<br />Dc}'orcst, 8rix, and Adams
<br />
<br />Debate/Commentary
<br />
<br />discussed below, some coldwater fish may be more scnsitivc Lhan warmwatcr
<br />fish. Another unccrtainty is lhc paucity of warl1lwatcr spccies fur which ovalY
<br />data are availablc. FUl'lher testing should bc conducted to understand thc
<br />variability in sensitivitics betwccn spccies.
<br />Whole Body. The whole body guidclinc of 4 mg/kg dw recommended by
<br />Lemlr (1993a) is based on mortality of rainbow trout cxposed to aqucous
<br />selcnHe (Hunn el at" 1987). In rcvicwing Ihc whole 1J0dy rcsiduc-bascd toxi-
<br />cological data, ,,:e nO,ted,t.wo potcntially significant rclationships that may cxist.
<br />Thc first potcntlally slglllhcant relationship is heavily dcpcndent on the chinook
<br />salmon c~ata from Hamiltun el at. (1990). As discussed ahovc, the toxicity
<br />rcsults altcr 90 days are clucstionable duc to the rapid dcclinc in sUlvival of
<br />~ontrollis~1 betwccn day 60 and 90 of the test. [f thc results aftel' 90 days are
<br />mdeed rehablc, and the obscrved mortality can be alll"ilmted to seleniulII, thc
<br />resu~l~ suggcsl coldwater fish (rcprescnted by chinouk salmon) may hc more
<br />~ensHlvc than warmwater fish (rcprcscntcd by IJlucgill) (Figure 4). Howcver,
<br />If thc results aftcr 50 days are more reOectivc of the sensitivity of chinuok
<br />salmon to selcniulll, there wonld IJc limited cvidcnce to suggest I hat coldwater
<br />fish are more sensitive than warlllwatcr fish. In lact, the sensitivity of coldwaler
<br />and warmwatcl' fish appear to he qnitc similar bascd on thc day GO data (Figlll'c
<br />4). The second potclllially significant rclalionship that becamc apparenl was
<br />that effects are observed in fish atlowcr whulc body selenium concentralions
<br />when they arc exposed to only waterborne seleniulll than when they arc
<br />expos,ed to dietary seleni~m. Cleveland el at, (1993), for example, exposed
<br />bluegtlls to an aqueous 6: I sclenatc:selenite mixture lor 50 days, and in a
<br />separate expel"im.ent, cxp?sed IJlucgills to dictalY organoselcninm lor 90 days.
<br />Obsctved mOl'lahty was htghcr at lowcr whole 1J0dy residnes in fish cxposcd
<br />onl,y to ~aterbOl:ne, inorganic selcnium (Figurc 5). This has impol'lant impli-
<br />callons m estabhslnng a whole body selcnium guidelinc bccausc, as discusscd
<br />previously, natural fish populations arc pl"imarily cxposcd to selcniulII via Iheir
<br />diet. Cleveland el at. (1993) support this ohscrvation, lIl>linH Ihat toxicily to
<br />, natural populations of bluegill and probalJly othcr aquatic OlWlIIislllS would
<br />probably occur at much lowcl' waterbol"lle concentratious than thosc tcstcd iu
<br />this study owing to food-chain exposure.
<br />~csscr el at. (1993) also obsclved thatlJlucgills from both aqucous and fooel-
<br />cha~n cxposures to seleno-mcthioninc accumulated greater selenium conccn-
<br />tl:allons tha~l bl~legills in cOlllparalJle exposures with inorganic selcniulII spc-
<br />CICS. .81uegtlls III fO,od-chain cxposures accumulated consistenlly greater
<br />selenium concentrallons than Ihose in aqncous exposurcs, and bluegills ex-
<br />posed to both aqucons and foodbornc selcnium accumulated most of Iheir
<br />total selenium residnes frollllheir food. Bertralll and Brooks ( I U8fi) mcasured
<br />diflcret~tnptakc ~inctics between walcrborne 5e1cniulII and dietalY selcniulII,
<br />snggesllng the eXlstcnce of two functional compartmenls for selcniulII accu-
<br />mulation in fish. Onc cOlllpartmentmay contain an unbound inorganic pool
<br />and thc othcr may contain organically bound compounds. The depuralion
<br />rate of the unbound inorganic pool is rapid. Thc aUlhol's noted Ihat acculllu-
<br />lation ofwaterbornc selenium and diclalY seleninlll is appl'llximately addilivc,
<br />
<br />and that in all cascs Ihe potcnliallJurdcn of selcnium in fish tissues is rclated
<br />to the conccntriltion of selenium in the fish diet.
<br />We hypothesize that waterlJorne inorganic selenium exposures result in
<br />toxicity at lower whole 1J0dy residucs than those based on dictary
<br />organoselcniulll cxposures bccausc thc inorganic selcnium in fish is rapidly
<br />depurated. [n natural populations, most of the selcnium exposure and result-
<br />ing bioacculllulalion is from dictalY organoselenium, resulting in higher
<br />whole body conccntrations, Comparison of whole body selenium concentra-
<br />tions in field-collected fish to laboratory-dcrivcd toxicity data lJased on water-
<br />bornc cxposnres to inorganic selcnium is, therefore, not a valid comparison,
<br />Givcn that Lemly's (1993a) I'ecolllmended whole body rcsidue guidelillt: of
<br />4 mg/kg dw is based on a study involving water only exposures, we recolII-
<br />mcnd guidelines based only on studics including dietary exposures.
<br />For coldwater fi5h, Ilamilton el at. (1990) is the only study in which fish were
<br />fed diets conlaining organoselenium. As discusscd above, the whole body
<br />chronic values in Lhis study are 7.4 mg/kg dw for reduced growLh and 7,6 IIIg/
<br />kg dw lor reduced slllvival following 60 and 90 days of exposure, respectively.
<br />Bascd on Ihe probit and logit models, the ECw. EC20, and EC50 values for
<br />mortality in Ihis study are 12, 17, and 27 mg/kg dw after 60 days and 1.7, 2,8,
<br />and 7,0 mg/kg dw after 90 days (Figure 6). Bascd on the growth endpoinl,
<br />ECI\l. EC2\l. and EC5\l values are 6, II, and 20 mg/kg dw aftcr 60 days of
<br />exposurc to selenium, and 5, 7, and 14 mg/kg dw aftcr 90 days (Fib'1.lre 7).
<br />The lowcst wholc body chronic value for warmwater fish from a dietary
<br />study, 10.5 mg/kg dw, is lJased onlalval mortality (Coyle el at" 1993). Similar
<br />to the ovary data discussed above, we poolcd thc wholc body selenium and
<br />Ialval mortalilY studies from all relcvant studies for wannwater fish (i.e" Ogle
<br />and Knight, J 989, Hcrmanllu el at" 1992, Coylc el at" 1993). Thc whole body
<br />ECIIl, EC2\l' and EC50 values, using bOlh thc pmbit and logh models, are 9, II,
<br />and 15 mg/kg dw, respectivcly (Figure 8). It should be noted thal the concen-
<br />tralion-response rclationship is not wcll defined as thcre are only two wholc
<br />body conccnlralions where lalval mortality was greater than 10% following
<br />Ahhou's correcliun for conlrul mortality. As for thc ovaries, we recommend
<br />Ihc ECIO valucs as whole body thrcsholds, Given that the results of the Hamilton
<br />elllt, (1990) stndy after 90 days are questionalJle and neeo 10 be confirmed, we
<br />rccommend that the ECIO for growth at (iO days (i,e" 6 mg/kg dw), the most
<br />sensitivc endpoint measured, be considered thc whole body threshold for
<br />lalval culdwater anadrolnous fish; Ihe EC,o of9 mg/kg dw is recommended as
<br />Ihe Ihreshold lor wannwatcr fish, 1L is important to notc that the coldwater
<br />ilnadromous lish threshold is for larval /ish, while thc wannwater threshold is
<br />for adults. The coldwater anaoromous lish threshold specifically does not
<br />apply to adultlish as maternal transfer of selenium has not becn evaluatcd in
<br />thesc lisla.
<br />/Jie/IllY. The Hamilton el (lL (1990) chinook salmon study is currcntly the
<br />most relevant lor delennilling a dictal)' threshold for coldwatcr fish lJecausc
<br />it is Ihe only coldwater study in which fish were fed organosclenium. I'rolJit
<br />
<br />HUIlI, Eeul. Risk Asscss. Vul. 5, Nu, 6, 1999
<br />
<br />1215
<br />
|