Health Canada
www.hc-sc.gc.ca
Common menu bar links
Institutional links
Priority Substances List Assessment Report for Releases from Primary and Secondary Copper Smelters and Copper Refineries - Releases from Primary and Secondary Zinc Smelters and Zinc Refineries
2.0 Summary of Information Critical to Assessment of "Toxic" under CEPA 1999
2.4.1.2 Releases to water
Information on the effects of direct release to water from CCR, CEZinc and CTO is summarized below. Further details are provided in Beak International (1999).
2.4.1.2.1 Whole-effluent toxicity
Whole-effluent toxicity test data indicate the percent concentration of effluent needed to kill half the test organisms (LC50), or the percent concentration needed to cut growth or reproduction in half (IC50), after a specified duration of test organism exposure to the effluent. Short-term (acute) tests are relevant for organisms, such as pelagic invertebrates, that will be exposed for short periods to the portion of the plume having elevated concentrations. Long-term (chronic) tests are more relevant for organisms, such as fish, that may hold position and reside at least partly in the plume for extended periods. Acute and chronic toxicity test data for the copper and zinc processing effluents considered in this assessment are summarized in Table 34.
The CCR facility contributes to the metal content of MUC-WWTP effluent. Toxicity tests were performed on grab samples of the MUC-WWTP treated effluent, collected over a five-day period in 1996 (MEF/EC, 1998). These tests indicated that the effluent was not acutely toxic to either Daphnia magna (a pelagic invertebrate) or rainbow trout (Oncorhynchus mykiss). However, chronic tests indicated effects on growth of fathead minnows (Pimephales promelas). Chemicals implicated as potentially contributing to chronic toxicity of MUC-WWTP effluent included surfactants (non-ionic and anionic), heavy metals (Cu, Cr) and ammonia. There are some questions about the representativeness of the grab samples tested and the quality of the chronic test results, and plans have been made to repeat these tests.
1 Different percentile critical loads are based on probabilistic evaluation of metal transport and fate, and assume that the ENEV is invariant.
Toxicity tests performed on treated CEZinc effluent in 1997 indicated both acute and chronic toxicity prior to pH adjustment of the test waters (pH 9.5-12). Acute tests with Daphnia magna, fathead minnows and rainbow trout exposed to pH-adjusted effluent in 1997 and 1998 indicated that effluent was non-toxic at pH 7.2-8.6. A 1998 effluent sample at pH 8.6 (unadjusted) was non-toxic to fathead minnows and to algae Selenastrum capricornutum in chronic tests. Thus, pH would appear to be a critical factor in determination of effluent toxicity. The pH is now better controlled in the wastewater treatment process.
Acute toxicity tests performed on treated Cominco-Trail effluents in 1994-95 indicated rainbow trout LC50s at 8-77% of effluent concentration for C-II and 11-26% of effluent concentration for C-III (Duncan and Antcliffe, 1996). The main contributions to toxicity were from Cd and Zn for C-II and from Cd, Zn, Tl, fluoride and ammonia for C-III. More recent (1998) toxicity test data (personal communication with facility operators) indicate substantially reduced toxicity, with LC50s of 71-100% for C-II and 76-100% for C-III following a cadmium reduction program. The C-IV outfall and Stoney Creek are not associated with current zinc operations, but, since they load upstream of C-II and C-III, they may contribute to effects observed in the river.
- MEF/EC (1998).
- Noranda-CEZinc data.
- Duncan and Antcliffe (1996) and recent Cominco data (personal communication with facility operators).
- pH adjusted to 7.2 - 8.1 using CO2 gas.
The implications of whole-effluent toxicity for aquatic organisms that reside in receiving waters can be judged by considering the spatial pattern of effluent dilution in receiving waters, as well as the effluent concentrations experienced by aquatic biota based on their likely locations and movements. Receiving water changes in pH and hardness, as compared to full-strength effluent, may also be important factors.
2.4.1.2.2 Derivation of CTVs and ENEVs
The potential effects of a chemical discharge to receiving water may also be estimated by comparing expected concentrations of individual chemicals in the plume to chemical concentration benchmarks (effect levels). These benchmarks, which represent estimates of low toxic effects to sensitive aquatic organisms, were used to define CTVs for aquatic organisms exposed to effluents. The application factor used to derive ENEVs was set to unity, to avoid having ENEVs within natural concentration ranges, and because the toxicity database is adequate for the chemicals considered. The ENEVs derived for application to lake waters of the Canadian Shield (Section 2.4.1.1.3) are in some instances not immediately suited for application to the harder, higher-pH waters of the St. Lawrence and Columbia rivers, the receiving bodies being considered in these site-specific assessments of aquatic releases. Further, values derived for generic application to the Canadian Shield considered only the metals Cu, Zn, Ni, Pb, Cd and As, while the aquatic releases assessment must consider a number of other components as well. As such, separate effect levels, tailored to the receiving environments, were derived for use in the assessment of aquatic releases. Relevant ENEVs for a variety of different aquatic biota, for heavy metals, Se, ammonia and fluoride, are listed in Table 35. Their derivation is briefly discussed below.
Chronic values for fish and epibenthic invertebrates were obtained as the lowest relevant chronic values from the U.S. Environmental Protection Agency (EPA) aquatic toxicity database (U.S. EPA, 1984, 1985a,b, 1986, 1987; Suter and Tsao, 1996) or from other literature in some cases. These values were utilized as ENEVs in this assessment for fish and epibenthic invertebrates. The values for Cd, Cu, Ni, Pb and Zn were adjusted to hardness levels of 50 and 100 mg/L, based on U.S. EPA (1995) equations, to obtain ENEVs suitable for application to the Columbia and St. Lawrence rivers, respectively. Ammonia values for fish were adjusted to pH levels of 7.8 and 8.3, respectively, based on equations from Broderius et al. (1985) and assuming a 20° C water temperature.
For short-term exposure situations, acute rather than chronic toxicity test data should be considered. Acute values for pelagic invertebrates (zooplankton) were obtained as the lowest relevant Species Mean Acute Values (SMAVs) in the EPA database (U.S. EPA, 1980a,b, 1984, 1985a,b, 1986, 1987; NJDEP, 1996) or from other literature in some cases. These values were used as ENEVs in this assessment for pelagic invertebrates that may receive short-term water column exposures. The values for Cd, Cu, Ni, Pb and Zn were adjusted to hardness levels of 50 and 100 mg/L, based on U.S. EPA (1995) equations, to obtain ENEVs suitable for application to the Columbia and St. Lawrence rivers, respectively.
Sediment quality benchmarks were considered for the assessment of potential effects on benthic invertebrates from metal-contaminated sediments. Both federal and provincial agencies have defined sediment quality guidelines. These are chemical concentrations in whole sediments below which adverse effects on benthic biota are considered to be unlikely.
- Chronic and acute values for Cd, Cu, Ni, Pb and Zn are given for hardness = 50 and 100 mg/L, using U.S. EPA (1995) equations for hardness adjustment; chronic values for ammonia are given for pH 8.3 and 7.8, 20°C.
- Chronic values based on original literature: Cd - Rombough and Garside (1982), Ni - Birge et al. (1983b), Pb - Davies et al. (1976), As - Birge et al. (1983b), Cr - Grande and Anderson (1983), Se - Hermanutz et al. (1992), Crane et al. (1992), Tl -Zitko et al. (1975), ammonia - Broderius et al. (1985).
- Chronic values from Suter and Tsao (1996) based on U.S. EPA database.
- Chronic values based on Species Mean Chronic Values from U.S. EPA (1984, 1985a,b, 1986, 1987).
- Acute values based on SMAVs from U.S. EPA (1980a,b, 1984, 1985, 1986, 1987).
- Acute values from NJDEP (1996) based on U.S. EPA database.
- OMEE (1993), Lowest-Effect Level (LEL) represents 10th percentile of species screening-level distribution.
- Data from CEPA PSL Assessment Report for Inorganic Fluorides (EC/HC, 1993).
The federal interim sediment quality guidelines (CCME, 1999) list both Threshold Effect Levels (TEL) and Probable Effect Levels (PEL). The incidence of adverse effects at metal concentrations below the TEL is estimated at 2-11%, depending on the metal. The incidence of adverse effects at concentrations above the PEL is estimated at 12-49%. The Ontario sediment quality guidelines (OMEE, 1993) are more precisely defined as percentiles of the distribution of benthic species impairment levels (species screening-level concentrations). The Lowest-Effect Level (LEL) represents the 10th percentile (10% of species impaired at this level), and the Severe-Effect Level (SEL) represents the 90th percentile. The LEL was used here as the ENEV for assessment of benthic invertebrates that may receive chronic sediment exposures. The Quebec sediment quality guidelines (MEQ/EC, 1992) were defined in similar fashion, but using a 15th percentile to represent the minimal effect level. These values are slightly higher than the LEL for most metals.
There has been an interest in development of water quality benchmarks for metals that specifically refer to biologically available forms. However, there is not a clear consensus as to which metal species are available. The free metal ion concentration is the best predictor of the metal's bioavailability (Campbell, 1995). In soft acidic waters, free ions comprise most of the dissolved metal, while in hard alkaline waters, other dissolved forms predominate and are less available.
For assessment of releases to water, it is assumed that the empirical metal toxicity versus hardness relationships, as represented by the U.S. EPA (1995) equations for hardness adjustment, adequately reflect the differences in availability of dissolved metals between the Columbia and St. Lawrence rivers. As a conservative measure, lower-bound hardness values have been used for both rivers.
The ENEVs are somewhat above regional background in most cases. Exceptions were the fish and epibenthic ENEVs for Cd in the Columbia River, which were slightly below background and were therefore overridden by the background concentration in river water.
Further work is needed on the subject of metal bioavailability. In the interim, it is assumed for the screening-level aquatic releases portion of these assessments that all dissolved metal is bioavailable, and the water ENEVs in Table 35 reflect this. It is recognized that somewhat higher benchmarks, particularly for Cu, Ni and Pb, may provide adequate protection for aquatic life due to reduced availability of some dissolved species.
2.4.1.2.3 Environmental effects monitoring
There has been no recent environmental effects monitoring (EEM) at the MUC-WWTP or CEZinc. Thus, there is no field survey contribution to the weight of evidence regarding ecological effects at these facilities.
There have been recent (1995) EEM studies in the Columbia River downstream of the Cominco facilities (Figure 1). However, chemical contributions to the effects observed in this region of the Columbia River arose from multiple sources, including the zinc plant, the lead plant, the fertilizer plant and historical landfill operations. Still more recent (1999) monitoring studies (not yet available) are suggesting environmental improvement related to substantially reduced Cominco loadings.
Columbia River water was not acutely toxic in April 1995 to either Daphnia magna (48 hours) or rainbow trout (96 hours) at locations downstream (d/s) of the major CTO outfalls (i.e., d/s Stoney Creek, d/s Island (C-III), New Bridge d/s C-II) (Duncan, 1997). Nor was there chronic toxicity in the Microtox bioassay. Water quality conditions have improved since this time, although more recent river toxicity data are not yet available.
Periphyton communities colonizing artificial substrates in 1995 were somewhat reduced in abundance at d/s Island and Old Bridge stations, as compared to reference stations at Birchbank and Waneta (Duncan, 1997). These communities also showed reduced diversity at d/s Stoney Creek, d/s Island and Old Bridge as compared to Birchbank and Waneta. Periphyton productivity, indicated by chlorophyll a and dry-weight biomass on sampling plates, was reduced at d/s Island and Old Bridge as compared to Birchbank and/or Waneta.
Microtox pore water bioassays in April 1995 indicated non-toxicity in sediments at New Bridge (as at Birchbank and Waneta). The New Bridge sediments consisted largely of historical slag deposits. The 14-day Chironomus tentans bioassay showed increased mortality and reduced growth at New Bridge as compared to Birchbank and Waneta, and also somewhat reduced growth at Waneta as compared to Birchbank (Duncan, 1997). These effects may have been related to the shard-like texture of the slag and/or reduced food supply in these deposits.
Sediments collected in a back-eddy pool near Beaver Creek (about 10 km downstream from the C-III outfall) were found by Godin and Hagen (1992) to be toxic to Daphnia magna in solid-phase tests. NECL (1993) could not duplicate these results for D. magna, but found similar results for Hyalella azteca. Sediments from reference stations further upstream at Ryan Creek and downstream at Waneta were not toxic to amphipods.
Macroinvertebrate communities colonizing artificial substrates in 1995 were somewhat reduced in abundance and diversity at the d/s Island and New Bridge stations as compared to Birchbank and Waneta (Duncan, 1997). The d/s Stoney Creek and Old Bridge stations were not so affected. The community effects at New Bridge were consistent with the Chironomus bioassay results.
Colonization of artificial substrates is usually considered to reflect current water quality, although habitat features in the vicinity, such as water velocity, availability of natural substrate and food availability, can all influence the status of the natural community and hence the colonization success. Both periphyton and macroinvertebrate results suggest a depressed community from the Island downstream to New Bridge or Old Bridge (i.e., in the area most influenced by releases from C-III and C-II outfalls). The community status here may have been influenced by historical slag deposits in these areas and/or water quality at the time of study.
2.4.2 Abiotic atmospheric effects
Substances are also assessed to determine whether they have the potential to cause harm to the environment on which life depends as defined in Section 64(b) of CEPA 1999. A substance may be found toxic under Section 64(b) if it is contributing significantly to atmospheric effects such as the formation of photochemical ozone, the depletion of stratospheric ozone or climate change. The discussion that follows begins by examining the typical profile of substances that are capable of causing adverse atmospheric effects. Next, the extent to which these types of substances are released from Canadian copper smelters and refineries and zinc plants is discussed in relation to releases from other Canadian sources. This is followed by an evaluation of whether releases from Canadian copper smelters and refineries and zinc plants are likely to be harming the atmosphere. In the case of emissions from copper smelters and refineries and zinc plants, the release constituents of relevance are SO2, PM, CO2 and VOCs.
2.4.2.1 Photochemical ozone creation
The formation of photochemical ozone is determined by several conditional parameters, but the relative importance of a typical precursor substance is dominated primarily by the reaction rate of the substance with tropospheric hydroxyl radicals (Bunce, 1996; Dann and Summers, 1997). In general, a substance capable of forming ozone must be a reactive compound and must also be volatile at ambient temperature and pressure (i.e., it is a VOC). Emission rates of VOCs, meteorological considerations such as temperature, and the degree of solar radiation present to drive the ozone-forming reactions are all important parameters (Bunce, 1996; Dann and Summers, 1997).
Based on the limited data available for 1995, only 16 tonnes (0.016 kilotonnes) of VOCs in total were reported released to the atmosphere from seven of the nine facilities of concern in these assessments (RDIS, 1995). The total reported VOC release for all of Canada in 1995 was 3575 kilotonnes (RDIS, 1995). The contribution from these smelters and refineries represents a very minor fraction of the total VOCs released from known sources in Canada. As a means of comparison, consider that the highest reported emissions of VOCs at any one location are the 4.62 tonnes released by Falconbridge-Kidd Creek (see Table 6).
Considering that the average light-duty gasoline vehicle emits about 33 kg of VOCs per year (based on sector emissions and number of vehicles in this class obtained from RDIS, 1995), VOC emissions from the Kidd Creek facility are roughly equivalent to those from 140 automobiles.
Therefore, emissions from copper smelters and refineries and zinc plants do not appear to contribute significantly to the formation of ground-level ozone.
2.4.2.2 Stratospheric ozone depletion
Certain compounds containing halogen atoms, such as chlorine, bromine, iodine or fluorine, are capable of depleting stratospheric ozone (WMO, 1998). A series of complex reactions occurs in the stratosphere, usually involving chlorinated or brominated molecules, leading to the creation of a reactive ion and ultimately to destruction of stratospheric ozone. Three of the four release constituents considered here - CO2, PM and SO2 - have no halogen atoms in their molecular structure; therefore, they play no role in the depletion of stratospheric ozone. VOCs can contain halogens, but, as noted above, total emissions of VOCs from Canadian copper smelters and refineries and zinc plants are very low.
It should be noted that SO2 is transformed in the troposphere into sulphate aerosols through a series of oxidative reactions. It is well known that sulphate aerosols facilitate the destructive reaction of chlorine and ozone in the stratosphere, by providing a reactive surface for the heterogeneous (gas-solid) reactions to take place (WMO, 1998). During episodes of volcanic activity, tonnes of sulphur-containing compounds may be injected into the stratosphere. Otherwise, very little of the tropospheric sulphate aerosols migrate to the stratosphere, as their lifetime is too short (4-5 days) to allow their transport to the upper atmosphere. Therefore, it is very unlikely that these stratospheric aerosols are derived from anthropogenic tropospheric sources such as smelting or refining.
Therefore, emissions of copper smelters and refineries and zinc plants do not appear to contribute to the depletion of stratospheric ozone.
2.4.2.3 Climate change
Typically, substances that influence or contribute to climate change must be able to absorb and re-emit radiant energy from the Earth's surface, within the wavelength range 7-14 mm (Wang et al., 1976). Such substances typically must be volatile and sufficiently long-lived to absorb and re-radiate this energy. In relation to emissions from copper smelters and refineries and zinc plants, CO2 and VOCs are the principal release constituents that fit the physical-chemical profile of substances that can contribute to climate change.
Releases of these substances and other greenhouse gases from the non-ferrous metal sector in Canada (including but not restricted to zinc and copper processing facilities) have been estimated for 1995 and reported as CO2 equivalents (Table 6). The non-ferrous metal sector is estimated to have released about 2790 kilotonnes of CO2 equivalents (Jaques, 1997).
The total reported from all Canadian sectors in 1995 was 619 000 kilotonnes. Therefore, the non-ferrous metals industry sector as a whole, which includes mining as well as lead and nickel smelting and refining, contributes about 0.5% of the Canadian total of greenhouse gas emissions (Jaques et al., 1997). Contributions from copper smelters and refineries and zinc plants would be significantly lower. Release data for CO2, N2O and CH4 were available from the RDIS for only four of the nine facilities being considered in these assessments. Emissions of these greenhouse gases from these four facilities in 1995 totalled 352 kilotonnes of CO2 equivalents.
Sulphur dioxide emitted from zinc and copper plants can result in the formation of sub-micrometre sulphate aerosols. Charlson et al. (1992) have described how this aerosol is able to scatter incoming shortwave (solar) radiation in clear sky conditions and improve the reflective properties of cloud surfaces (albedo) in cloudy skies or increase the lifetime of clouds. These ultimately have a cooling effect on the Earth's surface. At present, however, these cooling effects are not sufficiently well understood and fully integrated into global climate change models to quantitatively estimate their impacts on the Earth's climate.
Therefore, based on available information, emissions from copper smelters and refineries and zinc plants do not appear to contribute significantly to climate change.12
2.4.3 Effects on human health - epidemiological studies of populations in the vicinity of copper smelters and refineries and zinc plants
In this section, available epidemiological studies of health effects in human populations near copper smelters and refineries and zinc plants are reviewed.13 As discussed in Section 1.0, the studies considered have been restricted to those in which the population was exposed environmentally (i.e., non-occupationally), as it is these populations that are directly exposed to "releases" from these facilities.
2.4.3.1 Studies of mortality and of cancer incidence
Mortality of humans from various causes, both cancer and non-cancer, as well as cancer incidence in populations in the vicinity of copper smelters or zinc smelters and plants have been examined in a number of studies.
The endpoint most commonly studied was lung cancer. In a number of ecological (correlational) epidemiological studies, mortality from lung cancer was elevated above expectation (usually significantly) in populations near facilities smelting copper and/or zinc (Blot and Fraumeni, 1975; Newman et al., 1975; Pershagen et al., 1977; Cordier et al., 1983; Xiao and Xu, 1985; Semenciw and Manfreda, 1987); these included populations near Canadian smelters in Rouyn-Noranda, Quebec (Cordier et al., 1983), and Flin Flon, Manitoba (Semenciw and Manfreda, 1987). In contrast, lung cancer mortality was not related to residential proximity to U.S. copper smelters in several studies (Polissar et al., 1979; Mattson and Guidotti, 1980; Hartley and Enterline, 1981; Frost et al., 1987), and in a small study of the mortality experience of residents near the lead-zinc smelter in Trail, B.C., lung cancer mortality was not elevated compared to the province as a whole (British Columbia Cancer Agency, 1992).
In most of the ecological studies, some attempt was made to account for the effects of occupational exposure to smelting on lung cancer (usually by excluding smelter employees or by conducting separate analyses for females), although there was no control for occupation in two studies (Polissar et al., 1979; Xiao and Xu, 1985), and the excess observed in males in some studies was at least partly attributable to employment in copper smelting or mining (Newman et al., 1975; Pershagen et al., 1977). There was also no information on other potential confounders, particularly smoking, or on migration in most of these studies. The study populations were also generally quite small in size (i.e., on the order of several thousand), reflecting the remote location of many of the facilities studied.
Results of more robust case-control studies of lung cancer mortality or incidence in relation to residence near copper smelters or zinc smelters and plants, in which there was some attempt to estimate individual exposure, at least crudely, are also mixed. Residence near smelters, or cumulative exposure to smelter emissions, was associated with marginally increased relative risks of developing lung cancer, after controlling for occupational exposures, in several studies near copper (Pershagen, 1985; Frost et al., 1987; Xu et al., 1989) or zinc smelters (Brown et al., 1984). In contrast, there was no significant association between lung cancer risk and intervals at increasing distance from U.S. facilities smelting copper or zinc in three studies of similar design (Lyon et al., 1977; Greaves et al., 1981; Rom et al., 1982). There was also no significant association between lung cancer mortality and various measures of residential exposure to smelter emissions (including highest level of exposure, duration of exposure above background, or cumulative exposure above background) in two well-conducted studies in Arizona copper smelter towns (Marsh et al., 1997, 1998; Stone et al., 1997). In the latter studies, extensive efforts were made to reconstruct exposures and to account for possible confounders, using lifetime residential, occupational and smoking histories, time- and location-specific estimates of residential exposures to smelter emissions based on atmospheric diffusion modelling of ambient SO2 measurements, and application of multivariate statistical techniques.
Although case-control studies are generally considered to be of inherently stronger design than ecological (correlational) studies, the available case-control studies of lung cancer risk in smelter communities are quite limited in a number of respects. As in the ecological studies, exposure was inadequately characterized - few monitoring data were presented for any of the studies, and none from the earliest time periods when exposures would likely have been heaviest. The possible effects of smoking, migration and occupation were not accounted for, and the numbers of cases and controls in the areas nearest the smelters were very small in the studies of U.S. smelters by Lyon et al. (1977), Greaves et al. (1981) or Rom et al. (1982), all of which used a similar design. In addition, in two of these studies (Lyon et al., 1977; Rom et al., 1982), the method of analysis would have yielded a less powerful test than the standard design (Hughes et al., 1988). In the case of the studies by Marsh et al. (1997, 1998), the underlying data were limited by the small number of cases and controls estimated to have had residential exposure above background, and by substantial gaps in the residential and occupational histories for a large proportion of the decedents.
The inconsistent findings with respect to lung cancer risk across the epidemiological studies are perhaps not surprising, given the limitations inherent in the identified studies. While an association with lung cancer is plausible, based on the sufficient weight of evidence for several of the metals emitted from such facilities (Hughes et al., 1994a,b,c; Newhook et al., 1994), exposure of residents to smelter emissions would certainly have been much less than that in the occupational settings where significant excess risks for lung cancer have been observed, with the result that the increased risk, if any, would have been relatively small. In addition, the statistical power of all of the available studies is quite limited. In a review of epidemiological studies of health effects in communities surrounding arsenic-emitting industries, Hughes et al. (1988) estimated the minimum detectable risk for lung cancer near copper and zinc smelters, given the study design and the significance levels used, at 2.0 or more for 8 of the 13 studies they reviewed; the lowest minimum detectable risk was estimated at 1.18. The two case-control studies by Marsh et al. (1997, 1998) also had limited power to detect small increased risks, being designed to have greater than 80% statistical power to detect a relative risk of 2.0 for lung cancer mortality. Moreover, there was limited control for potential confounders, particularly smoking, in the available studies.
The weight of evidence for lung cancer as a result of environmental exposure to smelter emissions is, thus, inadequate. Although an association is plausible, there is little indication of consistency (although statistical power and accounting for potential confounders were limited or inadequate in all studies), strength of association, or an exposure-response relationship (although exposure was only crudely characterized, being most often limited to residence in the surrounding region).
Significant increases in cancers at some other body sites were reported in some studies (Polissar et al., 1979; Lauwerys and De Wals, 1981; British Columbia Cancer Agency, 1992; Kreis, 1992; Wong et al., 1992; Wulff et al., 1996a), but these results were most often based on very small numbers of cases, and there was no consistent increase in any specific type of cancer. Hence, there is no consistent convincing evidence of cancers for sites other than the lung either.
With respect to non-neoplastic causes of death, the only finding with any degree of consistency is mortality from respiratory disease, which was significantly increased in a few ecological studies (Mattson and Guidotti, 1980; Cordier et al., 1983; Semenciw and Manfreda, 1987). Pershagen et al. (1977) also observed a non-significantly increased standardized mortality ratio for respiratory disease mortality in both males and females residing near the Ronnskar copper smelter in northern Sweden. The category of respiratory disease that was affected was not consistent (i.e., acute disease mortality in some studies, chronic in others), although the reliability of the death certificate data on which these distinctions was made in these early studies is unknown. However, there is no reliable information on the levels of smelter-related substances to which these populations were exposed, and accounting for possible confounders, such as smoking, was inadequate in all of these studies.
2.4.3.2 Non-neoplastic effects
Non-neoplastic effects in populations near copper smelters or zinc smelters and plants have been investigated in numerous studies. Endpoints investigated have most often included blood lead levels and associated effects on the neurological and heme systems. Renal and respiratory effects have also been investigated in a number of studies.
Levels of lead in blood in populations in the vicinity of smelters have been investigated in a large number of studies. Blood lead levels were elevated in most studies of residents near copper smelters and zinc smelters and plants, reflecting the large quantities of Pb released to the environment (Landrigan et al., 1975a, 1976; Roels et al., 1976; Savoie and Weber, 1979; Ewers et al., 1985; Chenard et al., 1987; Cook et al., 1993; Gagné, 1993, 1994; Galvin et al., 1993; Trepka et al., 1997; Hilts et al., 1998). These included the Canadian facilities at Rouyn-Noranda, Quebec (Gagné, 1993, 1994), Murdochville, Quebec (Chenard et al., 1987) and Trail, B.C. (Hilts et al., 1998). [However, blood lead levels were not clearly increased in two surveys of populations residing near a number of U.S. smelters (Baker et al., 1977; Hartwell et al., 1983).] The increased Pb burden was typically most pronounced in young children (Landrigan et al., 1975a, 1976; Savoie and Weber, 1979; Hilts et al., 1998), as a result of such factors as increased contact with house dust or soil, increased hand-to-mouth activity and greater gastrointestinal absorption. The blood lead levels in a number of these studies, particularly the earlier ones, were markedly elevated, with a large proportion of the children near smelters having levels well in excess of 10 mg/dL (Landrigan et al., 1975a, 1976; Roels et al., 1976; Chenard et al., 1987; Cook et al., 1993; Gagné, 1993, 1994; Galvin et al., 1993; Hilts et al., 1998), the currently recommended intervention level (CEOH, 1994).
Reduced environmental exposure to Pb, as a result of one or more of reduced emissions, remediation measures, and education and intervention efforts, resulted in marked reduction in children's blood lead levels in a number of these studies (Yankel et al., 1977; Landrigan and Baker, 1981; Gagné, 1993, 1994; Hilts et al., 1998; Hilts, 2000), including near the Canadian facilities at Rouyn-Noranda (Gagné, 1993, 1994) and Trail (Hilts et al., 1998; Hilts, 2000). The most recent data from populations near copper smelters or zinc plants in Canada indicate that roughly 10-20% of children surveyed had blood lead levels greater than or equal to 10 mg/dL (Chagnon and Bernier, 1990; Gagné, 1993, 1994; Hilts, 2000). This compares favourably with earlier surveys at these locations, in which the majority of children studied had blood lead levels greater than or equal to 10 mg/dL (Chenard et al., 1987; Gagné, 1993, 1994; Hilts et al., 1998). No data on blood lead levels in populations in the vicinity of the remaining Canadian copper smelters and zinc plants were identified.
In the studies identified, levels of lead in blood were typically not related to a variety of other possible sources, including Pb from paint, local produce, drinking water or culinary pottery (Landrigan et al., 1975a, 1976; Baker et al., 1977; Cook et al., 1993; Hilts et al., 1998; Meyer et al., 1998), but were instead significantly associated with levels of Pb in ambient air, household dust or soil (Landrigan et al., 1975a, 1976; Roels et al., 1976; Yankel et al., 1977; Cook et al., 1993; Galvin et al., 1993; Hilts et al., 1998; Meyer et al., 1998).
In a number of these populations, elevated levels of lead in blood were accompanied by characteristic effects of Pb on the heme system, including reduced activity of delta-aminolevulinic acid dehydratase (Roels et al., 1976; Savoie and Weber, 1979), reduced hematocrit and increased levels of free erythrocyte protoporphyrin, although the latter effects occurred only at higher exposures (Landrigan et al., 1976; Roels et al., 1976; Savoie and Weber, 1979; Chenard et al., 1987). The nervous system was also affected in two early U.S. studies, in which extremely high blood lead levels in children living near smelters processing copper and/or zinc were accompanied by impaired performance in neuropsychological testing (effects on non-verbal cognitive and perceptual motor skills, fine motor skills and performance IQ) (Landrigan et al., 1975b) and reduced peroneal nerve motor conduction velocity (Landrigan et al., 1976).
The evidence of other non-neoplastic effects in populations with environmental exposure to substances released from copper smelters and zinc smelters and plants is more limited.
Effects on renal function and on calcium metabolism/bone mineralization were observed in three well-conducted cross-sectional studies in regions of Belgium and The Netherlands contaminated with Cd emitted from zinc smelters and plants. In these studies, Cd in urine (a measure of lifetime exposure to Cd) was significantly increased in the contaminated regions, and residence in these regions, proximity of residence to the zinc smelters and plants and/or urinary Cd excretion were significantly associated with increased excretion of various urinary markers of renal proximal tubular function, after adjustment for a wide range of potential confounders (Buchet et al., 1990; Kreis, 1992; Staessen et al., 1994; Hotz et al., 1999). There were also indications of Cd-related alterations in calcium balance in these populations (increased serum alkaline phosphatase activity, decreased serum calcium, increased urinary calcium excretion), perhaps secondary to the renal effects (Staessen et al., 1991a; Kreis, 1992). Urinary excretion of Cd or calcium or residence in the contaminated areas was associated with significantly decreased forearm bone density and increased risk of bone fractures and height loss, after adjustment for confounders (Staessen et al., 1999). There were no clear effects on blood pressure or on the prevalence of hypertension or cardiovascular disease in these populations (Staessen et al., 1991b; Kreis, 1992).
In another well-conducted study, children living near a copper-silver smelter in Germany had small but significantly increased urinary arsenic and cadmium and blood lead levels (Trepka et al., 1996, 1997; Ritz et al., 1998) and increased prevalences of respiratory diseases and allergies (including histories of bronchitis, allergy, eczema and various respiratory symptoms, as well as positive skin prick tests and increases in specific IgE in physical examination) (Heinrich et al., 1999). There were also increased prevalences of reported cough, but no significant effects on other respiratory symptoms or on lung function in children exposed to high levels of SO2 emitted from copper smelters in Arizona in two more limited studies (Dodge, 1983; Dodge et al., 1985).
There were no remarkable patterns in medical utilization/hospitalization studies of populations near Canadian copper smelters and zinc plants in Flin Flon, Manitoba (Anon., 1987) and Trail, B.C. (Fisk et al., 1994), although each of these studies was very limited in scope, and the utilization rates would have been affected by other factors in addition to disease prevalence.
2.4.3.3 Effects on reproduction and development
Epidemiological studies of effects on reproduction and development that were identified are limited to a small number of ecological (correlational) studies of populations residing near a copper smelter in northern Sweden. In these studies, residence in areas near the smelter was significantly associated with an increased frequency of spontaneous abortion (Nordstrom et al., 1978a) and with reduced birth weight in one study (Nordstrom et al., 1978b), but not with reduced weight in another study (Wulff et al., 1995), time to pregnancy (Wulff et al., 1999) or with the frequency of congenital malformations (Nordstrom et al., 1979; Wulff et al., 1996b). However, exposure characterization was limited to residence in a given area, and all of the studies were further limited by one or more of small numbers of cases and inadequate control for other factors that may affect the endpoints examined, such as smoking or parental employment at the smelter.
- 12. It is nevertheless recognized that efforts to reduce emissions of greenhouse gases from all relevant sectors should be undertaken as part of Canada's overall strategy to minimize climate change.
- 13. There are a large number of epidemiological studies of Japanese populations that were exposed environmentally to Cd emitted from facilities that produced copper and/or zinc (much of this work has been summarized by Tsuchiya [1978]). However, these studies are considered to be less relevant (the study populations were principally exposed via consumption of local rice grown in paddies contaminated by discharges from smelting and often also from mining) and have not been included.
