial fauna associated with this vegetation is primarily deer (Odocoileus virginianus), however, black bear (Ursus americanus) and moose (Alces alces) have also been observed. 1.5 km north of Magnola lies Burbank Pond. Typical mammals found around the pond include muskrats (Ondatra zibethicus), mink (Mustela vison), stoats (Mustela erminea), otters (Lontra Canadensis) and raccoons (Procyon lotor). Fifty species of common birds use the pond and the surrounding areas for nesting, summering and as a stop off during migrations. None of these birds are considered rare, vulnerable or in danger; however, only a few species of ducks, two species of heron, and the Canadian goose (Branta Canadensis) compose the majority of the bird population. We can also find a few species of amphibians and turtles in Burbank Pond. The two rivers near the plant contain a few species of fish. The main one's being trout (Salmoninae subfamily), bass (Micropterus salmoides), northern pike (Esox lucius) and yellow pike (Stizostedion vitreum).

Research Question / Hypothesis

What will be the impacts (if any) of Magnola's Hexachlorobenzene emissions on ecosystems and wildlife at both regional and local scales?

Methodology

To assess the potential environmental impacts of Magnola's HCB emissions, we first estimated the ambient concentrations of HCB in the soil, water and sediment of both the local and regional environments as a consequence of the factory's emissions into the air. For the regional assessment, we input Magnola's HCB emissions into a chemical fate model or Environmental Equilibrium Partitioning Model, which calculates how a chemical will partition itself between environmental 'compartments' such as air, water, water sediment and soil and the exchange fluxes between these compartments (Robson et al., 1999). For the local assessment, we obtained the results of a short-term industrial source dispersion model done for Magnola's Environmental Impact Assessment (Magnola, pers. Comm.). This model calculated ground air concentration levels of HCB in an area of 2000m2 surrounding the plant. They reported the maximum ground level air concentration of HCB over a period of one-year (1.37 ng/m3 (yr)). We input this value into the chemical fate model to find out what soil, water and water sediment concentrations will be with air concentrations at 1.37 ng/m3.
Next, we used biomagnification values for various organisms obtained from field measurements from previous studies on HCB in order to estimate HCB concentrations in the biota surrounding the factory. Finally, we referred to toxicity studies to see what effects, if any, these concentrations in the biota would have.

Analysis and Discussion

Regional Analysis
To assess any impacts of Magnola's emissions on a regional scale (on the order of 100,000's km2) we ran the chemical fate model, inputting HCB air emissions from the factory (21.4 kg/yr; BAPE report) added to our calculations of HCB emissions from the tailings pond (53.8 kg/yr), to get a total of 75.2 kg/yr. (See Appendix C7 for results). We didn't include the maximum ground concentration values calculated in Magnola's Environmental Impact Assessment because that concentration applies only on a local scale (100`s km2), not on a regional scale of several hundred thousand kilometers. The results show Magnola's contribution to the air, soil, water and sediment HCB concentrations in the Canadian environment as a consequence of long-range atmospheric transport. Ambient air concentrations of HCB have been shown to be similar (0.15 ng/m3) between urban Ontario, rural Ontario and the Canadian Arctic, indicating that concentrations are fairly even across Canada (except near point sources)(CEPA, 1993). The fate model's simulation indicates that Magnola's will increase this regional ambient concentration by about 2%.
As all equations in the fate model are linear, all the results are scaleable (Robson et al., 1999). Therefore, a 2% increase in ambient air HCB concentrations will cause a 2% increase in biota HCB concentrations. Most observations of animals in the wild with potentially dangerous levels of HCB in their tissues are the result of local contamination by a nearby point source. Animals that are only subject to long-range atmospheric transport of HCB tend to have undetectable or extremely low levels of HCB in their tissues. Biota concentrations that are not the result of a nearby point source, but only from long-range transport in the atmosphere and bioaccumulation in the food chain, have been obtained for remote regions of the Canadian North. One such study on mink in the Northwest Territories (Mustela vison) indicated an average HCB concentration of 0.41 ng/g wet weight (Poole et al., 1997). This is very low when compared to the tolerable daily intake of mink, which is 16,000 ng/kg body weight (Moore et al., 1996). A 2% increase in ambient concentrations will cause, through biomagnification, a 2% increase in tissue concentrations in mink. This will not lead anywhere near a daily dose of 16,000 ng/kg body weight, so the impact of Magnola's HCB emissions on a regional level appears minimal for mink.
A study on Beluga whales showed that they averaged 220 ng/g wet weight in eastern Hudson Bay and 930 ng/g wet weight in the St. Lawrence Estuary (Muir et al., 1990). A tolerable intake of beluga whales hasn't been established; however, there have been no obvious impacts on the population level. The St. Lawrence Estuary is considered rather highly contaminated with HCB as much is transported down the St. Lawrence River from the very industrial area of the Great Lakes. The level in Hudson Bay would largely be the result of long-range atmospheric transport, and if this level is increased by 2% it would lead to an average of 224.4, an increase that seems minimal when compared to the exposure beluga whales get in the St. Lawrence Estuary.
Due to HCBs high volatility, it tends to be widely and evenly dispersed when released into the air, except for the immediate vicinity of the plant. Because of this feature, wildlife on a regional level are not likely to be adversely affected by a 2% increase. Obviously, wildlife already highly exposed to nearby point sources will not benefit from any increases in ambient concentrations; however, those cases can most effectively be remedied by mitigating the most accountable point sources. Therefore, Magnola's impact on a regional scale will be minimal.

Local Analysis
Wildlife in areas affected by point sources of HCB emissions, often have much higher concentrations of HCB in their tissues (CEPA, 1993). Such is the case for the local wildlife surrounding the Magnola facility. To investigate the effects of emissions on the environment at this local level, up to 100 km2 surrounding the factory, we again used the chemical fate model. This time, however, instead of using emissions values, we input the maximum ground level air concentration values calculated in Magnola's Environmental Impact Assessment. The results of the simulations can be found in Appendix C8, and Table 8 summarizes the resulting water, sediment, and soil concentrations from the given air concentrations. The sediment concentration used comes from the results of the fate model (see table 11 in Appendix C8), not the diagram of the results as the diagram only includes the wet weight concentration.

To characterize risks to local wildlife, we used the quotient method used by Moore et al. (1996). This method uses very conservative point estimates to determine if HCB is potentially hazardous to mink (Mustela vison) at selected locations. We will use the same method for mink around the Magnola facility, particularly any living around Burbank Pond (about 1.5 km north of Magnola). Mink are top trophic level carnivores that eat small mammals and fish; thus, they are exposed to contaminants derived from both terrestrial and aquatic food webs. They are an opportunistic species, with aquatic organisms sometimes comprising up to 100% of their diet. Mink readily bioaccumulate environmental pollutants and are extremely sensitive to organochlorides such as HCB (Bleavins et al, 1986). Due to their sensitivity to organochloride contaminants, such as HCB, they make good test subjects. If there are no effects on mink, there are likely to be no effects on other organisms.
Offspring survivability of mink exposed to HCB in their diet was affected at a dose of 160,000 ng/kg body weight/day (Moore et al., 1996). This lowest effects dose was divided by a factor of 10 to derive a no-effects dose and to account for differences between laboratory and field conditions (Moore et al., 1996). This results in a tolerable daily intake (TDI) of 16,000 ng/kg body weight/day for mink.
The quotient method simply multiplies the concentrations of HCB in air, water and fish with the intake amount of each of these mediums by mink. The amounts of HCB accumulated from each medium are added together to get a total daily intake of HCB. This calculation is summarized with the equation:
TDI = CairNIRair + CwaterNIRwater + CfishNIRfish
Where TDI is total daily intake (ng/kg body weight/day), C is concentration, and NIR is intake rate normalized to a 1kg adult female mink.
For the air concentration, we used the maximum ground level concentration calculated in Magnola's Environmental Impact Assessment. The water concentration came from the results of the fate model. The fish concentration comes from multiplying the sediment concentration from the fate model with a sediment to fish biomagnification factor. Allen-Gil et al. (1997) have shown for Lakes in Alaska, that HCB concentrations increase from sediment (mean = 0.17 ng/g dry weight) to lake trout (Salvelinus namaycush) (mean = 1.15 ng/g wet weight of liver), a biomagnification factor of 6.76. If this biomagnification factor is used for the fish in Burbank Pond then with sediment concentrations of HCB at 0.72 ng/g dry weight, the HCB concentrations in fish would be 4.8 ng/g wet weight. We used the same intake rates as used by Moore et al. (1996). From Moore et al. (1996), the tolerable daily intake (TDI) for mink is 16,000 ng/kg body weight/day, 6.65 times higher than our calculated total daily intake. Given that the values used were hyperconservative (maximum exposure concentrations and diets consisting of maximally contaminated fish), it would appear unlikely that mink living near the Magnola facility will be adversely affected by exposure to HCB.
The risk characterization calculated above is deliberately hyper-conservative, as its intent is to create an upper-bound estimate of risk. It is highly unlikely that the entire lifetime diet of an individual mink would consist of the most contaminated individuals in a fish population from the most contaminated site known in the area. This is because the highest air HCB concentration was found in a 100m2 area right next to the plant. We also assume that Lake Trout are representative of other near-shore fish species and that the biomagnification factor measured in 1997 from four Alaskan arctic lakes is representative of current conditions in Quebec (since biomagnification factors are believed to be independent of concentration levels) (Norstrom et al., 1978).
Like mink, predatorial birds are organisms of a high trophic level and are therefore likely to be exposed to high concentrations of HCB (because of it's tendency to biomagnify). Boersma et al. (1986) showed that herring gull eggs (Larus argentatus) with tissue levels of 1,500 ng/g HCB wet weight had significantly reduced embryo weights. For many bird species, reduced embryo weights are associated with lower survival of chicks (CEPA, 1993). As for mink, this effects level was divided by a factor of 10 in order to derive a no-effects level and to account for potential differences in laboratory versus field conditions. Therefore, Environment Canada estimates the no-effect level for HCB in tissues of sensitive bird species to be 150 ng/g wet weight (CEPA, 1993).
Braune et al. (1989), investigated biomagnification factors in Lake Ontario herring gulls. The biomagnification factor between fish and tissue concentration in herring gull eggs is 20 (Braune et al., 1989). As biomagnification factors are not concentration dependent, we can apply this factor to gulls eating fish in the local Asbestos area around Magnola. If we take the HCB concentration in fish (calculated above in the analysis for mink) of 4.8 ng/g wet weight, multiply this by the biomagnification factor of 20, we get a potential tissue concentration of HCB in herring gull eggs of 96 ng/g wet weight.
Although this represents the upper bound value of possible tissue concentrations for herring gull eggs (using maximum concentrations), it does come fairly close to the tolerable tissue concentration of 150 ng/g. This is close enough to be of concern, not so much for herring gulls, but for other species of predatorial birds that may be more susceptible than herring gulls to HCB in the environment. For instance, HCB concentrations in peregrine falcon eggs (Falco peregrinus anatum) collected between 1980 and 1987 across Canada had a mean concentration of 279 ng/g wet weight (Peakall et al., 1990). There is no information on HCB bioconcentration factors for peregrine falcons so we cannot conclude if they will be affected by Magnola's emissions. However, if average tissue concentrations of HCB in peregrin falcons across Canada are greater than for other predatorial birds, it's likely that they tend to accumulate HCB more easily and are more susceptible to it. Furthermore, the potential for effects to peregrine falcons from exposure to HCB is considered to be a serious threat to the long-term survival of this species, given its current status as an endangered species in Canada (CEPA, 1993).
Ground level air concentrations of HCB would have to be about 2.25 ng/m, or about 65% higher for sediment concentrations to be high enough to get fish so contaminated that herring gulls could have the potential for tissue concentrations of 150 ng/g wet weight.

Limitations of analysis
The limitations of our research include the accuracy and the choice of the chemical fate model and the model used in Magnola's Environmental Impact Assessment. The accuracy of environmental models tends to be good for making general conclusions, not for specific outcomes. We chose the Fugacity-Based Environmental Equilibrium Partitioning model as our fate model as it was the most recent version available for download from Canada's Environmental Modeling Center. We believe that the choice of the Short Term Industrial Source Complex model used in Magnola's Environmental Impact Assessment to (estimate local ground level air HCB concentrations) is questionable. An article by Eschenroeder (1986) suggests using the Long Term version of the Industrial Source Complex model for modeling dispersion of HCB, not the Short Term version. This is because HCB is a persistent chemical and can last years in the environment.
The accuracy of Magnola's reported emissions is a concern, since this is only a projection of the emissions calculated in the experimental plant they operated in 1997 at Salabery-de-Valleyfied. The experimental plant was 250 times smaller than the actual plant, it only operated for seven full days and all predicted emissions are extrapolated from this test run. Also, due to Magnola's rights of confidentiality over their industrial process and the complexity of the process, there is no way for outside officials to calculate how much HCB will be emitted. Consequently there is a large potential for error in the published emissions.
We are also limited by the shortcomings of the literature, as there are no studies of impacts on wildlife populations, only impacts on individuals. Furthermore, the data we used comes from different ecosystems, as there is insufficient data for southern Quebec. We also had to generalize from one species to another, and even if closely related, their responses to HCB can differ greatly.

Recommendations

Magnola`s contribution to a 2% increase in ambient HCB concentrations on a regional level is likely to be minimal, since long-range atmospheric transport tends to disperse HCB fairly evenly and at low concentrations. The results of our local analysis, indicates little risk to piscivorous mammals, despite our choosing of a receptor known to by highly sensitive to HCB. However, for predatorial birds, HCB concentration in the local area may pose a risk to the most susceptable species, especially those that are already endangered.
We recommend to the Comité de Citoyens du Project Magnola that, when up-dates on Magnola's emission levels are available, they encourage Magnola to run a long-term dispersion model (instead of the short-term one used in their Environmental Impact Assessment) including hexachlorobenzene emissions from the tailings pond. Once obtaining the new concentration values, the Comité should re-run our analysis on piscivorous mammals and predatorial birds. This will provide a more accurate picture of any potential impacts on wildlife at both local and regional scales.