Selenium in freshwater and marine water
Toxicant default guideline values for protecting aquatic ecosystems
Extracted from Section 8.3.7 ‘Detailed descriptions of chemicals’ of the ANZECC & ARMCANZ (2000) guidelines.
The default guideline values (previously known as ‘trigger values’) and associated information in this technical brief should be used in accordance with the detailed guidance provided in the Australian and New Zealand Guidelines for Fresh and Marine Water Quality.
Description of chemical
Although the major source of selenium in the environment is weathering of rocks and soils (Rosenfeld & Beath 1964), anthropogenic sources such as emissions from burning fossil fuels may also contribute selenium to natural waters (Hart 1982, Health & Welfare Canada 1980). Selenium concentrations in natural waters are usually < 500 ng/L but can be considerably elevated by waste discharges such as those from coal fired power stations or drainage from seleniferous soils (Inhat 1989). Selenium is an essential element and is incorporated into living organisms as seleno-amino acids (analogues of sulfur-containing amino-acids). Excessive concentrations of selenium can be toxic. The difference between levels that cause toxicity and those that are required for nutrients is small; it can be beneficial in food below approximately 1 mg/kg but toxic above 5 mg/kg (Chapman 1999).
Summary of factors affecting selenium toxicity
- Selenium toxicity is dependent on valency state. Selenium (IV) is generally more toxic than selenium (VI). These predominant forms in natural waters exist as oxyanions selenate and selenite.
- Selenites are readily removed from the water column but selenates (Se VI) can be readily bioaccumulated.
- Food chain uptake, leading to secondary poisoning, is more significant than water uptake. Sediments can be a significant source of selenium in fish and invertebrates. Toxic effect threshold levels for selenium in freshwater, food chain organisms and fish have been reported as 2 µg/L, 3 mg/kg and 4 mg/kg (for whole fish), respectively (Lemly 1993).
- Factors affecting selenium uptake such as pH, hardness, sulfur and phosphate consequently affect toxicity. Selenate uptake increased in presence of calcium and magnesium.
- Selenate toxicity to microalgae was inversely proportional to sulfate and phosphate concentrations.
- Selenate uptake in algae was independent of pH between 5 and 9 but selenite uptake increased at low pH.
- Mercury and copper both ameliorate selenium toxicity.
- Binding of selenium to particulates does not necessarily reduce selenium bioavailability from food.
- Due to the transport and bioaccumulation of selenium, and the changes in form, whole hydrological units should be examined in any site-specific assessment. Lemly (1999) provides a framework for designating hydrological units. Lemly (1998) suggested that criteria should be adjusted by a fixed amount to account for the degree of biological hazard from bioaccumulation.
A variety of analytical methods are available for determining the speciation of selenium in water. These include techniques, such as selective hydride generation, chromatography and ion exchange (Howard 1989). Geochemical speciation modelling is of limited use, as the concentrations of selenium (VI) and selenium (IV) are rarely in true thermodynamic equilibrium because of biologically-mediated reactions. In addition, speciation modelling cannot predict the concentration of organoselenium species formed by the decomposition of organic matter. The current analytical practical quantitation limit (PQL) for total selenium is 0.03 µg/L in both fresh and marine water (NSW EPA 2000). The different valency states of selenium can be difficult to separate analytically. Hence, the suggested approach is to commence with analysis of total selenium and only consider proceeding to more complex analysis if the total selenium exceeds the trigger value.
Bioassays are typically used to determine metal-organism interactions. These can be used in conjunction with the measured speciation of selenium to define bioavailable selenium species.
Factors that affect the uptake and toxicity of selenium
Selenium chemistry in natural waters is very complex and is analogous to that of sulfur. Three oxidation states may occur in the water column, selenium (VI), selenium (IV) and selenium (II). Elemental selenium may also occur in reducing sediments (Maier & Knight 1994).
Selenium (IV) and selenium (VI) exist as the oxyanions selenite and selenate, respectively and do not form complexes with organic matter or inorganic ligands (Fergusson 1990). Selenites can form stable complexes with a number of cations, such as iron and aluminium, however, these are relatively insoluble and are readily removed from the water column (CCREM 1987). Selenites are reduced to elemental selenium under acidic and reducing conditions, the element having low solubility, which also acts to remove selenium from the water column. In alkaline and oxidising conditions, the formation of selenate is favoured (CCREM 1987) Selenate is not readily complexed by cations, is soluble, and may be easily accumulated by biota.
In seawater, the speciation of selenium is depth-dependent (Cutter & Bruland 1984). In surface waters, selenium is predominantly associated with organic matter. At greater depth, selenium (VI) predominates, and at still greater depths, some reduction occurs, such that both selenium (IV) and selenium (VI) are important (Cutter & Bruland 1984).
Selenium is an essential element and is incorporated into living organisms as seleno-amino acids (analogues of sulfur-containing amino-acids). Degradation of organic matter leads to the presence of significant quantities of organic-selenides in waters (Cutter & Bruland 1984). Between 30 and 60% of total dissolved selenium may be organically bound (Maier & Knight 1994). Degradation of organic selenides results in the formation of selenite, and ultimately, the regeneration of selenate (Cutter & Bruland 1984).
In general, the toxicity of selenium species follows the order: selenomethionine > selenite > selenate.
Natural freshwater plankton communities were found to accumulate selenite 4 to 5 times faster than selenate over a 24-hour period (Riedel & Sanders 1996). Further, organisms fed on a diet containing selenium where found to accumulate selenium (IV) at a significantly greater rate than selenium (VI) (Malchow et al. 1995). Selenite has also been found to be more toxic than selenate to both fresh and marine organisms (Hamilton 1995). For example, the 72-hour EC50 for selenium (IV) with the marine diatom Nitzschia closterium was 1 mg/L, whereas for selenium (VI) it was > 2 mg/L (Florence & Stauber 1991).
Most toxicity values are based on uptake from the water column but in natural populations this is insignificant compared to uptake through the food chain. Concentrations of selenium can build up to toxic levels in higher organisms even when selenium concentrations in the water column are low. Game fish populations suffered reproductive failure after bioaccumulation of selenium in lakes containing < 10 µg/L. Toxic effect threshold levels for selenium in freshwater, food chain organisms and fish have been reported as 2 µg/L, 3 mg/kg and 4 mg/kg (for whole fish), respectively (Lemly 1993). Sediments were found to be a significant source of selenium contamination in benthic infauna and in fish predators in Lake Macquarie, NSW (Peters et al. 1999). Selenium in water can be bioconcentrated by between 100 and 30,000 times in food organisms eaten by fish and wildlife (Lemly 1999), sometimes causing reproductive failure without affecting parents. Chapman (1999) outlines many of the issues to consider in site-specific assessments of selenium bioaccumulation and risk assessment.
Jarvinen and Ankley (1999) report data on tissue residues and effects for various forms of selenium for around 10 freshwater species. It is not possible to summarise the data here but readers are referred to that publication for more information.
Uptake of inorganic selenium is dependent on water chemistry, including pH, water hardness, phosphate and sulfate concentrations and these affect toxicity. Uptake of selenate in the green alga Chlamydomonas reinhardtii was independent of pH over the range 5 to 9, because selenate is completely dissociated over this pH range. Selenite uptake, however, increased markedly at low pH, corresponding to increasing dominance of the uncharged H2SeO3 species, which can rapidly penetrate the cell. Selenite uptake also increased when phosphate concentrations were low. This suggests that phosphate-limited algae accumulate more selenite, and phosphate concentrations should therefore be taken into account when modelling the fate of selenite in riverine systems (Riedel & Sanders 1996). Selenate uptake increased in the presence of calcium, magnesium and ammonium, and decreased at high sulfate concentrations particularly in soft waters. Additionally, selenate toxicity to microalgae and crustaceans was found to be inversely proportional to sulfate concentration (Williams et al. 1994, Ogle & Knight 1996).
Some studies have found that the combination of selenium and mercury, as well as selenium and copper, are less toxic than the individual metals. Both mercury and copper individually react with sulfhydryl groups, disrupting enzyme function and cell division. However, in the marine alga Dunaliella, mercury (II) and SeO3¯ react together to form a complex that cannot react with sulfhydryl groups, thereby ameliorating toxicity (Gotsis 1982).
Vaughan (1996) summarised recent data on the toxicity of selenium to freshwater and marine organisms. There is a wide range of sensitivity to selenium amongst freshwater biota, with the alga Chlorella pyrenoidosa being the most sensitive species (96-hour LC50 of 800 µg/L for both selenate and selenite). Amongst marine species, crustaceans have the widest range of acute toxicity values with 96-hour LC50 values from 1 mg/L for early life stages of the crab Cancer magister to 600 mg/L for adult mysid shrimp. Selenium toxicity to phytoplankton, molluscs and fish ranges from 0.25 to 10 mg/L. These toxicity values are based on uptake from the water column and not through bioconcentration in the food chain.
Acute toxicity data—Se (IV)
USEPA (1987c) compiled acute data for Se (IV) from freshwater fish and invertebrates species, with values ranging from 340 µg/L for an amphipod to 203,000 µg/L for a leech. Canton (1999) reported a re-evaluation of acute data for Se (IV), ranging from 550 µg/L for Daphnia magna to 48,200 µg/L for a midge Chironomus decorus.Canton (1999) calculated an acute criterion of 220 µg/L for selenite from these data by the US Environmental Protection Agency (EPA) method (Stephan et al. 1985). Hamilton and Lemly (1999) argued for a chronic water criterion (USEPA method) for Se of 2 µg/L to protect from bioaccumulation. Acute toxicity occurs predominantly from water exposure and chronic toxicity from food (Chapman 1999).
The US EPA (1987c) compiled acute toxicity data for Se (IV) for sixteen saltwater animals, (eight invertebrates and eight fish) ranging from 600 µg/L to 17,300 µg/L. Chronic values for mysid and sheepshead minnow ranged from 222 µg/L to 675 µg/L, resulting in acute-chronic ratios of 7 and 11 (USEPA 1987c).
Acute toxicity data—Se (VI)
Canton (1999) reported a re-evaluation of acute data for Se (VI), ranging from 750 µg/L for Daphnia magna to 115,000 for chinook salmon Oncorhynchus tschawytscha. The acute US criterion of 220 µg/L for selenite was considered to be adequate for selenate exposure (Canton 1999).
Freshwater guideline—Se (total)
Much of the screened freshwater chronic data included various forms of selenium, predominantly selenium (IV) but containing some Se (VI) and inorganic selenium compounds. Data were available for total selenium on four taxonomic groups (12 data points), as follows (the pH range was 7.30 to 7.98):
Fish: one species, P. promelas 14-day LC50, NOEC 600 µg/L. Acute 96-h LC50 for Oncorhynchus tshawytscha, 85,500 µg/L.
Crustacean: two species, 14-day NOEC, 14 to 86 µg/L (from LC50); 21-day NOEC, growth, 85 µg/L.
Insect: one species, Chironomus riparius, 30-day NOEC emergence, 252 to 303 µg/L.
Algae: one species, Selenastrum capricornutum, 3 to 6 day NOEC, growth of 13,000 to 19,800 µg/L (from EC50).
A freshwater high reliability trigger value of 11 µg/L was calculated for Se (total) using the statistical distribution method at 95% protection. This has not specifically considered bioaccumulation. The 99% protection level is 5 µg/L. The 99% protection level is recommended for slightly to moderately disturbed systems if there are no data to allow for adjustment for bioaccumulation at the specific site (Section 18.104.22.168 of the ANZECC & ARMCANZ 2000 guidelines).
Lemly (1998) suggested that criteria should be adjusted by a fixed amount to account for the degree of biological hazard from bioaccumulation.
Marine guideline—Se (Total)
A total of 34 screened data points (acute only) were available for 13 species for selenium (total) in marine waters but these were from only three taxonomic groups, as follows (the pH range was 6.8 to 7.93, but only 17 of 43 data points reported pH):
Fish: four species, 48 to 96-hour LC50, 1550 µg/L (Morone saxitilis) to 180,000 µg/L (O. tshawytscha).
Crustaceans: six species, 48 to 96-hour LC50, 738 µg/L (Cancer magister) to 82,000 µg/L (Artemia salina).
Molluscs: three species, 86-hour LC50, 255 µg/L (Argopecten irradians) to 2000 µg/L.
Algae: one species, 72-hour EC50, 1000 µg/L (Nitzschia closterium, Australian data). Although this did not survive the screening process it gives an indication of algal toxicity and only affects the size of the factor used.
A marine low reliability trigger value of 3 µg/L was calculated for Se (total) using an assessment factor (AF) of 100. This has not specifically considered bioaccumulation.
Freshwater guideline—Se (IV)
Screened acute toxicity data for selenium (IV) were available for only two taxonomic groups, as follows (it was not possible to screen the data in Canton, 1999, in this current revision): the pH range was 6.80-7.93.
Fish: six species, 48 to 120-hour LC50, 2250 µg/L (Colisa fasciata) to 46,500 µg/L (Carassius auratus).
Insect: one species, Tanytarsus dissimilis, 48-hour LC50, 42,500 µg/L. This was outside the pH range.
A freshwater low reliability trigger value of 11 µg/L was calculated for Se (IV) using an AF of 200 (data were chronic). This has not specifically considered bioaccumulation. In most cases it would be preferable to use the selenium (total) trigger value (same figure), given that it is moderate reliability.
Marine guideline—Se (IV)
The only data for Se (IV) in marine systems were from the US EPA. The only figure for Se (VI) was a 72-hour EC50 figure > 2 mg/L for Nitzschia closterium (Florence et al. 1994).
An environmental concern level (ECL) of 6 µg/L for Se (IV) could be derived from the US EPA data. However, it would be preferable to use the Se (total) trigger value (3 µg/L) as an indicative interim working level until more marine data can be obtained for Se (IV). No figure could be derived for Se (VI) in marine systems.
ANZECC & ARMCANZ 2000. Australian and New Zealand Guidelines for Fresh and Marine Water Quality, Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand, Canberra.
Canton S P 1999. Acute aquatic life criteria for selenium. Environmental Toxicology and Chemistry 18, 1425–1432.
CCREM 1987. Canadian water quality guidelines. Canadian Council of Resource and Environment Ministers, Ontario.
Chapman PM 1999. Selenium — A potential time bomb or just another contaminant? Human and Ecological Risk Assessment 5, 1123–1138.
Cutter GA & Bruland KW 1984. The marine biogeochemistry of selenium: A re-evaluation. Limnology and Oceanography 29, 1179–1192.
Fergusson JE 1990. The heavy elements: Chemistry, environmental impact and health effects. Pergamon Press, Oxford.
Florence TM & Stauber JL 1991. The toxicity of heavy metals to aquatic organisms. In Proceedings of IIR Conference on environmental monitoring, Sydney [9 pp].
Florence TM, Stauber JL & Ahsanullah M 1994. Toxicity of nickel ores to marine organisms. Science of the Total Environment 148, 139–155.
Gotsis O 1982. Combined effects of selenium/mercury and selenium/copper on the cell population of the alga Dunaliella minuta. Marine Biology 71, 217–222.
Hamilton SJ 1995. Hazard assessment of inorganics to three endangered fish in the Green River, Utah. Ecotoxicological and Environmental Safety 30, 134–142.
Hamilton SJ & Lemly AD 1999. Water-sediment controversy in setting environmental standards for selenium. Ecotoxicology and Environmental Safety 44, 227–235.
Hart BT 1982. Australian water quality criteria for heavy metals. Australian Water Resources Council, technical paper 77, Australian Government Publishing Service, Canberra.
Health and Welfare Canada 1980. Selenium. In Guidelines for Canadian drinking water quality 1978: Supporting documentation. Supply & Services, Canada, Hull, 541-554.
Howard AG 1989. Selenium combination techniques. In Environmental analysis using chromatography interfaced with atomic spectrometry, eds RM Harrison & S Rapsomanakis, Ellis Horwood, Chichester, 318-342.
Inhat M 1989. Occurrence and distribution of selenium. CRC Press, Boca Raton, FL.
Jarvinen A W & Ankley G T 1999. Linkage of effects to tissue residues: Development of a comprehensive database for aquatic organisms exposed to inorganic and organic chemicals. SETAC Technical Publication Series, SETAC Press, Pensacola FL.
Lemly AD 1999. Selenium transport and bioaccumulation in aquatic ecosystems: A proposal for water quality criteria based on hydrological units. Ecotoxicology and Environmental Safety 42, 150–156.
Lemly AD 1998. A position paper on selenium in ecotoxicology: A procedure for deriving site-specific water quality criteria. Ecotoxicology and Environmental Safety 39, 1–9.
Lemly AD 1993. Guidelines for evaluating selenium data from aquatic monitoring and assessment studies. Environmental Monitoring and Assessment 28, 83–100.
Maier KJ & Knight AW 1994. Ecotoxicology of selenium in freshwater systems. Reviews of Environmental Contamination and Toxicology 134, 31–48.
Malchow DE, Knight AW & Maier KJ 1995. Bioaccumulation and toxicity of selenium in Chironomus decorus larvae fed a diet of seleniferous Selenastrum capricornutum. Archives of Environmental Contamination and Toxicology 29, 104–109.
NSW EPA 2000. Analytical Chemistry Section, Table of Trigger Values 20 March 2000, LD33/11, Lidcombe, NSW.
Ogle RS & Knight AW 1996. Selenium bioaccumulation in aquatic ecosystems: I. Effects of sulfate on the uptake and toxicity of selenate in Daphnia magna. Archives of Environmental Contamination and Toxicology 30, 274–279.
Peters GM, Maher WA, Krikowa F, Roach AC, Jeswani HK, Barford JP, Gomes VG & Reible DD 1999. Selenium in sediments, pore waters and benthic infauna of Lake Macquarie, New South Wales, Australia. Marine Environmental Research 47, 491–508.
Riedel GF & Sanders JG 1996. The influence of pH and media composition on the uptake of inorganic selenium by Chlamydomonas reinhardtii. Environmental Toxicology and Chemistry 15, 1577–1583.
Rosenfeld I & Beath OA 1964. Selenium: Geobotany, biochemistry, toxicology and nutrition. Academic Press, New York.
Stephan CE, Mount DI, Hansen DJ, Gentile HJ, Chapman GA & Brungs WA 1985. Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. US Environmental Protection Agency, Washington DC. PB85-227049, 9811.
USEPA 1987c. Ambient aquatic life water quality for selenium. US Environmental Protection Agency. Environmental Research Laboratories Duluth, Minnesota. Narragansett, Rhode Island.
Vaughan GT 1996. Literature review of the chemical speciation and toxicology of trace elements in neutralised supernatant liquor. In Release pond discharge strategy, eds GT Vaughan & RP Lim, CSIRO Investigation Report CET/IR467, Appendix A.
Williams MJ, Ogle RS, Knight AW & Burau RG 1994. Effects of sulfate on selenate uptake and toxicity in the green alga Selenastrum capricornutum. Archives of Environmental Contamination and Toxicology 17, 449–453.