Barium selenite

Bioaccumulation: aquatic / sediment

Description of key information

Based on the combined data set for fish and aquatic invertebrates, the key bioaccumulation factor was calculated to be 4721 L/kg dw. This corresponds to a BAF of 944 L/kg ww when a conversion factor of 5 is taken into account (assumption of 20% dry weight for fish and aquatic invertebrates).

Key value for chemical safety assessment

BCF (aquatic species):

944 L/kg ww

Additional information

Overall, 172 studies were identified as potentially relevant for bioaccumulation in aquatic organisms. After screening, 125 studies were retained. Finally, 32 studies were considered reliable for bioconcentration (allowing calculation of a BCF or bioconcentration factor), bioaccumulation (allowing calculation of a BAF or bioaccumulation factor) and/or trophic transfer (allowing calculation of a BMF or biomagnification factor). The studies reporting reliable results only investigated bioconcentration, bioaccumulation and/or biomagnification in freshwater organisms, although some studies also included experiments with anadromous fish in brackish water. Reliable data are available for fish, aquatic invertebrates, aquatic plants (mainly algae) and amphibians. Reliable results were included from both field studies (ambient Se) and laboratory experiments (Na₂SeO₃, Na₂SeO₄, H₂SeO₄ and seleno-methionine).

Ranges of BCF, BAF and BMF values reported in this endpoint summary were derived based on Se concentrations determined in whole body only. Because of the importance of the dietary bioaccumulation route, BAF values were generally observed to be somewhat higher than BCF values. Overall, BCF and BAF values were observed to be higher with decreasing selenium concentrations. This is especially clear from typical bioconcentration studies conducted in the laboratory. Because of the essentiality of selenium and the very narrow border between concentrations resulting in deficiency and concentrations resulting in toxicity, BCF/BAF values observed at near-background concentrations were not considered relevant for the assessment of secondary poisoning and therefore not taken forward in further analysis.

Reliable BCF or BAF values (dry weight based) for freshwater fish ranged from 16.9 to 20765 L/kg dw. The lowest value represents a BCF value based on an experiment of Saiki et al. (1992a) in which chinook salmon fingerlings (Oncorhynchus tshawytscha) were exposed for 28 days to serial dilutions of Se-contaminated tile drainwater. The highest value represents a BAF value from the study of Hamilton et al. (2005a). In this study, razorback sucker larvae (Xyrauchen texanus) hatched from eggs obtained from adult fish from Se-contaminated areas were exposed to a combination of either site water/reference food, site water/site food, reference water/site food or reference water/reference food. The BAF value of 20765 L/kg dw resulted from an experiment in which larvae were exposed for 7 days to site water + site food. When considering BCF and BAF values separately, ranges were obtained of 16.9 to 9875 L/kg dw for the BCF and 492 to 20765 L/kg dw for the BAF. Based on the available data, BAF values can indeed be concluded to be generally somewhat higher than BCF values. For fish, reliable data are available for sodium selenate, sodium selenite, seleno-methionine, and ambient selenium. The data available do not allow strong conclusions on difference in BCF or BAF among these Se compounds tested. When comparing the data for selenate (23 to 574 L/kg dw) and selenite (41.5 to 2300 L/kg dw), selenate seems to have a lower capacity for bioconcentration/bioaccumulation, however, because there are no data allowing species-specific comparison of BCF values or BAF values for selenite with those for selenate, it is difficult to draw conclusions from this observation. It should however be noted that there is a possibility that selenate is bioaccumulated to a somewhat lesser extent than selenite because of the interaction of sulphate with the uptake of selenate. Only one value was available for seleno-methionine (2600 L/kg dw). This single value does not allow conclusions with regard to the comparative bioconcentration/bioaccumulation of seleno-methionine and selenite/selenate.

For aquatic invertebrates, reliable BCF or BAF values (dry weight based) ranged from 17 to 30204 L/kg dw The lowest value represents a BCF value from the study of Ogle and Knight (1996) in which adult daphnids (Daphnia magna) were exposed to increasing waterborne Se concentrations (added as Na₂SeO₄) in test media with increasing sulphate levels. The BCF value of 17 L/kg dw was obtained from the exposure treatment with the highest Se concentration (0.49 mg/L) and the highest sulphate concentration (325 mg/L). This could be expected since sulphate has been observed to protect against Se uptake and toxicity. The highest BCF value of 30204 L/kg dw originates from the study of Besser et al. (1993) in which adult daphnids (D. magna) were exposed for 96 h to a series of radiolabeled seleno-methionine concentrations. Extremely low concentrations (≤ 1.0 µg Se/L) were omitted because these concentrations fall within the range of typical background concentrations and may – due to the essentiality of Se – lead to extremely high BCF values. When omitting the data for seleno-methionine (highest BCF = 30204 L/kg dw, highest BAF = 28870 L/kg dw), BCF values ranged from 17 to 573 L/kg dw, whereas BAF values ranged from 72 to 12333 L/kg dw. Based on the available data it can be concluded that BAF values are generally somewhat higher than BCF values. For aquatic invertebrates, reliable data are available for sodium selenate, sodium selenite, seleno-methionine, and ambient selenium. As for fish, BCF and BAF values for selenate (17-388 L/kg dw) seem to be slightly lower than those for selenite (220-1087 L/kg dw). Some species-specific comparisons are possible here for Daphnia magna, and in most cases BCF and BAF values for selenite are indeed somewhat higher than those for selenate. However, not sufficent data are available to allow meaningful statistical evaluation. Because of the protective effect of sulphate on selenate uptake and toxicity, it would however be expected that the capacity for bioconcentration/bioaccumulation of selenate would be at least somewhat lower than that of selenite. Only three BCF or BAF values were available for seleno-methionine (1100 to 30204 L/kg dw). Seleno-methionine seems to be the selenium form with the highest capacity for bioconcentration and bioaccumulation, however, this may be highly species-specific, since for Helisoma sp., there was no large difference between the BAF for sodium selenite (950 L/kg dw) and the BAF for seleno-methionine (1100 L/kg), whereas for Daphnia magna the difference was much larger.

Based on the combined dataset for fish and aquatic invertebrates, a key BCF/BAF for risk assessment purposes was calculated. Because the overall difference between BCF and BAF values was not very high, all data were lumped. First, for each individual study yielding reliable BCF or BAF values based on whole body concentrations, an average value (geometric mean) for all observations was calculated. This was done at the species level and, in case data were available for different substances, a separate substance-specific average value was calculated for each species. For instance, for a field study in which two different species were sampled at different sampling sites, an average value was calculated for each species, and for a bioconcentration study conducted in the laboratory with a single species and different test substances, average values were calculated for each test substance.

Then, a single BCF/BAF value was derived for each species by calculating the average of study-specific and/or substance specific average values for each species. Averaging substance-specific average values was considered acceptable because differences among bioconcentration or bioaccumulation factors of the Se compounds tested was generally (selenite, selenate, ambient Se and seleno-methionine) were generally limited and not consistent at the species level. For fish, this approach yielded 8 data points (163 L/kg dw for Pimephales promelas, 2348 for Gambusia affinis, 4787 for Xyrauchen texanus, 1659 for Lepomis macrochirus, 42 for Oncorhynchus mykiss, 1550 for Micropterus salmoides, 26 for Oncorhynchus tshawytscha, and 35 for Morone saxatilis). For aquatic invertebrates, averaging at the species level was not possible because only a few studies focused on the species level. Other studies only specified the genus, the family, or an even broader group of organisms studied. Therefore, BCF and BAF values had to be averaged for the broadest groups. Four groups were considered: crustaceans (including data on – as specified in study – plankton, zooplankton, crustaceans, amphipods, Daphnia magna, and Artemia franciscana), insects (including data on – as specified in study – insects, Diptera, Coleoptera, Anisoptera, Zygoptera, Coroxidae, Chironomidae, Hydrophilidae, Helodidae, Haliphidae, Dytiscidae, damselfly, backswimmer, and boatman), mollusks (including data on – as specified in study – a snail, and Helisoma sp.), and annelids. This yielded 4 data points (1149 L/kg dw for crustaceans, 1237 for insects, 1289 for mollusks, and 4688 for annelids). Next, the fit of various distribution functions for these 12 data points were evaluated using goodness-of-fit statistics software (@RISK, Palisade Inc.). The best fitting distribution (log-beta distribution) resulted in a 10^th, 50^th and 90^th percentile of 28, 606 and 4721 L/kg dw, respectively. For risk assessment purposes, it is suggested to use the 90^th percentile as key value. It should be noted that this value takes also bioaccumulation due to dietary exposure into account (BAF data included) and is about equal to the largest BCF or BAF values observed (for annelids and the fish Xyrauchen texanus). Therefore, this value of 4721 L/Kg dry weight can be considered as a conservative estimate for bioaccumulation of selenium in aquatic organisms.

Although not used in standard risk assessment procedures, it is considered useful to include and discuss the reliable data for selenium bioconcentration in species other than fish and aquatic invertebrates as well as the reliable data for trophic transfer of selenium to aquatic invertebrates and fish.

Reliable BCF values for aquatic plants (mainly algae) and cyanobacteria varied from 31 to 16836 L/kg. The lowest BCF was obtained from a study in which blue-green algae of the species Anabaena flos-aquae were exposed for 10 days to a series of waterborne Se concentrations added as Na₂SeO₄, Na₂SeO₃ or seleno-L-methionine (Kiffney and Knight, 1990). The BCF of 31 L/kg was calculated for the exposure treatment at 0.995 mg Se/L, added as Na₂SeO₄. The highest BCF originates from the study of Besser et al. (1989) in which 28-day exposure experiments were conducted in closed-system microcosms containing natural sediment and reconstituted water, to which a background mixture of unlabelled sodium selenate, sodium selenite and seleno-methionine as well as a certain concentration of radiolabelled seleno-L-methionine was added. In these microcosms, aquatic macrophytes and zoo- and phytoplankton were present. The BCF under consideration was obtained for periphyton samples. For algae, reliable data were available for sodium selenate, sodium selenite, selenious acid, seleno-methionine, and ambient selenium. Lumping the data for sodium selenite and selenious acid, BCF values for selenite ranged from 263 to 2627 L/kg dw, whereas BCF values for selenate ranged from 31 to 1504 L/kg dw. BCF values for seleno-methionine varied between 1567 and 16836 L/kg dw. The order of increasing capacity for bioconcentration is clearly selenate < selenite < seleno-methionine. Following the same approach as for fish/aquatic invertebrates, five average values were obtained: 1834 L/kg dw for periphyton, 1172 for nectic unicellular green algae, 756 for diatoms, 440 for filamentous algae, and 234 for blue-green algae. Best-fit analysis resulted in a 10^th, 50^th and 90^th percentile of 247, 699 and 1973 L/kg dw, respectively.

Further, a reliable BCF of 9226 L/kg was calculated for tadpoles (species not reported) using data from a field study reported by Hamilton et al. (2005b).

For freshwater fish, reliable biomagnifications facors (BMF) ranged from 0.1 to 1.11 (dimensionless, dry weight based). Overall, higher factors were obtained from experiments in which fish were exposed to food collected from the field in contrast with experiments using Se-spiked diets. It is important to keep in mind that these BMF values are based on whole body Se accumulation and that Se accumulation in individual organs can be much higher, depending on the organ under consideration. The lowest BMF value was calculated for rainbow trout (Oncorhynchus mykiss) and originates from a study in which adult trout were exposed for 16 weeks to a series of dietary Se concentrations (added as Na₂SeO₃) (Hilton and Hodson, 1983). The highest BMF value orginates from a foodchain experiment with radiolabeled H₂SeO₄ in which fathead minnow larvae (Pimephales promelas) were fed rotifers (Brachionus calyciflorus) that had been feeding on Se-laden algae (Bennett et al., 1986). For fish, reliable data were available for selenious acid, sodium selenite, seleno-methionine and ambient selenium, however, no clear differences were observed between trophic transfer of individual substances. Following the same approach as for the BCF/BAF values, three values were obtained: 0.57 for Pimephales promelas, 0.71 for Oncorhynchus tshawytscha, and 0.38 for Oncorhynchus mykiss.

For aquatic invertebrates, reliable BMF values varied between 0.03 and 0.92 (dimensionless, dry weight based). The lowest value was obtained in a foodchain experiment with radiolabeled Na₂SeO₃ in which adult daphnids (D. magna) were fed Se-laden algae during 10 days (Lam and Wang, 2006, after wet weight to dry weight conversion assuming 20% dry weight). The highest value originates from a 14-day study by Besser et al. (1993) with radiolabeled Na₂SeO₄ in which adult daphnids (D. magna) were fed algae (C. reinhardtii) that had been exposed to a series of waterborne Se concentrations. For aquatic invertebrates, reliable data were available for sodium selenate, sodium selenite, selenious acid and seleno-methionine, however, here too, no clear differences were observed between trophic transfer of individual substances. Following the same approach as for the BCF/BAF values, two values were obtained: 0.31 for Daphnia magna and 0.73 for Paramecium putrinum.

Lumping all data for fish and aquatic invertebrates, nonparametric 10^th, 50^th and 90^th percentiles of 0.34, 0.57, and 0.72, respectively, were obtained for biomagnifications factors. Data for BMF were not taken into account for the assessment of secondary poisoning, because dietary exposure was included for the selection of the BAF value.