|
<br />.,'
<br />
<br />DeForest, 8rix, ami Adams
<br />
<br />Debate/Commentary
<br />
<br />INTRODUCTION
<br />
<br />cvaluale pOlcnlial selcnium conlaminalion at freshwaler silCS. In order to
<br />inlcq>rcl lhc signilicancc of lhc selcnium cunccnlralions measured under
<br />Ihesc programs, sevcral aUlhors have proposed selcnium guidelines fur vari-
<br />ous cnvironmcnlal comparlments (e.g., Lcmly, 1993a; Skorupa, Morman, and
<br />Selchick-Edmunds, EI!IG). Spcci/ically, guidelines havc bcen proposed for
<br />surfacc waler, scdimCnL and various lissues, including ovarics, whole body,
<br />diel, liver, eggs, and lesles based on the aUlhors' rcviews uf publish cd and
<br />nnpnblished litcratnrc,
<br />Thcsc guidelincs include recommended toxicily lhrcsholds for abiolic (wa-
<br />ler, sedimcnl) and biotic (various fish tissues and dict) compartmenls. In this
<br />paper we focus only on the toxicity thresholds for tissues. Given the site-
<br />specilic factors that influence selenium bioavailability, bioaccumulation, and
<br />toxicity in aquatic systcms, we feel that proposal of a single guideline value for
<br />selcnium in surfacc watcrs or sedimenls is inappropriate. Different sites will
<br />rcquirc different water or scdimenl selcnium conccnlrations to ensure lhal
<br />cunccntrations in tissues such as fish ovaries do nol exceed a lOxic threshold.
<br />Site-to-sile variability has been demonstraled for lish by Van DelVeer and
<br />Canton (19!17) and for lJirds by Adams et ai, (1 m.l8). Van DelVeer and Canlon
<br />(HIU7) used a scdimcnt-based lJiuaccumulation model lo dcmonstrate thal
<br />/ish in a lotic systcm in Colorado wcrc not at risk at water selcnium concentra-
<br />tions or appruximately 30 ~g/L, thrce timcs highcr than cuncentrations at
<br />which clkclswcrc obsclvcd in Uclews Lake. Adams et al. (1998) nsed a
<br />slalislical lJiuacculllulalion model lJascd 011 selenium data fur bird eggs and
<br />surfacc walcrs at 17 silcs in thc U.S.; the 90th and 10th percentilc waler
<br />selcnium cunccntralions associaled wilh a selcnium concentration of
<br />20 mg/kg dw in bird eggs ranged from 6,8 to 318 ~g/L in their model. Both
<br />of thcse studies support, site-specific water quality guidelines for selenium
<br />based on a biuaccumulation model that eSlimales selenium concentrations in
<br />the critical tissues fUI' toxicological effecls.
<br />For a watcr selenium threshold to bc appropriate it must be based on a
<br />bioaccumulation model that accounls for the site-specific facLOrs that inl1u-
<br />ence selenium cycling and lJioavailability to aqualic organisms, An effective
<br />watcr qualily critcrion must be lJascd on aqucous selcniulII concentrations
<br />that are snlliciently low to prevent accumulation in fish food urganisms, which
<br />in turn would rcsull in the accumulalion of seleniullllo high enough levels ill
<br />parcntallish to causc rcproductivc impairmcnt. Givcn this, idcntificalion and
<br />agrccmcnt un tissuc toxicity threshulds fur use in such a mudel is critical in
<br />predicting whethcr /ish populalions are at large-scale risk from selenosis.
<br />Wc critically rcvicwed lhc tuxicily thrcsholds proposed by Lemly (1993a)
<br />and Skorupa et al. (19911), and thc scientilic Iileralure on which lhey are based.
<br />Wc thcn formulated our own rccommendalions on the approprialc thresh-
<br />olds fiJllowing, lo Ihc cxtcnt possible, the USEPA gnidelines uscd to dcvelop
<br />watcr qnalilY crileria (SlCphan et ai" 1985), .Jalvincn ami Ankley (1999) re-
<br />(:cntly sunllnarilcdlish tissnc concclllralions and associaled effecls for several
<br />chcmicals, including sclcnium: Thcir intcrpretatiuns uf lhc loxicological data
<br />
<br />Selenium is a metalluid c1cmcnlthat hislorically has becn of inlcrcst duc 10
<br />ils potentialloxicily to livestock and ils ncccssity as a nUlrilional supplcmenl
<br />in livcSLOck feed in ccrtain arcas uf thc Unitcd Slates (Roscnleld and Hcath,
<br />) 9'16; 19(4). Inclusion of selcnium in I.hc list of 65 priurily' pollutants (subsc-
<br />(IUClllly cxpalldcd lo 129) in lhc mid-1970s providcd grealcr focns in dcvelop-
<br />ing aquatic toxicolo!,'}' dala for dcrivaliun uf a selcnium watcr-<lualilY crilerion
<br />(Adams, 1976; Cardwell et 111.,1976; llaltel', Adams, and.!olmson, 19811; Adams
<br />and Johnson, 1981). Thc first acute crilerion rccommended lJy thc USEPA fur
<br />freshwalcr organisms was 35 ~g/L (a ch roll ic criterion was nOll'ccummcndcd).
<br />The most reccnt acute and chronic criteria sel lJy the USEPA for proleclion
<br />of freshwaler fish and invertcbratcs are 20 and 5 flg/L, respcctivcly (USEPA,
<br />1987).
<br />These latest crileria were derived based 011 lield studies of fish populations
<br />in 8clews Lake, Nurth Carolina (CumlJie and Van I')orn, 1978), Hclcws Lakc
<br />was receiving fly ash from a coal-Imming e1cClric powcr facililY that increascd
<br />selenium conccnlratiuns in lhc lakc. Cumbie and Van Horn (1!17H) obsclvcd
<br />eficcls on lish populations at selcnilllll cUllccnlraliolls as low as 10 flg/l.,
<br />whereas populaliolls in OIlC ponioll of thc lakc appcarcd unallccled at <5 flglL
<br />(USEPA, 1!)87). Hcncc, thc basis for thc [II.TCllt li'cshw<ltcr chronic nilcrion
<br />of 5 ~g/L. Since the "c1cws l.ake sllulics wcrc rcportcd, selcnium poisoning
<br />offish poplllatiolls has bcclI dOClmlcnll:d al a ICw ,uldiliollallocalions aroulld
<br />lhe U.S., includillg l1ycoRcsclvoir in North Carolilla and Kcstcrson Reselvoir
<br />in Califomia.
<br />Selenium is unusual relalive lO most othcr mctals and melalloids in that
<br />many inorganic and organic forms uccur in thc aquatic cnvironment, and
<br />each form is differentially bioavailable and toxic to aquatic urganisms. Thc
<br />selenium forms presenl in an aqualic s)'stem are driven lJy thc biogeochemical
<br />cycling of selcnium that is slrongly coutrolled by site-spccilic environmcnlal
<br />factors such as redox, pH, and biulogical productivily (Lemly and Smith, 1987;
<br />Bowie and Grieb, 1991; Porcella el at., 1991).
<br />Rcducli~n of inorganic selenium species tcnds to immobilize selcnium in
<br />an aquatic syslcm, while olhcr processes, such as oxidaliun and biotransfurma-
<br />tion, tend to make selenium bioavailablc to aquatic organisms, Biological
<br />mecl...nisms such as uplake of scdimellt selcniulll by rooted plants, bcnlhic
<br />invertebrates, allll dctrillls-cating invcrtcbratcs, can aclto rcmobilizc selcnium
<br />into the aqualic foud welJ. Aecul'llingly, IClllie systcms lend to bioaeCUlllulal.c
<br />selenium much morc lhan lotic systems Ih,lI have highcr flushing ralcs allll
<br />lower productivity (Lillebo et ai" 1988; Vall DClvcer and Can LOn, ) 997). For
<br />examplc, Lillcbu et at. (1988) dcmonstralcd this by pluuing bioacculllulation
<br />dala fur impoundcd and flowing walcrs; /ish sclcuium rcsiducs wcre approxi-
<br />malely six limcs grcatcr in impoundcd watcrs than in l10wing walcrs al a W,lIcr
<br />selcnium conccnlratioll uf 10 flg/L.
<br />Based on incrcased awarencss of thc ccotoxicological effects of sclcnium, a
<br />number of walcr-qualilY monitoring programs havc bccn implcmcnlcd 10
<br />
<br />1111111. ECllI. Risk Assess, Vol. 5. Nil. 6, \!}!I!I
<br />
<br />1189
<br />
|