Water Quality Guideline Derivation
The Canadian water quality guidelines for aluminium were developed based upon CCME protocol (CCME 1991) and are provided for both Alim and Altot. Many studies have demonstrated that Alim is the most toxic form of aluminium (Parkhurst et al. 1990; Holtze and Hutchinson 1989; Baker and Schofield 1982; Driscoll et al. 1980) and that it is the best predictor of aluminium toxicity over a wide range of water quality conditions (Parkhurst et al. 1990; Gundersen et al. 1994; Clark and Hall 1985; Driscoll et al. 1980). In a field study, Baldigo and Murdoch (1997) used regression analysis to examine the effects of fluctuating chemical conditions to brook trout (Salvelinus fontinalis) in four Catskill Mountain streams. The regression analysis suggested that brook trout mortality was related to (in order of importance) Alim, pH, DOC, calcium, and chloride in stream water. Mean or median Alim concentrations accounted for 76-85% of the variability in brook trout mortality. A similar study by Parkhurst et al. (1990) examined the effect of pH, Altot, Alim, DOC, fluoride ion, and temperature on brook trout. The authors determined that the order of significance of these factors to brook trout fry survival in acidic solutions was Alim, pH, DOC, F-, and temperature. Alim reportedly had a much greater effect on the survival of brook trout than pH.
Calcium (Ca2+) can have a major impact on the toxicity of aluminium to some aquatic organisms (Cleveland et al. 1991; Brown 1983). Calcium has a known ability to reduce the permeability of biological membranes thereby reducing the loss of plasma ions due to both pH and the presence of metal ions, including Al (Freda and McDonald 1990; Sadler and Lynam 1988). Calcium acts at the level of the biological membrane, and specific responses will vary depending upon the organism (e.g., fish, invertebrates, plants). This guideline derivation did not use calcium as a contributing parameter because dose-response corelations were unavailable to be considered in the model relationship.
The impact of temperature on aluminium speciation has been addressed by separating the Alim and Altot guidelines into pH ranges based on the speciation of aluminium at 2°C (Lydersen et al. 1990; Howells et al. 1990). Alim, the most toxic form, is expected to be dominant over a broader pH range at 2°C than at 20°C.
The guidelines for Alim are based on the data presented in Figure 1. A first regression of guideline values is derived for pH values from 5.0 to 6.1. Data used to derive the guideline come from the studies of Buckler et al. (1995) on Atlantic salmon (Salmo salar); Parent and Campbell (1994) on a green algae (Chlorella pyreinoidosa), Baker and Schofield (1982) on brook trout (Salvelinus fontinalis) and white sucker (Catostomus commersoni). A second regression of guideline values is derived for pH values from 6.1 to 7.1. Data used to derive the guideline come from the study of Helliwell et al. (1983) on C. pyrenoidosa. Regression lines intercept at pH value 6.1, the lowest observed effect concentration (EC50 at Alim concentration of 5 μg-L-1; Helliwell et al. 1983). The combined regression lines follow the shape of the aluminium solubility curve (microcrystalline gibbsite) (OMOE 1988). Guideline regression equations for Alim are:
A safety factor of 0.1 is applied to the calculated Alim guideline values. Data on background concentrations of Alim in Canada were sparce and no conclusion could be drawn regarding the need for a safety factor as for total aluminum (see below). No Alim guideline was developed for freshwater in the range below pH 5.0 and above pH 7.1 because there are no data for aquatic organisms exposed to Alim in these ranges. At pH greater than 7.1, it is recommended to use the guideline for Altot.
The data requirements for the development of a full Altot water quality guideline for aquatic life have not technically been met due to the lack of one chronic invertebrate study as outlined in CCME (1991). Several authors (Vuori 1996; OMOE 1988; Havas 1985) have noted the general lack of invertebrate data. The data available to derive the guideline includes a large number of fish species, aquatic plant data and includes amphibian data. Amphibians tend be very sensitive to aluminium concentrations (Sparling and Lowe 1996; Freda 1991; Albers and Prouty 1987). According to CCME (1991) protocol, enough data is available for the development of an interim Altot guideline.
The toxicity of Altot is strongly influenced by pH and concentrations of dissolved organic carbon (Buckler et al. 1995; Wilson et al. 1994; DeLonay et al. 1993; Parkhurst et al. 1990; Palmer et al. 1988; Sadler and Lynam 1988; Parkhurst 1987; Baker and Schofield 1982). To take account of these variables in the derivation of guidelines for Altot, a multiple regression analysis was performed based on the data from Parkhurst (1987). The dependent variable was 21 day survival of juvenile brook trout transformed into logits and transformed again by adding 10 to eliminate negative numbers:
where p is the proportion of alevins survivi ng to 21 days. The independent variables were pH, DOC and Altot concentrations. In the regression analysis, total aluminium concentrations were log transformed, and power terms were added for log total aluminium and pH because these variables had non-linear relationships with survival. The analysis was performed in SAS® using PROC REG. Using the method of maximum likelihood, equation 4 was obtained:
Calculation details are provided following the reference section. All variables in the equation were significant (p<0.05). The correlation coefficient was 0.71. Using this equation, the concentrations of total aluminium causing 25% greater mortality than caused by pH and DOC alone were estimated for a number of pH and DOC combinations. Because the model is based on chronic exposure data involving an early life stage of brook trout, the estimated LC25s are considered to be chronic endpoints in this case because the sensitivity of the trout declined rapidly with age, lethality was the most sensitive response measured, and the IC25 has been considered in some cases to approximate the no observed effect concentration (see USEPA 1994). The protocol (CCME 1991) uses a safety factor of 10 on a chronic endpoint. However, there was concern that the resulting guideline would have been lower than natural background. To evaluate this, monitoring data on total aluminium water concentrations in non-industrialised areas were obtained from all regions in Canada (Caux et al. 2000). A common database of 75,000 data points was screened (post 1985 only) and compiled. The data were collected from as many Canadian "geological provinces" as possible (Wheeler et al. 1996). In total, 11 of the 13 geological provinces were represented. A frequency concentration plot for total aluminium demonstrated that the number of sites for which the concentration of total aluminium falls at or below 252 μg-L-1 was 80% (n=1196). The matrix for the proposed effects-based total aluminium guideline (pH and DOC dependent) in the absence of 10-fold safety factor, ranged from 61 to 252 μg-L-1, a perfect overlap for this proportion. In this case, a weight of evidence approach integrating effects thresholds, water chemistry relation-ships and background concentrations support a water quality guideline for total aluminium without the need of a safety factor.
At the site-specific level, in most Canadian geological provinces, there are proportionally only a few sites that have background concentrations of Altot below 61 μg-L-1. For these sites, if they are deemed ecologically sensitive, a site background concentration could be used as a site-specific objective value. Although natural levels of substances may have an adverse effect on certain organisms, defensible management options should consider the contribution of natural processes in order to focus on the sites and chemicals that are primarily influenced by human activities.
It is recognised that situations outside of the range of pH and DOC concentrations presented in Table 2 occur in the Canadian environment, and in fact, that the CWQG for pH alone is from 6.5 to 9.0 (CCME 1999). Parkhurst (1987), however, did not include treatments with pH higher than 6.5 or DOC concentrations higher than 10 mg-L-1. Hence it is recommended that the most conservative total aluminium guideline closest to the pH and DOC conditions in the water body of interest be applied (i.e., 74 μg-L-1). Users of the guideline should also be aware that rapid increases in pH, such as in mixing zones, may precipitate aluminium and result in over-saturated solutions that may be acutely lethal to aquatic life.
It is recommended that the choice of guideline, either Alim or Altot be based on the management goal and best professional judgement, on a case-by-case basis. The guideline for Alim contains a safety factor and is based on the putative toxic fraction, whereas the Altot guideline may be more practical, takes natural background concentrations into account and may be more useful for non-equilibrium conditions. In addition, highly coloured, low-pH lakes (dystrophic lakes) may need to be considered on a site-specific basis. Jurisdictions may choose to develop site-specific water quality objectives as outlined in CCME 1991.
The guidelines set a high standard for the protection of aquatic life across Canada. Their use may require consideration of local conditions. The guidelines may be updated as new information becomes available.
Page details
- Date modified: