The state of the water quality was judged by comparing values to Ministry of Environment, Lands and Parks' Approved and Working Criteria for Water Quality (Nagpal et al., 1997). Site-specific water quality objectives have not been set for the Columbia River upstream from the Keenleyside Dam.
MacLeod and Whitfield (1996) analyzed six years (1986-91) of water quality data from the Columbia River at Revelstoke to determine the processes controlling water quality and the presence of environmentally significant trends. No trends over time were found for water quality variables. However, upstream reservoirs had altered the natural seasonal patterns for most water quality indicators, and had removed the normal relationship between flow and sediment-related water quality indicators (e.g., total metals).
Variables plotted but not discussed below showed no clear trends over time and met all criteria. These variables are:
total arsenic, total barium, total beryllium, dissolved chloride, total chromium, dissolved fluoride, hardness, total lead, total lithium, magnesium, total molybdenum, total nickel, ammonia-N, nitrogen (nitrate/nitrite), total dissolved nitrogen, pH, total dissolved phosphorus, potassium, filterable residue, silica, sodium, specific conductivity, total strontium, dissolved sulphate, water temperature, and total vanadium.
Variables which exceeded criteria, exhibited trends over time, or were significant in some regard were:
Total alkalinity (Figure 3) and calcium (Figure 9) show that the river had a low sensitivity to acid inputs (the river was well buffered). Lower values were reported in July and August.
Total aluminum (Figure 4) values were collected between 1990 and 1996. One value exceeded the aesthetic criterion (0.2 mg/L) for drinking water in 1995. At least 89% of the values met the maximum criterion (0.1 mg/L dissolved) for aquatic life, and at least 48% of the values met the average criterion (0.05 mg/L dissolved) for aquatic life. However, total aluminum values are not directly comparable to dissolved criteria because dissolved aluminum values in the water column are much lower than total aluminum. Dissolved aluminum should be measured in any future monitoring to provide a direct comparison to the criteria.
Total cadmium (Figure 8) values were high between 1986 and 1991 due to preservative vial contamination. Since that time, all total cadmium values were less than the minimum detectable limit (0.0001 mg/L). The minimum detectable limits for cadmium (0.001 and 0.0001 mg/L) were 3 to 33 times above the aquatic life criterion (0.00003 mg/L). To compare the cadmium levels to criteria for aquatic life more effectively, the minimum detectable limit should be lowered to at most one-tenth of the criterion value, and dissolved cadmium should also be measured if high total cadmium values are found.
Total chromium (Figure 11) values were high due to suspected preservative vial contamination between 1986 and 1991. Since that time, all values met the aquatic life criterion (0.02 mg/L) for fish and the drinking water criterion (0.05 mg /L). Three values exceeded the criterion (0.002 mg/L) for plankton. Suspended sediments (non-filterable residues or turbidity) were low on these occasions, indicating that the chromium may have been biologically available. Total chromium values in excess of the 0.002 mg/L criterion have been observed at many of the Canada-B.C. water quality monitoring stations throughout the province. These occurrences may have been natural phenomena or due to network-wide artificial contamination (Pommen, 1996).
Total cobalt (Figure 12) exceeded the aquatic life criterion (0.0009 mg/L) in one sample collected in April, 1990. This sample had an elevated non-filterable residue value (55 mg/L), indicating that the cobalt was in a particulate form and probably not biologically available. Since 1990, all values have been near or below the detection limit of 0.0001 mg/L.
Total copper (Figure 14) had high values due to preservative vial contamination between 1986 and 1991. Since that time, two values exceeded the average criteria to protect aquatic life (0.002 mg/L to 0.003 mg/L, depending on hardness values) in 1991. The criteria levels have not been exceeded since then. However, these values were collected less frequently (less than five samples in 30 days) than required to evaluate the attainment of the criteria rigorously. There was no apparent relationship between copper and suspended sediments (non-filterable residues or turbidity) in samples collected after the period of vial contamination (Figure 15), suggesting that copper was not associated with particulate matter and may have been biologically available.
Hardness (Figure 17) values were lower than the optimum range (80 -100 mg/L) for drinking water at the Revelstoke site, but were still quite acceptable.
Total iron (Figure 18) values exceeded the 5 mg/L irrigation criterion on one occasion and the 0.3 mg/L drinking water (aesthetics) and aquatic life criterion in 8% of the samples between 1984 and 1996. Higher iron and non-filterable residue values occurred together (Figure 19), indicating that the iron was in a particulate form and probably not biologically available. Also, high iron values would be reduced by drinking water treatment to remove turbidity.
Total lead (Figure 20) had high values due to preservative vial contamination between 1986 and 1991. Since that time, all total lead values met the 30-day average criteria to protect aquatic life (0.004-0.006 mg/L).
Total manganese (Figure
24) exceeded the criterion for aquatic life (0.1 mg/L) once and two values
exceeded the aesthetic criterion for drinking water (0.05 mg/L). High manganese
and non-filterable residues or turbidity were reported in these samples collected
in 1989 and 1990, indicating that the manganese was in a particulate form and
probably not biologically available. Also, high manganese values would be reduced
by drinking water treatment needed to remove turbidity. Annual average
total manganese values decreased between 1986 and 1995 due to two factors:
· detection limits were lowered from 0.01 mg/L to 0.001 mg/L in 1986, and
· higher manganese values in 1989 and 1990 due to higher non-filterable residues or turbidity.
Consequently, the apparent decrease is deemed to be a measurement artifact, rather than an environmentally significant trend.
Dissolved Ortho-phosphorus (Figure 33) values were less than the minimum detectable limit (0.003 mg/L) before 1992. Since that time, 15% of the values were greater than the minimum detectable limit.
Total phosphorus (Figure 34) data were examined for trends using statistical methods due to concerns that upstream dams and reservoirs were reducing phosphorus input to the Arrow Lakes. Non-parametric tests indicated that there was strong evidence of a linearly decreasing trend in total phosphorus concentration data for 1986-96 (Regnier, 1998). B.C. Ministry of Environment Kootenay Region Fisheries has recommended beginning phosphorus fertilization in the Upper Arrow Lake Reservoir in the spring of 1999 (Columbia Basin Fish & Wildlife Compensation Program, 1999).
Non-filterable residue (Figure 37) values were greater than the minimum detectable limits (1 mg/L and 4 mg/L) in 14% of the samples collected between 1987 and 1995. Two values exceeded the criterion (25 mg/L) for good fisheries. The values were lower than would be expected for an unregulated river due to the settling out of non-filterable residues in Kinbasket Lake and Lake Revelstoke reservoirs upstream from the sample site.
Total selenium (Figure 38) values were greater than the minimum detectable limit (0.0001 mg/L) in 11% of the samples collected between 1984 and 1995. One value in 1995 slightly exceeded the maximum criterion (0.001 mg/L) for aquatic life. The low non-filterable residues and turbidity values for this sample indicate that the selenium was not in a particulate form and may have been biologically available. Bethlehem Resources Corporation's zinc mine in the Goldstream River Valley, which closed in January 1996, may have been a source of the elevated selenium. The mine was discharging selenium to the Goldstream River in 1995-96, resulting in levels of up to 0.01 mg/L in the Goldstream River (McDonald, 1996), and given the dilution for the Goldstream River in the Columbia River (about 22:1), it could have contributed significantly to the levels at Revelstoke.
Water temperature (Figure 46) values met the drinking water (aesthetics) criterion (15 °C) and the maximum criteria (22-24 0C) for adult and juvenile salmonids. All values were less than the optimum range (15-35 0C) for recreational use (swimming) between 1983 and 1995, and thus the water at this site was too cold for swimming.
Turbidity (Figure 47) values exceeded the 5 NTU aesthetics objective for drinking water (with disinfection only) in one sample collected in May 1989. The health-based drinking water criterion (1 NTU) was exceeded 29% of the time. Turbidity values were lower than in other natural rivers in the Kootenay area during freshet because of settling in the Kinbasket Lake and Lake Revelstoke reservoirs. The turbidity levels in the Columbia River water at Revelstoke are such that treatment processes to remove it are required prior to drinking.
Total zinc (Figure 49) had high values due to preservative vial contamination between 1986 and 1991. Since that time, all values met the maximum criterion for aquatic life (0.03 mg/L for fish and invertebrates), while one value in 1991 exceeded the criterion to protect algae (0.015 mg/L). This value was in a sample with low non-filterable residues or turbidity, indicating that the zinc was not in a particulate form, and may have been biologically available (Figure 50). All zinc values have met all criteria since 1991.