Stream ecosystem response to, and recovery from, experimental exposure to selenium

The effects of selenium on streamecosystems were studied in outdoor,experimental stream mesocosms during a dosingperiod in which sodium selenite was added atnominal concentrations of 30 µg/L,10 µg/L, and 2.5 µg/L. The durationof the high, medium, and low treatments were573 d, 972 d, and 311 d, respectively. Apost-dosing period of three years (hightreatment) and two years (medium, lowtreatments) also was studied. Seleniumconcentrations in water, sediment, plants, andmacroinvertebrates were measured throughoutthe dosing and recovery periods. Fatheadminnows and bluegill sunfish were periodicallyheld in the streams to measure seleniumaccumulation and its effects on fish survivaland reproduction. Quantitative samples ofmacroinvertebrates were collected to assessselenium effects on macroinvertebratecommunities.Mean selenium concentration inwater was quite close to the nominalconcentration. Selenium accumulated in thesediment in all three treated streams, but notin the control streams. Sediment seleniumdecreased slowly after dosing ceased, but wasstill significantly higher than in controlstreams three years (high treatment) and twoyears (medium treatment) later.Macrophytetissue selenium concentrations weresignificantly greater in all three treatmentsthan those in the control streams duringdosing. Macrophyte selenium bioaccumulationfactors (BAFs) ranged from about 300 to 1900. Tissue selenium decreased rapidly in all threetreatments after dosing ended.During dosing,selenium concentrations in animals from allthree treatments were significantly higherthan in those from control streams. The BAFsfor macroinvertebrates ranged from 1100 to2000. Isopods accumulated more, and amphipodsless, selenium than other invertebrates. Therewere no significant effects of selenium onmacroinvertebrate abundance, richness ordiversity. Several macroinvertebrates werenot affected by exposure to selenium, butisopod and Tubifex populations weredramatically reduced in the high and mediumtreatments. After dosing, mean seleniumconcentration in macroinvertebrates decreasedslowly.Bluegill sunfish accumulated seleniumduring dosing and after selenium additionsceased. Tissue selenium was highest in theliver, followed by the gonads, skeletalmuscle, and whole body. Tissue seleniumconcentrations one (high, medium) and two(high) years after dosing were lower thanduring dosing, but whole body, skeletal muscleand liver concentrations were high enough tobe considered potentially toxic.Recovery ofselenium contaminated streams includes bothreduction of tissue selenium concentration tonon-toxic levels in fish and their foodorganisms and recovery of populations of taxadeleteriously affected by selenium exposure. Our results suggest that when selenium iseliminated from the water in streams, seleniumconcentrations in sediment, plants,macroinvertebrates, and fishes will decreaseto levels that approach concentrationsconsidered to be non-toxic to fish andwildlife and that affected populations willrecover within several years. Based onselenium accumulation in the food chain andthe presence of real, but not statisticallysignificant, effects on fish mortality andreproduction in the low treatment streams, wesupport a selenium water quality criterion forthe protection of fishes and sensitiveinvertebrates of 2 µg/L or less.