2.1 NATURAL OCCURRENCE OF MANGANESE

Manganese comprises approximately 0.085% to 0.095% of the earth's crust and is a component of many rock types, particularly those of metamorphic and sedimentary origin (CCME, 1987). It is associated with iron ores of submarginal concentration; the predominant ores of manganese include pyrosulite (MnO₂), manganite (Mn₂O₃·H₂O), hausmannite (Mn₃O₄), psilomelane and rhodochrosite (MnCO₃) (CCME, 1987; Moore, 1991). Ferromanganese minerals such as biotite mica and amphiboles contain large amounts of manganese and manganese-rich nodules have been identified on the sea floor in conjunction with cobalt, nickel and copper (CCME, 1987; Moore, 1991). Important natural sources of manganese include soils, sediments and metamorphic and sedimentary rocks:

Manganese occurs in soil as a result of weathering of rock containing manganese during the process of pedogenesis. A broad range of naturally occurring manganese concentrations in soil has been observed. The BC Ministry of Environment, Lands and Parks (1998a) has collected data on uncontaminated British Columbia soils for various regions of the province. A summary of this data is presented in Table 2.1, as follows:

The data in Table 2.1 illustrate the broad range of concentrations of manganese that occur in British Columbia soils. Notable regional differences are apparent in the data, with concentrations in Vancouver Island soils significantly higher than those in other regions. Regional mean concentrations varied from 284 µg/g to 1359 µg/g while median concentrations (50th percentile) varied from 272 µg/g to 660 µg/g. Although the samples were obtained from a variety of locations within each region, samples were typically collected in or near areas of settlement and from native rather than fill soils. The size of and geologic variability within each region may limit the degree to which the data are representative on a region-wide basis. However, the data do provide valuable information regarding the range of manganese concentrations that occur in British Columbia.

The natural presence of manganese in rock and soil provides a source of manganese that may dissolve in ground and surface waters or may erode and deposit as sediment, with the subsequent potential for dissolution. Manganese accumulated in plant material will also provide a source for dissolution during decomposition. Manganese solubility increases at low pH and under reducing conditions and is most commonly in the 2+ and 4+ oxidation states in aquatic systems. The presence of high concentrations of chlorides, nitrates and sulphates may increase manganese solubility, increasing both aqueous mobility and uptake by plants (Clement Associates, 1985). Manganese precipitates out in sediment mainly as Mn4+ and re-solubilizes in the water column mainly as Mn2+ (Moore, 1991).

Dissolved concentrations of manganese in natural waters that are essentially free of anthropogenic sources/influences range from <0.01mg/L to >10 mg/L (McNeely et. al., 1979). Manganese concentrations in natural surface waters seldom reach 1.0 mg/L and are usually less than 0.2 mg/L, while seawater typically contains approximately 2 µg/L of manganese (McNeely et.al., 1979). Environment Canada data for the period of 1980- 1985 for the Pacific Region (CCME, 1991) and data from the BC Ministry of Environment, Lands and Parks (1998b) are summarized in Table 2.2:

Total manganese concentrations in surface water showed a typical seasonal trend, with the highest annual manganese concentrations observed during high runoff periods (e.g. spring snow melt period for Interior streams) and lower concentrations observed during periods of stable stream flow. Concentrations in stream waters were higher than concentrations in lakes and concentrations in streams downstream of lakes were lower than concentrations in other streams. These trends are in keeping with expected results as higher suspended sediment (and consequent higher manganese) loads typically occur during higher runoff periods and in flowing water. Concentrations in excess of 1.0 mg/L were rare in the BC Environment data set.

2.2 MAN-MADE SOURCES OF MANGANESE

Manganese is used in industrial processes and in various consumer products. The major man-made sources of environmental manganese include municipal wastewater discharge, sewage sludge, emissions generated during alloy, steel and iron production, and to a lesser extent by emssions from the combustion of fuel additives (Moore, 1991; Jaques, 1987). Worldwide anthropogenic input of manganese to freshwater is summarized in the following table (Nriagu et. al., 1988):

The primary man-made sources of atmospheric manganese worldwide are secondary non-ferrous metal production, coal burning and municipal waste incineration (Moore, 1991). Incineration of sewage sludge was estimated to be the third largest worldwide anthropogenic source of manganese emissions to the atmosphere in 1983. Environment Canada estimated that 1984 emissions of manganese in Canada totaled 1225 tonnes, of which 47% resulted from ferromanganese and silico-manganese production (all in Quebec), 28% resulted from iron and steel production (mainly in Ontario and to a lesser extent Quebec) and 17% resulted from gasoline-powered motor vehicle emissions (Jaques, 1987). In British Columbia, total emissions were estimated at 31 tonnes, with 27 tonnes originating from gasoline powered vehicles (Jaques, 1987).

Although it is not known whether manganese emissions from sources other than gasoline powered vehicles have increased significantly in British Columbia since 1984, it seems probable that vehicle emissions continue to be the major source of manganese emissions in the province. Manganese additives in gasoline are the source of manganese in vehicle emissions. Methylcyclopentadienyl manganese tricarbonly, or MMT, is the main additive containing manganese (approx. 24.4% by weight); the additives LP62 (containing 62% MMT) and LP 46 (containing 46% MMT) are also common (Jaques, 1987). The main benefits of MMT addition to gasoline are octane enhancement and suppression of smoke during combustion. The recommended Canadian limit for MMT in gasoline is 18 mg Mn/L. Based on the emissions information provided by Environment Canada, it would not appear that MMT is a significant source of environmental manganese. This may be borne out by the soils data for Greater Vancouver and the Lower Mainland (see Table 2.1), the area with the greatest urban population and concentration of automobiles. Manganese concentrations in the upper 60 cm of soil from these areas had the lowest median concentrations in the province and maximum individual sample concentrations were low as compared to many other regions.

2.3 FUNCTIONS/ESSENTIALITY OF MANGANESE IN BIOTA

CCME¹ reports that manganese is an essential trace element for microorganisms, plants and animals and is therefore present in almost all organisms. Manganese in plant tissues mainly occurs in nuts, seeds, whole grains (particularly the bran and germ), legumes, dark leafy green vegetables and alfalfa; egg yolks, black tea and coffee beans also contain significant manganese (Haas, 1998; Klassen, 1996). Manganese content in plant tissue is largely dependent on sufficient manganese content in the soils in which the plants grow.

Manganese activates an essential part of enzyme systems that metabolize proteins and energy in all animals; manganese is also involved in the formation of mucopolysaccharides needed for healthy joint membranes (Haas, 1998). It concentrates in the mitochondria and is present in higher concentrations in tissues rich in mitochondria. Manganese concentrations in fish tissue were found to be higher in liver and gill tissue than in muscle tissue (Legoburu et. al., 1988)). In humans, manganese is involved in the digestion and absorption of food through peptidase activity, in the synthesis of cholesterol and fatty acids, in glucose metabolism and in the use of biotin, thiamine, vitamin C and choline. In the divalent state (Mn++), it also appears to provide protection against oxygen free radicals as part of the enzyme superoxide dismutase (Haas, 1998). A daily allowance of 1.2 mg of manganese has been recommended for humans and information appears to indicate that insufficient manganese may result in inhibited carbohydrate metabolism and impaired insulin production, while excess manganese may inhibit iron absorption (Moore, 1991).

2.4 FRESHWATER AQUATIC TOXICITY DATA IN LITERATURE

Studies pertaining to the aquatic toxicity of manganese to various fresh water organisms were researched to determine the breadth and applicability of existing data. Although the number of studies that evaluated manganese toxicity to aquatic organisms was not extensive, a few studies provided important information to supplement the new information generated by BC Environment and are presented in Sections 3 and 4 of this thesis. For ease of presentation, studies that are applicable to species that exist in BC waters have been separated from species not present in BC waters.

2.4.1. Studies on Species Present in B.C. Waters

A summary of aquatic toxicity literature data for species present in B.C. fresh water is provided in Table 2.4, as follows:

Note: 1 - Baird et. al., 1991
2 - Davies and Brinkman, 1994
3 - Stubblefield et. al., 1997

IC₂₅ - Statistically derived concentrations at which 25% of organisms are inhibited for the exposure endpoint in the study (e.g. growth, reproduction, hatching success) vs. controls

Baird et. al. (1991) evaluated six clones of Daphnia magna to determine the differences in acute toxic response between genotypes to nine different chemicals, including manganese. Measured ionic concentrations used in the studies ranged from 1-100 mg/L for Mn 2+ as manganese chloride. Daphnia species 14 day reproduction testing methodology outlined by OECD (Organization for Economic Cooperation and Development), which included an acute immobilization test, was employed and the measured effect was lethality as evidenced by immobility. Hard water, as defined by the American Society for Testing and Materials (ASTM, 1980), was used to culture the organisms. No hardness value was reported.

The resulting EC₅₀ (concentrations at which effects were observed in 50% of organisms vs. controls) data were converted to normal density functions with relative frequency (i.e. fraction or percentage of occurrence) plotted against concentration. The EC₅₀ values represented the midpoints of the density functions. The EC₅₀ values presented in the report for the six genotype clones ranged from a minimum of 4.7 mg/L to a maximum of 56.1 mg/L. In the absence of raw data, the lowest concentration for which a toxic response was observed was extrapolated from the probability density plots. A value of approximately 3 mg/L resulted and this value was associated with the plot having an EC₅₀ of 4.7 mg/L. In the context of a freshwater aquatic life guideline, the relevance of particular genotypes may be little more than recognition of the most sensitive genotype and the associated EC₅₀ concentration, thus ensuring a conservative and ecologically protective approach.

Baird et. al. (1991) concluded that genotypes of Daphnia magna exhibited a considerable range of EC₅₀ concentrations with no concordance between genotype response between the different chemicals. Genotypes that were the most sensitive to one chemical may have been the least sensitive to another chemical and lie near the middle of the response results for another chemical, with no pattern emerging.
The recent study by Stubblefield et. al. (1997) focussed on brown trout, a species that is present in British Columbia in localized waters. The objectives of the study were to "determine the toxicity of manganese to early life stages of brown trout, to evaluate the hardness-toxicity relationships and to provide data useful in developing a protective manganese criterion. The hardness-toxicity relationship was evaluated by testing several manganese concentrations at water hardness values of 30, 150 and 450 mg/L CaCO₃. The life stages utilized in the study included fertilized eggs and larvae/fry. A summary of the materials and methods applied during this study follows.

Measured amounts of manganese chloride (Mn CL₂·4H₂0) were dissolved in de-ionized water to prepare the test solutions. Reservoir water with a hardness of 30 mg/L CaCO₃, well water with a hardness of 450 mg/L CaCO₃, and a mixture of the two water sources to obtain a hardness value of 150 mg/l CaCO₃ were used. Seven nominal manganese concentrations were tested at each of the three hardness values, with dissolved concentrations analyzed weekly. The toxicity testing methodology was based on ASTM Method E1241-92 (ASTM, 1993). Mean dissolved concentration ranges to which organisms were exposed were 0.43 to 15.15 mg/L for a hardness of 30, 2.84 to 71.95 mg/L for a hardness of 150 and 2.41 to 93.36 mg/L for a hardness of 450. The dissolved manganese concentrations used for the control groups were all <0.02 mg/L.

For each test, fifteen randomly chosen embryos were placed in 2.2 litres of test solution contained in a glass aquarium. Each test was repeated four times for a total of sixty organisms per treatment. Temperature was maintained at 12±1 degrees C and the total duration of the tests was sixty-two days. Mean organism wet weights were measured and statistically evaluated to compare hatching success, survival and growth versus controls for each of the tests. The lowest observable effect concentration or LOEC was established as the lowest concentration for which a statistically significant effect was observed versus controls. The no observable effect concentration or NOEC was established as the highest concentration for which no statistically significant effect was observed. Although some discussion regarding statistical testing applied during the study was provided in the text, it was not clear what constituted "statistically significant."

The main findings reported by Stubblefield et. al. (1997) were as follows:

1. Hatching success varied from 86.6% to 98.2% and was not generally affected by exposure to manganese at the test concentrations used. The mean time to hatch was decreased for the highest manganese concentrations at hardness values of 150 and 450 mg/L CaCO₃.
2. Survival of larvae decreased with increasing manganese concentrations for each of the test hardness values. Dissolved manganese LOEC values for organism survival (not growth) were determined to be 7.38 mg/L for a hardness of 30 mg/L CaCO₃, 8.81 mg/L for a hardness of 150 mg/L CaCO₃ and 16.21 mg/L for a hardness of 450 mg/L CaCO₃
3. For each water hardness tested, organism mortality was observed sooner at higher dissolved manganese concentrations and in general, increased manganese concentration equated to increased mortality.
4. Reductions in growth, as indicated by decreased body weights, were observed at significantly lower dissolved manganese concentrations than the concentrations affecting survival and thus growth was determined to be a more sensitive exposure endpoint. Dissolved manganese LOEC values based on organism body weight (not survival) were 4.41 mg/L for a hardness of 150 mg/L CaCO₃ and 8.68 mg/L for a hardness of 450 mg/L CaCO₃.
5. IC₂₅ values (interpolated concentrations at which a measurable biological response would be anticipated in 25% of organisms) for dissolved manganese were determined to be 4.67 mg/L at a hardness of 30 mg/L CaCO₃, 5.59 mg/L at a hardness of 150 mg/L CaCO₃ and 8.68 mg/L at a hardness of 450 mg/L CaCO₃.

In the discussion section of the paper, the authors stated that the current study results confirmed previous results that indicated a relationship between water hardness and manganese toxicity. Brown trout embryos were found to be tolerant of dissolved manganese at the concentrations analyzed. Although some decreases in mean time to hatching were observed, the ecological importance of this observation was not clear to the authors and hardness did not appear to have affected hatching success.

IC₂₅ concentrations were found to increase with increasing water hardness and were greater than the statistically derived NOEC values for all three water hardnesses and less than the LOEC values for water hardnesses of 150 and 450 mg/L CaCO₃. A LOEC value was not determined for a hardness of 30 mg/L CaCO₃ due to a statistically insignificant difference between the test organisms and the control groups. The authors recommend the use of IC₂₅ values over NOECs and LOECs. They based this recommendation on the fact that, by definition, the NOEC and LOEC values must be two of the test solution concentrations and the values are dependent on statistical testing which may or may not determine a biological response to be significant. Use of interpolated values such as an IC₂₅ provides a means of evaluating concentration response data based on an acceptable level of effect without the constraints of pre-set concentrations where the effect concentration is determined by the initial concentrations chosen for the test.

An equation to calculate hardness-based IC₂₅ values is provided by the study. The equation, which was determined by plotting the IC₂₅ values from the study against the natural logarithms of the water hardness values, is shown below:

IC_{25(at specified hardness)} = e^{0.2064(ln hardness) + 7.7092}
The regression analysis used to develop the equation had a positive correlation of r² = 0.88. The authors concluded that "the data presented here provide a basis upon which to estimate the potential adverse effects of chronic manganese exposure to salmonid species" and "in conjunction with acute and chronic data from other species, can be used to derive standards protective of aquatic organisms."

The study also quotes unpublished toxicity test data from which IC₂₅ values of 5.71 and 5.15 mg Mn/L were derived for C. dubia at a water hardness of 50 mg/L CaCO₃. These values that are fairly consistent with the 4.67 mg/L dissolved manganese IC₂₅ concentration the authors determined for brown trout at a hardness of 30 mg/L CaCO₃.

Davies and Brinkman (1995) studied the acute toxicity of manganese to brown trout in hard water using 96 hour LC₅₀ tests. Eggs from Colorado's Delaney Butte Reservoir and fingerlings from LaPorte Colorado's Bellevue Research Hatchery were collected for the study. Fish were placed in 92 litre aquaria filled with water sourced from a well, with water quality characteristics determined using American Public Health Association (1985) methodology. Manganese as MnSO₄·H₂O was used in the testing, with nominal concentrations of 0.0, 15.0, 27.0, 54.0, 84.4, 112.5 and 150.0 mg Mn/L chosen for analysis. The summarized materials and methods presented in the referenced document indicated that dissolved oxygen was measured using a YSI Model 58 section meter; the number and/or frequency of dissolved oxygen measurements were not identified. Manganese concentrations were measured on a daily basis using grab samples and atomic absorption spectrophotometry. Water hardness was measured in control tanks only, the authors citing interferences from manganese in the other tanks as the reason. Organism mortality was evaluated every second hour during the day (what constitutes "the day" was not defined) during the first 96 hours. Median LC₅₀ concentrations were estimated by applying probit analysis and the Spearmen-Karber method (Hamilton et. al., 1978).

The mean water quality characteristics determined from the control water sampling are summarized below:

Hardness 454 mg/L CaCO₃ (1 sample)
Alkalinity 311 mg/L CaCO₃ (7 samples)
pH 8.00 (7 samples)
Dissolved Oxygen 7.65 mg/L (7 samples)

Temperature 16.76 degrees C (7 samples)

The average fork tail length and weight of brown trout used in the study were 6 mm and 18.91 gm, respectively. Measured manganese concentrations and 96 hour acute mortality data are presented in the Table 2.5.

The median 96 hour LC₅₀ concentration estimated from the experiment was 49.9 mg Mn/L for the probit analysis and the Spearman-Karber method. The 95% confidence intervals about the mean were 43.6-57.4 mg Mn/L for probit analysis and 43.5-57.3 mg Mn/L for the Spearman-Karber method. Very good agreement between the two methods of estimating mean LC₅₀ values was noted by the researchers.

The reported hardness value of 454 mg/L CaCO₃ was based on a single measurement. However, seven alkalinity measurements resulted in a mean concentration of 311 mg/L CaCO₃, with a standard deviation of 2.60. Based on the low standard deviation value of 2.60, it is probable that water hardness values did not deviate significantly from the measured value.

Davies and Brinkman (1994) also completed acute and chronic studies of the effects of manganese on rainbow trout and brown trout in soft water. Exposed and unexposed test organisms were utilized to determine what effect pre-exposure to low levels of dissolved manganese may have on tolerance during acute and chronic exposures at higher concentrations. Eyed rainbow trout eggs were placed in "relatively soft water" (actual water hardness was not defined by the researchers) at a temperature of 6 degrees C for a four day period to acclimate. Brown trout fingerlings were similarly placed in aquaria containing 6 degrees C soft water and allowed to acclimate for two weeks. The "exposed" test organism groups were subjected to manganese (added as manganous sulphate) concentrations of 0.14 mg/L through Day 2, 0.36 mg/L through Day 5 and 0.80 mg/L for four months. Water quality conditions were the same for the "unexposed" test organisms, with the exception that no manganese was added to the water. Rainbow egg and sac fry mortality were observed daily; the researchers reported no difference in egg and sac fry mortality between the "exposed" and "unexposed" groups. For brown trout, mortality in both groups was reported as negligible. No numerical data (i.e. mortality or survival rates) were presented in the report.

Following the initial exposure period, 96 hour LC₅₀ acute and four month chronic toxicity tests were conducted on surviving organisms; exposure endpoints for the chronic tests included mortality and length/weight of survivors. For "exposed" and "unexposed" rainbow trout, separate aquaria containing water with very similar characteristics were used to conduct the testing. Seven nominal dissolved manganese concentrations, including a control solution containing no detectable manganese, were used in the experiment. For "exposed" and "unexposed" brown trout, sub-groups of twenty fish were placed in each of seven aquaria, with the adipose fin clipped from the "exposed" fish for identification. Each aquarium contained a different dissolved manganese concentration. Fish were not fed during the acute toxicity testing and dissolved manganese concentrations in the aquaria waters were confirmed daily by analyzing samples using atomic absorption spectrophotometry. Temperature, alkalinity, pH, conductivity and dissolved oxygen levels were measured using American Public Health Association methods¹⁹. Hardness was measured in the control aquaria only (the researchers cited manganese interference in the other aquaria waters).

For the chronic tests, sub-groups of twenty "exposed" and twenty "unexposed" rainbow trout were placed in separate aquaria for each of the nominal dissolved manganese concentrations evaluated. For brown trout, fish were placed in the same aquarium for each of the manganese test concentrations, with the twenty "exposed" fish having their adipose and right pelvic fins clipped, distinguishing them from the twenty "unexposed" fish. During the initial acute phase of the studies, water quality data were collected as described above. After the 96 hour period had elapsed, samples were collected on Day 7 and weekly thereafter. Hardness was again only measured in the control aquaria. Fish were fed based on weight of control fish and numbers of survivors in each aquarium. The weights of surviving fish were recorded at the end of the four month period.

The 96 hour LC₅₀ concentrations were determined using the Spearman-Karber (Hamilton et. al., 1978) method where 100% mortality occurred and by the probit method where less than 100% mortality occurred. The acute toxicity test results for rainbow trout are summarized in Table 2.6.

Note: Water hardness, pH, temperature and dissolved oxygen values are "corrected" values obtained from an addendum to Davies and Brinkman 1994, which was appended to Davies and Brinkman 1995
95% Confidence Intervals based on six to seven Mn concentrations and groups of twenty organisms exposed at each concentration

The 96 hour LC₅₀ value for exposed rainbow trout was 3.32 mg Mn/L, with a 95% confidence interval range of 2.97 to 3.72 mg Mn/L. The 96 hour LC₅₀ concentration for unexposed rainbow trout was 4.83 mg Mn/L and the 95% confidence interval range was 4.18 to 5.58 mg Mn/L. The values for the pre-exposed group were lower than those for the unexposed group, despite the smaller mean length and weight of the unexposed organisms; no explanation as to the cause or significance of these findings was provided and the authors stated that the "96 hour LC₅₀s were only slightly different in the exposed and unexposed groups."

The 96 hour LC₅₀ results for exposed and unexposed brown trout are presented in Table 2.7.

Note: Water hardness, pH, temperature and dissolved oxygen values are "corrected" values obtained from an addendum to Davies and Brinkman 1994, which was appended to Davies and Brinkman 1995
95% Confidence Intervals based on six to seven Mn concentrations and groups of twenty organisms exposed at each concentration

The 96 hour LC₅₀ concentrations for exposed and unexposed brown trout were 9.06 and 3.77 mg Mn/L, respectively, demonstrating a significant difference between the two groups. Surviving organisms mean weights and lengths in each of the groups were very similar. The 95% confidence interval range for the unexposed group was 3.17 to 4.41 mg Mn/L, which fell between the confidence interval ranges for exposed and unexposed rainbow trout.

Chronic toxicity test results were calculated from the geometric means of the effect and no effect concentration data generated from the tests. The values for exposed and unexposed rainbow trout and brown trout are presented in Table 2.8 and Table 2.9, respectively.

The exposed group chronic toxicity test value was 1.57 mg Mn/L while the unexposed group chronic value was 0.79 mg/l. The exposed group chronic value was twice that for the unexposed group and the effect/no effect ranges for the two groups also differed by a factor of about two. Water quality characteristics were very similar and lengths and weights of surviving organisms were also similar. The data suggested an increase in tolerance for the exposed group relative to the unexposed group at a water hardness value of 36.8 mg/L CaCO₃.

The test results presented in Table 2.9 for brown trout showed a pattern similar to rainbow trout. The chronic value calculated for the exposed group was 4.19 mg Mn/L while the chronic value for the unexposed group was 2.70 mg Mn/L. Water quality characteristics were identical and mean lengths and weights of surviving fish were very similar. The data suggested that the exposed group exhibited increased tolerance to manganese as a result of pre-exposure versus the unexposed group. This may indicate that organisms inhabiting surface waters may be naturally more tolerant of or acclimated to the manganese levels present in those waters. Such natural tolerance to local water conditions may not be observed in test organisms utilized in toxicity testing and should be given consideration when interpreting data generated from such tests.

Rouleau et. al. (1996) investigated the relationship of manganese uptake in brown trout tissue to pH of water. Groups of five fish (weighing 9.0 plus or minus 1.8 g) were exposed to 0.1 micrograms Mn/L for 21 days, with ⁵⁴Mn used as a tracer. Fish were exposed at pH of 7.5 plus or minus 0.2 and at pH 4.9-5.0. The 4.9-5.0 pH rose to 5.3-6.0 after 24 hours and was readjusted daily. The authors attributed this rise to ammonia excretion and determined that the average pH during the experiment was 5.3. Water in the aquaria was changed every three to four days and total Mn concentrations were determined in newly prepared water and in water before replacement; Mn concentrations were found to be constant.

At the end of the study, manganese concentrations in tissue were analyzed and the following results were presented:

The study found that manganese concentrations were similar in the rest of the body tissue (excluding viscera, brain and eyes) at both pH values. The authors concluded that "the uptake of ⁵⁴Mn(II) increased significantly at low pH but the mechanisms by which this occurred remain unclear. There was also no indication if this pattern of uptake would occur at much higher manganese concentrations; the reported experimental concentration of 0.1 micrograms/L would represent a very low concentration of total manganese relative to naturally occurring levels observed in BC surface waters.

2.4.2 Other Studies

Wepener, Van Vuren and Du Preez (1992) studied the non-lethal effects of manganese on the banded tilapia (Tilapia sparrmanii) of South Africa. The effects of a manganese chloride concentration of 4.43 mg/L (manganese concentration of 1.93 mg/L) at pH values of 5 and 7.4 in 96 hour flow-through tests were evaluated with respect to red and white blood cell counts, hemoglobin concentrations, mean corpuscular volume and hematocrit. At pH of 5, significant decreases in all the exposure endpoints parameters were observed, while at pH of 7.4, the white blood cell count, hemoglobin concentration and mean corpuscular volume decreased significantly. A slight increase in the activity of delta-aminolevulinic dehydratase was noted at both pH values. Overall, manganese was observed to cause a greater stress at a pH of 7.4 versus a pH of 5 and chronic sub-lethal concentrations were observed to be detrimental to the organisms at non-lethal levels.

Studies on the potential toxic effects of manganese on aquatic plants are not extensive. Unni, Santhakumar and Nair (1995) researched the effect of manganese on growth and physiology of rice (Oryza sativa L.). Concentrations studied ranged from 2 to 200 ppm (parts per million - assumed to be mg/L) for a period of 40 days under hydroponic conditions. Exposure endpoints included seed germination, growth retardation, and total chlorophyll, soluble sugar and protein contents. Table 2.11 summarizes the main results of the study.

Seed germination was not affected by the presence of manganese at the concentrations used in the study. The results of the study indicated a progressive reduction in chlorophyll, sugar and protein contents with increased exposure time at all three study concentrations. It was not clear why a concentration of 2 ppm (mg/L) resulted in greater reductions in several of the exposure endpoints. For example, reductions in sugar content on Day 10 and Day 40 were greatest at 2 ppm (mg/L) versus 100 or 200 ppm (mg/L). In all cases, however, reductions increased with increasing exposure time.

Wang (1986) conducted 4 day acute and 7 day sub-chronic tests on the effects of manganese on the growth of duckweed (Lemna minor). The study was conducted using tap water at a pH of 7.5 (no hardness or temperature data provided) and the exposure endpoint was growth as indicated by the number of fronds initially and at the end of the exposure period. Twenty colonies of duckweed were studied and an EC₅₀ (reduction in frond growth in 50% of test organisms vs. controls) of 31 mg/L was derived.

Stauber and Florence (Stauber and Florence, 1987) demonstrated the ameliorating effect of manganese on copper toxicity to the marine diatom Nitzschia closterium. Copper affects the organism's ability to defend against hydrogen peroxide and oxygen-free radicals, while manganese aids in the complexation of these compounds. Kaitala (1988) determined that the presence of copper ions increased the uptake of manganese in blue mussels (Mytilus edulis) and burrowing clams (Macoma baltica). Concentrations of copper (0.2 mg/L) and manganese (2 mg/L) were evaluated as individual applications and in combination along with zinc (0.4 mg/L). The author concluded that a 100% increase in manganese accumulation and a 25% increase in zinc accumulation was apparent in mussels when copper was present. For clams, manganese accumulated but zinc did not, suggesting that copper has a significant effect on the accumulation of manganese in these organisms.

Sinha et. al. 1993) studied the effect of chromium and manganese interaction on the aquatic plant Hydrilla verticalla. Manganese uptake was enhanced while chromium uptake was inhibited when the metals were combined versus uptake of the individual metals when tested separately.