Polycyclic aromatic compounds (PACs) in the Canadian environment: The challenges of ecological risk assessments.

Ecological risk assessments (ERAs) of polycyclic aromatic compounds (PACs), as single congeners or in mixtures, present technical challenges that raise concerns about their accuracy and validity for Canadian environments. Of more than 100,000 possible PAC structures, the toxicity of fewer than 1% have been tested as individual compounds, limiting the assessment of complex mixtures. Because of the diversity in modes of PAC action, the additivity of mixtures cannot be assumed, and mixture compositions change rapidly with weathering. In vertebrates, PACs are rapidly oxygenated by cytochrome P450 enzymes, often to metabolites that are more toxic than the parent compound. The ability to predict the ecological fate, distribution and effects of PACs is limited by toxicity data derived from tests of a few responses with a limited array of test species, under optimal laboratory conditions. Although several models are available to predict PAC toxicity and rank species sensitivity, they were developed with data biased by test methods, and the reported toxicities of many PACs exceed their solubility limits. As a result, Canadian Environmental Quality Guidelines for a few individual PACs provide little support for ERAs of complex mixtures in emissions and at contaminated sites. These issues are illustrated by reviews of three case studies of PAC-contaminated sites relevant to Canadian ecosystems. Interactions among ecosystem characteristics, the behaviour, fate and distribution of PACs, and non-chemical stresses on PAC-exposed species prevented clear associations between cause and effect. The uncertainties of ERAs can only be reduced by estimating the toxicity of a wider array of PACs to species typical of Canada's diverse geography and environmental conditions. Improvements are needed to models that predict toxicity, and more field studies of contaminated sites in Canada are needed to understand the ecological effects of PAC mixtures.

Polycyclic aromatic compounds (PACs) in the Canadian environment: Exposure and effects on wildlife.

Polycyclic aromatic compounds (PACs) are ubiquitous in the environment. Wildlife (including fish) are chronically exposed to PACs through air, water, sediment, soil, and/or dietary routes. Exposures are highest near industrial or urban sites, such as aluminum smelters and oil sands mines, or near natural sources such as forest fires. This review assesses the exposure and toxicity of PACs to wildlife, with a focus on the Canadian environment. Most published field studies measured PAC concentrations in tissues of invertebrates, fish, and birds, with fewer studies of amphibians and mammals. In general, PAC concentrations measured in Canadian wildlife tissues were under the benzo[a]pyrene (BaP) guideline for human consumption. Health effects of PAC exposure include embryotoxicity, deformities, cardiotoxicity, DNA damage, changes to DNA methylation, oxidative stress, endocrine disruption, and impaired reproduction. Much of the toxicity of PACs can be attributed to their bioavailability, and the extent to which certain PACs are transformed into more toxic metabolites by cytochrome P450 enzymes. As most mechanistic studies are limited to individual polycyclic aromatic hydrocarbons (PAHs), particularly BaP, research on other PACs and PAC-containing complex mixtures is required to understand the environmental significance of PAC exposure and toxicity. Additional work on responses to PACs in amphibians, reptiles, and semi-aquatic mammals, and development of molecular markers for early detection of biological responses to PACs would provide a stronger biological and ecological justification for regulating PAC emissions to protect Canadian wildlife.

Morphological and molecular effects of two diluted bitumens on developing fathead minnow (Pimephales promelas).

Canada has experienced a significant increase in the transport of diluted bitumen (dilbit), a predominant oil sands product that combines bitumen with diluents derived from oil-gas condensates and other proprietary compounds. The toxicity of dilbit to fish embryos, which are immobile and thus at a high risk of exposure to oil in the event of a spill, remains largely unknown for most species. This study assessed the toxicity of water accommodated fractions (WAF) and chemically enhanced water accommodated fractions (CEWAF) of two winter dilbit blends, Access Western Blend (AWB) and Cold Lake Blend (CLB), to fathead minnow (Pimephales promelas) embryos. The TPH-F EC50s for malformations were 834 and 1058 µg/L for AWB WAF and CEWAF, respectively, and 500 and 715 µg/L for CLB WAF and CEWAF, respectively. Levels of cyp1a mRNA increased up to 46- and 69-fold, respectively, reflecting increasing exposure to polycyclic aromatic compounds (PACs) in AWB and CLB. Similarly, levels of gst mRNA were elevated up to 3.8-fold and 2.7-fold with increasing total concentrations of PACs in AWB and CLB, respectively. However, there were no significant changes in mRNA levels of p53, sod, cat, and gsr. These results suggest that the expression of cyp1a and gst may serve as biomarkers for dilbit exposure in fathead minnow, furthering our understanding of dilbit-responsive indicators of toxicity in fish species native to North America. This study is important as it utilizes the same exposure methodology to examine the toxicity of two commonly used Canadian dilbits, facilitating comparison of dilbit toxicity.

Diluted bitumen causes deformities and molecular responses indicative of oxidative stress in Japanese medaka embryos.

This study characterized the toxicity and physiological effects of unweathered diluted bitumen (Access Western Blend dilbit; AWB) to fish. Embryos of Japanese medaka (Oryzias latipes) were exposed for 17 days to dilutions of physically-dispersed (water accommodated fraction; WAF) and chemically-dispersed (chemically-enhanced WAF; CEWAF) dilbit. AWB dilbit exposure was not lethal to medaka, but resulted in a high prevalence of blue sac disease (BSD), impaired development, and abnormal or un-inflated swim bladders at hatch. Physiological effects were indicated by the relative mRNA levels of key genes associated with, among others, cell cycling and the response to mutations (p53), xenobiotic metabolism (ahr, arnt2), phase I (cyp1a) and II processes associated with oxidative stress (cat, g6pdh, hsp70, gst, gpx, gsr, nfe2, and sod). AWB dilbit treatment increased p53 and cyp1a transcript levels (1.5-fold and >15-fold, respectively), with significant, but less pronounced changes in indicators of oxidative stress and metabolism. The exposure-related changes in embryotoxicity and mRNA synthesis were consistent with metabolism of polycyclic aromatic hydrocarbons (PAHs) to reactive and toxic metabolites. Medaka embryos responded similarly to WAF and CEWAF treatments, but CEWAF was about 100 times more efficient in delivering toxic concentrations of PAHs. The toxicity of chemically-dispersed nujol, a non-toxic mineral oil used as an experimental control, suggested that a portion of the observed effects of AWB could be attributed to excess dispersant in solution. This first study of the physiological effects of dilbit toxicity to fish embryos provides a baseline to compare toxicity between dilbit and conventional crude oils, and the groundwork for the development of molecular biomarkers of the sensitivity and level of risk of native Canadian fish species to dilbit exposure.

Field studies using fish biomarker--how many fish are enough?

Significant efforts are invested in field studies using fish, and it is important to optimize the number of organisms collected to evaluate the possible impacts of contamination. This paper provides ecotoxicologists with the approximate numbers of fish needed to identify statistically significant differences among samples using physiological indices and biochemical markers of fish health. The numbers of fish to collect are reported for ethoxyresorufin-o-deethylase (EROD); ethoxycoumarin-o-deethylase (ECOD), serum sorbitol dehydrogenase (sSDH), stress proteins, gonadosomatic index, liver somatic index, condition factor, and biliary metabolites of polycyclic aromatic hydrocarbons. The number of fish to collect was as few as four for ECOD activity (with a power of 80%), but as high as 106 for CF (with a power of 95%). Achieving statistical significance between sites does not help in the interpretation of the biological significance of a parameter, but well-planned field samplings will maximize the chances of correctly identifying areas of concern.

Biological responses to PCB exposure in shorthorn sculpin from Saglek Bay, Labrador.

A local source of polychlorinated biphenyls (PCBs) in Saglek Bay, Labrador, has contaminated marine sediments and the coastal food web. As part of a larger assessment of ecological risks in the Bay, we evaluated biological responses to PCB concentrations in a northern fish species, the shorthorn sculpin (Myoxocephalus scorpius). Biological endpoints, including ethoxyresorufin-O-deethylase (EROD) activity in liver tissue, fish body condition, lipid content, and relative liver mass were examined in 35 sculpin collected during August-September 1999. Across a wide range of PCB concentrations (5.1-6920 ng/g wet weight (ww) in whole fish excluding liver), sculpin showed significant EROD induction (as much as 25-fold in the most exposed group). Responses varied directly with PCB concentrations but there was also an apparent threshold for induction at about 50 ng/g ww (whole fish excluding liver). A strong relationship between sculpin PCB concentrations and the concentrations of PCBs in the marine sediments of Saglek Bay suggests that concentrations above this threshold can arise from very low concentrations in sediments (2.3 ng/g dry weight). Other biological endpoints did not show significant responses to PCB concentrations, nor were they related to the observed EROD activity. Although PCDF compounds were present in trace amounts (primarily 2,3,4,7,8-PnCDF), mono-ortho and non-ortho substituted (coplanar) PCBs appeared to contribute the majority of the total dioxin toxic equivalent (TEQ) concentrations. Overall, the results indicate that biological responses occur in shorthorn sculpin with relatively low PCB concentrations (approximately 50 ng/g), which are not unrealistic for even mildly contaminated areas in northern Canada.

In vitro and in vivo comparisons of fish-specific CYP1A induction relative potency factors for selected polycyclic aromatic hydrocarbons.

Induction of cytochrome P450 (CYP1A), as measured by liver ethoxyresorufin-O-deethylase (EROD) activity in juvenile rainbow trout (Oncorhynchus mykiss), was used to derive relative potency factors (RPFs) for several polycyclic aromatic hydrocarbons (PAHs), chosen for their induction potency in a rainbow trout liver cell line (RTL-W1). Potency for causing induction was estimated as the median effective concentration (EC50) from exposure-response curves. With the exception of phenanthrene, all PAHs tested induced EROD activity in juvenile trout, ranked as: benzo[k]fluoranthene>benzo[b]fluoranthene>benzo[b]fluorene>beta-napthoflavone>retene (7-isopropyl-1-methylphenanthrene). When induction potency was expressed relative to benzo[k]fluoranthene, RPFs ranged from 0.02 to 1, and the rank order in vivo was identical to the rank order with RTL-W1-derived values. The additivity of PAHs in mixtures in RTL-W1 cells was compared to whole-fish results from a previous study. EROD induction showed additive interactions for PAHs with exposure-response curves of similar slopes. This study demonstrates that assays of CYP1A induction using rainbow trout liver cells in culture would be a convenient substitute for assays with whole fish as part of testing programs for risk assessment of PAHs.

MFO induction in fish by filtrates from chlorine dioxide bleaching of wood pulp.

Spent bleaching liquors from pulp bleached with chlorine dioxide were assessed for their potency to induce hepatic mixed function oxygenase enzymes (MFO) in rainbow trout, as indicated by activity of ethoxyresorufin-o-deethyase (EROD). Filtrates were collected from two kraft mills in Central Canada to assess the potency of filtrates from hardwood and softwood bleaching. All mill-scale bleaching filtrates induced MFO activity, and filtrates from softwood pulp bleaching appeared more potent than filtrates from hardwood bleaching. Filtrates from the final bleaching stage were most potent, and filtrates from the first stage were the least potent. In laboratory bench-scale bleaching experiments, pulp from softwood and hardwood kraft mills in Eastern Canada was bleached via an industry-standard 5-stage chlorine dioxide bleaching sequence. The filtrates were collected and used in fish bioassays to assess EROD-inducing potency. Potency of bench-scale filtrates varied depending on wood furnish (i.e. softwood vs. hardwood) and the bleaching stage, with all bench-scale filtrates being much weaker EROD inducers than mill-scale filtrates. Recycled paper mill washwater is a possible source of compounds causing increased potency of the mill-scale filtrates.

Aquatic pollution-induced immunotoxicity in wildlife species.

The potential for chemicals to adversely affect human immunologic health has traditionally been evaluated in rodents, under laboratory conditions. These laboratory studies have generated valuable hazard identification and immunotoxicologic mechanism data; however, genetically diverse populations exposed in the wild may better reflect both human exposure conditions and may provide insight into potential immunotoxic effects in humans. In addition, comparative studies of species occupying reference and impacted sites provide important information on the effects of environmental pollution on the immunologic health of wildlife populations. In this symposium overview, Peter Hodson describes physiological changes in fish collected above or below the outflows of paper mills discharging effluent from the bleaching process (BKME). Effects attributable to BKME were identified, as were physiological changes attributable to other environmental factors. In this context, he discussed the problems of identifying true cause and effect relationships in field studies. Mohamed Faisal described changes in immune function of fish collected from areas with high levels of polyaromatic hydrocarbon contamination. His studies identified a contaminant-related decreases in the ability of anterior kidney leukocytes to bind to and kill tumor cell line targets, as well as changes in lymphocyte proliferation in response to mitogens. Altered proliferative responses of fish from the contaminated site were partially reversed by maintaining fish in water from the reference site. Peter Ross described studies in which harbor seals were fed herring obtained from relatively clean (Atlantic Ocean) and contaminated (Baltic Sea) waters. Decreased natural killer cell activity and lymphoproliferative responses to T and B cell mitogens, as well as depressed antibody and delayed hypersensitivity responses to injected antigens, were identified in seals fed contaminated herring. In laboratory studies, it was determined that rats fed freeze-dried Baltic Sea herring had higher virus titers after challenge with rat cytomegalovirus (RCMV) than rats fed Atlantic Ocean herring; perinatal exposure of rats to oil extracted from Baltic herring also reduced the response to challenge with RCMV. Keith Grassman reported an association between exposure to polyhalogenated aryl hydrocarbons and decreased T cell immunity in the offspring of fish-eating birds (herring gulls and Capsian terns) at highly contaminated sites in the Great Lakes. The greatest suppression of skin test responses to phytohemagglutinin injection (an indicator of T cell immunity) was consistently found at sites with the highest contaminant concentrations. Judith Zelikoff addressed the applicability of immunotoxicity studies developed in laboratory-reared fish for detecting altered immune function in wild populations. She presented data from studies done in her laboratory with environmentally relevant concentrations of metals as examples. Although the necessity of proceeding with caution when extrapolating across species was emphasized, she concluded that published data, and results presented by the other Symposium participants, demonstrate that assays similar to those developed for use in laboratory rodents may be useful for detecting immune system defects in wildlife species directly exposed to toxicants present in the environment.

Measuring the potency of pulp mill effluents for induction of hepatic mixed-function oxygenase activity in fish.

A bioassay protocol was optimized for measuring the potency of effluents or waterborne chemicals for inducing mixed-function oxygenase (MFO) activity of rainbow trout (Oncorhynchus mykiss). Measurements of ethoxyresorufin O-deethylase (EROD) can be made with an established endpoint assay using large volumes of reagents and tissue. However, a new kinetic microplate assay offers significant savings in time, reagents, and sample volumes. Data are distributed lognormally and must be log transformed before statistical analyses. EROD activity increases with exposure time to pulp mill effluent, and a 4-d exposure provides a near-maximal response. Optimum fish size conforms to standard practices in fish toxicology; loading rates should not exceed 1 g of fish per liter of test solution per day. Feed should be withheld from test fish 48 h before testing to reduce the variance of measured activity, and anaesthetizing fish with MS-222 does not affect their response to MFO inducers. Pulp mill effluents do not lose their potency during 2-3 wk of exposure at temperatures ranging from -20 to 13 degrees C, whether stored in plastic or glass. Steel containers were associated with slight losses in potency. Bioassays of MFO induction in fish exposed to liquid effluents are practical and conform to standard practice for testing the lethality of waterbone chemicals. The results are sufficiently precise that differences among means based on live fish per treatment can be discriminated statistically when activity changes by threefold or more.

Rainbow trout hepatic mixed-function oxygenase induction by polychlorinated dibenzo-p-dioxins (PCDDs) as a function of time and tissue concentration.

Rainbow trout, dosed orally with 0.060-84 micrograms/kg of [3H]-2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), [14C]-1,2,3,4,7,8-hexachlorodibenzo-p-dioxin (HxCDD), or [14C]-1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD) showed dose-dependent increases in hepatic ethoxyresorufin O-deethylase (EROD) activity, up to 250-fold, after 2-16 d. Induction of EROD activity was a sensitive and rapid indicator of exposure to PCDDs. The effects of time after exposure were not dramatic, but generally EROD activity after 2 d was lower than EROD activity after 16 d. Slopes of dose-response curves relating EROD activity to hepatic concentrations of PCDDs significantly decreased and intercepts significantly increased with increasing time of exposure. When fish were grouped by oral doses, only the low doses of TCDD and HpCDD exhibited time-dependent EROD induction, with significantly greater activity 16 d after dosing compared to 2 d. In most cases, hepatic and muscle concentrations of PCDDs did not significantly change over time. Concentrations of PCDDs in liver and muscle accounted for up to 7% of the orally administered dose. Hepatic levels of PCDDs ranged from 20-100 pg/g at the lowest doses to about 1000 pg TCDD/g, 2000 pg HxCDD/g, or 40,000 pg HpCDD/g at the highest doses. Of the PCDD in the liver, approximately one-third to one-half was associated with the postmitochondrial supernatant (PMS). PCDD concentrations in muscle did not span as wide a range of concentrations as did PCDD in liver; fish given low doses had 10 pg/g in muscle whereas fish given high doses had several hundred picograms per gram in muscle for all three PCDDs.

Polycyclic aromatic compounds in cod (Gadus morhua) from the Northwest Atlantic and St. Lawrence Estuary.

There is limited information available on levels of polycyclic aromatic hydrocarbons (PAH) in major fish populations including populations from the Northwest Atlantic. The cod (Gadus morhua) stocks off eastern Canada form the basis of one of the world's most important fisheries. Muscle tissues of cod collected from Northwest Atlantic Fisheries Organisation (NAFO) management Divisions 2J, 3K and 3Ps off Newfoundland and Labrador, as well as three contaminated sites in the Gulf of St. Lawrence were analyzed for total polycyclic aromatic (PAC) by fluorimetry. Concentrations were determined in terms of crude oil and chrysene equivalents in line with recommendations of the International Oceanographic Commission. Overall, relatively low concentrations of PAC were found, the highest values generally being below 0.1 microgram/g (dry weight) in terms of chrysene equivalents. Similarly, only trace levels of a few PAH were detected in composite samples analyzed by gas chromatography-mass spectrometry (GC-MS). It is of interest that the highest levels of PAC were found in fish from NAFO Division 3K, while concentrations in fish from the two contiguous zones, 2J and 3Ps, as well as the Gulf of St. Lawrence, were similar. Division 3K is a major fishing zone and it is important to determine if trawler fleets are important sources of hydrocarbons derived from fossil fuels, in this and similar fishing areas of the world's oceans.

Indicators of ecosystem health at the species level and the example of selenium effects on fish.

Chemical monitoring of aquatic ecosystems describes the chemical exposures of aquatic biota and measures the success of pollution control. However, meeting water quality criteria cannot assure that aquatic biota are protected from the effects of unexpected chemicals, mixtures and interactions between toxicity and environmental stressors.Biological monitoring is an obvious solution since aquatic biota integrate spatial and temporal variations in exposure to many simultaneous stressors. Top predators, typical of specific ecosystems (e.g. lake trout in cold water oligotrophic lakes) indicate whether environmental criteria have been met. The presence of naturally reproducing, self-sustaining and productive stocks of edible fish demonstrates a high quality environment. If these conditions are not met, there is a clear sign of environmental degradation. Specific changes in population structure and performance may also diagnose which life stage is affected and the nature of the stressor.Unfortunately, environmental managers cannot rely solely on populations, communities or ecosystems to indicate chmical effects. The lag between identifying a problem and finding a cause may destroy the resource that we wish to protect, particularly where chemicals are persistent.A solution to this dilemma is the measurement of primary or secondary responses of individual organisms to chemical exposure. Since toxicity at any level of organization must start with a reaction between a chemical and a biological substrate, these responses are the most sensitive and earliest sign of chemical exposure and effect.Application of this idea requires research on molecular mechanisms of chemical toxicity in aquatic biota and adaptation of existing mammalian diagnostic tools. Since relevance of biochemical responses to populations and ecosystems is not obvious, there is a need to study the links between chemical exposure and responses of individuals, populations and ecosystems.The recognition of chemical problems and cause-effect relationships requires the integration of chemical and biological monitoring, using the principles of epidemiology to test the strength of relationships and to identify specific research needs. The contamination of a reservoir with selenium and impacts on fish populations provide an excellent example of this approach.

Accumulation of waterborne selenium by rainbow trout (Salmo gairdneri), eggs, fry and juveniles.

The effect of waterborne selenite levels on selenium accumulated by different developmental stages of rainbow trout (Salmo gairdneri) was studied using(75)SeO 3 (=) as a tracer. All stages readily accumulated selenium at both high and low concentrations, but the rate of accumulation increased as the trout developed from the egg to the juvenile feeding stage. The low rate of selenium accumulation by embryos seemed to be related more to a lack of gills than to the presence of a chorion. The bioconcentration factor (the ratio of tissue-to-water concentrations) declined in all groups with increased waterborne selenium levels; accumulation rates appeared limited by cell permeability. At low levels of waterborne selenium (0.4µg/l), the total accumulated was small relative to existing body burdens and unlikely to contribute significantly to the nutritional requirement for selenium. High levels (45.6µg/l), however, caused considerable selenium accumulation and may be sufficient to overcome the effects of low dietary selenium.

A comparison of the acute toxicity of chemicals to fish, rats and mice.

The acute toxicity of chemicals to rainbow trout, as shown by intraperitoneal injections (IP LD50), oral dosing (oral LD50) and aqueous exposure (LC50) was compared with published values for IP LD50S and oral LD50S of mice and rats. The method of comparison was by simple linear regression analyses of log-transformed data, modified to recognize that X (fish toxicity) was neither fixed nor measured without error. Within-species comparisons demonstrated very strong linear correlations (r = 0.866-0.998) between IP and oral LD50S. Variability was least for the fish data since it was all generated in one laboratory. Comparisons between species of IP and oral LD50S gave correlation coefficients ranging from 0.59 to 0.95 with the majority over 0.80. Correlations were best (r = 0.83-0.94) between fish LD50S and rat and mice IP LD50S. Correlations were poorest between fish and mammalian oral LD50S (r = 0.59-0.66) because the sample sizes and the ranges of values were very small. In all cases, the slopes were close to, or equalled, 1.0. Comparisons of fish LC50S to fish or mammalian LD50S were not as successful. Correlation coefficients ranged from 0.19 to 0.83. Presumably the cause was the aqueous exposure. Interactions of the chemicals with water (e.g. dissociation) and with lipid membranes (partitioning) should cause considerable variations in uptake efficiency. However, adjustments of LC50S for dissociation constants and partition coefficients did not improve these correlations, probably because there were few chemicals for which all data were available. These comparisons demonstrate a potential for a wider use of surrogate species in toxicity testing and for adapting existing data from mammalian toxicology to aquatic hazard assessments.

Effect of increased dietary carbohydrate on selenium metabolism and toxicity in rainbow trout (Salmo gairdneri).

Juvenile trout were reared on either a high available carbohydrate (HC) or low available carbohydrate (LC) diet supplemented with from 0 to 10 micrograms selenium per gram of diet for 16 weeks, to determine if excess liver glycogen deposition affected the metabolism and toxicity of dietary selenium. Trout reared on the HC diet with 10 micrograms selenium per gram diet first demonstrated signs of selenosis and had significantly higher (P less than 0.05) liver selenium levels than trout reared on the LC diet with 10 micrograms selenium per gram diet after 16 weeks, indicating that excess dietary carbohydrate enhances dietary selenium toxicity in trout. The mechanism of the interaction is unclear since neither selenium elimination rates nor carcass and kidney selenium levels were affected by the dietary carbohydrate level. Trout reared on high dietary selenium diets (10 micrograms/g) had an increased incidence of renal calcinosis. In addition, liver copper levels were significantly affected by both dietary selenium and liver glycogen content indicating a significant copper-selenium and copper-glycogen interaction in trout. The development of renal calcinosis and the copper interactions suggest a variety of toxic effects of selenium on trout that may all be responsible for the observed changes in growth and feed efficiency.

Absorption, distribution, half-life and possible routes of elimination of dietary selenium in juvenile rainbow trout (Salmo gairdneri).

1. The influence of different levels of dietary selenium on the metabolism of selenium in rainbow trout was studied using 75Se as an indicator. 2. Gastric absorption of selenium by the trout appeared to be very efficient. 3. Highest tissue concentrations of selenium were noted in the liver and kidney. 4. Blood did not concentrate selenium and the plasma was the major transport medium. 5. The liver and kidney appeared to be involved in selenium excretion based on high tissue concentrations and variations in half-lives with selenium loading. 6. The biological half-life of selenium in the tissues decreased with increased selenium loading except in the liver, which at toxic dietary selenium concentrations became longer, suggesting a rate-limiting metabolic transformation of selenium for excretion in this organ.

The requirement and toxicity of selenium in rainbow trout (Salmo gairdneri).

This study measured the dietary selenium requirement of rainbow trout and their response to excessive levels of dietary selenium. A dietary selenium level of 0.07 microgram/g dry feed with a waterborne selenium level of 0.4 +/- 0.2 microgram/liter and a dietary vitamin E level of 0.4 IU/g dry diet was sufficient to prevent frank selenium deficiency symptoms. Maximal plasma GSH.px activity was obtained at a dietary selenium level between 0.15 and 0.38 microgram/g dry feed which is less than the average selenium concentration of commercial diets. Chronic dietary selenium toxicity occurred at 13 microgram selenium/g dry feed. Major effects of selenium toxicity were reduced growth rate, poor feed efficiency and a high number of mortalities. No histopathological lesions or significant deviation in the investigated blood parameters or liver somatic index were detected in trout raised on diets containing 13 microgram selenium/g dry feed. Tissue selenium analysis indicated that trout can maintain homeostasis with dietary selenium levels up to 1.25 microgram/g dry feed. The selenium uptake and accumulation in tissues of trout reared on diets containing in excess of 3 microgram/g dry feed may ultimately be toxic to trout if maintained over long periods of time.