United States                                    FINAL DRAFT
               Environmental Protection                              ECAO-CIN-G058
               A°encv                                       September, 1989
4>EPA       Research and
               Development
                HEALTH AND ENVIRONMENTAL EFFECTS DOCUMENT
                FOR SELENIUM AND COMPOUNDS
               Prepared for
               OFFICE OF SOLID WASTE AND
               EMERGENCY RESPONSE
               Prepared  by
               Environmental Criteria and Assessment Office
               Office of Health  and Environmental Assessment
               U.S. Environmental  Protection Agency
               Cincinnati, OH 45268
                           DRAFT: DO NOT CITE OR QUOTE
                                  NOTICE

               document 1s a preliminary draft.  It has not been formally released
               .S. Environmental Protection Agency and should not at this stage be
               I to represent Agency policy.  It  1s being  circulated for comments
               ichnlcal accuracy and policy Implications.
                                                                      f

-------
                                  DISCLAIMER

    This report  Is  an external draft  for  review purposes only and  does  not
constitute  Agency  policy.   Mention of  trade names  or  commercial  products
does not constitute endorsement or recommendation for use.
                                      11

-------
                                   PREFACE
    Health and  Environmental  Effects  Documents  (HEEOs) are prepared  for  the
Office of Solid  Waste  and  Emergency Response (OSWER).  This  document  series
1s Intended to support listings under  the  Resource  Conservation  and  Recovery
Act (RCRA) as  well  as to  provide  health-related  limits and  goals for  emer-
gency and  remedial  actions under  the Comprehensive  Environmental Response,
Compensation   and Liability  Act  (CERCLA).   Both   published  literature  and
Information obtained  for  Agency Program Office files  are evaluated  as  they
pertain to potential human health, aquatic  life and environmental  effects  of
hazardous waste  constituents.   The  literature searched for In this  document
and  the  dates  searched  are  Included In  "Appendix: Literature  Searched."
Literature search material  Is  current up  to 8  months previous  to the  final
draft date  listed  on  the front  cover.   Final  draft  document  dates  (front
cover) reflect the date the document 1s sent to  the  Program Officer (OSWER).

    Several  quantitative  estimates are  presented  provided  sufficient  data
are available.   For systemic toxicants, these Include  Reference  doses  (RfDs)
for  chronic   and subchronlc  exposures  for  both   the  Inhalation and  oral
exposures.  The  subchronlc or  partial lifetime  RfO,  1s  an  estimate of  an
exposure  level   that  would not  be expected  to cause  adverse  effects  when
exposure  occurs  during a  limited  time Interval  I.e.,  for an Interval  that
does  not  constitute a  significant portion  of  the  Hfespan.   This  type  of
exposure  estimate  has  not been  extensively used,  or  rigorously defined  as
previous risk assessment efforts have  focused primarily  on lifetime  exposure
scenarios.  Animal  data   used  for  subchronlc  estimates  generally   reflect
exposure  durations  of 30-90  days.   The general  methodology   for  estimating
subchronlc RfDs  1s  the same as  traditionally  employed  for  chronic  estimates,
except that subchronlc data are utilized  when available.

    In the case  of  suspected  carcinogens,  RfDs  are  not  estimated.   Instead,
a  carcinogenic   potency  factor,  or  q-|*  (U.S.  EPA,  1980b),  1s provided.
These potency estimates  are  derived  for both oral   and  Inhalation exposures
where possible.   In addition, unit risk estimates for  air  and drinking water
are presented  based on Inhalation and  oral  data, respectively.

    Reportable quantities  (RQs)  based on both chronic toxlclty  and  carclno-
genldty are derived.   The RQ  Is used  to determine  the quantity  of a hazard-
ous substance for  which  notification  1s required  In the  event  of a release
as  specified  under  the  Comprehensive  Environmental Response,  Compensation
and Liability  Act  (CERCLA).    These  two  RQs  (chronic  toxldty  and  carclno-
genldty) represent two  of six scores developed (the  remaining  four reflect
1gn1tab1lHy,  reactivity,  aquatic  toxlclty,  and  acute mammalian  toxldty).
Chemical-specific RQs  reflect the lowest of  these six  primary criteria.   The
methodology for  chronic  toxldty and  cancer based  RQs  are defined   In  U.S.
EPA, 1984 and  1986b, respectively.
                                      111

-------
                               EXECUTIVE SUMMARY

    In  the  United   States,   selenium  Is  produced  by  recovering  1t   from
by-products generated during  the electrolytic refining of  copper  ore  (Elkln,
1982; U.S. Department of  Interior,  1982,  1988).   In 1984,  559,078 pounds  of
selenium were  recovered  from these  by-products  by  three domestic  producers.
Also, selenium  Is  recovered domestically from previously  manufactured  xero-
graphic  drums and selenium  rectifiers  (USDI,  1982).  Approximately 1  million
pounds  of  selenium  were  Imported  1n 1987  (USDI).   The  following  are  esti-
mates of  selenium consumption  by  end-use categories  In  1987 (USDI, 1988):
electronic and  photocopier  components, 43%; glass manufacturing,  20%; chemi-
cals  and  pigments,  20%;  other  (Including agriculture  and  metallurgy),  17%.
Because  of Us  photoelectric  and semiconducting properties,  selenium  1s  used
primarily  In the  production  of  photocopying components,  photoelectric cells,
photovoltaic cells and semiconducting rectifiers  (Elkln,  1982).
    Selenium Is  an  element;  therefore, It  does  not degrade  In  the environ-
ment, It  simply  changes  from one  form  to another.   The  ma.lor   features  of
selenium chemistry that affect  Its  fate and transport  1n  the environment are
associated with changes In  Us  oxidation  state and  the resulting differences
1n  chemical  properties   (Callahan  et  al.,   1979).   Because  selenium  only
changes  form and does not  degrade  In  nature, 1t undergoes  an environmental
cycling  process  Involving soil,  rocks,  plants,  animals,  water and air  (NAS,
1976).  When  released  to the  atmosphere  In  partlculate-phase,   selenium  1s
removed  by physical  processes  such as  rainfall  and  settling   (NAS,  1976).
Volatile  selenium compounds  are  released to  air  by  plant  and  mlcroblal
transformations  (Zleve  and Peterson,  1984;  Chau  et al.,  1976),  where  they
react with sunlight-formed  HO radical to  form  selenium  dioxide  and  other
                                      1v

-------
compounds  that  can  be  transported physically  to water and  soil.   In  both
water and  soil,  the  blomethylatlon of selenium to release  volatile  selenium
to air  1s  an  Important  part of the environmental cycling process  (Cooke and
Bruland, 1987;  Z1eve and  Peterson,  1984).  Selenium  In  oxidation  state  +6
(selenate)  1s  the  most  soluble and mobile selenium species In the  soil  and
water environments (Gruebel et al., 1987;  Callahan et  al.,  1979).  Because of
the solubility,  stability  and  ability of selenate to  be taken  up  by plants,
1t 1s considered the most  dangerous form of  selenium 1n relation  to  environ-
mental pollution 1s (NAS, 1976; Callahan et al., 1979).
    Selenium  Is  distributed widely  1n  the  earth's  crust,  and with  suffi-
ciently  sensitive  analytical  techniques,  selenium   can   be  detected  In
virtually all rocks and  soil on the earth's  surface  (NAS,  1976).   Because of
this  ubiquitous  distribution   1n  the  environment,  the  detection of  selenium
1n air,  water,  soil,  food  and vegetation  1s expected  (IARC,  1975).  Selenium
1s released to  the environment from natural sources such as  volcanoes,  rock
and soil erosion,  sea sprays  and volatile emissions  from  plants  and  micro-
flora  (NAS,  1976; Arlmoto et al.,  1985).  Humans  release selenium  to  the
environment  through  Incineration   of  coal,  fuel  oil  and  solid waste,  from
emissions  and waste  streams  generated  through  mining, refining and  Indus-
trial  applications  (NAS,  1976;   U.S.  EPA,  1987a).   More   than  half  of  the
human releases  are attributed  to  the combustion of  coal  (NAS,  1976).   Based
upon  available  monitoring  data,   the  average  selenium content of  drinking
water  Is  ~0.2   ppb  and  the  mean concentration  In urban  air  1s   ~3  ng/m3
(Bennett,  1986).   These  concentrations  can  be  used  to  estimate  average dally
Intakes  of 400  ng/day for  water  and  60  ng/day  for  Inhalation.   The average
dally  Intakes  for  water  and Inhalation  are  small  1n comparison to the esti-
mated average dally Intake of  139 yg/day  for food (Gartrell  et al., 1986a).

-------
    Acute toxldty  data  for selenium  1n  aquatic  vertebrates  were available
for  21   species  of  fish  and  1  amphibian.   Among  freshwater  species,  the
96-hour  LCcgS  for  selenium as  sodium selenlte  ranged  from 0.62  mg/l  for
the  fathead  minnow,  Plmephales promelas  (Elsler,  1985)  to 35  mg/l for  the
common carp,  CypMnus  carplo  (Etnler  et  al., 1987).  Fathead  minnows were
also  the least  tolerant   vertebrate  species  to  selenate-selenlum,  with  a
96-hour  LC5Q  of   2  mg/l   1n  a  flowthrough   test  (Etnler  et   al.,   1987).
Zebraflsh, Brachyodanlo rerlo.  and  juvenile striped bass,  Morone saxatnis.
were  the most  tolerant   of  exposure  to  selenate-selenlum,   with  96-hour
LC5Qs  of 82  and  85.8  mg/l,  respectively  (N11m1  and Laham,  1976; Klauda,
1986).   The  toxldty  of  selenium as  selenium dioxide  ranged  from a 96-hour
LC5Q   of  7.3  mg/l   for   fathead   minnows   to   20  mg/l   for   zebraflsh
(Cardwell et al.,  1976;  N11m1  and Laham,  1976).
    Fathead  minnows were equally sensitive to the  sodium  salts of selenlte
and  selenate,  while zebraflsh were  equally sensitive to  the  potassium and
sodium salts  of  selenlte  and selenate.   The  toxlclty  of selenium as  sodium
selenate to striped bass decreased nearly  10-fold as fish developed from the
prolarva  to  juvenile stage {Klauda,  1986).   Toxldty values  for  saltwater
fish  ranged  from a  96-hour LC5Q  of  0.6  mg/l for  haddock  larvae, Helano-
grammus   aegHflnus.   to    96-hour  LC5_s   of  14.2-15.1   mg/l  for   winter
flounder  larvae,  Pseudopleuronectes   ameMcanus.   exposed  to   unidentified
forms of selenium (Elsler,  1985).
    In a  series of  behavioral  and physiological studies, goldfish,  Carasslus
auratus. exhibited behavioral  Impairment  at 0.25  ppm (Weir and  Mine,  1970),
while  fathead  minnows,  Plmephales   promelas.  did  not   avoid  selenate  at
concentrations  of  0.3-11.2  mg  Se/l  (Watenpaugh   and  BeHlnger,   1985).
                                      v1

-------
Oxygen consumption by  fathead minnows  was  not affected by exposure to 60 mg
Se/8,  for  24 hours (Watenpaugh  and BeHlner, 1985).   Oxygen  consumption by
the  hermit  crab.Cllbanarlus  vlttatus.  exposed  to  100  ppm  selenium  was
depressed at 16°C  1n  10  o/oo  salinity test medium, although  oxygen consump-
tion  rates  were  generally  elevated at lower  salinities  and  higher tempera-
tures (Wolfenberger,  1987).
    Acute toxldty data for selenium  In aquatic  Invertebrates were available
for a total of 22  species,  Including  3 cladocerans, 2  copepods, 4 amphlpods,
4  decapods, 1  mysld,  3  Insects   and  5  molluscs.    The  48-hour  LC5Qs of
selenlte-selenlum  to  daphnlds  ranged  from 0.098-3.87 mg/a,  (Etnler  et  al.,
1987;  Reading  and  Bulkema, 1983).   The  48-hour  EC^s  for  unfed  and  fed
daphnlds exposed  to  sodium selenlte  were  0.47  and  1.5  mg/l,  repsectlvely.
The  48-hour NOEC  for  unfed  daphnlds  was  1.0 mg/8.  (Adams  and Heldolph,
1985).   The  96-hour   LC5Qs   for   amphlpods  ranged  from   2.88-6.17   mg/8,
(Etnler et  al.,  1987).  The toxldty  of  selenlte-selenlum to Daphnla magna
was  4-fold  lower  when assays  were conducted  In soft  water;  no  significant
differences were  observed  for  Chlronomus  plumosus.  based  on water hardness
(Mayer and Ellersleck, 1986).
    The toxldty of  selenium as  sodium selenate to 5th  Instar  Daphnla magna
(48-hour LC,-ns  of 0.55  and 0.75  ppm) was comparable to  that  observed for
sodium  selenlte  (Johnston,  1987).   In contrast,  the  toxldty  of  selenate-
selenlum  to Hyallela  azteca  (96-hour LC5Q  of   0.76 mg/l)  was  4- to  8-fold
greater than that  observed  with  sodium selenlte (Etnler et al.,  1987).  The
toxldty of selenium as selenium  dioxide  was  represented  by 96-hour LC   s
of  33,  1.90 and  0.25  mg/8.   for   the  crab,  Scylla   serrata.  bay  scallop,
Arqopecten  1rrad1ans.  and  surf  clam,   Splsula  sol1d1ss1ma.   respectively
(KMshnaja,  1987;  Nelson et  al.,   1988).   The  96-hour  LC5Q  of  selenium as

-------
selenium  dioxide  to  crabs,  oysters  and  mussels  was  >10 mg/a  (Martin  et
al.t  1981).   The  96-hour  LC5Q  for  the  midge, Tanytarsus  d1ss1m111s.  was
42.4  mg/a  (Elsler,  1985).   Exposure  of  embryos of  the zebraflsh,  Brachy-
danlo  rerlo,  to <10  mg/a   selenium  as  selenium  dioxide for  168 hours  did
not affect hatching and mortality.  Exposure of  zebraflsh  larvae  resulted In
significantly  higher  levels  of  mortality  at  concentrations  >3  mg/a  after
168 hours and at 10 mg/a after 96 hours  (N11m1  and  Laham, 1975).
    The  168-hour LC5Q  for  fathead minnows,  Plmephales  promelas.  was  2.9 mg
Se/a.   The  336-hour   LC5Qs   for  goldfish,  Carrasslus  auratus.   and  blue-
gills,   Lepomls  macrochlrus,  were   8.8   and   17.6   mg  Se/a,   respectively
(Cardwell  et  al.,  1976).   The  113-hour   LC™  for  embryos  and  the  7-day
LC5Q  for  tadpoles  of  the  frog,  Xenopus  laevls.  were  2.0  and 1.5 mg  Se/a,
respectively  (Elsler,   1985).   The 48-day LC5Qs  for  bluegllls,  L.  macro-
chlrus.  and   fathead  minnow,  P.  promelas.  were 0.4-2.0  and  1.1  mg  Se/a,
respectively.   The  10-day   LC™   for  perch,   Perca  flavescens.   was  4.8
mg/a,  while   the  43-day LC,n for  coho  salmon, Oncorhynchus  klsutch.  was
0.16  mg/a.   The 96-day LC5Q  for  rainbow  trout,  Salmo qalrdnerl. was  0.29
mg Se/a  (Elsler, 1985).
    Exposure  of Daphnla  maqna  to  0.2-0.8 mg selen1te-selen1um/a  for  28
days  did  not  affect  total  eggs  per  daphnld  or  live  young per  daphnld;
however, mean  brood  size was  affected  at  0.8  mg Se/a  (Reading and  Bulkema,
1983).   In  another  study,  the  7-,  14- and   21-day   EC5Qs  for  D.  maqna
exposed  to selenlte-selenlum  were  0.38,   0.38  and  0.35 mg/a,  respectively.
The  NOEC  based on  survival  and  reproduction  was  0.24  and  >0.24  mg/a,
respectively.   The  28-day  LC,.   for   D.  maqna  1n a third  study was  0.24 mg
Se/a  (Adams and Heldolph, 1985).

-------
    Exposure  of  the  amphlpod  Allorchestes  compressa  to  44  yg  selenlte-
selenlum/i  did  not  significantly   affect  growth  or  survival  of  treated
Individuals  after  4  weeks.   Animals  exposed  to  93 yg  Se/8,  experienced
reduced  growth  and a  significant  level  of mortality  over  the  same  period
(Ahsanullah  and  Brand,  1985).  The 14-day LC5Q  for  another  scud,  Hyallela
azteca. was 0.07 mg Se/8. (Elsler,  1985).
    Fathead  minnow  larvae,  Plmephales  promelas,   experienced  significant
reductions  In  final  dry  weight  when  larvae  were  fed rotifers  cultured  on
selenium-contaminated algae.   The  mean selenium  concentrations In  minnows
were  43 and  51.7  yg/g  In  two  experiments  (Bennett  et  al.,  1986).   High
tissue  concentrations of  selenium  1n blueglll  sunflsh,  Lepomls macrochlrus.
resulted  1n  significant  differences In percent fertilization  and hatch  when
the ovaries  of  females  contained  from 5.79-8.0 mg  Se/kg.   These levels  were
16-21  times  higher  than   those  1n  uncontamlnated sunflsh   (Glllesple  and
Baumann, 1986).
    Exposure of  early life stages of the  striped bass,  Morone saxatllls.  to
low  (0.089-0.099  mg  Se/l)  and  high   (1.217-1.360  mg Se/i)  concentrations
of  selenate-selenlum resulted  1n  a  significant  reduction   In  exponential
growth  rates for  fish   1n both  treatments for  days  3-15.   There were  no
differences  In  growth rates between days 19 and  60 but  there  was a signifi-
cant  Incidence  1n  developmental  abnormalities  of  the lower  jaw and  severe
blood  cytopathology (Klauda, 1986).   Exposure of  rainbow trout fry,  Sal mo
galrdnerl,  to  >47 yg  selenlte  selenium/a resulted  1n significant  reduc-
tions  1n survival  and  fry weight  and  length  after  90 days.   Survival  and
growth  were not  affected  at <21  yg Se/l.  BCFs  were Inversely related  to
exposure  concentrations  but did  not  exceed 200.    Exposure of  fry  to  >12  yg
Se/a  resulted  1n  significant  reductions   1n  the  calcium  concentrations  1n
                                      1x

-------
bone  (Hunn  et  al.,  1987).  Percent  hatch  of sheepshead minnows,  CypMnodon
varlegatus.  was  <4X  at  concentrations  of  <3.6  mg  selenlte-selenlum/i.
Juveniles experienced 4, 24 and 90X  reductions  In  survival  at  0.47,  0.97  and
>0.97  mg  Se/l,  respectively.   Growth   of  juvenile  sheepshead  minnows  was
reduced by 8X at 0.47 and 0.97 mg/l (U.S. EPA,  1987).
    Exposure  of  four  consecutive  generations  of  Cer1odaphn1a  afflnls   to
sodium selenlte resulted  1n a reduction 1n  the  tolerance of C.  afflnls over
succeeding  generations.   The  NOEL was  reduced  from  0.2-0.1  mg/l after  two
generations  (Owsley  and  McCauley,  1986).  There were  no  significant  differ-
ences  1n  growth of  Daphnla maqna  exposed  to <0.05  ppm  selenlte-selenlum or
In mean numbers of eggs  produced,  the percent  maturation  or mortality levels
for daphnlds exposed  to  <0.025 ppm selenlte-selenlum  (U.S. EPA,  1987).   The
number  of  offspring  produced  and  survival  of  first  generation  myslds,
Mysldopsls  bahla. exposed  to  selenlte-selenlum  was  significantly  reduced at
0.32  mg/i.   No  statistically  significant  differences   were  observed   In
myslds exposed to 0.14 mg/l (U.S. EPA,  1987).
    The  no-effect  level  for  dietary  selenium as  either  sodium  selenlte  or
selenomethlonlne  In  blueglll  sunflsh,  Lepomls  macrochlrus. over  324  days Is
-13-30  yg  Se/g.   The  maximum  BCF  for  blueglll   sunflsh,  L.  macrochlrus.
exposed  to  120 yg  Se  (as   selenous add)/l  for  28 days  was  20  and  the
tissue half-life during a  7-day  depuration  period  was between  1  and 7 (Woock
et al.,  1987).   BCFs for  rainbow  trout, Salmo  qa1rdner1.  exposed to  <10 yg
selen1te-selen1um/kg  ranged  from  3.1   1n  embryos  to  104  1n  livers   of
juveniles.   BCFs  for  trout  exposed to  ~100 yg selen1te-selen1um/kg  ranged
from  1.6  1n sac-fry  to  31.6 1n livers of Juveniles after 96 hours (Hodson et
al.,  1986).  The  BCF  for selenium  as  sodium selenlte  In   carp,  Cvprlnus
carplo,  for an  unspecified exposure duration  and  selenium  concentration  was

-------
0.6-6.0.   The BCF  for  selenlte-selenlum  1n  minnows,  Plmephales  promelas.
exposed to 83 yg/8. for 28 days was 4443 (Etnler et al..  1987).
    Exposure  of   fathead  minnows,   Plmephales   promelas.   to   waterborne
selenate-selenlum (10-40  ng  Se/ml)  resulted  In a maximum body burden  level
of  -0.5-0.8  yg Se/g  fish after  20-30  days.  Selenium  was  depurated  to  ~3
yg/g after  30 days.   Exposure  of minnows to  selenium-contaminated  daphnlds
(1.33-7.32  yg Se/g)  resulted  In  selenium  body  burden  levels  of  -0.3-1.2
yg  Se/g fish after  80 days.   Depuration  of  selenium  was  slower  than  that
observed  1n  fish  exposed  to  waterborne  selenium.   Fish  exposed  to  both
sources of  selenium  had  body-burden levels of selenium  that did  not plateau
during  the  56-day exposure  phase,  reaching 0.4-2.0  yg Se/g  fish.   Depura-
tion of  selenium  from these  fish  was  slow  (Bertram and  Brooks,  1986).   Fed
and starved juveniles  of  the  striped bass,  Horone saxat111s. exposed to 1.29
mg  selenate-selen1um/a  for  60  days   accumulated  selenium  at   comparable
rates,  producing  BCFs of 0.68  and 0.69.   Fed  juveniles exposed  to  90  yg
Se/a  showed  no  Increase  1n  whole body  levels  of  selenium,  while starved
juveniles  had a  BCF  of   11.78  (Klauda,  1986).   The  half-lives  In  days  for
selenate,  selenlte  and selenometh1on1ne administered orally via  gavage at a
level of 20 ng Se/g  to fathead  minnows,  £. promelas. ranged from 3.9 (liver)
to  487  (adipose  tissue),  2.2  (liver)   to  64  (heart) and 1.0  (liver)  to  69
(adipose tissue)  (Klelnow and Brooks,  1986).
    Mussels   exposed   to  50   yg   selenium/4  for   15-50  days   accumulated
selenlte-selenlum at  a rate of 0.12 ng Se/g/day.   The  presence  of  Inorganic
(30  yg  Hg/a) and  organic   (3  yg Hg/a)  mercury  Increased  the uptake  of
selenlte-selenlum  to  0.24  and  0.40   ng   Se/g/day,  respectively.   Organic
selenium   (CQH,00Se)0 accumulated  at   a   rate  of  0.15   ng  Se/g/day  In
            Old  C
the presence  of mercury  (Pelletler, 1986).
                                      x1

-------
    There were no apparent effects observed In growth of  natural assemblages
of  phytoplankton  or  In  growth  of  a species  Isolated  from  field  samples
Thalassloslra aestevalls  exposed to  <1000 nM selenium (Holllbaugh et al.,
1980).  There were  no effects on  growth  of either  the green  or red  algae,
Dunallella   pMmolecta   and   Porphyr1d1um  cruentum.   exposed   to  10   ppm
selenHe-selenlum (Gennlty et  al., 1985).  The  48-hour LC5Q for  Oedogonlum
cardlacum  was  <0.1   mg/a,  while  the 96-hour  LC™  for Anabaena  var1ab111s
was  15-17  mg/l   (Elsler,  1985).   The  96-hour  EC5Q  for  duckweed,   Lemna
minor, exposed to selenium was  2.4 mg/B,  (Wang,  1986).
    Population growth  of  Tetrahymena pyMformls  was   Inhibited by 1.4  ppm
selenium  and  stopped  completely  at 140  ppm  (Tang  et  al.,  1985).    T.
pyrlformls  also   experienced   a  dose-dependent  Inhibition  1n  division   of
synchronized cells  exposed to  10,  20,  50 and  100 ppm  selenium   (as  SeO?)
(Cao  and  Tang,  1985).   Growth of  cultures  of  the  algae,  Chlorella  vulgarls
and Phorm1d1um foveolarum. was reduced by  -40%  In the presence of  0.25  ppm
selenate-selenlum.  Growth of  £.  vulgarls  1n  a 4.0  ppm solution of selenium
was 97%  of that  observed  1n controls and completely Inhibited 1n  P.  foveo-
larum  (Trlpathl and  Pandey,  1985).   Selenium  retarded  growth  of the  marine
dlnoflagellate,  Prorocentrum mlcans.  1n cultures Incubated with 100 and 1000
ppm selenium.   Growth was slightly  enhanced  at 10  and 50 ppm selenium  for
the first  15 days of treatment but was Increasingly Inhibited compared with
controls  after 15 days.   Growth  of the marine dlnoflagellate, Crypthecodln-
1um cohnll.  was  severely  Inhibited  by  selenium at >10 ppm selenium  within
1.5 days of the Initiation of  treatment  (Prevot and Soyer-Goblllard, 1986).
    Concentrations  of  selenium  1n  soil  and  earthworms  ranged  from  trace
levels  to 1.3  mg/kg at  an Industrial site and from trace quantities  to  7.6
mg/kg,  respectively  (Bayer and  Cromartle,  1987).   In  an  experimental  field

-------
study,  the  concentration of  selenium  1n  earthworms,  Aporrectodea  tubercu-
lata. Aporrectodea turglda. and Lumbrlcus  rubellus.  from  control  plots  (<0.1
mg Se/kg  soil)  was  16 mg/kg,  while concentrations of  selenium  1n  worms  from
treated plots were 36, 43, 51, 36  and  78  mg/kg.   Uptake  of selenium by  worms
was  not  Influenced  by pH  or  other  soil  variables  (organic matter  content,
phosphorous,  potassium  or  magnesium)  (Beyer  et al.,   1987).   The 8-hour
LD5Qs  for   selenlte   and  selenate   In  earthworms,  Lumbrlcus   terrestMs.
treated  Intraperltoneally  were  31  and  60  mg/kg,  respectively   (Serda  and
Furst, 1987).
    Exposure  of  American  coot,   Fullca   amerlcana,  mallard,   Anas platy-
rhynchos.  northern  pintail,  Anas  acuta.  cinnamon  teal,  Anas  cyanoptera.
gadwall,  Anas  strepera.  black-necked stilt, Hlmantropus  mexlcanus.  American
avocet,  Recurvlrostra  amerlcana.  and  eared  grebe,  Podlceps nlgrlcolHs.  to
selenium-contaminated   Irrigation dralnwater  ponds (~300  ppb Se)  resulted  1n
an  Increase  1n  reproductive  Impairment.   Frequency  of  dead  and  abnormal
embryos  ranged  from 2.5  (ducks)  to 31.7%  (grebes)  and  4.0 (ducks)  to  8.8%
(coots),  respectively.   Overall,  19.6% of  the 347  nests  monitored  produced
at least  one  embryo or chick with an  abnormality.   There were  no abnormali-
ties  In  embryos of  birds from  92  nests  1n  an  uncomtamlnated  area over  a
2-year period.  Average  selenium concentrations  In bird  livers  and eggs  from
nests  1n  contaminated areas  ranged  from  9.1-81.4 ppm dry weight,  compared
with  4.1-6.1  ppm   1n  livers  of   birds  from  an  uncontamlnated  area.    The
no-effect level  for dietary  selenium as  sodium  selenlte  or selenometh1on1ne
1n  mallard  ducklings,  Anas   platyrhynchos.  1s  ~10  ppm  (Ohlendorf  et  al.,
1986).
    Concentrations of  selenium In  water  and  sediment  from two bays on  the
Gulf  Coast  of  Texas were 0.11  and  1.44 mg/kg.   Concentrations  of selenium 1n
                                     xlll

-------
barnacle, Balanus  aburneus.  crab,  CalUnectes sapldus.  oyster,  Crassosstrea
v1rq1n1ca. clam, Ranqla cuneata. and  polychaete,  Nereis  sp.,  from those bays
were 0.77, 0.08, 0.14, 0.54  and  0.49  mg/kg  wet weight,  respectively (Guthrle
et  al.,  1979).   The  level  of selenium  In  algae, Euglena sp.,  from  tailing
areas of  the  Elliot  Lake  mining district,  Canada, was  2700  ng/g dry  weight,
compared  to  a concentration of  0.2 ng/g 1n world river water  (Mann  et al.,
1988).   Selenium  concentrations  1n  water,  sediment,  and  clam valves  and
viscera of Asiatic clams, Corblcula flumlnea.  from a  control  site In  the New
River, Virginia, were 0.11,  0.88, 0.29  and  3.90  ppm,  respectively.   Selenium
concentrations  In  samples from  sites receiving  thermal  effluent discharges
were 0.10, 0.60, 0.50 and 16.5 ppm,  respectively (Rodgers et  al., 1980).
    The  concentrations  of   selenium  In  water  and   sediment  1n  the  upper
Mississippi  River  were   below   detection   (1  yg/i  and  ~0.2  yg/g,  respec-
tively)  but  concentrations  of  selenium  In fillets,  livers  and kidneys  of
common  carp   ranged  from  0.161-0.356,   0.858-2.17  and  0.943-1.62  yg/g  wet
weight, respectively.  Concentrations of selenium 1n  fillets  from smallmouth
bass  and sauger  ranged   from  0.36-0.425 and 0.284-0.369 yg/g   wet  weight,
respectively.   Concentration factors  for   the edible  tissue  of  common carp
ranged from 322-712 (Boyer, 1984).
    Concentrations of  selenium  1n  surface  and  bottom  waters  of  a  cooling
water  reservoir ranged from 
-------
levels >5  yg/g wet weight  after  20 weeks  (Woock  and  Summers, 1984).   In  a
separate study,  concentrations  of  selenium 1n the  water  column of  a  power
plant cooling  reservoir  were  20-30 times higher than  background  levels  with
a  mean   value  of  10  yg/l,  while  concentrations   1n  flora  and fauna  were
10-15 times  background  levels. Concentrations  of  selenium 1n  surface  sedi-
ments were  ~4 yg  Se/g wet weight.   Selenium was  519- and 3975-fold higher
1n perlphyton and fish, respectively, than 1n water (Lemly, 1985a).
    Concentrations  of  selenium 1n  water from  utility wastewater  treatment
basins  were   7.0,   3.0,  <2.0,   <2.0  and  <2.0  yg/l.   Concentrations  of
selenium  In  black  crapple,  Pomoxls  nlgromaculatus.  pumpklnseed,  Lepomls
aurltus.   brown  bullhead,  Ictalurus  nebulosus.  and  carp, Cyprlnus  carplo.
from  these  basins  ranged  from 2.7-37.6  mg/kg (Skinner,  1985).   Concentra-
tions of selenium  1n carcasses of  bluegllls,  Lepomls macrochlrus.  and large-
mouth bass, Mlcropterus  salmoldes  from cooling  water reservoirs that receive
ash  pond  effluent  ranged  from  4-9  ppm.   Selenium concentrations  1n gonads
varied  between  sexes  within  reservoirs  with  significantly  higher  levels
(<2-fold)  In  ovaries  (<1  to  -12  ppm)  than  In  testes  (<1  to  ~7  ppm).
Selenium concentrations  In carcasses of bluegllls  and  bass  from  a  cooling
water reservoir  that did not  receive ash pond effluent was ~1 ppm.   Selenium
concentrations  1n carcasses  of  bluegllls  and  bass  from  a  municipal  water
reservoir were <0.5 ppm  (Baumann and Glllesple, 1986).
    Concentrations  of   selenium   1n oysters,   Crassostrea  vlrqlnlca.   and
sediment from  a  site In  Lake Pontchartraln,  Louisiana,  were  0.013  and 0.007
yg/g  dry   weight,   respectively.    Concentrations   of  selenium  In  clams,
Rangla cuneata.  and sediment  from a second  site  were 0.032  and  0.031  yg/g
dry  weight, respectively.   Concentrations  of  selenium  1n clam tissue and
sediment  from a  third  site  were  0.041   and 0.05  yg/g  dry  weight,  respec-
tively (Byrne and DeLeon, 1986).
                                      xv

-------
    A  selenium  concentration of  14 yg/g  was  measured at  the northern  end
of Lake Macquarle, New  South Wales, but the concentration decreased  rapidly
to  a  plateau  level   of 4  yg/g.    Concentrations  of  selenium In  seagrass,
Zostera capMcornl. from the northern and  southern sections of the  lake  were
~3.9  and   ~0.6  yg/g,  respectively.   Concentrations  of  selenium  1n algae
Enteromorpha sp. were higher  In  samples collected from the northern  section
of  the lake,  ranging  from  0.3-1.6  yg/g.   Whole  body  concentrations  of
selenium In  mussels,  Trlchomya  hlrsuta. and cockles,  Anadara  trapezia,  from
the  northern  end  of   the   lake  were  3.3  and  6.4  yg  Se/g,  respectively
(Bailey, 1987).
    Phytoplankton from Xiamen Bay,  Fujlan  Province,  China contained  1.24 ppm
selenium,   while  seaweed  concentrations   ranged  from  0.08-0.61  ppm.    The
concentration of selenium 1n zooplankton ranged  from 2.16-6.30 ppm.   Concen-
trations  of selenium  In other   Invertebrates  and  fish  were  approximately
equal to those found  In zooplankton  on  a dry weight  basis  but  concentrations
were dependent  on the tissue analyzed (L1u  et al.,  1987).
    Selenium concentrations  1n  the  top 10  cm of sediment  from 14  Ontario
lakes  ranged from <1-16  yg/g.   Concentrations  of  selenium  1n  tissues  of
lake  trout, whHeflsh,  common   sucker,   yellow  perch,  northern  pike   and
walleye from the  study  lakes were  0.78, 0.84,  0.55, 0.38,  0.37 and  0.25 yg
wet we1ght/g, respectively (Johnston, 1987).
    Concentration  ranges  for selenium  1n   water  and  sediment  from  the  San
Luis Drain,  Kesterson ponds  and Volta waterways  In California  were  0.29-0.33
and  65-100,  0.009-0.32  and  1.8-67,  and 0.0002-0.0014 and <0.2-0.5,  respec-
tively.   Concentration  ranges  for   selenium In algae and net plankton  In
these  systems  were  63-72 and  not  measured,  12-330  and  58-120,  and below
detection-!.4 and  1.4-2.8,  respectively.   Concentration ranges  for  selenium
                                      xvl

-------
In all  aquatic  Insects and  mosquHoflsh  In  these  systems  were 170-330 and
140-370,  16-290  and 104-290,  and 0.68-3.0 and  1.2-1.4,  respectively.  The
Investigators noted  that  selenium  concentrations  generally  Increased  from
water to sediments to plants  to animals  (Salkl and Lowe, 1987).
    Mean  selenium  concentrations  In  kidney,  liver and muscle  of  dolphins,
Laqenorhynchus  alblrostrls.  were  5.85,  8.15  and 1.91  mg/kg  dry weight,
respectively.  Selenium concentrations  1n  blubber, kidney,  liver and  muscle
of pilot  whales,  Globlcephala melaena, from  two different  collection  sites
were 0.49  and 0.59  (mg/kg  wet weight), 13.0 and  11.3,  50.5  and  31.4, and
1.22 and 2.94 mg/kg dry weight, respectively (Mulr et  al., 1988).
    BCFs for phytoplankton, perlphyton and plants exposed to selenium  1n the
field  ranged  from  237-1320,  158-1070,  and  166-24,400,  respectively.    BCFs
for  zooplankton,   Insects,  annelids,  crustaceans  and  molluscs  exposed  to
combined waterborne and dietary sources of  selenium under natural conditions
1n  the   field  ranged  from   176-2080,  371-5200,  770-1320,   420-1975  and
600-2550,  respectively.   BCFs  for carnivorous,  planktlvorous and omnivorous
fish  1n  the  field  ranged  from  590-35,675,  445-27,000   and  364-23,000,
respectively (Lemly, 1985b).
    Addition  of  selenium  to  field mesocosms  resulted  1n  a  replacement  of
chrysophytes  by  chlorophytes  In low  dose  enclosures  and  cyanophytes  1n the
high  dose enclosure.   There  were  no  significant  differences  between  zoo-
plankton  communities  among  1, 10  and 100  yg  selenlte-selenlum  treatments
(Salkl et al., 1985).
    Studies  1n dogs  (Welssman  et  al.,  1983)  and  rats  (MecMnsky  et al., 1981)
Indicate  that selenium metal and  selenlous  add  are absorbed  readily follow-
ing  Inhalation  exposure.   Gastrointestinal  absorption and b1oava1lab1l1ty  of
selenium  Is  greater with  organic  selenium compounds  (e.g., SeMet) compared
                                     xv11

-------
to  Inorganic  compounds.   Among  Inorganic  selenium compounds, selenltes  and
selenates  are  absorbed  more  readily  than  metal  selenldes  and  elemental
selenium (Venugopal and Luckey, 1978).
    Following absorption,  selenium  Is distributed throughout  the  body,  with
higher  levels  found In  the  liver and  kidneys.   McAdam and  Levander  (1987)
found higher muscle  selenium  levels  In rats  fed  diets  with  organic selenium
compounds (D- or L-SeHet) compared to rats fed sodium selenate or  selenlte.
    Reduction  of   selenium  compounds  to   hydrogen   selenlde  followed  by
methylatlon Is  the  major  transformation pathway leading to  the excretion of
selenium  (Mushak,   1983).   Methylatlon occurs  predominantly  In  the  liver.
Tr1methylselenon1um, excreted  1n the  urine,  1s  the major excretory  product
at  low  doses  of  selenium.   At higher  doses  of  selenium,  larger  amounts of
dlmethylselenlde  are  produced.   Olmethylselenlde   Is  exhaled  through  the
lungs.   Metal  selenldes  and   metal-protein  selenlde compounds  may also be
formed from hydrogen selenlde.
    Selenium  1s  an essential   element  that  1s a  part of glutathlone  peroxl-
dase.   The  dally  safe and  adequate  level  of selenium  Intake for  adults 1s
considered  to be  50-200 yg   (0.05-0.2 mg)   (NAS, 1980).   Selenium 1s  also
quite toxic,  with  selenosls  (brlttleness of  nails,  loss of  nails  and hair,
dermatitis, nervous  symptoms)  reported  In  humans following  chronic  dietary
Intake of 3.2-6.69 mg/day (Yang et al., 1983).
    Adverse  effects   reported  1n   animals  treated   orally  with  selenium
compounds  Include  effects  on  growth,  reduced  survival,   hlstopathologk
changes  1n the  liver and  testes  and  Immune  system effects.   Subchronlc NOELs
that  have  been  Identified  In  laboratory  animals are  0.16  mg Se/kg/day of
sodium  selenlte  or selenlferous wheat  1n  rats (Halverson et  al.,  1966) and
0.63 mg  Se/kg/day  of sodium selenlte 1n hamsters  (Beems and  van Beek,  1985).
Limited  chronic studies of selenium compounds do not Identify NOELs.
                                     XV111

-------
    Selenium toxlclty  following  Inhalation  exposure  has  not  been studied In
animals.   Symptoms  reported  In  humans  following  occupational  exposure to
selenium Include a  strong  garlic  odor  1n the breath, sweat and urine, acute
sore  throats  and cold-Uke  symptoms,  gastrointestinal  Irritation,  lacrlma-
tlon,  and  a metallic  taste  In the  mouth  (Hamilton and  Hardy,  1974).   The
development of  garlic breath  and  cold-Uke  symptoms  Is  thought  to  result
from the hepatic production of dlmethylselenlde which 1s exhaled through the
lungs  (01sk1n et al.,  1979).
    CarclnogenlcHy  studies  of  selenium  are  not  conclusive.   Selenium
sulflde has  tested  positive  for  cardnogenlclty In  rats  and  female mice In
an NCI/NTP  (1980a)  bloassay.   However,  the  Identify  of  the test compound 1s
unclear and  thus  the  study Is judged  Inadequate for derivation  of  a cancer
potency  factor  for  selenium  sulflde.   Nelson  et  al.  (1943)  reported
Increased tumor Incidences In rats treated orally with selenlferous  corn and
wheat  (diets  were  suboptlmal  In protein).   Schroeder  and MHchener  (1971a)
also reported  tumor Incidences 1n  rats  treated with  sodium  selenate;   how-
ever,  because  treated  rats lived longer than  controls,  the  tumor  Incidence
may not be  related  to  the  cardnogenlclty of  the compound.  Negative  results
were reported 1n rats treated with sodium selenate or selenlte (Harr  et  al.,
1967;  Tlnsley  et al.,  1967)  and  1n mice  treated  with  sodium  selenate or
selenlte (Schroeder and MHchener,  1972).
    Epidemiology studies  (Wlllett and  Stampfer,  1986;  Salonen  et al., 1984,
1985;  Kok  et al.,  1987)  and studies  1n  animals  (Shamberger,  1985)  suggest
that selenium has  antlcardnogenlc activity.   This  effect may  be due to the
ability of  selenium to protect  against  cellular damage  by  peroxldatlon of
fat  (Shamberger, 1985), or 1t  may be a  result  of  the  effect  of selenium on
the Immune system (Koller  et al.,  1986).
                                     x1x

-------
    Selenium  1s  teratogenlc to  livestock  (Barlow and  Sullivan,  1982), but
developmental  effects   reported   1n   laboratory  rodents  are  limited  to
decreased  fetal  body  weight  (Nobunaga  et  al., 1979).   A  limited  multi-
generation study  (Schroeder  and Kitchener,  1971b)  found that  by  the  third
generation,  CD-I  mice  treated with  selenate In  the  drinking  water  at 0.57
mg/kg/day failed to breed,  or produced  a  large  proportion  of runts.
    The  evidence   that   selenium  Is an  animal  carcinogen  1s conflicting.
Ep1dem1olog1cal data  did  not  suggest  that  selenium  1s  a   carcinogen  to
humans.  Mutagenldty data  were mixed,  but  there Is  evidence that  certain
compounds of selenium are mutagenlc and clastogenlc.
    There were  no data   regarding  the  cardnogfenldty of selenium compounds
In  humans  and the  data  In  laboratory animals  for  selenium  are considered
Inadequate.  Therefore,  selenium,  based  on  U.S.  EPA classification  scheme,
1s  considered  a Group  D substance:  not  classifiable as to cardnogenldty
In  humans  (U.S.   EPA,   1986b).   Neither  cancer  potency  estimates  nor  a
cancer-based  RQ  were derived  for  selenium.  However,  because  of  positive
evidence of  cardnogenldty  1n both rats and  female  mice (NCI/NTP,  1980a),
selenium sulflde,  based  on the U.S. EPA  (1986b)  classification scheme,  could
be considered as Group  B2 - probably carcinogenic to humans.
    Inhalation data were Inadequate for  derivation  of RfD values  for sub-
chronic or  chronic  Inhalation  exposure.   An oral  RfD of 0.003 mg Se/kg/day
was derived  by  applying  an  uncertainty  factor  of 10  and a modifying factor
of  1.5 to  the LOAEL  of 3.2  mg/day  associated  with selenosls In  a   high-
selenium region 1n  the Peoples Republic  of  China (Yang et al., 1983).  This
derivation conforms to the  derivation  of verified RfD values  for selenourea
and  selenlous  acid  available  on  IRIS  (U.S.  EPA,  1985c,d).   The  RfD for
                                      xx

-------
selenium of 0.003 mg/kg/day was adopted for both subchronlc and chronic oral
exposure.   An  RQ  of  10  for   chronic  (noncancer)  toxldty  was  based  on
selenosls  (severe  CNS disturbances)  1n  humans  In the  Yang  et  al.  (1983)
study.
                                     xxl

-------
                              TABLE  OF  CONTENTS

                                                                       Page

1.  INTRODUCTION	1-1

    1.1.   STRUCTURE AND CAS NUMBER	1-1
    1.2.   PHYSICAL AND CHEMICAL PROPERTIES 	   1-1
    1.3.   PRODUCTION DATA	1-1
    1.4.   USE DATA	1-6
    1.5.   SUMMARY	1-13

2.  ENVIRONMENTAL FATE AND TRANSPORT	2-1

    2.1.   AIR	2-1
    2.2.   WATER	2-3

           2.2.1.   Chemical Speclatlon 	   2-3
           2.2.2.   Photolysis	2-4
           2.2.3.   Blomethylatlon and Volatilization 	   2-4
           2.2.4.   Sorptlon	2-5
           2.2.5.   B1oconcentrat1on	2-5

    2.3.   SOIL	2-6

           2.3.1.   Leaching	2-6
           2.3.2.   Blotransformatlon 	   2-6

    2.4.   SUMMARY	2-7

3.  EXPOSURE	3-1

    3.1.   WATER	3-2
    3.2.   FOOD	3-4
    3.3.   INHALATION	3-9
    3.4.   DERMAL	3-10
    3.5.   SUMMARY	3-10

4.  ENVIRONMENTAL TOXICOLOGY	4-1

    4.1.   AQUATIC TOXICOLOGY 	   4-1

           4.1.1.   Acute Effects on Fauna	4-1
           4.1.2.   Chronic Effects on Fauna	4-12
           4.1.3.   Effects on Flora	4-25
           4.1.4.   Effects on Bacteria 	   4-27

    4.2.   TERRESTRIAL TOXICOLOGY 	   4-27

           4.2.1.   Effects on Fauna	4-27
           4.2.2.   Effects on Flora	4-36

    4.3.   FIELD STUDIES	4-36
    4.4.   AQUATIC RISK ASSESSMENT	4-43
    4.5.   SUMMARY	4-44
                                    xx11

-------
                         TABLE  OF  CONTENTS  (cont.)

                                                                       Page
5.  PHARMACOKINETCS	5-1

    5.1.   ABSORPTION	5-1
    5.2.   DISTRIBUTION	5-7
    5.3.   METABOLISM	5-10
    5.4.   EXCRETION	5-12
    5.5.   SUMMARY	5-14

6.  EFFECTS	6-1

    6.1.   SYSTEMIC TOXICITY	6-1

           6.1.1.   Inhalation Exposure 	   6-1
           6.1.2.   Oral Exposure	6-2
           6.1.3.   Other Relevant Information	6-11

    6.2.   CARCINOGENICITY	6-14

           6.2.1.   Inhalation	6-14
           6.2.2.   Oral	6-14
           6.2.3.   Other Relevant Information	6-16

    6.3.   MUTAGENICITY	6-20
    6.4.   TERATOGENICITY	6-20
    6.5.   OTHER REPRODUCTIVE EFFECTS 	   6-25
    6.6.   SUMMARY	6-26

7.  EXISTING GUIDELINES AND STANDARDS 	   7-1

    7.1.   HUMAN	7-1
    7.2.   AQUATIC	7-2

8.  RISK ASSESSMENT	8-1

    8.1.   CARCINOGENICITY	8-1

           8.1.1.   Inhalation	8-1
           8.1.2.   Oral	8-1
           8.1.3.   Other Routes	8-2
           8.1.4.   Weight of Evidence	8-2
           8.1.5.   Quantitative Risk Estimates 	   8-3

    8.2.   SYSTEMIC TOXICITY	8-4

           8.2.1.   Inhalation Exposure 	   8-4
           8.2.2.   Oral Exposure	8-5
                                    XX111

-------
                           TABLE  OF  CONTENTS  (cont.)

                                                                        Page
 9.  REPORTABLE QUANTITIES 	   9-1

     9.1.   BASED ON SYSTEMIC TOXICITY 	   9-1
     9.2.   BASED ON CARCINOGENICITY	   9-6

10.  REFERENCES	10-1

APPENDIX A: LITERATURE SEARCHED	   A-l
APPENDIX B: SUMMARY TABLE FOR CROTONALDEHYDE 	   B-l
                                     xxlv

-------
                               LIST OF TABLES
No.                               Title                                Page
1-1     Synonyms, CAS Numbers, Molecular Heights, Empirical
        Formulas and Structures of Selenium and Compounds 	  1-2
1-2     Physical Properties of Selenium and Compounds 	  1-4
1-3     1977 Production Data for Selenium and Compounds 	  1-7
3-1     Average Dally Intake (yg/day) of Selenium 1n Fiscal
        Years 1978-1981/1982	3-3
3-2     Selenium 1n the Adult American Diet for Fiscal Years
        1981-1982	3-5
3-3     Selenium 1n the Infant American Diet for Fiscal Years
        1981-1982	3-6
3-4     Selenium In the Toddler American Diet for Fiscal Years
        1981-1982	3-7
3-5     Average Selenium Content of Some Foods In the American
        Diet	3-8
4-1     Median Response Concentration for Aquatic Vertebrates
        Exposed to Selenium	4-2
4-2     Median Response Concentration for Aquatic Invertebrates
        Exposed to Selenium	4-7
4-3     Metal Concentration 1n Soil and Earthworms from
        Contaminated and Natural Sites	4-28
5-1     Calculated Internal Absorption of 75Se 1n Rats Expressed
        as Fraction of Administered Dose	5-3
5-2     Calculated Internal Absorption of 7SSe 1n Rats Expressed
        as Fraction of Administered Dose	5-4
5-3     Concentrations of Selenium In Tissues from Rats Fed Diets
        Containing Added D-Selenometh1on1ne (SeMET), L-SeMet,
        Selenlte or Selenate	5-9
6-1     Selenium Levels In Hair, Blood and Urine of Residents of
        High- and Adequate-Se Areas of China	6-8
6-2     Safe and Adequate Ranges of Dally Selenium Intake 	  6-12
6-3     Acute Oral Toxlclty of Selenium Compounds 	  6-13
                                     xxv

-------
                           LIST  OF TABLES  (cont.)
No.                              Title                                Page
6-4     Incidence of  Tumors 1n F344  Rats and 86C3F1 Mice Treated
        by Gavage with  Selenium  Sulflde for 103 Weeks	6-17
6-5     Mutagenldty  Testing of  Selenium Compounds	6-21
9-1     Toxldty Summary  for Oral Exposure to Selenium	9-2
9-2     Composite Scores  for Selenium  	  9-5
9-3     Selenium: Minimum Effective  Dose (MED) and Reportable
        Quantity (RQ)	9-7
                                    xxvl

-------
                             LIST OF ABBREVIATIONS
AADI                    Adjusted acceptable dally Intake
AIC                     Acceptable  Intake chronic
ATP                     Adenoslne tMphosphate
AWQC                    Ambient Water  Quality CMtlera
BCF                     Bloconcentratlon factor
B-G                     Beta glucuronldase
CAS                     Chemical Abstract Service
CMC                     Carboxymethylcellulose
CS                      Composite score
DNA                     Deoxyrlbonuclelc acid
ECso                    Concentration  effective  to 50%  of  recipients
                        (and all other subscripted concentration  levels)
ECG                     Electrocardiogram
GGT                     Gamma-glutamyl transpepsldase
                        Concentration  lethal to  50% of  recipients
                        (and all other subscripted dose levels)
                        Dose lethal to 50% of recipients
LDH                     Lactate dehydrogenase
LOAEL                   Lowest-observed-adverse-effect  level
LTso                    Exposure time  lethal to  50% of  recipients
MATC                    Maximum allowable toxicant concentration
MED                     Minimum effective dose
NK                      Natural killer
NOAEL                   No-observed-adverse-effect level
NOEC                    No-observed-effect concentration
NOEL                    No-observed-effect level
                                     xxvll

-------
                         LIST  OF ABBREVIATIONS  (cent.)
PEL                     Permissible exposure limit
ppb                     Parts  per  billion
ppm                     Parts  per  million
RBC                     Red blood  cell chollnesterase
RfD                     Reference  dose
RMCL                    Recommended maximum contaminant level
RQ                      Reportable quantity
RV(j                     Dose-rating value
RVe                     Effect-rating value
SDH                     SorbHol dehydrogenase
TLV                     Threshold  limit value
TWA                     Time-weighted average
                                    XXV111

-------
                               1.   INTRODUCTION
1.1.   STRUCTURE AND CAS NUMBER
    The  CAS   Registry   numbers,   molecular  weights,   empirical   formulas,
structures  and  common synonyms for  selenium and Us compounds are  presented
In Table 1-1.
1.2.   PHYSICAL AND CHEMICAL PROPERTIES
    The physical  properties  of selenium and selected selenium compounds  are
listed 1n Table  1-2.  The  Important oxidation states of selenium are  -2,  0,
+2,  +4  and +6.   There   1s  no  evidence  that  the +2  state  occurs   In  nature
(Elkln, 1982).
    Selenium  1s  positioned  between  sulfur and .tellurium 1n group 6A  on  the
periodic  table  of  elements.   The  chemical   properties  of  selenium  salts
resemble 1n behavior  those  of the  corresponding sulfur  and tellurium salts
(Elkln, 1982).   Selenium can  act  as  both an  oxldant  and  reductant In many
reactions  because of  Us  variable  oxidation states.   Strong  oxldants  can
convert  selenium dioxide (+4  oxidation  state)  and Us  derivatives   to  the
hexavalent (*6) state (Elkln, 1982).
    The selenium halldes and  oxyhalldes  hydrolyze  upon  contact with  water,
and  are  therefore  not   stable 1n  moist  environments   (Elkln,  1982).   The
organic  selenium  compounds  are  frequently   susceptible  to  photolysis   or
oxidation when exposed to light or  air.   This  Instability  1n  light  or  air  Is
characterized  by the selenium compound turning red  (Elkln,  1982).
1.3.   PRODUCTION DATA
    In  the United  States,  selenium  Is  produced  by  recovering  It from
by-products generated during the electrolytic  refining of  copper ore  (Elkln,
1982;  USDI,   1982,  1988).    Copper  refinery anode  slimes  contain  selenium,
0141d                               1-1                              03/30/89

-------
                                                     TABLE  1-1


       Synonyms.  CAS Numbers.  Holecular  Heights,  Empirical  Formulas  and  Structures  of  Selenium and Compounds
Chemical/Synonyms
Selenium (Element)
Hydrogen selenlde
selenium hydride


Selenlc acid

Selenlous acid
monohydrated selenium dioxide
selenous acid
Selenium bromide
dlselenlum dlbromlde
Selenium chloride
dlselenlum dlchlorlde
Selenium dlethyldlthlocarbamate
CAS Number
7782-49-2
7783-07-5


7783-08-6


7783-00-8
7789-52-8

10025-68-0
5456-28-0
Holecular
Weight
78.96
80.98


144.98


128.98
317.73

228.83
672.08
Empirical Formula
Se
H2Se


H204Se

•
H203Se
Br2Se2

Cl2Se2
CpOHiONiSeSft
Structure
Se
H-Se-H
0
II
HO-Se-OH
II
0
0
II
HO-Se-OH
Br-Se-Se-Br

Cl-Se-Se-Cl
CHpCHi
     tetrakls(dlethyldlthlocarbamato)
     selenium ethyl selenac
GO
CO
o
oo
Se-Sc-N
                                                                                                     \
                                                                                                      CH2CH3

-------
                                                   TABLE 1-1 (cont.)
Chemical/Synonyms
Selenium dioxide
selenium oxide
selenium anhydride
selenium (IV) dioxide
selenlous acid anhydlrlde
Selenium dlsulflde
Selenium oxychlorlde
selenlnyl chloride
_, selenlnyl dlchlorlde
^ selenium chloride oxide
selenium oxydlchlorlde
Sodium selenate
selenlc acid, sodium salt
Sodium selenlte
selenlous acid, sodium salt
CAS Number Molecular Empirical Formula
Weight
7446-08-4 110.96 02Se
7488-56-4 143.08 SeSe2
7791-23-3 165.87 Cl2OSe
13410-01-0 188.94 Na2Se04
10102-18-8 172.95 Na2Se03
Structure
0=Se=0
0=Se=S
Cl-Se-Cl
II
0
0
II
Na-0-Se-O-Na
II
0
0
II
                                                                                                  Na-0-Se-O-Na
o
CO
CO
o
CD

-------
                                                                                TABLE 1-2

                                                              Physical Properties of Selenium and Compounds
Chemical
Selenium
(element)
Hydrogen
selenlde
Selenlc
acid
Selenlous
acid
Selenium
bromide
Selenium
chloride
Selenium
dlethyl-
dlthlo
carbamate
Selenlous
dioxide
Selenium
dlsulflde
Selenium
oxychlorlde
Description
red amorphous powder or
colloidal crystal to
gray-black crystal*-'*
gas with a disagreeable
odor6
white hygroscopic
solid5
deliquescent prisms6
dark red liquid with
unpleasant odor6
deep red oily liquid6
orange-yellow powder b

lustrous tetragonal
needles6
bright red-yellow
compound
nearly colorless or
yellow compound*1
Neltlng
Point
217"Ca
-65.73°CC
58"CC
decomposes
at 70°Cd
NA
-85-C*:
63-71 °Cb

340°CC
<100°Cd
5eCe
Boiling
Point
685°C
-41.3eCe
260°C
decomposes*1
NA
227
decomposes*1
127°C
decomposes0
(733 mm Hg)
NA

NA
decomposes
180°C6
Density
4.81 (red 20°C)C
2.2(-42°C)c
2.9508 (15/4°C)6
3.004 (15/4"C)6
3.604
(15/4°C)e
2.7741 (25/4"C)6
1.32 (20/20°C)b

3.95 (15/15°C)d
NA
2.44(16/4eC)e
Water
Solubility
Insoluble*1
270 mg/100 tag/i
567 ml/100
mlc (20°C)
167 g/100 ccd
(20°C)
decomposes*1
decomposes*1
Insoluble5

38.4 g/100 cc
(14'C)
Insoluble In
cold water*1
decomposes*1
Nonaqueous Solubility
soluble In H2S04d
soluble 1n CS2 and
COC2
soluble In H2S04e
decomposes In alcohol
very slightly soluble
In alcohol
decomposes In alcohol,
soluble 1n CS2d
soluble In benzene,
chloroform. CS2C
soluble 1n benzene,
chloroform. CS2

soluble In benzene;
66.7 g/100 cc alcohol
(14°C)d
decomposes Vti aqua regla
and HN03d
mlsclble In CC14,
chloroform, CS2
Vapor Pressure
1 mm Hg (343.7°C)a
4.5 atm (0.2°C)e
12.0 atm (30.8°C)e
NA
2 mm Hg (150C)6
4.5 mm Hg (35°C)
NA
NA
NA

NA
NA
NA
                                                                                                                    benzene and toluene6
00
CD

-------
                                                                           1ABLE 1-2 (cont.)
Chemical
Selenium
selenate
Selenium
selenlte
Description
white crystals6
white crystals'*
Melting
Point
NA
NA
Boiling Density Water
Point Solubility
NA 3.213 (17.4T.)*1 43.5 g/100 cc
(20°C)d.f
NA NA soluble4
Nonaqueous Solubility Vapor Pressure
NA NA
Insoluble In alcohold NA
            (1982)
     "Hawley (1981)
     'Dean (1985)
     <*Ueast (1985)
     ewindholz  (1983)
     fHater solubility  for  the  decahydrate
     NA =  Not available
o
CO
i
00
vO

-------
tellurium, gold and silver.  The  selenium  recovered  from  these  slimes  (known
as primary selenium)  Is  recovered by various soda smelting  or  soda  roasting
operations.
    In  1983   and  1984,  U.S.  production  of primary  selenium  amounted  to
780,115 and  559,078 pounds, respectively  (USDI,  1988).   Production  volumes
of  primary  selenium   for   1985-1987  were  withheld  by  the  Department  of
Interior  to  avoid  disclosing  company  proprietary  data.   During 1987,  the
U.S. producers of primary  selenium were  Asarco  at  Amarlllo,  TX,  Phelps Dodge
Refining  Corp.,  at  El Paso,  TX, and  BP Minerals  America  Corp.  (formerly
Kennecott Corp.) at Magna,  UT (USDI,  1988).
    Secondary  selenium  Is  selenium   recovered   from   products   previously
manufactured  from  selenium.   In  1981,  -100,000 pounds of  secondary  selenium
was  recovered (USDI,  1982).  Xerox  Corporation   1n  Webster,  NY,  recovered
secondary  selenium  from  used  xerographic   drums  and  Selenium  Inc.,  1n
Maplevllle, RI,  recovered  1t from both  xerographic  drums and  used  selenium
rectifiers  (USDI,  1982).    More  recent  production  volumes   of  secondary
selenium were not available from the USDI (1988).
    U.S.  Imports  and  exports of  selenium  1n 1987 amounted  to  1,093,170 and
357,621 pounds, respectively (USDI, 1988).
    Available  production  data  for  the  selenium   salts  are  presented  1n
Table 1-3.
1.4.   USE DATA
    The  following  are  estimates  of  selenium  consumption  by   end-use  cate-
gories  1n 1987  (USDI,  1988):    electronic and  photocopier  components,  43%;
glass  manufacturing,   20%;  chemicals  and  pigments, 20%;   other  (Including
agricultural and metallurgy), 17%.
0141d                               1-6                              06/15/89

-------
                                  TABLE 1-3



               1977  Production Data  for  Selenium and  Compounds*
Company/Location
Selenium (Element)
Asarco Inc.,
Amarlllo, TX
Denver, CO
Kennecott,
Salt Lake CHy, UT
U.S. Metals Refining Co.,
Carteret, NJ
Cerac, Inc.,
Milwaukee, MI
United Mineral & Chemical Corp.,
New York, NY
Robeco Chemical,
New York
Leonard 0 Buck & Co Inc. ,
Morris town, NJ
Comlnco American Inc.,
Spokane, WA
Henley & Co. , Inc.,
New York, NY
North American Minerals Co.,
PHtsburg, PA
Owens-Illinois Inc.,
Toledo, OH
IMC,
Llbertyvllle, IL
Copalco International Ltd.,
New York
Manufacturer/
Importer
manufacturer
manufacturer
manufacturer
manufacturer
manufacturer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Production Range
(Ibs/year)
10,000-100,000
<1,000
100,000-1,000,000
100,000-1,000,000
confidential
<1000
none
10,000-100,000
<1000
confidential
10,000-100,000
10,000-100,000
none
1000-10,000
0141d
1-7
03/30/89

-------
                              TABLE 1-3 (cont.)
Company/Location
Selenium (Element) (cont.)
Alloychem, Inc.,
New York, NY
Brandels Goldschmldt & Co.,
New York, NY
NL Industries Inc.,
New York, NY
Phillips Brothers Division of
Engelhard Minerals & Chem.,
New York, NY
Hercules Inc.,
Wilmington, DE
Holtrachem Inc. ,
Natlck, MA
Xerox Corp. ,
Rochester, NY
Degussa Corporation,
Theodore, AL
Indussa-D1v1s1on of African
Metals, New York, NY
Kaweckl Berylco Ind., Inc.,
Reading, PA
Ametalco Inc.,
New York, NY
Manufacturer/
Importer

Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Importer
Production Range
(Ibs/year)

1000-10,000
none
10,000-100,000
confidential
confidential
none
10,000-100,000
none
10,000-100,000
1000-10,000
confidential
Hydrogen Selenlde

Synthatron Corp.,
  Parslppany, NJ

Scientific Gas Products Inc.,
  South Pla1nf1eld, NJ
    manufacturer


    manufacturer
1000-10,000


1000-10,000
0141d
1-8
         03/30/89

-------
                              TABLE 1-3 (cont.)
        Company/Location
    Manufacturer/
      Importer
Production Range
   (Ibs/year)
Hydrogen Selenlde (cont.)

II-VI Inc.,
  Saxonburg, PA
    manufacturer
<1000
Selenlc Add

Engelhard Industries D1v.,
  Union, NJ
    manufacturer
confidential
Selenlous Add

Chemical Products Plant,
  Cleveland, OH

United Mineral & Chemical,
  New York, NY

Falrmount Chemical Co., Inc.,
  Newark, NJ

The Harshaw Chemical Co.,
  Cleveland, OH
    manufacturer


    Importer


    Importer


    Importer
<1000


1,000-10,000


<1000


<1000
Selenium Chloride

Cerac, Inc.,
  Milwaukee, WI

Falrmount Chemical Co., Inc.,
  Newark, NJ

United Mineral & Chemical,
  New York
    manufacturer


    manufacturer


    Importer
confidential


<1000


<1000
Selenium Dioxide

Cerac, Inc.,
  Milwaukee, WI
    manufacturer
confidential
0141d
1-9
         03/30/89

-------
                              TABLE  1-3  (cont.)
Company/Location
Selenium Dioxide (cont.)
United Mineral & Chemical,
New York, NY
Henley & Co., Inc.,
New York, NY
Alloychem., Inc.,
New York, NY
SST Corp.,
Clifton, NJ
Manufacturer/
Importer
Importer
Importer
Importer
Importer
Production Range
(Ibs/year)
<1000
confidential
1000-10,000
<1000
Selenium Dlsulflde

Robeco Chemicals, Inc.,
  New York, NY

ICD Group, Inc.,
  New York, NY
    Importer


    Importer
confidential
none
Sodium Selenate

Cerac, Inc.,
  Milwaukee, WI

Falrmount Chemical Co. Inc.,
  Newark, NJ
    manufacturer


    Importer
confidential


<1000
Sodium Selenlte

Cerac, Inc.,
  Milwaukee, WI

United Mineral & Chemical Corp.,
  New York, NY

Falrmount Chemical Co. Inc.,
  Newark, NJ
    manufacturer


    Importer


    Importer
confidential


1000-10,000


<1000
0141d
1-10
         03/30/89

-------
                              TABLE 1-3 (cont.)
        Company/Location
    Manufacturer/
      Importer
Production Range
   (Ibs/year)
Sodium Selenlte (cont.)

EM Laboratories,  Inc.,
  Elmsford,  NY

Deka Minerals, Inc.,
  New York,  NY
    Importer


    Importer
1000-10,000
none
*Source:  U.S.  EPA,  1977
OUld
1-11
         03/30/89

-------
    The photoelectric and  semiconducting  properties of selenium are  respon-
sible  for  Us primary  use  In the  electronics  and optical Industry  (Elkln,
1982).  The  electrical  conductivity of  selenium  Is  ~3  orders of  magnitude
higher  during exposure  to  light  than  In the dark.   Because  of these  prop-
erties,  selenium  Is  used  1n photovoltaic  cells,  photoelectric  cells  and
semiconducting rectifiers.  Xerography  1s  the most highly developed  form of
electrophotography  or   photocopying  and  Is  a   primary  application  of  the
photoconductlve properties of  selenium  (Elkln, 1982).
    The glass  manufacturing Industry  uses  selenium to Impart  various  tints,
such  as red  and  bronze,  and to  neutralize or  de-colorize  greenish  tints
caused  by  Iron  Impurities  (IARC,  1975;  Elkln,  1982).   The ceramics,  paint
and plastics  Industries use a variety of selenium pigments  to  produce colors
ranging from yellow to maroon.
    Selenium  1s  used In  metallurgy as  a  property  modifier  to form  alloys
with Iron,  copper, nickel, cobalt  and  lead  (Elkln, 1982).
    Sodium selenate and selenlc add have been  added  to  chrome-plating  baths
since  1959  to   Improve  corrosion  protection  from  pitting,  blistering  and
rusting (Elkln,  1982).   An Important application  of  this Is  the  plating of
automobile parts exposed to road  salt  In the snow belts.
    Selenium  dlethyldHhlocarbamate  Is  used by  the  rubber   Industry  as  a
vulcanization accelerator, antloxldant  and  UV stabilizer  (Elkln, 1982).
    Selenium  dioxide  1s an Important  oxidizing  agent and  catalyst  for  the
synthesis  of organic chemicals  and drug  products (Elkln, 1982).   Selenium
sulflde  1s  used  1n topical  pharmaceutical  products  (such as Selsun  from
Abbott  Laboratories)  to control  seborrhelc  dermatitis  of the  scalp  (Elkln,
1982).   Sodium  selenlte  and  various  organoselenlum compounds  are used  In
photographic  printing  solutions  and  bleach  fixing baths  (Elkln,  1982).
0141d                               1-12                             03/30/89

-------
Tablets containing several mlcrograms  of  sodium selenlte are sold without  a
prescription In the United States as dietary supplements  (Elkln,  1982).
    In agriculture,  selenium  (sodium  selenate and  selenlte)  1s  used  as  a
feed additive to compensate for  animal feed  grown  1n  selenium-deficient  soil
(USOI,  1988).    In  April  1988,  the  U.S.   Food   and  Drug  Administration
             \
published new  regulations  allowing the level  of  selenium In animal  feed  to
be Increased from 0.1-0.3 ppm (USOI, 1988).
1.5.   SUMMARY
    In  the  United  States,   selenium  Is  produced  by  recovering  H   from
by-products generated during  the electrolytic  refining of  copper  ore  (Elkln,
1982; USOI, 1982,  1988).   In  1984,  559,078  pounds  of  selenium  were recovered
from  these  by-products  by  three  domestic  producers.    Also,  selenium  1s
recovered  domestically  from  previously  manufactured  xerographic drums  and
selenium  rectifiers  (USOI,   1982).   Approximately   1   million   pounds   of
selenium  were   Imported  1n  1987  (USDI).   The  following are  estimates  of
selenium consumption by end-use  categories  In  1987  (USDI,  1988):   electronic
and  photocopier components,  43%;   glass  manufacturing,   20%;  chemicals  and
pigments,  20%;  other (Including agriculture and metallurgy), 17%.   Because
of  Us  photoelectric  and   semiconducting   properties,   selenium  1s   used
primarily  In the production of photocopying  components,  photoelectric cells,
photovoltaic cells  and semiconducting rectifiers (Elkln,  1982).
0141d                               1-13                             03/30/89

-------
                     2.  ENVIRONMENTAL FATE  AND TRANSPORT

    Selenium 1s  an  element;  therefore, 1t  does  not degrade  In  the  environ-
ment, but  simply changes  from  one form to  another.   The major  features  of
selenium chemistry that affect  Its fate and  transport  1n the  environment  are
associated with changes 1n Us  oxidation state and  the resulting differences
1n chemical properties  (Callahan  et  al.,  1979).   The  chemical  speclatlon  of
selenium Is discussed  In  more  detail 1n Section  2.2.   Because  of  the  chang-
ing  forms  of  selenium  compounds  1n  nature,  selenium  undergoes an  environ-
mental  cycling  process  that  Involves  soil, rocks,  plants,  animals,  water
(oceans,  lakes,  etc.)  and  the atmosphere  (NAS,  1976).   Figure 2-1  1s  a
simplified diagram of the cycling of  selenium 1n  nature.
2.1.   AIR
    Selenium can  exist  In  the atmosphere  1n both  the partlculate  and  the
vapor phase.   Partlculate-phase selenium enters  the atmosphere  from sources
such  as  wind-blown  dusts  (from soil  and  rock erosion),  fly ash from  coal
combustion, and  volcanic emissions  (Chapter  3).    No  data are  available  to
suggest that chemical  transformation processes  are Important to  the  fate  of
partlculate-phase  selenium In  air.    Removal  of partlculate-phase  selenium
from  air  occurs  physically  by wet  and  dry deposition processes,  such  as
rainfall and  settling  (NAS,  1976).    A  mean  selenium concentration of  210
ng/i  was  detected  In  rain  and  snow  collected  In   Cambridge,  MA,  from
December  1964  to  March  1965   (Stahl,  1969).   Lower  mean  levels of  11-21
ng/i  were  detected  1n rainfall collected  at the  Enewetak  Atoll  (Marshall
Islands) 1n 1979 (Arlmoto et al., 1985).
    Vapor-phase  selenium enters  the atmosphere  by volatile emissions  from
mlcroflora  and  plants.   Important  selenium compounds  emitted  to air  from


0142d                               2-1                              06/15/89

-------
                                                  SEDIMENTS l>
                                                  SEDIMENTARY
                                                  MOCKS
                                    EARTH'S
                                    CORE
                                   FIGURE 2-1
    The  Cycling of Selenium  1n Nature.   (For  simplicity, microorganisms  are
not Included  1n the above scheme, although they are Important  to many of  the
processes Involved  1n  the cycle.)
Source:  NAS  (1976)
0142d
2-2
03/30/89

-------
mlcroflora  Include  dimethyl selenlde  and dimethyl  dlselenlde  (Callahan  et

al.,  1979;  Olson et  al.,  1983;  Cooke and  Bruland,  1987).   These  selenium

compounds are  expected  to react with  sunlight-produced hydroxyl  radicals  In

air,  assuming  their atmospheric  chemistry  Is  similar  to  that  of  dimethyl

sulflde and dimethyl dlsulflde.   Based upon  experimentally measured  reaction

rate  constants  of  the  sulfldes  with  H0»  and an  average  concentration  of

5xlOf5  HO*/cm3   1n air   (Atkinson,  1985),  the  half-lives  of   dimethyl

selenlde and dimethyl dlselenlde  In  air can  be  estimated  to  be  3.75  days  and

1.9 hours,  respectively.   Based  upon  observed sulflde reactions  (Atkinson,

1985), the major products of hydroxyl  reaction will  Include  selenium dioxide

and  methyl   hydrogen  selenlte.   These   selenium  degradation  products   are

probably susceptible to  physical removal from air.

2.2.   WATER

2.2.1.   Chemical Spedatlon.   Selenium  Is  stable  In  four  valence  states

(-2,  0,  +4,  +6), which  are discussed  Individually  below (Callahan  et  al.,

1979;   NAS,  1976).   In aerobic  water  conditions,  selenium  can  exist  In  the

form of the selenate, selenlte  or hydrogen  selenlte  anlons  (Callahan et  al.,

1979;  NAS, 1976).


    Selenate  selenium  (+6  oxidation  state):  the  selenates  are  rela-
    tively  soluble  1n  water,  stable  over  the  environmental pH  range,
    and  readily  taken   up   by  plants.   Because  of  this   solubility,
    stability and  plant   uptake  potential,  selenate appears  to be  the
    most  dangerous   form  of  selenium  In  relation  to  environmental
    pollution.

    Selenlte  selenium  (+4  oxidation  state):  most  selenlte  salts  are
    less  soluble  than  the corresponding  selenates;  especially  relevant
    to  the  aquatic  environment  Is   the  very low  solubility of  ferric
    selenltes.   Another  characteristic   (Important   to  environmental
    cycling) 1s  the property of  selenlte  to rapidly become  reduced  to
    elemental selenium under addle  conditions  by  mild  reducing  agents,
    such  as  ascorbic  add  or  sulfur  dioxide.   The  probability  that
    selenlte  will  form   Insoluble   compounds,  absorbates  with  ferric
    oxides,  or be reduced to Insoluble elemental  selenium minimizes  the
    possibility for  Its  transport  1n the aquatic environment.



0142d                               2-3                              03/30/89

-------
    Elemental  selenium  (0  oxidation  state);   because  of  Us  extreme
    Insolubility  1n  water,  elemental  selenium  appears to  be a  major
    sink for selenium  that can be  considered  to  be  Inert 1n the aquatic
    environment.   Elemental   selenium  burns  1n  air   to form  selenium
    dioxide.   In  the  combustion  of   fossil  fuels   or  solid  wastes
    containing selenium, the  selenium  1s  converted  to  selenium dioxide,
    which  can   then  be  reduced   to  elemental  selenium by  the  sulfur
    dioxide  that  1s  also  present.  This  mechanism can be  responsible
    for  the occurrence  of  elemental  selenium  1n  fly ash  from  coal
    combustion.

    Selenlde selenium  (oxidation  state  -2):   the  metal selenldes  are
    relatively  Insoluble 1n  water and  do not  appear  to  be  taken  up
    readily  by  plants.   Thus,   the  metal  selenldes  (and  elemental
    selenium) may  represent  a useful Inert sink  for  the detoxification
    of selenium  In areas where  these compounds occur  or are  deposited.
    In  semlarld and  arid regions,  selenlde-selenlum  appears  to  have
    been oxidized, over  geologic time, to  selenate.   Hydrogen  selenlde
    decomposes  rapidly  to  form  elemental  selenium  and   1s  of  minor
    Importance to the overall fate of selenium 1n aquatic environments.

    Organic  selenium  (oxidation   state  -2):   essentially  all  of  the
    relevant  organic  selenium   compounds   In  the  aquatic  environment
    contain  selenium  1n the  -2  oxidation state.  These compounds  will
    decompose 1n the  environment  to eventually form elemental  selenium.


2.2.2.   Photolysis.    Although   selenium   exhibits   photo-conductive  prop-

erties,  no  data are  available   to  suggest that  photolysis  Is  Important  In

determining  the  environmental fate  of selenium  1n the aquatic  environment

(Callahan et al., 1979).

2.2.3.   Blomethylatlon   and   Volatilization.    The   blomethylatlon    and

volatilization  of  selenium are  considered  together because the  products  of

blomethylatlon  Include  the  volatile  dimethyl selenldes.  Chau  et  al. (1976)

demonstrated 1n  laboratory studies  that  microbes present 1n  lake  sediments

(from  Sudsbury,  Ontario)  could  methylate  organic and Inorganic  selenium

compounds  with  the  formation  of  gaseous  dimethyl   selenlde  and  dimethyl

dlselenlde.   In addition, Chau  et al.  (1976)  found  that  4 of  12  sediment

samples were able  to evolve  methylated  selenldes  without the addition of any

selenium  compounds;   the   naturally  occurring  selenium   content   of  the

sediments ranged from 0.48-20.48 mg/kg.
0142d                               2-4                              03/30/89

-------
    Cooke  and  Bruland  (1987)  examined  surface water  from  three  sites  In
California to  determine  the chemical  species  of  selenium that were  present
In  the  natural  waters.   Six  dissolved  species  were  Identified:  Inorganic
selenates  and  selenltes,  nonvolatile  organic selenldes  (Including  seleno
ami no adds  and  a  d1methylselenon1um  Ion),  and  volatile dimethyl  selenlde
and dimethyl dlselenlde.   The  presence of the  volatile methylated  selenldes
was  considered  Important  In the  environmental cycling  of  selenium and  a
major process by which selenium Is removed from the aquatic environment.
2.2.4.   Sorptlon.   The  sorptlon  of  selenium  to  aquatic  sediments   and
suspended  matter depends  on  a  variety  of   factors,  such  as  the  chemical
species  of  selenium that Is present,  the pH of the water,  and  the  presence
of  minerals  and  clays   (Callahan   et al.,   1979;  Gruebel  et  al.,  1987;
BarYosef, 1987).   The mobility of selenium 1n  both  surface water  and ground-
water has  been  found  to depend  strongly upon oxidation  state  (Gruebel  et
al., 1987).  Mobility  1n  groundwater 1s associated with oxidation  state +6,
while mobility 1n surface waters may  be  associated with both the *4  and  +6
oxidation  states.   This  seems  reasonable,   since  the f6  oxidation  state
refers  to  the  selenates,   which   are the  most   soluble  of  the  selenium
compounds.  Apparently,  selenium  can undergo  an adsorptlon-desorptlon cycle
In the aquatic environment,  which  1s determined primarily by  the changes  In
the oxidation  states of the  selenium species  that  are present.   Partitioning
to sediments Is  likely to occur as  the oxidation  state 1s  reduced from +6-0;
desorptlon  (with subsequent dissolution  1n  the water  column)  Is likely  to
occur when the oxidation state reaches +6.
2.2.5.   B1oconcentrat1on.   Data  pertaining   to   the   bloconcentratlon   of
selenium  In  aquatic organisms  has  been compiled  by U.S.  EPA  (1987a).  F1sh
(fathead  minnow,  striped bass) exposed  to sodium  selenate  (oxidation state

0142d                               2-5                              09/20/89

-------
+6) were reported to exhibit BCFs  ranging  from  0.68-52.  Fish  (rainbow  trout,
fathead  minnow,  blueglll,  largemouth  bass)  exposed to  sodium selenlte  or
selenlous acid  (oxidation state +4)  were  reported  to  exhibit BCFs ranging
from 2-470.
2.3.   SOIL
2.3.1.   Leaching.   Data  specific  to  the  leaching  of  selenium 1n soil  were
not  located  1n  the  available  literature;   however,  mobility  1n  soil  Is
probably analogous  to movement  In  groundwater and  surface water as discussed
1n  Section   2.2.4.   It  Is  expected  that  selenium  In  oxidation  state  +6
(selenate)   will  be  the most  mobile of  the  selenium species  1n soil,  while
selenium 1n  oxidation state 0 will  be  generally  Immobile.
2.3.2.   Blotransformatlon.  Selenates and selenltes are  taken up by  plants
and are generally reduced  within the plant to oxidation  state -2  In  the form
of soluble  amlno adds or  protein-bound  ami no adds  (Callahan et  al.,  1979).
As  plants decay,  the selenium  1n  the  plant  tissue  1s  returned to  the  soil
where  It can  be  oxidized  to other  oxidation  states.  It has  also  been shown
that   living  plants  can   release volatile  selenium   compounds  Into  the
atmosphere  (Z1eve and  Peterson, 1984).   Z1eve and  Peterson (1984) suggested
that  plants,  1n  addition   to soil  microorganisms,  play  an Important role  1n
selenium  volatilization   to the   atmosphere  as  part  of  the  environmental
cycling process.
    The  mlcroblal   formation  of   volatile selenium  compounds  (principally
dimethyl selenlde)  has  been reported to be  widespread  (Z1eve  and Peterson,
1984).   Reamer  and  Zoller (1980)   demonstrated  that microorganisms  1n  soil
and   sewage   sludge  converted   Inorganic   selenium  compounds   to  volatile
methylated  spedes.   These  volatile   species   Included  dimethyl  selenlde,
dimethyl dlselenlde, dimethyl selenone and  methyl methylselenlte.

0142d                               2-6                              03/30/89

-------
2.4.   SUMMARY
    Selenium  Is  an  element;  therefore, It  does  not degrade  In  the  environ-
ment,  1t  simply changes  from one  form  to another.   The  major features  of
selenium chemistry that affect Its  fate and  transport  1n  the  environment  are
associated with changes In Its oxidation  state and  the resulting differences
1n  chemical   properties   (Callahan  et  al.,  1979).    Because  selenium only
changes form  and  does not degrade  In nature, It undergoes an  environmental
cycling process  Involving  soil,  rocks,  plants,  animals, water and air  (NAS,
1976).  When  released to  the atmosphere  In  partlculate-phase, selenium  1s
removed by  physical  processes  such  as  rainfall  and  settling   (NAS,  1976).
Volatile  selenium   compounds  are  released to  air   by plant  and mlcroblal
transformations  (Z1eve and Peterson,  1984;  Chau  et al.,  1976), where they
react  with   sunlight-formed   H0«   to   form  selenium  dioxide  and   other
compounds  that  can  be  transported  physically  to water  and  soil.   In both
water and soil,  the blomethylatlon of selenium to  release  volatile  selenium
to air 1s an  Important part of  the  environmental cycling  process  (Cooke  and
Bruland,   1987;  Z1eve and  Peterson,  1984).   Selenium  1n  oxidation  state  *6
(selenate) Is  the  most soluble  and  mobile selenium species  In  the  soil  and
water environments  (Gruebel et al.,  1987;  Callahan  et  al.,  1979).  Because of
the  solubility,  stability  and ability of  selenate  to  be  taken  up  by plants,
It 1s considered the  most  dangerous  form  of selenium 1n relation to  environ-
mental pollution (NAS, 1976; Callahan et al., 1979).
0142d                               2-7                              06/15/89

-------
                                 3.   EXPOSURE

    Selenium  1s  widely distributed  1n  the  earth's crust,  and with  suffi-
ciently  sensitive  analytical   techniques,   selenium   can   be   detected   In
virtually  all  rocks  and  soil   on  the  earth's  surface  (NAS,  1976).   Host
estimates  of   Us  average  concentration  1n  the  earth's  crust  range  from
0.03-0.8  ppm;  several  fall  around  0.1  ppm  (NAS, 1976).   Because of  this
ubiquitous  distribution  1n  the environment,  the  detection  of selenium  1n
air, water, soil, food and vegetation Is  expected (IARC, 1975).
    Distribution of selenium to the air and  water environments occurs  from
both natural  and anthropogenic   sources.    Natural  sources of release  to  the
atmosphere  Include volcanic  emissions, wind-blown  dusts,  sea  sprays  from  the
oceans,  and release  of  volatile  selenium compounds  from  both  plants  and
animals  as a  result   of  biological processes  (NAS,  1976;  AMmoto et  al.,
1985).    Anthropogenic sources   of  release  to  the atmosphere  Include  the
mining,  smelting and  refining of selenium-containing ores,  the  processing of
Industrial  products   from  refined   selenium,  and  the  Incineration of  coal,
fuel oil  and  solid waste  (NAS,  1976;  U.S. EPA,  1987a).   More than half  of
the anthropogenic  releases  are  attributed to  coal combustion  (NAS,  1976).
The average selenium  content 1n U.S. coal  1s  ~2-3 ppm;  the selenium content
of  fuel  oil or  crude  oil may range  from 0.006-2.2 ppm  (NAS,  1976).   Combus-
tion sources release  selenium In fly ash particles and  as vapors (flue-gas),
which will primarily become partlculate material.
    Selenium  Is  released  naturally to  the  water  environment  by  weathering
and erosion of  rocks   and  soil  (U.S.  EPA,  1987a).  Man  releases  selenium to
water  by wastewater   discharges  from mining,  refining  and  Industrial  appli-
cations  processing.    In  addition,  selenium  Is   released  to water  through
0143d                               3-1                              06/15/89

-------
physical  removal  (wet and  dry deposition)  of participate  matter  from  the
atmosphere (U.S. EPA, 1987a).
3.1.   WATER
    Based  upon  available monitoring  data,  1t  has been  estimated  that  the
average selenium content  of  drinking  water 1s ~0.2 ppb  (Bennett,  1986).   It
1s  possible,  however,  that  some  drinking  waters  may  contain much  higher
selenium  levels at   various  geographical  locations   where   the  content  of
naturally  occurring  selenium 1s high.   For  example,  In a high selenlferous
area of  South  Dakota, selenium levels of  50-330  ppb  were found 1n 10  of  44
drinking waters  from sampled wells (U.S.  EPA,  1980a).  A study of home  tap
water  samples  collected  from  3676 residences -located 1n 35  geographically
dispersed  areas  found 9.96X of the  samples  with selenium  levels above  the
detection  limit of  1  ppb and a maximum  positive  detection of  36.8  ppb (U.S.
EPA, 1980a).
    Assuming an average  drinking water  selenium concentration  of 0.2  ppb
(200  ng/a,)  (Bennett,  1986)  and  an adult  consumption  of   2  I/day,  the
average  dally  Intake  from  drinking  water  can be  estimated  to  be 400  ng,
which 1s much lower than the average dally Intake from food (Table 3-1).
    An  analysis  of  monitoring  data from   211  sampling   stations  1n  the
National  American  Stream  Quality Accounting  Network  and  National  Water
Quality Surveillance  System  found mean selenium  concentrations of <1  ppb 1n
river water from throughout  the United States (Smith  et al., 1987);  monitor-
Ing  was  conducting between  1974  and  1981 with a detection  limit of  1  ppb.
An  analysis of  1062  groundwaters  and  590   surface  waters  from New  Jersey
found  selenium  In  all water  samples,   with a  median  concentration  of  2.0 ppb
In  both groundwater  and surface water   (Page,   1981).   Preliminary  results
from the U.S. EPA's  Nationwide  Urban  Runoff  Program detected selenium levels
of  2-77 ppb  1n  stormwater runoff  from three  U.S.  cities (Cole  et al., 1984).
0143d                               3-2                              06/15/89

-------
                                  TABLE 3-1
  Average  Dally  Intake  (yg/day) of Selenium  1n Fiscal Years 1978-1981/1982*
Fiscal Year
1981/1982
1980
1979
1978
Adult
139
141
152
156
Toddler
54
58
45
52
Infant
22
19
15
18
*Source:  Gartrell et al.. 1986a,b
0143d
3-3
03/31/89

-------
It has been estimated that  the average  selenium  concentration  of  waters  from
the open oceans (not affected by  localized contamination)  Is -0.09  ppb  (NAS,
1976; Elkln, 1982).  The  principal  species  In the ocean  1s the  selenate  1on
(Elkln, 1982).
3.2.   FOOD
    The  U.S.  Food  and  Drug Administration  has  conducted  the  Total   Diet
Studies  (also  referred  to as Market  Basket  Programs) since the  early  1960s
(Gartrell et al., 1986a).  These  studies  survey  the  American food supply  and
measure  dietary  Intake   of  selected  pesticides,  Industrial  chemicals  and
elements.  The Adult Total Diet Study monitors the market food supply 1n  the
diets  of 16-  to  19-year-old  males,  who  generally   consume more food  than
other  age  groups.  The  results   of  the most  recent study with respect  to
selenium  are  presented   In  Table 3-2.   For   this  study, 27  market basket
samples  (consisting of   -120  Individual  food  Hems)  were  collected   from
retail markets  from the  northeast,  south, west  and  north central regions of
the United States.  Collection occurred  between  October 1980 and  March  1982,
referred  to  as  Fiscal   Years  1981/82.   The  collected  food  Hems   were
separated  Into   the   various   food   groups   and  were   then   blended   Into
homogeneous composites for  analysis  by food  group.   Tables 3-3  and  3-4 11st
similar  results  for the  Infant  (6  months old)  and Toddler  (2 years  old)
Total  Diet  Studies.  The average dally  Intake of selenium for  the  American
adult, toddler and  Infant  (see Table  3-1)  Is  based upon results of the  Total
Diet Studies for Fiscal Years 1978 to 1981/82.
    Table 3-5 lists reported selenium  levels  1n  various  staple food Hems of
the American diet.   The  selenium content of  food  grown  In high-selenium vs.
low-selenium  soils, however,  may vary  by  up to  three  orders  of  magnitude
(U.S. EPA, 1980a).

0143d                               3-4                              06/15/89

-------
                                  TABLE 3-2
        Selenium  1n  the  Adult  American  Diet for Fiscal Years 1981-19823
Food Group
Dairy products
Meat, fish, poultry
Grain and cereal
products
Potatoes
Leafy vegetables
Legume vegetables
Root vegetables
Garden fruits
Fruits
011s and fats
Sugar and adjuncts
Beverages
TOTAL
Average
Concentration0
(ppm)
0.008
0.209
0.183

0.004
0.002
0.003
0
0.001
0
0.002
0
<0.001
NA
Number
Positive0
11
27
27

5
5
8
0
3
0
2
0
1
89e
Concentration
Range0"
trace-0.02
0.14-0.30
0.10-0.34

trace
trace
trace
0
trace
0
trace
0
trace
trace-0.34
Average
Intake
Ug/day)
6.17
54.9
77.0

0.590
0.114
0.245
0
0.071
0
0.106
'0
0.272
139
^source:  Gartrell et al. (1986a)
DBased on all composites analyzed
C0ut of 27 composites for the Individual food group
dTrace Indicates at the detection limit
eOut of 324 composites
NA = Not applicable
0143d
3-5
03/31/89

-------
                                  TABLE 3-3
       Selenium In the Infant American  Diet  for  Fiscal  Years  1981-19823
Food Group
Whole milk
Other dairy products
Meat, fish, poultry
Grain and cereal
products
Potatoes
Vegetables
Fruits and fruit
juices
011s and fats
Sugar and adjuncts
Beverages
Drinking Water
TOTAL
Average
Concentration13
(ppm)
0.009
0.015
0.112
0.192
0.002
0.007
0
0.005
0
0
0
NA
Number
Positive0
8
7
13
13
1
7
0
2
0
0
0
51e
Concentration
Ranged
trace
trace-0.09
0.05-0.20
0.12-0.34
trace
trace
0
trace
0
0
0
trace-0.34
Average
Intake
(yg/day)
5
2
5
6
0
0
0
0
0
0
0
21
.63
.94
.34
.80
.021
.690

.110



.5
aSource: Gartrell et al., 1986b
bBased on all composites analyzed
C0ut of 13 composites for each Individual  food group
dTrace Indicates at the detection limit
eOut of 143 composites
NA = Not applicable
0143d
3-6
03/31/89

-------
                                  TABLE 3-4
       Selenium  In  the Toddler American Diet for Fiscal Years 1981-19823
Average
Food Group Concentration1*
(ppm)
Whole M1lk
Other dairy products
Meat, fish, poultry
Grain and cereal products
Potatoes
Vegetables
Fruits and fruit juices
011s and fats
Sugar and adjuncts
Beverages
Drinking Water
TOTAL
0.009
0.022
0.208
0.189
0.003
0.003
0
0.036
0.005
0
0
NA
Number
Positive0
8
11
13
13
1
3
0
9
4
0
0
62e
Concentration
Ranged
trace
trace
0.06-0.33
0.12-0.31
trace
trace
0
trace-0.17
trace
0
0
trace-0.33
Average
Intake
(vg/day)
4.63
1.51
24.8
22.1
0.138
0.199
0
0.605
0.153
0
0
54.1
aSource: Gartrell et al., 1986b
bBased on all composites analyzed
C0ut of 13 composites for each Individual food group
dTrace Indicates at the detection limit
eOut of 143 composites
NA = Not applicable
0143d
3-7
03/31/89

-------
                                  TABLE 3-5
         Average Selenium Content of Some  Foods  1n  the  American  D1eta
                     Food
             Selenium Content
              (vg/g  wet  wt.)
         Vegetables,  canned and freshb
         Fresh garlic
         Mushrooms,  canned and fresh
         Fruits,  canned and fresh
         Cereal products0
         Corn flakes
         Rice cereal
         Egg white
         Egg yolk
         Brown sugar
         White sugar
         Cheeses
         Table cream
         Whole milk
         Meat (excluding kidney)
         Seafood
           0.010  (0.004-0.039)
           0.249
           0.118
           0.006  (<0.002-0.013)
           0.387  (0-0266-0.665)
           0.026
           0.028
           0.051
           0.183
           0.011
           0.003
           0.082  (0.052-0.105)
           0.006
           0.012
           0.224  (0.116-0.432)
           0.532  (0.337-0.658)
aSource:  U.S. EPA, 1980a
bMean excluding mushrooms and garlic
cMean excluding corn flakes and rice cereal
0143d
3-8
06/15/89

-------
3.3.   INHALATION
    A  review  of  available  data  concerning  the   concentration   of   trace
elements  1n  ambient  atmospheric  participate  matter   has   found   selenium
concentrations of  0.0056-0.19,  0.01-3.0  and 0.2-30  ng/m3  In  the  air  from
remote  locations,  rural  locations and  U.S.  urban  locations,  respectively
(Schroeder et al., 1987).  The mean selenium  concentration 1n  air from urban
locations  1s  ~3 ng/m3  (Bennett,  1986).    A concentration  of  1.07   ng/m3
was detected  In  air  near the  Delaware  coast  In  August,  1982;  this was  higher
than  levels  of  0.16-0.43  ng/m3  detected  In  air near  the  coast of Bermuda
(Gibson  et  al.,  1986).  The  5-year  (1978-1982) average selenium concentra-
tion  1n the  air of  El Paso,  TX, has  been reported  to  be  1  ng/m3,  with
highest  levels  reaching  205  ng/m3   (Wlersema  et   al.,   1984).    Selenium
levels  of  2.9-4.3 ng/m3 have been monitored  Inside of automobiles travel-
Ing the  California freeways  (WHz et  al., 1986); these  levels may  be  higher
than the outdoor  ambient levels  by a factor of 2 or  3.
    A major  use  of  selenium Is  1n the manufacture of photocopier components
(USDI,  1988).  Measurement  of selenium 1n the air near  working  photocopiers
has been reported to  range  from  10-600  ng/m3  (Hansen  and Andersen,  1986).
The level of  selenium In various tobacco, leaves has been reported  to range
from  0.03-2.28  mg/kg,  with  the  selenium content  of  associated  cigarette
smoke ranging from 7.7-63 ng/clgarette (Jenkins, 1986).
    Assuming  an   average  atmospheric  selenium  concentration  of   3   ng/m3
(Bennett,  1986)   and  an  adult  air   Inhalation  of  20   mVday,  the average
dally  Intake from Inhalation can be  estimated  to  be  60  ng, which Is  much
lower than the average dally  Intake from food (see Table 3-4).
0143d                               3-9                              06/15/89

-------
3.4.   DERMAL
    Selenium sulfldes are  used  In  various  anti-dandruff  shampoos  (concentra-
tion  of  1-2.5%) and  topical pharmaceutical  products  to control  seborrhek
dermatitis  of   the  scalp   (U.S.  EPA,  1980a;  Elkln,  1982).   It  has  been
reported  that  ordinary  application  of  selenium-containing  shampoo does  not
significantly elevate  selenium  levels  1n  blood when compared with controls
(U.S. EPA, 1980a).
    Selenium has many  Industrial  uses,  and most  significant  dermal exposure
would be confined primarily  to  occupational settings  (U.S.  EPA,  1980a).   Eye
Irritations and  skin  burns have occurred  from direct contact with selenium
dioxide at a plant  refining copper  by the electrolytic method  (Robin,  1984).
3.5.   SUMMARY
    Selenium 1s  distributed widely  1n   the  earth's crust, and  with  suffi-
ciently  sensitive  analytical  techniques,   selenium  can  be   detected   In
virtually all rocks and soil on the  earth's  surface (NAS,  1976).   Because of
this  ubiquitous  distribution In the  environment, the detection of selenium
1n air, water,  soil,  food  and vegetation  Is  expected  (IARC,  1975). Selenium
1s released  to  the  environment  from natural  sources  such as  volcanoes,  rock
and  soil  erosion,  sea sprays and  volatile emissions from  plants  and  micro-
flora  (NAS,  1976;   AMmoto et  al.,  1985).   Humans  release selenium   to  the
environment  through Incineration   of  coal, fuel  oil and  solid  waste,  from
emissions  and  waste  streams  generated   through  mining,  refining  and  Indus-
trial  applications  (NAS,  1976; U.S.  EPA, 1987a).   More  than  half   of  the
human releases are  attributed  to  the combustion  of  coal  (NAS, 1976).   Based
upon  available  monitoring  data,   the  average  selenium  content  of drinking
water  1s   ~0.2  ppb  and  the  mean  concentration  1n  urban  air  1s ~3  ng/m3
(Bennett, 1986).  These concentrations  can  be used  to estimate average dally
0143d                               3-10                             03/31/89

-------
Intakes of  400  ng/day for water  and 60 ng/day for  Inhalation.   The  average
dally  Intakes  for  water  and  Inhalation are  small  In  comparison with  the
estimated  average  dally  Intake  of  139 ug/day  for  food  (Gartrell  et  al.,
1986a).
 0143d                                3-11                              06/15/89

-------
                         4.  ENVIRONMENTAL TOXICOLOGY
4.1.   AQUATIC TOXICOLOGY
4.1.1.   Acute Toxic  Effects  on  Fauna.   The acute toxldty  of selenium  to
aquatic  vertebrates  as  expressed by  the LC™  1s  presented  1n Table  4-1.
LC5Qs  were   reported  for  numerous   species   of   freshwater  and   marine
vertebrates exposed  to  selenium dioxide,  sodium and potassium  selenlte,  and
sodium and potassium selenate.   Exposure  of rainbow trout, Salmo  qalrdnerl.
to  selenium   as  sodium  selenlte  produced  the  lowest  96-hour  LC™  (1.80
mg/l)  Identified  In  this  review  for a freshwater  vertebrate  (Hunn  et  al.,
1987).   Carp, Cyprlnus  carplo.  zebraflsh,  Brachyodanlo  rerlo.  and  white
sucker,  Catostomus   commersonl.  were among  the least  sensitive  freshwater
vertebrates,   with   96-hour   LC50S   ranging  from  23-35  mg/l   (N11m1   and
Laham, 1976;   Etnler  et al., 1987).   U.S.  EPA (1987a) reported the  results  of
additional studies 1n which the toxldty  of  selenium as  selenlte ranged from
0.620 mg/l for fathead minnows to 35  mg/l  for the common  carp.
    Fathead minnows,  Plmephales  promelas. appeared  to  be the most  sensitive
freshwater species   to  selenium  as  sodium  selenate   (96-hour   LC50  of  2.0
mg/l  under  flowthrough  conditions)  but  there  were no  studies reported  1n
which  the  sensitivity of  rainbow trout  to  selenium as  sodium  selenate  was
assessed.  Zebraflsh,  Brachyodanlo  rerlo. and juvenile  striped  bass,  Horone
saxatms.  were   the  most  tolerant  of   exposure  to  selenium  as   sodium
selenate,  with  96-hour  LC5Qs  of  82  and  85.8 mg/l,  respectively  (N11m1
and Laham, 1976;  Klauda, 1986).
    The  toxldty of selenium  as selenium  dioxide ranged   from  a  96-hour
LCcQ  of   7.3  mg/l  for  fathead  minnows,   Plmephales  promelas   (Cardwell  et
al.,  1976)  to 20 mg/l  for zebraflsh,  Brachyodanlo rerlo (N11m1  and  Laham,
1976).   The   sensitivity  of  fathead  minnows to selenium either  as  sodium
0144d                               4-1                              04/01/89

-------
                                TABLE 4-1



Median Response Concentration for Aquatic Vertebrates Exposed to Selenium
Oi 	 — 	 	 — 	 	 	
a. Median Response Concentration








i






o
^^
o
oo
Species
Rainbow trout
Rainbow trout
Salmo qalrdnerl


Salmo qalrdnerl
Salmo qalrdnerl
Salmo qalrdnerl
Salmo qalrdnerl
Rainbow trout
Salmo qalrdnerl
Brook trout
Salvellnus fontanalls
Northern pike
' Esox luclus
Northern pike
Esox luclus
Goldfish
Carasslus auratus
Goldfish
Carraslus auratus
Carp
Cvprlnus carplo
Cvprlnus carplo
Chemical
Se
Se
Se
Se
Se
Na2Se03
Na2Se03
Na2Se03
Na2Se03
Na2Se03
Se02
Se
Na2Se03
Se02
Se02
Se
Na2Se03
Test
Method
S
NR
NR
NR
NR
FT.M
FT.H
S.M
FT.H
FT
FT
NR
FT.H
S
FT
NR
S
24-Hour
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
36.3 mg/l
NR
NR
NR
71.3 mg/t
(65.8-77.2)
72 mg/t
NR
48 -Hour
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
23.8 mg/l
NR
NR
12 ppm
(7.9-17.0)
46.5 mg/l
(42.6-50.7)
NR
NR
96 -Hour
1.8 mg/l
8.1 mg/l
4.2-4.5 mg/l
12.5 mg/l
7.2-8.8 mg/l
8.8 mg/l
8.2 mg/l
4.35 mg/l
7.2 mg/l
1.80 mg/l
(1.29-2.51)
14.3 mg/l
NR
NR
NR
36.6 mg/l
(26.7-50.2)
35 mg/l
35 mg/l
Comments
LC5Q. hardness = 40 yg/l
"50
"50
"50
"50
LC5Q. fish not fed
LCso. fish fed every other
day
"50
LC50. fish fed dally
LC5o, temperature = 12°C
LC5Q. temperature = 15.5°C
75-hour LC50 =11.1 mg/l
75. 5 -hour LC50 =11.1 mg/l
LC5Q based on mortality after
a 7-day recovery period
LC5Q. temperature = 25.4°C
"50
"50
Reference
Palawskl et al.. 1985
Elsler. 1985
Elsler. 1985
Hsler. 1985
Elsler. 1985
Etnler et al., 1987
Etnler et al.. 1987
Etnler et al.. 1987
Etnler et al.. 1987
Hunn et al.. 1987
Cardwell et al.. 1976
Klaverkamp et al..
1983
Etnler et al.. 1987
Weir and Nine, 1970
Cardwell et al.. 1976
Elsler. 1985
Etnler et al.. 1987

-------
TABLE 4-1 (cont.)
o
o.





i







04/01/8
Species
Fathead minnow
Plmephales promelas
Fathead minnow
Plmephales promelas
Fathead minnow
Fathead minnow
Plmephales promelas
(Juveniles)
Plmcphales promelas
Fathead minnow
Plmcphales promelas
Fathead minnow




Zebraflsh
Brachyodanla rerlo
(larvae)


White sucker
Catostomus coonersonl
Chemical
Se
Se
Se
Se02
45. 7X
Na2Se04
Na2Se04
Na2Se04
Na2Se03
Na2Se03
Na2Se03
Se02
Na2Se03-5H20
(selenlte)
Na2Se04
(selenate)
(selenlte)
K2Se04
(selenate)
Se
Test
Method
NR
NR
S
FT
S
S
S.M
FT.H
S.H
S.H
S.H
S
S
S
S
S
NR
Median
24 -Hour
NR
NR
NR
24.3 mg/l
(16.4-36)
NR
82 mg/t
(76-89)
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
Response Concentration
48 -Hour
NR
NR
NR
NR
NR
NR
NR
NR
«R
NR
NR
NR
NR
NR
NR
NR
48.6 mg/t
96-Hour
0.62-0.97
mg/l
2.2-12.5
mg/l
10 mg/l
7.3 mg/l
(5.7-9.2)
10 mg/t
(5.9-17)
NR
11.8 mg/t
2.0 mg/l
10.9 mg/l
6.7 mg/t
2.8 mg/t
20 mg/l
23 mg/l
82 mg/l
15 mg/l
81 rag/I
31.4 mg/l
Comments Reference
LC50 Elsler. 1985
LCso Elsler. 1985
LCso, hardness = 40 mg/t Palawskl et al., 1985
LCso, temperature = 24.7°C Cardwell et al.. 1976
LC$o Mayer and E Hers leek,
1986
LC5Q Watenpaugh and
Bellinger. 1985a
LCso Etnler et al.. 1987
LCso Etnler et al.. 1987
LCSO Etnler et al.. 1987
LCso Etnler et al.. 1987
LCso Etnler et al.. 1987
LCso Nllml and Laham. 1976
LCso Nllml and Laham. 1976
LCso Nllml and Laham. 1976
LCso Nllml and Laham. 1976
LCso Nllml and Laham. 1976
LCso Elsler. 1985

-------
TABLE 4-1 (cent.)
2 Median Response Concentration
*fc
Q.




4*






O
O
00
Species
White sucker
Catostomus c earner son 1
Catostomus comersonl
Channel catfish
Ictalurus punctatus

Mosqultof Ish
Gambusla afflnls
Striped bass
Ttorone saxltllls


Striped Bass
Moronc saxatllls
(prolarva)
Horone saxatllls
(larva)
Horone saxatllls
(juvenile)
Blueglll sunflsh
Blueglll
Lepoals macrochlrus
Frog
Xenopus laevls
(embyro)
Xenopus laevls
(tadpole)
Chemical
Na2Se03
Na2Se03
Se02
45. 7X
Se
Se
Se
Se
Na2Se04
Na2Se04
Na2Se04
Se
Se02
Se
Se
Test
Method
Fl
Fl.H
FT
S
NR
S
S
S
FT
FT
FT
S
FT
NR
NR
24-Hour
NR
NR
46.7 mg/t
(42.1-51.8)
NR
NR
NR
NR
NR
NR
NR
NR
NR
77.3 mg/t
(72.1-82.8)
NR
NR
48 -Hour
NR
NR
NR
NR
>6 mg/t
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
96 -Hour
31.4 mg/l
29 mg/t
19.1 mg/t
(17.1-21.4)
6.1 mg/t
(4.1-9.0)
12.6 mg/t
1.32 mg/t
(1.09-1.69)
2.40 mg/t
(2.89-3.05)
1.55 mg/t
(0.97-2.47)
9.8 mg/t
(8.3-11.3)
13.0 mg/t
(11.6-14.5)
85.8 mg/t
(81.6-90.0)
4.5 mg/t
NR
4.0 mg/t
NR
Comments
"50
"50
LC5Q temperature = 24.9°C
"50
"50
LCjQ, hardness = 40 mg/t
LC50, hardness = 285 mg/t
LC50. salinity = 1°/00
hardness = 455 mg/t
"50
"50
"50
LC5Q. hardness = 40 mg/t
LC50, temperature = 24.9°C
"50
72-hour. LC5o = 8 mg/t
Reference

Etnler et al.. 1987
Klaverkamp et al
1983
Cardwell et al..
• •
1976
Mayer and E Hers leek,
1986
Elsler. 1985
Palawskl et al.,
Palawskl et al.,
Palawskl et al.,
Klauda. 1986
Klauda. 1986
Klauda. 1986
Palawskl et al..
Cardwell et al..
Elsler. 1985
Elsler. 1985

1985
1985
1985



1985
1976



-------
                                                                           TABLE 4-1  (cont.)
I
en
Species Chemical
Haddock Se
Melanoqrammus
aegllflnus (larvae)
Sheepshead minnow Se
Cyprlnodon varlegatus
Atlanta stlverslde Se
Menldla menldla
Four spine stickleback Se
Apeltes quaduacus
Plnftsh Se
Laqodon rhomb lodes
Mullet Wa2Se03
Mug 11 auratus
(larvae)
Summer flounder Se
Parallchthys dentatus
(larvae)
Winter flounder Se
Pseudopleuronectes
amerlcanus
(larvae)
Striped gouraml SeOp
Collsa fasclatus
Median
Test
Method 24-Hour
NR NR
NR NR
NR NR
NR NR
NR NR
S NR
NR NR
NR NR
S 9.99 mg/t
Response Concentration
48-Hour 96-Hour
NR 0.6 rog/t
NR 7.4-67.1 mg/t
NR 9.72 mg/t
NR 17.4 mg/t
NR 4.4 mg/t
(2.9-6.7)
6.2 ppm NR
(±1.08)
NR 3.5 mg/t
NR 14.2-15.1
mg/t
5.77 mg/t 2.65 mg/t
Comments
1050
LC50
LCso
LC50
LC50
LC50, temperature = 18°C
LCSO
LCSO
LCjQt temperature = 25°C
120 -hour LCso =2.25 mg/t
Reference
Elsler. 1985
Elsler. 1985
Elsler. 1985
Elsler. 1985
Elsler. 1985
Unsal. 1987
Elsler. 1985
Elsler. 1985
Srlvastava and
Tyagl. 1985
     FT = Flowthrough; N = measured; NR = not reported; S = static
 oo
 lO

-------
selenlte or selenate was comparable  (Etnler et al.,  1987)  and  zebraflsh were
equally sensitive to the potassium and  sodium salts  of  selenlte  and  selenate
(N11m1 and  Laham,  1976).   Klauda (1986)  reported  a nearly 10-fold  Increase
1n  tolerance  of  striped bass, Horone  saxatlHs.  to sodium selenate as fish
developed from the prolarva to juvenile stages.
    Studies assessing the  toxldty of  selenium to  saltwater vertebrates were
less  numerous  than  those  dealing  with  freshwater  vertebrates.   Toxldty
values  ranged  from  a  96-hour  LC5Q  of  0.6  mg/ft,   for  haddock  larvae,
Melanogrammus  aegllflnus.  to  96-hour  LC5Qs  of  14.2-15.1 mg/i  for  winter
flounder  larvae,   Pseudopleuronectes  amerlcanus.   exposed to  unidentified
forms of selenium (Elsler,  1985).
    In  another  study,  Hlraoka  et  al.  (1985)  assessed  the  toxldty   of
selenium as  sodium selenlte to  red  medaka,  Oryzlas latlpes.   Investigators
exposed 8-day-old  fry  to  1, 10  and  100 ppm selenium  for  24   hours  at 25°C.
Test  solutions  were prepared  In a  50:50 mix  of  tap  and delonlzed  water.
Observed  mortality  levels  In  the  test  solutions  were  0,  30  and  100%,
respectively.
    Klauda  (1986)  assessed the  toxldty  of  selenate as  the   sodium salt  to
embryos of  the striped bass,  Morone saxatllls.   Embryos  were exposed to  a
series  of   selenium  concentrations   ranging  from  6-2000   mg/i  1n   1  1
volumes.  Solutions were aerated and renewed  dally during the 96-hour test.
The  Investigator  reported  a  93.3X  hatch  and  survival for  eggs exposed  to
2000 mg Se/l for 4 days.
    The acute  toxldty  of  selenium  to aquatic  Invertebrates as  expressed  by
the  LC5Q  1s  presented  1n  Table 4-2.   Ninety-six  hour  LC5Qs  of  33, 1.90
and  0.25  mg/a  were  reported  for  the crab,  Scylla  serrata. bay  scallop,
Argopecten  Irradlans.   and  surf  clam,  Splsula  so11d1ss1ma,   respectively,
0144d                               4-6                              04/01/89

-------
                                                                              TABLE 4-2



                                             Nedtan Response Concentration For Aquatic Invertebrates Exposed to Selenium
Species
Water flea
Daphnla magna
Daphnta magna
Oaphnla magna
Water flea
Oaphnla magna
(5th Instar)
Water flea
Daphnla magna
(5th Instar)
Water flea
Oaphnla magna
(juveniles)
Oaphnla magna
Oaphnla magna
Daphnla pulex
Daphnla pulex
Daphnla pulex
Daphnla pulex
Cladocera
Cerlodaphnla afflnls
Copepod
Acartta clausl
Acartla tonsa
Chemical
Se
Se
Na2Se03
Ma2Se03
Na2Se04
Na2Se04
45. 7X
Na2Se03
45. 7X
Na2Se03
Ma2Se03
Na2Se03
Na2Se03
Na2Se03
Na2Se03
Se
Se
Test
Method
S
NR
FT
SR
SR
SR
S
S
NR
S
S
S
S
NR
NR
Median
24 -Hour
0.66 mg/t
(0.53-0.87)
NR
NR
1.65 ppn
1.51 ppm
NR
NR
NR
NR
NR
NR
NR
0.76 mg/t
(0.64-0.90)
NR
NR
Response Concentration
48-Hour
0.43 ng/t
(0.35-0.57)
<0.25 mg/t
0.71 mg/t
0.68 ppn
0.75 ppm
0.55 ppm
(0.50-0.60)
4 mg/t
(3.0-5.2)
1 mg/t
(0.5-1.9)
3.87 mg/t
(3.5-4.46)
1.37 mg/t
0.61 mg/t
0.098 mg/t
0.60 mg/t
(0.51-0.71)
NR
NR
96-Hour
NR
0.71 mg/t
NR
0.44 ppm
0.58 ppm
NR
NR
NR
NR
NR
NR
NR
NR
1.74 mg/t
0.80 mg/t
Comments
"50
"50
"50
"50
"50
"50
EC5Q, temperature = 18°C
hardness = 280 mg/t
EC5Q, temperature = 18°C
hardness = 40 mg/t
"50
"50
"50
"50
LC5Q. temperature = 23°C
"50
"50
Reference
LeBlanc. 1980
Elsler. 1985
Etnler et al.. 1987
Johnston. 1987
Johnston. 1987
Johnston, 1987
Mayer and
Ellersleck. 1986
Mayer and
Ellersleck. 1986
Reading and Bulkema,
1983
Etnler et al.. 1987
Etnler et al.. 1987
Etnler et al.. 1987
Owsley and McCauley.
1986
Elsler. 1985
Elsler. 1985
O9

-------
TABLE 4-2 (cont.)
o
— ' Median Response Concentration
^
Q.






4*
00





04/01/89
Species
Copepod (cont.)
Hyallela azteca
Hayallela azteca
Notocalllsta sg.
Allorchestes
compressa
Allorchestes
compressa
Cyclaspts usltata
Scud
Hyallela azteca
Blue crab
Calllnectes sapldus
Dungeness crab
Cancer maglster
(larvae)
Crab
Cancer maglster
(zoeae)
Crab
Scylla serrata
Shrimp
Hysldopsls bah la
(adult)
Hysldopsls bahla
(juvenile)
Brown-shrimp
Penaeus aztecus
Chlronomus plumosus
Chemical
Se
Na2Se03
Na2Se03
Na2Se03
Na2Se03
Na2Se03
Na2Se04
Se
Se
Se02
SeOp
Se
Se
Se
45. 7X
Na2Se03
Test
Hethod
NR
FT
FT.H
FT.H
FT.H
FT.H
FT.H
NR
NR
S
S
NR
NR
NR
S
24 -Hour
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
68 mg/t
NR
NR
NR
NR
48 -Hour
NR
0.94 mg/t
NR
NR
NR
NR
NR
NR
NR
NR
46 mg/t
(39.5-52)
NR
NR
NR
50 mg/t
(41-61)
96-Hour
0.34 mg/t
NR
2.88 mg/t
6.17 mg/t
4.77 mg/t
6.12 mg/t
0.76 mg/t
4.6 mg/t
(2.7-7.8)
1.04 mg/t
>10 mg/t
33 mg/t
(28.4-38.3)
1.5 mg/l
0.6 mg/t
1.2 mg/t
(0.8-1.8)
NR
Comments Reference
IC50 Elsler. 1985
LC5Q Etnler et al..
LC5Q Etnler et al. .
LCso Etnler et al. ,
LC5Q Etnler et al..
LC5Q Etnler et al..
LC§o Etnler et al..
lC5o Elsler. 1985
LC$o Elsler. 1985
LC5Q. temperature = 15°C Martin et al..
LCso Krlshnaja et al
1987
LCso Elsler. 1985
LC50 Elsler. 1985
LCso Elsler. 1985
ECso, temperature = 22*C Mayer and
hardness - 280 mg/t Ellersteck. 198


1987
1987
1987
1987
1987
1987


1981
••


6

-------
                                                                          TABLE 4-2 (cont.)
Median Response Concentration
Species Chemical Test
Method 24-Hour 48-Hour 96-Hour
Brown-shrimp (cont.)
Chlronomus plumosus 45. 7X S NR
Mosquito larvae Se NR NR
Culex fatlgans
Nidge Se NR NR
Tanytarsus dlsslmllls
Snail Se NR NR
Physa
Oyster SeO? S NR
Crassostrea glgas
(embryos)
Mussel Se02 S NR
Mvtllus edulls
(embryos)
Bay scallops SeOp S NR
Argopecten Irradlans
(juvenile)
Surf clam SeOp S NR
Splsula solldlsslma
(Juvenile)
36.5 mg/t NR
(25.2-52.9)
<3.10 mg/t NR
NR 42.4 mg/t
>10 mg/t 24 mg/t
>10 mg/t NR
>10 mg/t NR
NR 0.25 mg/t
(0.23-0.28)
NR 1.90 mg/t
(1.79-2.01)
Comments Reference
ECjQ, temperature = 22°C Mayer and
hardness = 39 mg/t Ellersleck. 1986
LC50 Elsler. 1985
LCso Elsler. 1985
LCso Elsler. 1985
EC50. temperature = 20°C Martin et al.. 1981
EC50. temperature = 14°C Martin et al.. 1981
LC50. temperature = 20°C Nelson et al.. 1988
LC5Q. temperature = 20°C Nelson et al.. 1988
    FT = Flowthrough;  H = measured;  NR =  not  reported;  S =  static;  SR  =  static  renewal
o



03

-------
exposed to  selenium  as  selenium dioxide  (Krlshnaja  et  al., 1987; Nelson et
al.,  1988).   Martin et   al.   (1981)  reported  that  the  96-hour   LC5Q  of
selenium when  derived  from selenium dioxide  for  crabs,  oysters and mussels
was >10 mg/l.
    The toxldty of selenium as sodium selenlte was  reported  for  water  fleas
and amphlpods  (Reading and  Bulkema,  1983; Owsley and McCauley, 1986;  Etnler
et  al.,  1987;  Johnston,  1987).  The  48-hour LC5Qs  for  Daphnla  magna were
0.68  and  0.71  mg/l,  while those  for  Daphnla pulex ranged from 0.098-3.87
mg/i.   The   48-hour  LC,.  of   selenium  as  sodium  selenlte   to  another
cladoceran,  Cerlodaphnla  afflnls.  was  comparable with that  for  species of
Daphnla   (0.60  mg/l).     Amphlpods   were   slVghtly   less   sensitive  than
cladocerans  to selenium  as sodium  selenlte. The 96-hour  LC5Qs for several
species ranged from 2.88-6.17  mg/a.  Mayer  and  Ellersleck (1986)  reported
a  48-hour  EC5Q  that  was 4-fold   lower  for  D.  magna  when assays  were
conducted  1n  soft  water,  but  no  significant   differences  for  Chlronomus
plumosus  when  assays  were  conducted  In hard  and  soft   water.   U.S.  EPA
(1987a) reported the results  of additional studies   1n which  the  toxldty of
selenium  as selenlte ranged  from 0.21  mg/l for  D_. magna  to  203  mg/a  for
the leech, Nephelopsls obscura.
    The toxldty of  selenium  as sodium selenate  to  5th  Instar  Daphnla  magna
was  comparable to  that  observed for  sodium selenlte (Johnston,  1987).   The
48-hour LC   s  for  selenium as  sodium selenate were 0.55  and 0.75  ppm.   In
contrast, the  toxldty of  selenium as  sodium selenate to  Hyallela azteca  was
4- to 8-fold  greater than  that  observed  with sodium selenlte (Etnler et al.,
1987).  The 96-hour LC5Q was 0.76 mg/i.
    Some  Investigators   did  not  Identify  the  form of  selenium  to  which
aquatic  Invertebrates  were  exposed  (LeBlanc,   1980;  Elsler,  1985).   The
0144d                               4-10                             06/15/89

-------
amphlpod, Hyallela azteca. was the most  sensitive  Invertebrate  Identified  1n
this  category,  with  a  96-hour  LC5Q of  0.34 mg/i.   The  midge,  Tanytarsus
d1ss1m111s.  was  the  most  tolerant  Invertebrate  Identified,  with a  96-hour
LC5Q of 42.4 mg/1.
    In  other   studies,  Adams  and  Heldolph  (1985)  conducted  static  tests
exposing  fed  and unfed  Daphnla  magna  to selenium  as  sodium  selenlte.  The
48-hour  EC5_s  for unfed  and  fed  daphnlds  were 0.47  and  1.5  mg/a,  respec-
tively.  The 48-hour NOEC for  unfed daphnlds was 1.0 mg/l.
    Pagano et al. (1986) assessed  the toxlclty  of  selenium  to embryos of  the
sea urchin,  Paracentrotus  11v1dus.  Gametes  were collected  from sacrificed
adults.  Fertilized embryos were  exposed for the duration of  their  develop-
ment.    Investigators  observed  a  blockage  of  development  at  the  gastrula
stage for embryos exposed to 55.3 ppm selenium.
    Wolfenberger  (1986)  assessed  the  survival  of  the hermit  crab,  CUban-
arlus  vlttatus. after a 24-hour period of exposure  to  100 ppm selenium under
varying  conditions  of  salinity  and temperature.   Mean survival  time  was
greatest  In  20  °/    salinity  (~80  and  140  hours  for  two  populations)
when assays were  conducted  at 16°C.   At 20 and 24°C,  mean  survival  time  was
greatest  for   crabs   exposed   to   selenium  1n  10   /    salinity  solutions
(~80 and 215 hours,  and 130 and 70 hours, respectively).
    Johnston  (1987)  reported LT5Qs for   the  water  flea,  Daphnla  magna.
exposed  to  selenium  as  sodium selenlte and selenate.   LT5_s for  daphnlds
exposed  to  3.0,  1.0,  0.8, 0.6 and 0.5  ppm selenium as  sodium  selenate were
17, 31,  55,  86 and 150  hours, respectively.  LT5Qs for daphnlds exposed  to
the same concentrations of selenium, but as  sodium selenlte,  were 9.5, 26.5,
50, 74 and 99  hours, respectively.  The  IT™ for  daphnlds  exposed  to  0.4
ppm selenium  as  sodium  selenlte  was 130  hours.  The  Investigators  reported
0144d                               4-11                             04/01/89

-------
that  sodium  selenlte  was  1.42  times  more  toxic  to  daphnlds  than  sodium
selenate  based  on  the   Incipient  LC  s  of  0.31  and  0.44  ppm  selenium,
respectively.
    In a  series  of behavioral  and  physiological  studies,  several  groups  of
Investigators  assessed  the  effects of  selenium on  avoidance  responses  by
fish and  respiration  by  crabs and  fish.  Weir  and  H1ne  (1970) reported that
goldfish,  Carasslus  auratus,  exhibited  behavioral  Impairment  at   0.25  ppm.
Watenpaugh  and  Beltlnger  (1985a)  reported  that fathead minnows,  Plmephales
promelas.  did not  avoid  selenate  at  concentrations  of  0.3-11.2 mg  Se/l.
Watenpaugh  and Beltlnger  (1985b)  reported  that  oxygen  consumption  by  fathead
minnows was not  affected by  exposure  to 60  mg. Se/l  for  24  hours,  although
minnows  at  this   level  may  have  physiologically  compensated  to  maintain
oxygen consumption rates.  Wolfenberger  (1987)  reported  that  oxygen consump-
tion by  the hermit crab, C11banar1us  vlttatus. exposed to 100  ppm selenium
was  depressed at  16°C  1n   10   /    salinity  test medium,  although  oxygen
consumption  rates  were  generally  elevated  at  lower  salinities  and  higher
temperatures.
4.1.2.   Chronic Effects on Fauna.
    4.1.2.1.   TOXICITY — N11m1  and  Laham (1975)  assessed  the toxlclty  of
selenium  as  selenium  dioxide  to  embryos   and  larvae   of  the  zebraflsh,
Brachydanlo  rerlo.   Embryos  and  larvae  were  exposed  to selenium  at  26°C  1n
covered petrl dishes 1, 2.5,  7  and  27  hours after  fertilization (correspond-
ing to Stages 3-5, 10,  15 and 21  of normal embryonic  development).  Exposure
media were  renewed dally.  Hatching and  mortality  of zebraflsh embryos  was
unaffected  by exposure to <10 mg/8, for  168 hours of  treatment  for  any  of
the embryonic  stages.   Mortality of larvae after hatching was significantly
affected  by   selenium  at   concentrations  >3  mg/a  after   168   hours   of
treatment and at 10 mg/i after 96 hours of treatment.
0144d                               4-12                             04/01/89

-------
    Cardwell  et  al.  (1976)  assessed  the  toxlclty of  selenium as  selenium
dioxide   to   fathead   minnow,  Plmephales   promelas.   goldfish,   Carrasslus
auratus.  bluegUl   sunflsh,  Lepomls  macrochlrus.   and  flagflsh,  Jordanella
florldae.  Fish  were  exposed to selenium under  Intermittent  flow  conditions
at  24-25°C  for   120-336   hours.    The 120-,  144- and  168-hour  LC5 s  for
fathead  minnows  were  4.5,  3.2  and  2.9  mg Se/l,  respectively.   The  120-,
216-,  144-,  168-,  264- and   336-hour  LC5()s for  goldfish  were 32.7,  22.3,
17.2,  13.0,  11.5  and  8.8  mg/l,  respectively.    The  168-,  192-,  240-,
288- and  336-hour  LC5Qs for  bluegllls were  30.7,  27.7, 23.6,  20.5  and 17.6
mg   Se/a,  respectively.    Investigators   reported   LT,. s   (1n    Heu   of
LC5Qs) of  83,  67,  68,  55  and 44  hours  for  flagflsh exposed  to 11.2,  16.9,
21.8,  27.9 and  37.6  mg  Se/l, respectively.   Cardwell et  al. (1976)  also
examined  the  latent  effects  of exposure  to selenium 1n fathead minnows  and
flagflsh.  Fish  exposed to  <33.2  mg  Se/a  for  24  hours were  then  monitored
1n  uncontamlnated  water for  28  days.  No  effects  on survival or  growth  of
any of the treated groups of fish were reported.
    Klaverkamp et  al.  (1983) assessed the  toxlclty of selenlte selenium  to
yellow  perch, Perca  flavescens.   1n  an  Intermittent  flowthrough  exposure
study.  Perch were  collected by  beach seine from  a mercury-polluted lake  1n
northwestern Ontario.   A 240-hour  LC-n of 4.8 mg Se/J. was  reported.
    Reading and  Bulkema (1983)  assessed  the chronic  toxlclty  of selenium  as
sodium selenlte  to  Daphnla magna  In a 28-day  static-renewal  test.   Exposure
concentrations were 0.2,  0.4, 0.6  and 0.8 mg Se/i.  Daphnlds  were  fed  dally
a  50,000  cells/mi,  algal suspension.   No  effects on  total  eggs per  daphnld,
live  young per  daphnld and  mean brood size were reported, except at 0.8  mg
Se/l.
0144d                               4-13                             04/01/89

-------
    Adams  and  Heldolph (1985)  assessed  the toxldty  of  selenium as  sodium
selenlte  to  Daphnla  magna 1n 21-day static-renewal  tests.   Selenium  concen-
trations  were  measured throughout the study.   Daphnlds  were fed  trout  chow
at  a  concentration  of 30  mg/8.  and  test  solutions were  renewed 3  times/
week.    The Investigators  reported  7-,   14- and  21-day EC^_s  of  0.38,  0.38
and 0.35  mg/l, respectively.   The  NOEC  based  on  survival  for  all  observa-
tion  periods  was  0.24  mg/l.    The  NOEC  based  on  reproduction  for   all
observation periods was >0.24 mg/l.
    Ahsanullah and Brand  (1985)  assessed the toxldty of  selenium as  sodium
selenlte  to  the  amphlpod,  Allorchestes  compressa.   First  Instar juveniles
were exposed  to  a continuous flow  of  sodium selenlte stock  solutions mixed
with  seawater  (32-35  °/00)  for  4  weeks.   Exposure  to  44  yg  Se/8.   did
not significantly  affect  either  growth  or  survival  of  treated  Individuals.
Amphlpods  exposed to  193  yg  Se/8.  experienced 100%  mortality  by  the  end
of  the   study, while  animals  exposed   to  93  yg  Se/B.  experienced   reduced
growth  and a  significant level  of mortality  over  the  4-week  study.   The
estimated  MATC  for  Allorchestes  compressa  juveniles   exposed  to   sodium
selenlte  1s 44-93 yg Se/l.
    Elsler (1985)  reported  the  results  of studies  assessing  the  toxldty  of
selenium  to  several  aquatic  Invertebrates.   The  14- and  28-day LC  s  for
the waterflea, Daphnla magna.  exposed  to selenium were 0.43  and  0.24  mg
Se/i,  respectively.    The  14-day  LC5Q   for  scud,  Hyallela  azteca.   exposed
to  selenium  was  0.07  mg  Se/l.   Elsler  (1985)  also  reported the  results  of
studies   assessing the toxldty  of selenium  to  aquatic vertebrates.   The
113-hour  LC5Q  of  selenium  to  embryos  and  the  5- and  7-day LC  s  Of
selenium  to  tadpoles  of the  frog,  Xenopus  laevls. were  2.0, 2.6  and  1.5  mg
Se/l,  respectively.    The  48-day  LC5Qs   for  selenium  1n blueglll  sunflsh.
0144d                               4-14                             04/01/89

-------
Lepomls macrochlrus.  and fathead minnow,  Plmephales  promelas. were  0.4-2.0
and  1.1  mg  Se/a,  respectively.   The 10-day  LC5Q In  perch,  Perca  flaves-
cens,  was  4.8  mg/a,  while the  43-day  LC™  In  coho  salmon,  Oncorhynchus
klsutch. was  0.16 mg/a.   The 9-,  21- and 96-day  LC5Qs  for  rainbow  trout,
Sal mo  qalrdneM.  exposed to  selenium were 5.4-7.0,  0.46  and 0.29 mg  Se/a,
respectively.
    Bennett  et  al.  (1986)  assessed   the  uptake  and  transfer  of  selenium
through an  aquatic  food  chain (algae,  Chlorella  pyrenoldosa. to  rotifers,
Brachlonus   calydflorus)   and  Its   effects   on   fathead  minnow   larvae,
Plmephales promelas.  Algae  were  cultured  1n  the presence of  selenium  for  3
days,  with a maximum concentration of selenium  1n the culture media of  2.5
mg  Se/a.   Rotifers  were permitted  to  feed  for 5  hours  In  selenium-laden
algae   before  harvesting.    The   algae:rot1fer   ratio   (yg   algae/maiyg
rotifer/ma)  was  25:50.  The concentration of  selenium 1n rotifers  reached
a  plateau  at  ~40  yg Se/g  rotifer.   Three  feeding  experiments  were  con-
ducted with  larval  fathead  minnows.   The  first experiment entailed  feeding
minnows  contaminated  rotifers  for   7   days  commencing  4  days  posthatch,
followed  by   19  days  on a  control   diet.   The second experiment  entailed
control diets for  8  days posthatch, followed  by  a  selenium-contaminated diet
for  9  days.   The  third   experiment entailed  control  diets  for  2   days
posthatch  and  a   selenium-contaminated  rotifer   diet  for  7   days.    No
mortalUes among  larval  fish  could  be  attributed  to dietary selenium,  but
there  were significant  reductions In  final dry  weight of treated  larvae In
the   first   and   second  experiments.   The   mean   final   larval   selenium
concentrations   1n   these   two   experiments   were   43  and  51.7   yg/g,
respectively.
    G11lesp1e  and   Baumann   (1986)   assessed   the  effect   of  high   tissue
concentrations of  selenium  on the reproductive  success of blueglll  sunflsh,
0144d                               4-15                              06/15/89

-------
Lepomls macrochlrus.  Adult bluegllls were  collected  by  electroshocklng  from
a  cooling water  Impoundment  and  a  municipal  water  supply  lake  In North
Carolina.  Males  and  females  were mated  with  fish both from  their own  lake
and the other water body.  Fertilization  success was  determined  by  stripping
eggs from females and  combining with sperm J[n vitro.   Percent fertilization
was determined  from the proportion  of  mltotlcally active zygotes  2-3 hours
after   fertilization.   Hatching success was  estimated by  holding  fertilized
eggs In  closed  aquaria with redrculatlng  dty  lake water at  22-25°C.   The
Investigators reported  significant  differences  1n percent fertilization  or
hatch   among  parent combinations.   Larvae  from  crosses Involving  selenium-
contaminated  females   had  gross  abnormal  morphologies   characterized  by
general  edema.   Concentrations of  selenium 1n  the ovaries  of the  females
ranged from  5.79-8.0 mg/kg  wet weight;  these levels were  16-21  times  higher
than those 1n uncontamlnated fish.
    Klauda (1986) assessed  the effects of  exposure of early  life  stages  of
the  striped  bass,  Morone  saxatllls.   to  selenium   as   sodium   selenate.
Prolarvae,  postlarvae   and  Juveniles  were  exposed  to  low (0.089-0.099  mg
Se/8.)   and  high  (1.217-1.360   mg  Se/l)  concentrations  of  selenium under
constant  flow conditions  for   60  days.  A  significant  reduction  In expo-
nential  growth  rates  as  determined  by  total  length  was  reported for  both
selenium treatments for days 3-15  compared  with  controls,  but  no significant
difference based  on dry weight.  There were no  differences  observed  between
days 19  and  60.  A significant Incidence In developmental  abnormalities  of
the lower jaw and severe blood cytopathology were also reported.
    Owsley and  McCauley (1986) assessed  the  toxldty of  sodium selenlte  to
four consecutive  generations  of CeModaphnla afflnls.   Treatments  consisted
of  exposure  to  0.0, 0.05, 0.1, 0.2,  0.4  and 0.8 mg Se/l  until  three  broods


0144d                               4-16                             06/15/89

-------
were  produced  In the  controls  (~8 days).   Test  solutions were  renewed  and
organisms were fed  every  other  day.  The assays demonstrated  a  reduction 1n
the  tolerance  of Cer1odaphn1a aff1n1s  In succeeding generations.   The  NOEL
was  reduced from  0.2  to  0.1  mg/a  after   two  generations  of   exposure  to
selenium.
    Hunn et  al.  (1987) assessed  the  chronic toxldty of  selenium  as  sodium
selenlte  to  rainbow  trout  fry,   Salmo  galrdnerl.  1n  partial  life-cycle
studies.  Fry were  exposed  to selenium under flowthrough  conditions at  12°C
for  90  days.   F1sh were  fed  dally a commercial diet supplemented  with  live
naupl11  of  brine  shrimp.  Concentrations  of selenium  1n exposure  vessels
were determined by hydride  generation atomic  absorption.   Exposure  of  fry to
>47  yg  Se/a significantly  reduced survival  and  fry length  after   90  days.
Survival  and  growth  were   not  affected   at  <21   yg   Se/a.    Weight   was
significantly  reduced  at  100   yg  Se/a  by the  end  of  the  study.    The
highest  body burden   levels  of  selenium were  achieved  after   30  days  of
treatment and declined thereafter.   BCFs  were Inversely  related  to exposure
concentrations  but  did  not  exceed  200  yg/a.   Exposure  of  fry  to >12  yg
Se/a  significantly   reduced   the  calcium   concentrations   1n   bone.    The
Investigators suggested   that  12 yg  Se/a represents  a  NOEL  for  Inorganic
selenium 1n rainbow trout.
    Johnston (1987)  assessed  the  subacute  toxldty of  selenium as  sodium
selenate In  daphnlds  (12±6 hours  old  neonates)  exposed  to selenium at  22°C
for  15  days.  Test  medium was   renewed dally.   Daphnlds were provided  with
cultures  of  Chlorella   pyrenoldosa   (108   cells/ma)  as  a  food  source.
Daphnlds   demonstrated   a   concentration-response   relationship    between
Increasing  concentrations  of  selenium (0.025-0.5  ppm) and  reductions  In  the
percent Increase In  Initial length  and  mean  numbers  of  eggs produced.   There

0144d                               4-17                             04/01/89

-------
were no significant differences regarding growth between groups at  <0.05  ppm
or  regarding  mean  numbers of  eggs  produced  at <0.025 ppm.  There  were  also
no  differences  observed  In  the  percent maturation  or  mortality  levels  for
daphnlds exposed to 0.025 ppm selenium compared with control  daphnlds.
    U.S. EPA  (1987a)  reported the  results  of  chronic  studies with myslds,
Mys1dops1s  bahla.  and  sheepshead  minnows, CypMnodon vaMegatus, exposed  to
selenlte  selenium.  Number  of  offspring  produced  and   survival  of   first
generation   myslds  were  reduced   significantly  at  0.32  mg/a.  Nonstatlstl-
cally  significant  reductions  1n  survival  (18%)  and reproduction (22%)  were
observed  In  myslds  exposed   to  0.14  mg/a.    Percent  hatch  of  sheepshead
minnows was  <4%  at  concentrations  of 0.97,  1.9 and  3.6 mg/a.   Juveniles
experienced 4,  24  and  90% reductions In survival at 0.47, 0.97 and >0.97  mg
Se/a,  respectively.   Growth of  juvenile sheepshead  minnows  was  reduced  by
8% at 0.47  and 0.97 mg/a.
    Woock et  al.  (1987) assessed the  effects  of dietary  selenium  as  either
sodium  selenlte  or  selenomethlonlne  to  blueglll   sunflsh,  Lepomls  macro-
chlrus.  Effects of dietary  selenium on fish were also assessed  In conjunc-
tion with  waterborne  selenium as  sodium selenlte.   Treatments consisted  of
dietary selenium  as  selenomethlonlne  at  concentrations  of  3,  13 and 30  yg
Se/g,  dietary  selenium  as  sodium  selenlte at 13  and  30  yg Se/g,  and  a
combined treatment  of  dietary selenomethlonlne  at   13 yg  Se/g and  a  water-
borne  concentration  of  sodium selenlte  of  10  yg   Se/a.  Selenium   1n  water
and  food  was  determined  by  hydride  generation  atomic  absorption  spectro-
scopy.    Exposure  of   adult  bluegllls   to  these  treatments  for  260  days
resulted 1n significantly  higher  levels  of mortality and  significantly lower
final  body  weights and  lengths  among fish fed  a  diet supplemented with  30
yg  Se/g.   Exposure  of   adult   bluegllls  to   selenium   for   287-324  days
0144d                               4-18                             04/01/89

-------
resulted  In  reproductive  Impairment  manifested  by  a  reduction  In  larval
survival  and   the  percent  of   larvae  with  teratogenlc  conditions.   Larval
survival  was  reduced  In  groups  that  received a  diet supplemented with  30
yg  Se/g  1n  either   form  and  adults  fed  a  diet  containing 13  yg Se/g  as
sodium  selenlte  combined with a  waterborne  exposure   of   10  yg  Se/l  as
sodium  selenlte.    Exposure  of  larvae  to  the  dletary/waterborne  selenium
treatment and  30  yg/g dietary  selenium  as selenomethlonlne  produced  50  and
100%  of  larvae  with  terata,  respectively.    There  were no  effects  on  the
percentage hatch for  eggs  from adults  treated as described  above.   The  NOEL
for  dietary  selenium 1n  either  form  In  blueglll  sunflsh appears  to  be
between 13 and  30 yg Se/g.
    Boyum and  Brooks  (1988) assessed  the effects  of  selenium In water  and
diet on  Daphnla  magna.  Oaphnlds were exposed to  0.05,  0.1, 0.5 and  1.0  mg
Se/l  as  sodium  selenate  In   culture  water  at   23°C  for   28  days.    Test
solutions were  renewed twice weekly.  Daphnlds were fed  Chlamydomonas reln-
hardtH dally.    In  another part  of  the  experiment, algae exposed  to  sodium
selenate  at  the concentrations  listed  above  were fed  to  daphnlds for  the
28-day  treatment  period.   In  daphnlds  exposed to  1.0 mg dissolved Se/l  or
fed algae exposed  to 1.0 mg Se/l  10054 mortality  occurred In 16 and 7 days,
respectively.    Daphnlds   given  algae  exposed  to  0.75  mg  Se/l  experienced
100% mortality  after 18 days.   Mortality  was  80-90%  1n  daphnlds  exposed  to
0.05,  0.1  and  0.5   mg dissolved  Se/l  for  28 days.   Daphnlds  given algae
exposed to 0.05-0.5 mg Se/l experienced  30-50% mortality during  the  28-day
study.  Control mortality  1n both  studies  ranged  from  10-20%. There were no
significant   differences   1n  numbers of  offspring  produced   among  controls.
Total  number   of  offspring  produced  by  daphnlds  fed  selenium-laden  algae
exceeded that of control  daphnlds  In all  treatment  groups except  for the 1.0
0144d                               4-19                             04/01/89

-------
mg  Se/8,  treatment  groups.   The  Intrinsic  growth   rate  for  daphnlds  was
always  significantly  greater  In  daphnlds  fed  selenium-laden  algae  and
declined  with  an  Increase  1n  the  concentration  of  selenium  1n which  the
algae were cultured.
    4.1.2.2.   BIOACCUMULATION/BIOCONCENTRATION — Barrows   et   al.   (1980)
assessed  the bloconcentratlon and  elimination  of  selenium (as selenous add)
by blueglll sunflsh, Lepomls macrochlrus.   F1sh were  exposed In open aquaria
under  flowthrough  conditions with  a 5-6  volume  turnover/day.   Sample  size
for  the  28-day  exposure portion  of  the study was 100  fish.   Fish remaining
at the  end of  the  exposure phase  were transferred  to  clean  systems  for  a
7-day depuration  phase.   The mean water  concentration for  selenium  1n  this
study was  120  yg/l.   The  maximum BCF  was 20 and  the tissue  half-life  was
between 1 and 7  days.
    Flnley  (1985)   assessed  the  uptake  of  selenium  by  blueglll  sunflsh,
Lepomls macrochlrus.  fed  mayfly nymphs, Hexagenla llmbata.  collected  from  a
selenium-contaminated  lake.   Control sunflsh  were  fed  a low-selenium  diet
consisting  of  mealworms.   F1sh were  fed  to  satiation  twice  dally  for  44
days.   There  were  four  fish In  each  treatment  group.   F1sh  were held  In
glass  aquaria   at   21PC   under  constant   flow  conditions.   Three  of  four
bluegllls  fed  selenium-contaminated  mayflys  died  on  days  17,  35  and  44.
Selenium  concentrations  1n skeletal  muscle of  those fish were  5.1,  5.4  and
7.9  yg  Se/g wet weight,  respectively.   Liver  concentrations were  8.5,  15.1
and  86  yg  Se/g,  respectively.   Concentrations   of  selenium  1n muscle  and
liver of  the remaining  fish were  7.5  and  69 yg/g,  respectively.   Concen-
trations  of  selenium  1n  muscle  of  control fish  ranged   from  1.8-2.1  vg/g.
Concentrations   1n  the livers of  two fish were 5  and 7 yg/g.   The  Investi-
gators  also  reported   that  treated  fish had  a variety of hlstopathologlcal
abnormalities.

0144d                               4-20                             04/01/89

-------
    Bertram  and  Brooks  (1986)  assessed  the accumulation  of  selenium  from
food and  water by  fathead  minnows, Plmephales  promelas.   F1sh were  either
exposed to waterborne selenium  (as  sodium selenate),  offered  daphnlds  raised
on  selenium-contaminated  algae, Chlamydomonas,  or  subjected to both  treat-
ments  under  flowthrough conditions  at  25°C.  Selenium  water concentrations
were  10,   20  and  40  ng  Se/ml.    Concentrations  of  selenium  1n  the  food
supply were  1.33,  3.66  and  7.32 jjg Se/g.   F1sh were exposed  to  waterborne
selenium  and  both  treatments   simultaneously   for   8  weeks.   Duration  of
treatment for  fish fed  a  selenium-contaminated diet  was  11  weeks.   A minimum
depuration period  of 2  weeks  followed the  termination  of treatment  1n  all
cases.   Uptake of  selenium by  minnows  exposed ,to  10 and 20  ng/mi  reached a
plateau  at  a  body  burden  level  of -0.5  vg Se/g   fish  after 25-30  days.
Uptake of  selenium by  minnows  exposed to  40 ng/mi  reached  a plateau at a
body burden  level  of ~0.8  ^g  Se/g  fish  after  20  days.  All  groups  demon-
strated  a  linear  decrease 1n  body burden   levels  of   selenium  to ~3  yg/g
after  a  30-day depuration period.   Whole body burden levels of  selenium 1n
fish  offered  a  selenium-contaminated   diet   demonstrated  a   linear  Increase
over the  course  of  the treatment  period for all  treatments  reaching ~0.3,
0.5  and  1.2  jjg   Se/g   fish  for  the   low,   moderate and  high  treatments,
respectively,  after  80  days.   Depuration of  selenium was  observed  following
the conclusion of  the exposure  phase but  at  a slower rate than that observed
1n  fish  exposed to  waterborne  selenium.  Fish exposed  to   both sources  of
selenium  also demonstrated  a  linear  Increase  1n   body  burden  levels  of
selenium  without   reaching  a   plateau  during   the   56-day   exposure  phase,
reaching  0.4,  0.9  and  2.0  pg  Se/g fish  for  the   low,  moderate  and  high
treatments, respectively.  Depuration of  selenium from these fish was slow.
0144d                               4-21                             04/01/89

-------
    Hodson  et  al.  (1986)  assessed  the  accumulation of  waterborne  radio-
labeled selenlte by rainbow trout, Salmo galrdnerl. eggs,  fry  and  juveniles.
Fertilized eggs were obtained from a laboratory brood  stock  and  Incubated  at
10-12°C.   Eggs  were  exposed   to   selenium   In   cylindrical  mesh-bottomed
containers  In  30 i  tanks.   Eggs   were  collected  each  day  for  4  days.
Sac-fry were  held In  the same  containers as  described  for  the eggs,  while
sw1m-up fry  swam freely  through  the 30 l  tanks.   Fry were collected  dally
for  4  days.   Eggs   and   fry   were exposed   to   0.4  and   45.6   yg  Se/a.
Juvenile  trout were  fed  a caseln-torula  diet  supplemented with 0.4  yg  Se/g
of dry  feed.   After 12 weeks on this diet,  fish  were held without  feeding
for 4  days  and exposed to  the  low  and  high  treatments  as  described  above.
Fish were subsequently sampled  over the  ensuing  96 hours.  Eggs  exposed  to
the  high-dose   treatment  accumulated   10-100  times  more   selenium  (~100
yg/kg)  than  eggs  1n  the  low-dose  treatment  (<10  yg/kg).   The majority  of
the uptake  occurred  In the Initial  24  hours.  Swim-up  fry  accumulated  more
selenium  than  sac-fry.  Fry  experienced  a linear   Increase In  the  concentra-
tion of  selenium over the 96-hour  exposure period with no  Indication  of  an
approaching  plateau.   Concentrations  of selenium  1n sw1m-up  fry from  the
high  dose approached  1000  yg/kg by the end  of  the  study.   Concentrations
of selenium In juveniles  Increased as a  function  of both exposure  concentra-
tion  and  time.   Concentrations  1n  liver  were  the  highest  (-1000  yg/kg)
among  tissues  examined  (liver,  viscera,  gill, muscle  and  gill) after  96
hours.   BCFs  for  the  low-dose   fish ranged  from  3.1 1n  embryos  to  104  In
livers of juveniles.   BCFs  for  the high-dose  fish  ranged from 1.6  1n sac-fry
to 31.6 1n livers of juveniles.
    Klauda  (1986)  measured  the  uptake  of selenium by prolarvae,  postlarvae
and juveniles  of  striped  bass,  Horone saxatlHs.  exposed  to sodium selenate

0144d                                4-22                             06/15/89

-------
for 60 days under  flowthrough  conditions.   Fed  and  starved  Juveniles  exposed
to  1.29  mg/l  accumulated  selenium at  comparable  rates,  producing BCFs  of
0.68  and  0.69.   Fed juveniles  exposed  to 90  jig  Se/l  showed no  Increase
In whole body levels of selenium, while starved juveniles had a BCF  of  11.78.
    Klelnow and  Brooks  (1986) assessed  the uptake, distribution and  elimi-
nation  of  orally  administered  selenium  as the  radlolabeled  [75Se]  sodium
salts  of   selenate  and   selenlte,  and   l-selenometh1on1ne   In  the  fathead
minnow,  Plmephales  promelas.    A  single  oral  dose  of  test  solution  was
administered  to  fish by  gavage  at a  level  of  20 ng Se/g tissue wet  weight.
Fish were held In  flowthrough  aquaria  at  25°C  and collected at various times
over a 40-day period.  Selenium  accumulated to  a  greater  degree 1n  liver  and
kidney  compared  with  muscle,  adipose tissue  or  gonads.    The  pattern  of
accumulation  was  generally  similar   for  each  of  the  compounds  tested,
although   selenate    preferentially    accumulated    1n    the   blood   and
selenomethlonlne 1n  the  heart.   Other  significant  differences  1n  magnitude
of uptake during the absorption and early  distribution  phases  were observed
for  the  liver, bile,  kidney,  spleen  and gill.   The  half-lives In days  for
selenate,   selenlte  and   selenomethlonlne   ranged  from  3.9  (liver)   to  487
(adipose tissue),  2.2  (liver) to  64  (heart) and 1.0 (liver)  to 69 (adipose
tissue), respectively.
    Pelletler  (1986) assessed  the  accumulation  of  selenium by the  mussel,
HytHus  edulls.   In  the   absence and  presence  of  Inorganic  and  organic
mercury.  Forms of  selenium  to which  mussels were exposed Included  Inorganic
dissolved     selenlte     Na2SeO~     and     adsorbed     organic     selenium
(CghLOpSe)-.   Mussels   were   exposed   to  50   yg  selenlum/l  for   15-50
days  In  tanks  provided  with  continuously circulating seawater.   Mussels
accumulated Inorganic selenium but not  organic  selenium at  a rate  of 0.12 ng
Se/g/day  In  the  absence  of  mercury.   The presence  of  Inorganic   (30  ^g

0144d                               4-23                             06/15/89

-------
Hg/a)   and   organic   (3  yg   Hg/1)  mercury   Increased   the   uptake   of
Inorganic  selenium  to  0.24  and  0.40  ng  Se/g/day,  respectively.   Organic
selenium became available  to  the mussels 1n the presence  of methyl  mercury,
and accumulated at a rate of 0.15 ng Se/g/day.
    Etnler  et  al.  (1987)  reported  the  results  of  studies  assessing  the
accumulation of   selenium  In  carp,  Cyprlnus  carplo,  and  fathead  minnows,
Plmephales promelas.  The BCF  for  selenium  as  sodium  selenlte 1n carp for an
unspecified exposure  duration and  selenium concentration was 0.6-6.0.   The
BCF  for  selenium as  sodium  selenlte 1n  minnows  exposed  to 83  yg/ft,  for  28
days was 4443.
    Boyum  and  Brooks   (1988)  assessed  the  uptake  of  selenium  as  sodium
selenate by Daphnla magna.  Daphnlds  were exposed  to  radioactive selenium at
a  concentration   of  0.5  mg/8,  In  algae-free  water  and  1n solutions  with
selenium-laden  algae.   A 72-hour duration  exposure phase was followed  by a
96-hour  depuration  phase.   The  rates  of  uptake  and  depuration  and  total
amount of selenium Incorporated  by  the daphnlds were  reduced In  the presence
of  selenium-laden  algae.   There were no significant  differences among three
treatments of selenium-laden algae  In terms  of  uptake and depuration, except
for  a  single  sample period  for  algae with  the  lowest  ratio of  selenium to
algae.
    Boyum and  Brooks  (1988) also  assessed  the uptake of  selenium  as sodium
selenate  by Daphnla  magna  In the  presence of selenomethlonlne.   Daphnlds
were  exposed  to   selenium  at  a concentration  of  0.50 mg/8.  In one  treatment
augmented  with  1  and  2  mg  Se-methlonlne/i In two  other  treatments.   The
results  Indicated  that  the uptake and depuration  of  selenium were  decreased
by the presence of Increasing concentrations of selenomethlonlne.
0144d                               4-24                             04/01/89

-------
4.1.3.   Effects on Flora.
    4.1.3.1.   TOXICITY — HolUbaugh et al.  (1980)  assessed  the effects of
selenium  on  growth  of  natural  assemblages  of   phytoplankton  In  enriched
seawater.  Water samples  from Sanlch Inlet, B.C., Canada, were  collected  by
peristaltic pump  and augmented with N, P  and  SI.   Water was  then  filtered
through  a  102  ym  pore  mesh  and  transferred   to  experimental  vessels.
Samples were held  1n an  outdoor waterbath  maintained at  12-16°C.  Growth was
monitored  dally  as  changes  In the  In vivo  fluorescence of chlorophyll  a.
There were  no  apparent  effects observed 1n growth of  natural assemblages  of
phytoplankton  or   1n  growth  of a  species  Isolated from  field  samples  of
Tha1ass1os1ra aestevalls exposed to <1000 nM selenium.
    Elsler  (1985)  reported  the results  of studies assessing the  toxlclty  of
selenium  to algae.   The  48-hour  LC5Q  for  Oedogonlum   cardlacum was  <0.1
mg/a, while the 96-hour LC_. for Anabaena var1ab111s  was  15-17 mg/a.
    Gennlty et  al.  (1985)  assessed  the  effects  of selenium  on the  I1p1d
content  of green  and  red  algae,  Dunallella  pMmolecta  and  Porphyr1d1um
cruentum.  respectively.   Axenlc  cultures  of   these species were grown  1n
artlflcal  seawater at  22°C.    Selenium  as Na9SeO_  was   added  to achieve  a
                                              L.    0
concentration of 10 ppm.  Growth was  not  affected  1n either  of  these species
at  this   concentration.    No   selen1te-1nduced   changes  were   observed  1n
cultures  of  £_._ cruentum  after 35  days of  treatment.    Significantly  lower
levels  of  certain  fatty adds  of  polar llplds were observed 1n  cultures  of
£. pMmolecta exposed to 10 ppm selenium after 35  days of treatment.
    Tang  et al.  (1985) reported that  1.4 ppm selenium  (as  the  sodium salt)
Inhibited  population  growth of Tetrahymena  pyrlformls.   A  concentration  of
140 ppm selenium completely stopped  growth.  Cao and Tang  (1985)  reported a
dose-dependent  Inhibition  In  division  of  synchronized  T.  pyrlformls  cells
exposed to 10, 20,  50 and 100 ppm selenium (as SeO?).

0144d                               4-25                             04/01/89

-------
    TMpathl and  Pandey  (1985)  assessed the  effects  of selenium  (as  sodium
selenate)  on  growth  of   the  algae,   Chlorella   vulgaMs   and   Phormldlum
foveolarum.  Tests  were  conducted  1n   150  ma of  culture  solution  1n  250
ma  conical  flasks.   Assays  were  conducted  for   30  days  at  temperatures
ranging  from 25-29°C.   Growth  was  determined by  fresh  and  dry  weight  of
algae.   Growth  of cultures  In  the presence  of  the lowest  concentration  of
selenium  tested   (0.25  ppm)  was  only  60-61% of  that observed  In  control
cultures.   Growth  of cultures  In higher  concentrations  of   selenium  demon-
strated  a  concentration-dependent  response.  Growth of  C.  vulqaMs 1n a  4.0
ppm solution of selenium was  97%  of  that observed  In  controls and  completely
Inhibited 1n P.  foveolarum.
    Prevot  and  Soyer-Goblllard  (1986)  assessed  the  effects  of selenium  on
growth of  two  marine dlnoflagellates, Prorocentrum mlcans and  Crypthecodln-
1um  cohnll.   Investigators  reported that  selenium  retarded  growth  of  P.
mlcans  1n  culture  Incubated with  100   and  1000  ppm  selenium.  Growth  was
slightly  enhanced at  10  and 50  ppm  selenium  for  the  first 15  days  of
treatment, but was Increasingly Inhibited  compared  with controls beyond  that
point  1n time.   Growth  of  C. cohn11 was  severely  Inhibited by selenium  at
all test  concentrations  (10, 100, 1000  ppm) within 1.5 days of the  Initia-
tion  of  treatment.   A variety of  ultrastructural  changes  were noted  by  the
Investigators In cells exposed to  selenium.
    Wang (1986) assessed  the toxldty of selenium  to  duckweed, Lemna  minor.
Tests  were conducted In  200 ma  fruit   Jars  with   200  ma  of plant  nutrient
solution and 20 colonies  or 40  fronds of  duckweed  1n  each jar.  Assays  were
conducted  for  4  days  at  27°C.   The test  endpolnt was a  reduction  1n  the
number  of  fronds  In  comparison  with  control samples.   Frond numbers  were
determined  dally.   The  96-hour   EC™  for  duckweed exposed   to selenium  as
determined by this assay was 2.4 mg/a.
0144d                               4-26                             04/01/89

-------
    4.1.3.2.   BIOCONCENTRATION — Mann et al.  (1988)  measured  the level of
selenium In algae, Euglena  sp.,  from  tailing  areas  of  the Elliot Lake mining
district, Canada.  A concentration of  2700  ng/g dry weight was  reported  In
algae  from  this  area.  The  Investigators  cited a  concentration  of  0.2  ng/g
1n world river water  for comparison.
4.1.4.   Effects  on   Bacteria.    Pertinent  data  regarding  the  effects of
exposure of  aquatic  bacteria to  selenium  were not located  1n  the available
literature cited 1n Appendix A.
4.2.   TERRESTRIAL TOXICOLOGY
4.2.1.   Effects  on   Fauna.    Beyer   and   Cromartle  (1987)   surveyed   the
concentrations of  selenium  In earthworms  and  soil  from a  variety of sites.
Their  findings  are presented  In Table 4-3.   Concentrations 1n  soil  ranged
from  trace  levels to  1.3  mg/kg at  an  Industrial  site.   Concentrations  1n
worms ranged from trace quantities to 7.6 mg/kg.
    Beyer et  al.  (1987) assessed the potential  for  earthworms, Aporrectodea
tuberculata.  Aporrectodea  turglda  and  LumbMcus  rubellus.  to  concentrate
selenium.   A  moderately  eroded  plowed field  was  divided  Into 9 m2  plots.
Five plots  were  treated with selenous acid 1n  the  autumn  of 1980 to achieve
concentrations  of  0.44,  0.64,   0.89,  3.9  and  6.7 mg  selenium/kg  of  soil.
Clltellate earthworms were  collected  19 months  after treatment  of the plots.
Earthworms  were refrigerated  on  moist paper  towels  for  3 days  and  then
sacrificed  by Immersion  Into boiling  water.   Worms  were  cut  Into  1-3  cm
sections  before  alimentary  canals  were flushed  with  distilled  water.   The
concentration of  selenium In worms  from  control plots  (<0.1 mg  Se/kg soil)
was  16  mg/kg.   Concentrations of  selenium 1n worms from  treated plots  were
36, 43,  51,  36  and 78 mg/kg, respectively.   The  uptake  of  selenium by worms
was reportedly  not Influenced by pH  or other  soil  variables (organic matter
content, phosphorous, potassium or magnesium).
0144d                               4-27                             04/01/89

-------
                                                     TABLE 4-3

                  Metal Concentration In Soil and Earthworms from Contaminated and Natural Sites*
          Site
                                Species
                                                               Selenium
                                                         mg kg"1 (dry weight)
                                                         Earthworms
                                                                     Soil
                               Comments
OO
Natural

  Natural A



  Natural B



  Natural B



  Natural C



  Natural D


  Natural E


  Natural F
                           Elsenoldes  carollnenls  (c)
                           £.  loennbergl  (c)
                           Lumbrlcus  rubellus  (c)
                           E_.  loennbergl  (c)
                           £.  loennbergl  (c)
                           Aporretodea  turglda  (c)
                           £.  loennbergl  (c)
1.4
2.5



1.0


t


1.8
OO
us
Hell drained soil on
Patuxent Wildlife Res. Ctr.,
Laurel, ND

Moderately well drained soil
on Patuxent Wildlife Res.
Ctr., Laurel, MD

Moderately well drained soil
on Patuxent Wildlife Res.
Ctr., Laurel, MD

Poorly drained soil on
Patuxent Wildlife Res. Ctr..
Laurel, MD

Stream bed on Patuxent Wild-
life Res. Ctr., Laurel, MD

Mountalntop In Shenandoah
National Park, Skyland VA

Beach Swamp by Pocomoke
River, Whaleysvllle, MD

-------
TABLE 4-3 (cont.)
rvj
10
o
0
^
GO
Site
Natural
Natural F
Natural G
Mining
Mining A
Mining B
Mining B
Mining Cl
Mining C2
Mining C2
Selenium
mg kg"1 (dry weight)
Species Comments
Earthworms Soil
E. loennbergl (a) 1.5 t Beach Swamp by Pocomoke
River. Whaleysvllle, MD
Sparganophllas elsenl (c) 0.48 t Border of Potomac River.
Me lean. VA
L. terrestrls (a) t 0.58 Soil near Mineral mil
copper mine, Louisville, MD
A. tuberculata (c) 2.1 t Soil near New Galena lead
mine. New Galena, PA
A. Tonga (c) 0.48 t Soil near New Galena lead
mine, New Galena, PA
Pheretlma sp. (c) 0.19 t Dump near Ecton lead-copper-
zinc mine. Audobon, PA
L. terrestrls (c) t t Dump near Ecton lead-copper-
zinc mine, Audobon, PA
A tuberculata (ate) 0 60 t Dumo near Ecton Ip^rl-ronnpr

zinc mine Audobon, PA

-------
                                                  TABLE 4-3 (cont.)
OJ
o
Site Species
Mining
Mlnlnq C3 L. terrestrls (a*c)
M1n1nq C4 L. terrestrls (a)
Industrial
Industrial A Apporrectodea spp. (c)
Industrial B L. terrestrls (a)
Industrial C A. Tonga (c)
Industrial D A. trapezoldes (c)
Industrial D A. trapezoldes (a)
Selenium
mg kg"1 (dry weight)
Comments
Earthworms Soil
t t Dump near Ecton lead-copper-
zinc mine Audobon, PA
t t Dump near Ecton lead-copper-
zinc mine Audobon, PA
4.0 1.3 Cherry Hill Park, near city
Incinerator. Baltimore. MD
t t Greening Park, In an Indus-
trial area. Baltimore. MD
1.1 t Park near Wagners Point,
In an Industrial area,
Baltimore. ND
2.0 t Fort Arml stead mineral pig-
ments plant, Baltimore, MD
t t Fort Arml stead Park, near
o
-p-
^x
O
oo
vO
mineral pigments  plant,
Baltimore, MD

-------
                                                 TABLE  4-3 (cont.)
          SHe
         Species
      Selenium
m«j kg"1 (dry weight)

Earthworms     Soil
                    Comments
    Industrial

      Industrial E



      Industrial E



      Industrial Fl



      Industrial F2



      Industrial F2



      Industrial G



      Industrial G
A. trapezoldes (c)



A. trapezoldes (a)



A. Tonga (c)



A. Tonga (c)



A. tuberculata (c)



1^. rubellus (c)



A. trapezoldes (c)
    0.50
    0.83
    1.6
    1.7
CD
10
t         Park on Caroline Street,
          In an Industrial area.
          Baltimore. ND

t         Park on Caroline Street.
          In an Industrial area,
          Baltimore. MD

0.39      Brldesburg Recreation
          Center, In an Industrial
          area, Philadelphia, PA

t -        Brldesburg Recreation
          Center. In an Industrial
          area. Philadelphia. PA

t         Brldesburg Recreation
          Center, In an Industrial
          area, Pylladelphla, PA

t         Soil from road right-of-way,
          near mineral pigments plnnl.,
          Beltsvllle. MD

t         Soil from road right-of-way.
          near mineral pigments plant,
          Beltsvllle. ND

-------
                                                 TABLE 4-3 (cont.)
          Site
                                Species
      Selenium
mg kg"1 (dry weight)

Earthworms     Soil
                               Comments
Industrial

  Industrial HI        A. trapczoides (c)
      Industrial  H2
CO
ro
                       A.  trapezoides (c)
    *Source:   Beyer and Cromartle,  1987

    t = trace, c = clltellate, a =  aclltellate
                                                              1.8
2.7
               0.20
Soli ft oni grounds of coal-
fired power plant, Vienna.
MD

Soi I f i om yi  /ids of can I -
fired power plant, Vienna,
MD
Galvanized Towers
Tower A
Con. A
Tower B
Con. B

E_. loennbergl
E. loennbergl
E. loennbergl
E. loennbergl

(a)
(a)
(c)
(a)

2.6 t Soil from beneath power 1
tower. Laurel, MD
7.6 t Soil 20 m from power line
tower. Laurel, MD
5.5 t Soil from beneath power 1
tower. Laurel, MD
2.5 t Soil 20 m from power line
tower. Laurel, MD

Ine

Ine


-------
    Serda  and  Furst  (1987)  assessed  the toxldty  of the  sodium salts  of
selenlte  and  selenate  to  the earthworm,  LumbMcus  terrestMs.  Worms  were
Injected with doses  ranging  from 40-160 mg/kg behind  the  clHellum  Into the
hemocoel  and  observed  for  24  hours.   The  8-hour  LO-.s  for  selenlte  and
selenate   1n   worms   treated  Intraperltoneally   were  31  and   60   mg/kg,
respectively.
    Ohlendorf  et  al. (1986) monitored  the reproductive efforts of  American
coot, Fullca amerlcana.  mallard, Anas platyrhynchos.  northern  pintail,  Anas
acuta, cinnamon  teal,  Anas cyanoptera. gadwall, Anas  strepera.  black-necked
stilt, Hlmantropus mexlcanus.  American avocet, Recurvlrostra amerlcana.  and
eared grebe, Podlceps nlgrlcollls.  from  Irrigation dralnwater ponds  contami-
nated with high levels of  selenium  for a  4-month  period.   Nests were checked
weekly  during  the Incubation  periods.  Eggs  were  periodically examined  to
determine  the  condition  of  the embryo.   Frequency  of   dead  and  abnormal
embryos  ranged  from  2.5-31.7%  1n  ducks  and  grebes,   respectively,   and
4.0-8.8%  1n  ducks  and  coots,  respectively.   There  was   some  evidence  to
Indicate  that mortality  of hatchllngs was high.   Overall, 19.6%  of  the 347
nests monitored  1n this  study  produced at least one embryo or  chick  with an
abnormality.   There  were no  abnormalities  In embryos of birds  from  92 nests
In an uncontamlnated area  over a 2-year  period.   Average  selenium  concentra-
tions 1n  bird  livers and  eggs  from nests 1n contaminated  areas ranged  from
9.1-81.4 ppm dry weight compared  with  4.1-6.1  ppm  In  livers of  birds from an
uncontamlnated  area.   Concentrations  of  other  heavy  metals (Ag, As,  Cd,  Hg,
Pb and  Zn) In  livers and  embryos of  birds  from  the contaminated site  were
equal to  or less  than levels  1n  birds from  uncontamlnated sites.  Levels of
selenium measured 1n the water column  were reported  to be  -300  ppb.   Most of
0144d                               4-33                             04/01/89

-------
the selenium  1n  plants, Invertebrates and  fish  1s 1n organic  forms  such  as
selenomethlonlne  and  selenocystlne,   which  have  a  greater  potential  for
toxlclty to birds.
    Heinz et  al.  (1987)  assessed the effects of selenium  on  reproduction  1n
mallard ducks,  Anas  platyrhynchos.   Ducks  of  both sexes were  fed  selenium-
contaminated duck mash at concentrations  of  selenium  ranging  from 1-100 ppm.
Sources  of  selenium  Included   the   sodium  salt  and  seleno-DL-meth1on1ne.
Females were  offered  treated  mash for a  total of  4 weeks  and prevented from
laying  eggs  during   that  time  until  paired  with  males   offered  the  same
contaminated  diet for   that  period.   Eggs  were  held  1n an  Incubator  at
37.6°C.  All  but one  bird  offered mash  contaminated with 100 ppm selenium
died  before  the end  of  the study (16-39  days)  without producing  any  eggs.
One male died after  57  days  on  a diet containing 25 ppm  selenium.  Percent
fertility of  adults  or  hatchabllHy  of  eggs  was  not affected among groups
treated with  selenium  as the  sodium salt.   Eggs from females fed seleno-DL-
methlonlne  (10  ppm  selenium)  exhibited very  low hatching  success.   Survival
of  ducklings   from  groups  fed  25  ppm  (as sodium  salt)  and  10  ppm  (as
seleno-DL-meth1on1ne) was lower  than  In other  groups.   Incidence  of abnormal
embryos was significantly higher  1n ducks  fed mash contaminated with >10 ppm
selenium  (from  either  source).   Abnormal  embryos   from  sodium  selenlte
treated ducks  exhibited embryotoxlc  manifestations  rather than  the  terato-
genlc  abnormalities  exhibited  by embryos  from ducks  treated  with seleno-OL-
methlonlne.   Duckling  weights  were lower  In 25  ppm  (as sodium salt) and  10
ppm   (as  seleno-DL-meth1on1ne)  groups.   Eggs  from  ducks  fed   seleno-DL-
methlonlne exhibited a 5- to 10-fold  higher  level  of  selenium than eggs from
ducks  fed selenium as the sodium salt.
0144d                               4-34                             04/01/89

-------
    Heinz and  Gold  (1987) assessed  the  behavior  of mallard  ducklings,  Anas
platyrhynchos.  produced  by  adults   treated  with  selenium  (as  seleno-DL-
methlonlne).   Birds  were  treated for  4  weeks  with  1,  2,  4  or  8 ppm  of
selenium as part of their  dally  feed.   Of  the 95  females treated, 10 did not
produce any ducklings for  testing.   Ducklings were  maintained at 21-25°C for
6 days  after  hatching.   Behavior was  assessed  by  frightening  the  ducklings
with a  rotating  alternating black and white paddle.   The  distance  traversed
by  a  duckling  for  2 seconds  after  experiencing  the  stimulus  was  compared
among  ducklings  from different  groups.   There were  no significant  differ-
ences  1n  behavior  of  surviving  ducklings  from  different groups  despite the
fact that  ducklings  from  the  8  ppm  selenium group experienced  teratogenlc
effects, embryo mortality, and death  and reduced growth of  hatchllngs.
    Heinz et al. (1988) assessed  the  toxldty of  selenomethlonlne and  sodium
selenlte  to  mallard  ducklings,  Anas  platyrhynchos.  Groups  of  10  ducklings
were randomly  assigned  to  each  of  20 pens.  Birds  In 4 pens  each  received
dietary additions of 0, 10,  20,  40 and 80  ppm selenium mixed Into commercial
duck starter mash.  Treatments  continued for  ~6 weeks.  Birds  receiving the
80  ppm  selenium treatment  1n either  form  experienced significant  levels  of
mortality (60-80X) during  the  first  2 weeks of treatment  and 100X  mortality
by  the  end  of  the  study.  Significant  mortality  was observed  In  birds
treated with 40  ppm  selenium as  sodium  selenlte after 2 weeks  (~10X)  and  as
selenomethlonlne after 3 weeks  (~5X).   There were  no  mortalities among birds
offered diets  containing  0,  10  and 20  ppm selenium.   Sodium selenlte  at >20
ppm selenium  reduced  food consumption  after  1  week;   treatment  with  seleno-
methlonlne at  >40 ppm selenium  also  resulted  1n a  reduction 1n food consump-
tion after  1  week.   For   the most  part,  food  consumption  by  birds  treated
with 10  ppm  selenium  did  not differ  from  that  of  controls.   Body weights  of
0144d                               4-35                             04/01/89

-------
ducklings  treated  with  >20  ppm  were  significantly  lower  than  those  of
controls, while body  weights  of  birds treated with 10 ppm  selenium  were not
significantly lower than those for control birds,  except  for  one  observation
period.  Ducklings  fed 10  ppm selenium as sodium  selenlte  had  larger  livers
than  control  birds.   This  effect was  not  observed  1n  birds  fed 10  ppm
selenium  as  selenomethlonlne.   Selenium  accumulated more  rapidly  1n  the
livers of birds treated with  selenomethlonlne  than  those  treated  with  sodium
selenlte.   The   Investigators  concluded  that  10  ppm  dietary  selenium  as
sodium  selenlte  or  selenomethlonlne  appeared  to  be  close  to  a no-effect
level In mallard ducklings.
4.2.2.   Effects  on  Flora.   Pertinent   data '  regarding  the  effects   of
exposure of terrestrial  flora  to selenium were not located  In  the available
literature cited 1n Appendix A.
4.3.   FIELD STUDIES
    GuthMe  et  al.   (1979)  determined  the  concentrations   of  selenium  1n
water, sediment and biota  from two bays  on  the Gulf Coast of Texas.  Samples
were  analyzed  by  neutron  activation.  Concentrations of selenium  1n  water
and  sediment  were  0.11   and  1.44  mg/kg,  respectively.   Concentrations  of
selenium  1n  barnacle, Balanus aburneus.  crab, CalUnectes  sapldus. oyster,
Crassosstrea  vlrglnlca.  clam,  Rangla cuneata.  and  polychaete,  Nereis  sp.
were 0.77, 0.08, 0.14, 0.54 and 0.49  mg/kg wet weight,  respectively.
    Rodgers et  al.  (1980)  determined  the  concentrations  of  selenium  In the
shells and  viscera of Asiatic clams,  Corblcula  flumlnea. from two  sites  1n
the New  River,  Virginia,  and the water and  sediment  from those  sites.  Clams
were collected from one  site  that  received  the thermal  effluent  from a coal-
fired electric generating  plant  and a  second  site that  was  not  Influenced by
thermal  effluents.   Selenium  concentrations  were  determined  by  neutron
0144d                               4-36                             04/01/89

-------
activation.  Selenium concentrations 1n water,  sediment, and  clam valves  and
viscera  from  a  control  site were  0.11,  0.88,  0.29 and  3.90 ppm,  respec-
tively.   Selenium concentrations  1n  samples  from  sites  receiving  thermal
effluent discharges were 0.10, 0.60, 0.50  and 16.5 ppm,  respectively.
    Boyer  (1984)  determined  the concentrations  of selenium  1n   the  water,
sediment  and  fish of  the  upper  Mississippi River.   The  concentrations  of
selenium  1n  water  and  sediment were  below detection  limits (1 yg/i  and
-0.2  yg/g, respectively).   Concentrations   of  selenium  In  fillets,  livers
and  kidneys   of   common  carp  ranged  from   0.161-0.356,   0.858-2.17   and
0.943-1.62 yg/g   wet  weight, respectively.   Concentrations   of   selenium  In
fillets   from  smallmouth   bass   and  sauger  .ranged  from   0.36-0.425   and
0.284-0.369 yg/g  wet weight,  respectively.   Concentration  factors for  the
edible  tissue of common   carp  collected  from  the upper  reaches  of  the
Mississippi River reportedly ranged from 322-712.
    Sorensen  et al.  (1984)  assessed  a  variety of  factors  relating to  changes
associated with  selenium accumulation  1n  green  sunflsh,  Lepomls cyanellus,
from Belews Lake, North  Carolina.   Selenium  concentrations 1n  tissues  ranged
from  2.3-22.3  ppm  from contaminated  fish   and  1.3 ppm  1n  reference  fish.
Investigators   reported   capillary   changes  and  resultant  edema   1n  hepato-
pancreas and skeletal muscle, swollen  and vacuolated gill  lamellae, signifi-
cantly  Increased  hepatopancreas  weight-to-body weight  ratios, and signifi-
cantly  lower  hematocrlts.   Additional hlstopathologlcal  effects  that  were
not observed  1n  reference  fish  were noted  1n hepatopancreas, kidney,  heart
and ovaries.
    Woock  and  Summers  (1984)  monitored the  levels  of selenium 1n the  water,
sediments  and  biota  from a  North Carolina reservoir and assessed the  uptake
of  selenium  by caged golden shiners, Notemlqonus  crysoleucas.   In the  same


0144d                               4-37                             04/01/89

-------
reservoir.  Concentrations  of  selenium In  surface  and  bottom waters over  a
2-year  period  ranged  from  5  yg/g  wet
weight after 20 weeks of exposure.
    Lemly  (1985a)  determined the  concentrations  of selenium  In  water,  sedi-
ment and  biota  of  a North Carolina  power plant  cooling reservoir.   Selenium
concentrations  were  determined   by  differential  pulse polarography  or  by
flameless  atomic  absorption spectrophotometry.   Concentrations of  selenium
In  the  water column  were 20-30   times  higher  than background levels  with  a
mean  value  of  10  yg/t,  while   concentrations   In  flora  and  fauna  were
10-15  times  background  levels.    Concentrations  of  selenium  1n  surface
sediments  were  ~4  yg Se/g  wet weight.  Selenium concentration  was  519- and
3975-fold  higher In  perlphyton and fish, respectively,  than  In  water.   Fish
were exposed  to dietary concentrations of  selenium that were 519-1395 times
the concentration of selenium  In  water.
0144d                               4-38                             06/15/89

-------
    Lemly  (1985b)  summarized  the results of several  field  studies  assessing
the  Impact  of  selenium on  the  endemic  fauna.   Durations of  exposure  ranged
from 56  days  to 8 years.   Concentrations  of selenium In water  and  sediment
ranged   from   1-360   and   0.2-17.1  yg  Se/l,  respectively.    The   highest
concentrations of selenium  In the  biota  of  these  systems  was  generally found
1n  fish.   Observed   effects   Included   reproductive  failure,   teratogenlc
effects and high levels of mortality.
    Lemly (1985b) also assembled the results of a  series  of studies  address-
Ing  the  concentration of  selenium In  aquatic  fauna  and wildlife  following
exposure to combined  waterborne  and  dietary  sources under  natural  conditions
In  the  field.   BCFs  for   phytoplankton,  perlphyton  and  plants  ranged  from
237-1320,  158-1070  and  166-24,400,  respectively.    BCFs   for   zooplankton
ranged from 176-2080.  BCFs for Insects, annelids, crustaceans  and  molluscs
ranged from  371-5200, 770-1320, 420-1975  and  600-2550,  respectively.   BCFs
for  carnivorous,  planktlvorous  and  omnivorous  fish  ranged from 590-35,675,
445-27,000 and 364-23,000, respectively.
    Salkl et al.  (1985)  examined the Influence of  selenium on  the  flora and
fauna  of field  mesocosms.  Mesocosms  (10  m  diameter)  consisted  of  poly-
ethylene-walled  experimental  enclosures  set Into  a   sheltered  bay In  Clay
Lake,  Ontario.   The  mesocosms  enclosed -79  m2 of  sediment  and  130 m3  of
water.   Radlolabeled  sodium selenlte was added to the enclosures  to achieve
concentrations  of  1,  10  and   100  yg  Se/a.    The  study  duration  was  ~6
weeks.   Among  phytoplankton, chrysophytes  Initially  dominated  all  communi-
ties.  By  the end of  the study, chlorophytes  dominated  1n control  and  low-
dose  enclosures, chrysophytes  were the  dominant  group  1n  the 10 yg  Se/ft,
enclosure, and  cyanophytes  dominated 1n  the  high-dose enclosure.  There  were
no   significant  differences   among   zooplankton  communities  within   the
0144d                               4-39                             06/15/89

-------
enclosures 2 days after the addition  of  selenium  to  the  enclosures.   Changes
In total abundance  and  blomass  of zooplankton were  similar  among  enclosures
through the course of the  experiment.
    Skinner  (1985)   determined   the  concentrations   of   selenium  In  black
crapple,  Pomoxls  nlgromaculatus. pumpklnseed,  Lepomls  aurUus. brown  bull-
head, Ictalurus nebulosus. and carp,  CypMnus  carplo. collected from utility
wastewater  treatment  basins.   F1sh  were collected  by  electro-shocking  and
sacrificed within 1  hour  of collection.   F1sh tissues  were held  on  dry  1ce
enroute to  the  laboratory  and  stored at 0°C  until analysis.   Concentrations
of selenium  1n  water  from basins A, B, C, D and  E were 7.0,  3.0,  <2.0f <2.0
and  <2.0   iig/8,,  respectively.    Carp  and  crapple from  basin  A had  tissue
levels  of  selenium  ranging  from  11.2-15.8  mg/kg.   The  concentration   of
selenium 1n carp from basin B was 22.7  mg/kg.   The concentration of  selenium
1n a  male and a  female carp  from  basin  C  was 33.1  and  37.6  mg/kg,  respec-
tively.   The concentration of   selenium  In  carp,  bullhead and  pumpklnseed
from basin  D  ranged from  2.7-8.9 mg/kg and  1n  carp  from  basin E  ranged from
2.9-3.7 mg/kg.  The  highest-levels  of selenium 1n fish  tissue were  found In
fish collected 1n basins receiving fly ash water.
    Baumann  and  G1llesp1e  (1986)  assessed  the accumulation of selenium by
largemouth  bass,  Hlcropterus  salmoldes. and  bluegllls,  Lepomls macrochlrus.
from one municipal and  three power  plant  cooling  water  reservoirs.  Selenium
concentrations 1n carcasses of bluegllls and bass from  a  municipal reservoir
were <0.5  ppm.   Selenium  concentrations  In  carcasses  of bluegllls  and bass
from a  cooling water  reservoir that did not  receive  ash  pond effluent was ~1
ppm.   Concentrations  of  selenium 1n  carcasses  of  bluegllls  and bass from
cooling water  reservoirs  that did receive ash  pond  effluent ranged  from  4-9
ppm.   There were no  apparent differences   1n  selenium  levels  of  carcasses


0144d                               4-40                             06/15/89

-------
between sexes  or  species,  except for  a  2-fold difference 1n species from  a
reservoir   receiving  ash pond  effluent.   Selenium  concentrations  In gonads
varied  between sexes  within  reservoirs,  with  significantly  higher levels
(<2-fold)  1n ovaries (<1 to ~12 ppm) than In testes  (<1  to ~7 ppm).
    Byrne  and  DeLeon  (1986)  determined the  concentration  of  selenium  In
oysters,  Crassostrea  v1rg1n1ca. clams,  Rangla cuneata.  and sediments   from
three sites 1n Lake Pontchartraln,  Louisiana.  The  concentration of  selenium
In mollusc  tissue  and  sediment was  determined by gas hydride-atomic  absorp-
tion  spectrophotometry.    Selenium  was  not  concentrated from  sediment  by
these organisms.   Concentrations of selenium  In  oyster  tissue and  sediment
from  one   site  were   0.013  and   0.007  yg/g  dry  weight,   respectively.
Concentrations  of  selenium 1n  clam tissue  and  sediment  from a second  site
were  0.032 and  0.031  yg/g  dry weight,   respectively.   Concentrations  of
selenium  1n clam  tissue and sediment  from  a  third  site were 0.041 and  0.05
yg/g dry weight,  respectively.
    Batley  (1987)  determined  the   concentrations  of selenium  In  sediment,
seagrass,   Zostera   capMcornl.  green  algae,  Enteromorpha  sp.,  the  hairy
mussel,  TMchomya   hlrsuta.  and  the  cockle,  Anadara  trapezia,  from   Lake
MacquaMe,  New South  Wales.   Selenium concentrations  1n  sediment and  biota
were determined by neutron activation  analysis.  A  selenium  concentration  of
14 yg/g  was measured  at  the  northern end  of the  lake,  but the  concentra-
tion  decreased rapidly  to a  plateau level  of  4  yg/g.   Concentrations  of
selenium  1n seagrass   from the  northern and  southern  sections  of the  lake
were  ~3.9  and  ~0.6   yg/g,  respectively.   Concentrations  of  selenium  In
algae  were higher  1n  samples  collected from the  northern  section  of  the
lake, ranging  from 0.3-1.6 yg/g.   Whole body  concentrations of selenium  1n
mussels and cockles  from  the  northern end  of the  lake were 3.3 and 6.4  yg
Se/g, respectively.

0144d                               4-41                              04/01/89

-------
    L1u et  al.  (1987)  determined  the concentrations  of  selenium 1n  phyto-
plankton,   seaweed,  zooplankton,  other  Invertebrates  and  fish  from  Xiamen
Bay,  Fujlan Province,  China.   Flora concentrated  selenium  3-4  orders  of
magnitude above  the ambient  concentration  In seawater.   Phytoplankton  con-
tained 1.24 ppm selenium, while  seaweed  concentrations  ranged  from 0.08-0.61
ppm.  The  concentration of  selenium  1n  zooplankton  was 3- to 4-fold  higher
than  that  found  In  marine flora.  Concentrations ranged  from  2.16-6.30  ppm.
Concentrations  of  selenium  In  other  Invertebrates  and  fish  were  approxi-
mately  equal  to  those  found  In  zooplankton  on a  dry  weight  basis,  but
concentrations  were dependent on the tissue analyzed.
    Johnson  (1987)  determined  the  concentrations of  selenium In  sediments
and  fish  from 14  Ontario  lakes.   Sediment was  collected from the  deepest
part  of  each lake.   Selenium  concentrations  In sediment  and fish  samples
were  determined  by  coloMmetrlc techniques.  Selenium  concentrations  In  the
top  10  cm of sediment  ranged  from  <1-16 yg/g.  Concentrations of  selenium
In  tissues  of  lake trout,  whlteflsh,  common sucker, yellow perch,  northern
pike  and walleye  from the study lakes were  0.78, 0.84,  0.55,  0.38,  0.37  and
0.25  vg wet welght/g,  respectively.
    Sa1k1   and  Lowe  (1987)  determined  the  concentrations of  selenium  In
water, sediment  and biota  from  the  San  Luis Drain, Kesterson Reservoir  and
the  Volta  Wildlife  Area of  California.   Samples  were  analyzed  for  total
selenium  by  hydride   generation-atomic   absorption  spectrophotometry.   The
detection  limits  for  selenium  1n  water,  sediments  and  tissues were  0.00017
mg/a,  0.2  vg/g   wet   weight  and  0.05   yg/g,  respectively.   Concentration
ranges for  selenium 1n  water  and sediment from  the  San Luis Drain, Kesterson
reservoir   and   Volta  waterways  were  0.29-0.33  mg/i   and  65-100  vg/g,
0.009-0.32  mg/8,  and   1.8-67  yg/g,  and  0.0002-0.0014  mg/l  and  O.2-0.5

0144d                               4-42                             04/01/89

-------
yg/g,  respectively.   Concentration  ranges  for  selenium  In algae  and  net
plankton  1n  these  systems  were  63-72 yg/g  and not  measured,  12-330  yg/g
and   58-120   yg/g,   and   below   detectlon-1.4   yg/g   and   1.4-2.8   wg/g,
respectively.  Concentration  ranges  for  selenium In all aquatic  Insects  and
mosqultoflsh  In  these  systems  were 170-330  yg/g  and  140-370  yg/g,  16-290
yg/g  and  104-290  yg/g,  and  0.68-3.0   yg/g   and   1.2-1.4 yg/g,   respec-
tively.   The  Investigators   noted   that   selenium  concentrations  generally
Increased from water to sediments  to plants to animals.
    Mulr et al. (1988)  determined the  concentration of  selenium  In  dolphins,
Lagenorhynchus alblrostMs.  and pilot  whales,  Globlcephala melaena.  from the
coast of  Newfoundland,  Canada.   Dolphin  tissues  were obtained from  dolphins
killed  by entrapment  1n  1ce  and whale  tissues were  obtained  from  whales
stranded on beaches In  Newfoundland.   Processed  tissue  samples were  analyzed
for their selenium  content by  flameless atomic absorption  spectrophotometry.
Mean  selenium  concentrations  1n  kidney,  liver and muscle  of dolphins  were
5.85, 8.15 and 1.91 mg/kg dry  weight,  respectively.   Selenium concentrations
1n blubber, kidney, liver and muscle of whales from  two different  collection
sites were 0.49  and 0.59 (mg/kg  wet weight), 13.0  and 11.3, 50.5 and  31.4,
and 1.22 and 2.94 mg/kg dry  weight,  respectively.
4.4.   AQUATIC RISK ASSESSMENT
    There were  no significant  studies Identified  during the compilation  of
this document to refute the conclusions reached  or  the  criteria  generated 1n
U.S.  EPA  (1987a).   Since the National Water  Quality Criteria Approach  used
by U.S. EPA  (1987a) to  generate fresh  and  saltwater  criteria for  selenium Is
the approach  that  has  been  adopted for analysis of aquatic  toxldty data 1n
HEEDs,  an analysis  of  the  data  presented  1n  this  document  will not  be
conducted.
0144d                               4-43                             04/01/89

-------
4.5.   SUMMARY
    Acute  toxIcHy  data  for selenium  1n  aquatic  vertebrates were available
for  21   species  of  fish  and  1  amphibian.   Among  freshwater  species,   the
96-hour  LC5Qs  for  selenium as  sodium selenlte  ranged  from 0.62  mg/l  for
the  fathead  minnow,  Plmephales promelas  (Elsler,  1985)  to  35  mg/l  for  the
common carp,  Cyprlnus  carplo  (Etnler  et  al.,  1987).   Fathead  minnows were
also  the  least  tolerant   vertebrate  species  to   selenate-selenlum,  with a
96-hour  LC5Q  of   2  mg/l   In  a  flowthrough  test   (Etnler  et   al.,   1987).
Zebraflsh, Brachyodanlo rerlo.  and  juvenile striped bass, Horone saxatlHs.
were  the  most  tolerant  of   exposure  to  selenate-selenlum,   with  96-hour
LC5Qs  of 82  and  85.8  mg/l,   respectively  (N1tm1  and Laham,  1976;  Klauda,
1986).   The  toxldty  of  selenium as selenium  dioxide  ranged from a 96-hour
LC50  of  7.3  mg/l   for   fathead  minnows  to   20  mg/8.   for  zebraflsh
(Cardwell et al.,  1976;  N11m1  and  Laham,  1976).
    Fathead minnows were equally  sensitive  to  the  sodium  salts of selenlte
and  selenate,  while  zebraflsh were  equally sensitive to  the  potassium  and
sodium salts  of  selenlte  and  selenate.   The toxldty  of selenium as  sodium
selenate to striped bass decreased nearly 10-fold as fish developed from  the
prolarva  to  juvenile stage  (Klauda,  1986).   Toxldty  values  for saltwater
fish  ranged   from   a   96-hour  LC5Q  of   0.6  mg/l  for   haddock  larvae,
Melanogrammus  aegl1f1nus.   to  96-hour  LC5Qs of  14.2-15.1   mg/l  for   winter
flounder  larvae,   Pseudopleuronectes   amerkanus.   exposed   to   unidentified
forms of selenium (Elsler,  1985).
    In a  series of behavioral  and physiological studies, goldfish, Carasslus
auratus. exhibited behavioral  Impairment  at 0.25  ppm (Weir  and H1ne,  1970),
while  fathead  minnows,  Plmephales  promelas.  did  not   avoid   selenate  at
concentrations  of  0.3-11.2  mg  Se/l   (Watenpaugh  and  Beltlnger,  1985).
0144d                               4-44                             04/01/89

-------
Oxygen consumption by  fathead minnows  was  not affected by exposure to 60 mg
Se/8,  for  24 hours (Watenpaugh  and  Beltlner, 1985).   Oxygen  consumption by
the  hermit  crab,  CllbanaMus   vlttatus.  exposed  to  100  ppm  selenium was
depressed  at  16°C  1n  10  °/    salinity   test  medium,  although  oxygen
consumption  rates  were  generally  elevated  at  lower  salinities  and higher
temperatures (Wolfenberger,  1987).
    Acute toxldty data for selenium 1n aquatic Invertebrates were available
for a total of 22  species,  Including 3 cladocerans,  2  copepods, 4 amphlpods,
4  decapods,  1  mysld,  3  Insects  and  5  molluscs.    The  48-hour  LC5Qs of
selenlte-selenlum  to  daphnlds  ranged  from 0.098-3.87 mg/a,  (Etnler  et  al.,
1987;  Reading and  Bulkema, 1983).   The  48-hour  EC5-s  for  unfed  and fed
daphnlds exposed  to  sodium selenlte were  0.47 and  1.5 mg/a,  repsectlvely.
The  48-hour  NOEC  for  unfed  daphnlds  was  1.0 mg/8,  (Adams  and Heldolph,
1985).   The  96-hour   LC5Qs   for   amphlpods  ranged  from   2.88-6.17   mg/8.
(Etnler et  al.,  1987).  The toxldty  of  selenlte-selenlum to  Daphnla  magna
was  4-fold  lower  when  assays  were  conducted  In  soft water;  no  significant
differences were  observed  for  Chlronomus  plumosus.  based on water  hardness
(Mayer and Ellersleck,  1986).
    The toxldty of  selenium as  sodium selenate to  5th  Instar  Daphnla  maqna
(48-hour LC5Qs of  0.55 and 0.75 ppm)  was  comparable with that observed for
sodium  selenlte  (Johnston,  1987).   In contrast,  the  toxldty of  selenate-
selenlum  to Hyallela  azteca  (96-hour  LC™  of  0.76  mg/l)  was 4- to 8-fold
greater than  that  observed  with sodium selenlte (Etnler et al.,  1987).  The
toxldty of selenium as selenium  dioxide  was  represented  by  96-hour  LC5Qs
of  33,  1.90 and  0.25  mg/l   for  the  crab,  Scylla  serrata.  bay  scallop,
Arqopecten  Irradlans.   and  surf  clam,   Splsula   solldlsslma.   respectively
(Krlshnaja,  1987;  Nelson et  al.,  1988).   The 96-hour  LC™  of selenium as
0144d                               4-45                             06/15/89

-------
selenium  dioxide  to  crabs,  oysters  and  mussels  was  >10 mg/l  (Martin et
al.,  1981).   The 96-hour  LC,.-  for  the  midge, Tanytarsus  dlsslmHls. was
42.4  mg/l  (Elsler,  1985).   Exposure of  embryos  of  the zebraflsh,  Brachy-
danlo  reMo,  to <10  mg/l  selenium  as  selenium  dioxide for  168  hours did
not affect hatching and mortality.  Exposure of zebraflsh  larvae resulted  1n
significantly  higher  levels  of  mortality  at  concentrations  >3  mg/l  after
168 hours and at 10  mg/l after 96 hours  (N11m1  and  Laham, 1975).
    The 168-hour LC5_  for  fathead minnows,  Plmephales  promelas.  was 2.9 mg
Se/l.    The    336-hour   LC5Qs   for   goldfish,   Carrasslus   auratus.   and
bluegllls, Lepomls  macrochlrus.  were   8.8 and  17.6 mg Se/l, respectively
(Cardwell  et  al.,  1976).   The  113-hour  LC  •  for  embryos  and  the  7-day
LC5_  for  tadpoles  of  the  frog,  Xenopus  laevls.  were 2.0  and  1.5 mg  Se/l,
respectively   (Elsler,   1985).   The  48-day LC  s  for bluegllls,  L. macro-
chlrus.  and   fathead  minnow,  £.  promelas.  were  0.4-2.0  and 1.1  mg  Se/l,
respectively.   The  10-day   LCcn  for  perch,  Perca  flavescens.  was  4.8
                               5U               	 	
mg/l,  while   the  43-day  LC5Q for  coho  salmon,  Oncorhynchus  klsutch. was
0.16  mg/l.   The 96-day LC5Q  for  rainbow trout,  Salmo  galrdnerl. was  0.29
mg Se/l (Elsler, 1985).
    Exposure   of Daphnla  magna  to  0.2-0.8 mg  selen1te-selen1um/l  for  28
days  did  not  affect  total   eggs  per   daphnld or  live  young  per  daphnld;
however, mean  brood  size was  affected  at  0.8  mg Se/l (Reading and  Bulkema,
1983).   In  another   study,   the   7-p   14- and  21-day  EC5Qs  for   D.  maqna
exposed  to  selenlte-selenlum  were  0.38,  0.38  and  0.35 mg/l,  respectively.
The  NOEC  based on  survival  and  reproduction was  0.24  and  >0.24  mg/l,
respectively.   The  28-day  LC™  for  ]).  magna  1n  a  third  study  was  0.24 mg
Se/l  (Adams and Heldolph, 1985).
0144d                               4-46                             04/01/89

-------
    Exposure  of  the  amphlpod  Allorchestes  compressa  to   44  yg  selenlte-
selenlum/a.  did  not  significantly  affect   growth  or  survival  of  treated
Individuals  after  4  weeks.   Animals  exposed  to  93  yg  Se/8.  experienced
reduced  growth  and a  significant  level  of mortality  over  the  same  period
(Ahsanullah and  Brand,  1985).  The 14-day  LC5   for  another scud,  Hyallela
azteca. was 0.07 mg Se/a, (Elsler,  1985).
    Fathead  minnow  larvae,  Plmephales  promelas.   experienced  significant
reductions  1n  final dry  weight  when  larvae were  fed  rotifers  cultured  on
selenium-contaminated algae.   The  mean  selenium  concentrations 1n  minnows
were  43  and  51.7  yg/g  In  two  experiments  (Bennett  et  al.,   1986).   High
tissue concentrations of  selenium  1n blueglll  sunflsh,  Lepomls macrochlrus.
resulted  1n  significant  differences 1n percent  fertilization and hatch  when
the ovaries of  females  contained  from 5.79-8.0  mg  Se/kg.  These levels  were
16-21   times  higher than  those  In  uncontamlnated sunflsh  (Glllesple  and
Bauamann, 1986).
    Exposure of early life stages of the striped bass,  Horone  saxatllls.  to
low  (0.089-0.099   mg  Se/a)  and  high  (1.217-1.360 mg  Se/a)  concentralons
of  selenate-selenlum resulted  In  a  significant  reduction  In  exponential
growth  rates  for  fish  1n  both  treatments  for  days  3-15.   There were  no
differences 1n  growth rates  between days 19 and  60 but  there was a signifi-
cant  Incidence  1n  developmental  abnormalities  of  the  lower jaw and  severe
blood  cytopathology (Klauda, 1986).   Exposure  of  rainbow  trout fry,  Salmo
qalrdnerl.  to  >47  yg  selenlte  selenium/a  resulted  1n significant  reduc-
tions  1n survival  and  fry weight  and  length after 90 days.   Survival  and
growth were not  affected  at <21  yg Se/a.  BCFs were Inversely related  to
exposure  concentrations  but  did not  exceed  200.  Exposure of  fry  to  >12  yg
Se/8.  resulted  1n  significant  reductions  1n  the  calcium concentrations  1n
0144d                               4-47                             04/01/89

-------
bone  (Hunn  et  al.,  1987).  Percent  hatch  of sheepshead minnows,  Cyprlnodon
varlegatus.  was  <4%  at  concentrations  of  <3.6  mg  selenlte-selenlum/i.
Juveniles experienced 4, 24 and 90%  reductions  In  survival  at  0.47,  0.97  and
>0.97  mg  Se/i,  respectively.  Growth   of  juvenile  sheepshead  minnows  was
reduced by 8% at 0.47 and 0.97 mg/s, (U.S. EPA,  1987a).
    Exposure  of  four  consecutive  generations  of  CeModaphnla  affInls   to
sodium selenlte resulted  In a reduction 1n  the  tolerance of C.  afflnls over
succeeding  generations.   The  NOEL was  reduced  from  0.2-0.1  mg/B, after  two
generations  (Owsley  and  McCauley,  1986).  There were  no  significant  differ-
ences  In  growth of  Daphnla magna  exposed  to <0.05  ppm  selenHe-selenlum or
1n mean  numbers of eggs  produced,  the percent  maturation  or mortality levels
for daphnlds exposed to  <0.025 ppm selenHe-selenlum (U.S.  EPA,  1987a).   The
number  of  offspring  produced  and  survival  of  first  generation  myslds,
Mysldopsls  bahla. exposed  to  selenlte-selenlum  was  significantly  reduced at
0.32  mg/a.   No  statistically  significant  differences  were  observed   In
myslds exposed to 0.14 mg/i (U.S.  EPA,  1987a).
    The  no-effect  level  for  dietary selenium as  either  sodium  selenlte or
selenomethlonlne  In  blueglll  sunflsh,  Lepomls  macrochlrus. over  324  days Is
-13-30  yg  Se/g.   The  maximum BCF  for  blueglll   sunflsh,  L.  macrochlrus.
exposed  to  120 yg  Se  (as   selenous add)/ft.  for  28 days  was  20  and  the
tissue half-life during  a  7-day depuration  period  was between  1  and 7 (Woock
et  al.,  1987).   BCFs for  rainbow  trout, Salmo  qalrdnerl.  exposed  to  <10 yg
selen1te-selen1um/kg  ranged  from  3.1   In  embryos  to  104  1n  livers   of
Juveniles.   BCFs  for  trout  exposed to  -100 yg selenlte-selenlum/kg  ranged
from  1.6  In  sac-fry  to  31.6  In  livers  of juveniles after 96 hours (Hodson et
al.,  1986).  The  BCF  for selenium as  sodium selenlte  1n  carp,  Cyprlnus
carplo.  for  an  unspecified exposure duration  and  selenium  concentration  was
0144d                               4-48                             06/15/89

-------
0.6-6.0.   The  BCF  for  selenlte-selenlum  1n  minnows,  Plmephales  promelas.
exposed to 83 yg/a for 28 days was 4443 (Etnler et al.,  1987).
    Exposure   of   fathead  minnows,   Plmephales   promelas.   to   waterborne
selenate-selenlum  (10-40  ng  Se/ma) resulted  In  a maximum  body  burden  level
of  -0.5-0.8  yg Se/g  fish after  20-30  days.  Selenium  was depurated  to  ~3
yg/g after  30 days.   Exposure  of minnows to  selenium-contaminated  daphnlds
(1.33-7.32  yg Se/g)  resulted  1n  selenium  body  burden  levels  of  -0.3-1.2
yg  Se/g  fish  after  80 days.   Depuration  of  selenium  was  slower  than  that
observed  1n  fish  exposed  to  waterborne  selenium.   Fish  exposed  to  both
sources of  selenium  had  body-burden levels of selenium  that did not plateau
during  the  56-day exposure  phase,  reaching 0.4-2.0  yg Se/g  fish.   Depura-
tion of  selenium  from these  fish  was  slow (Bertram and  Brooks,  1986).   Fed
and starved juveniles  of  the  striped bass,  Horone saxatllls. exposed to 1.29
mg  selenate-selenlum/ 8.  for   60  days  accumulated  selenium  at  comparable
rates,   producing  BCFs of 0.68 and  0.69.    Fed  Juveniles  exposed  to 90  yg
Se/8.  showed   no  Increase 1n  whole body  levels   of  selenium,  while  starved
juveniles  had a  BCF  of   11.78  (Klauda,  1986).   The  half-lives  In  days  for
selenate,  selenlte  and selenomethlonlne  administered orally via gavage  at  a
level of 20 ng  Se/g  to fathead  minnows,  P.  promelas.  ranged from 3.9 (liver)
to  487  (adipose tissue), 2.2  (liver)  to  64  (heart) and  1.0  (liver)  to  69
(adipose tissue) (Klelnow and Brooks,  1986).
    Mussels   exposed   to   50  yg  selenium/8,  for   15-50   days   accumulated
selenlte-selenlum at  a rate  of  0.12 ng Se/g/day.   The  presence  of Inorganic
(30  yg  Hg/l)  and  organic   (3  yg Hg/a)  mercury  Increased   the  uptake  of
selenlte-selenlum  to  0.24   and  0.40  ng  Se/g/day,  respectively.   Organic
selenium   (CQH,00Se)0  accumulated  at   a   rate   of  0.15  ng  Se/g/day  1n
            o  / c   e.
the presence of mercury (Pelletler, 1986).
0144d                               4-49                             04/01/89

-------
    There were no apparent effects  observed  In  growth  of  natural  assemblages
of  phytoplankton or  In  growth of  a  species  Isolated  from field  samples
Thalassloslra aestevalls  exposed  to  <1000 nM  selenium (HolUbaugh et  al.,
1980).   There were  no effects  on  growth of either  the green or  red  algae,
Dunallella   prlmolecta   and   Porphyrldlum   cruentum.   exposed   to  10   ppm
selenlte-selenlum (Gennlty  et al., 1985).   The 48-hour LC5Q  for Oedoqonlum
cardlacum  was  <0.1  mg/l, while  the  96-hour  LC5Q  for Anabaena  var1ab111s
was  15-17  mg/8.   (Elsler, 1985).   The  96-hour  EC5Q  for  duckweed,   Lemna
minor, exposed to selenium was 2.4 mg/SL (Wang,  1986).
    Population growth  of  Tetrahymena pyrlformls  was   Inhibited  by 1.4  ppm
selenium  and  stopped  completely   at  140  ppm- (Tang  et  al.,   1985).    T.
pyrlformls  also   experienced  a dose-dependent  Inhibition   1n  division  of
synchronized cells  exposed  to  10, 20,  50 and  100 ppm  selenium  (as  SeOp)
(Cao and  Tang,  1985).   Growth  of  cultures  of  the  algae,  Chlorella vulqarls
and Phormldlum foveolarum,  was reduced  by  ~40% 1n  the presence  of  0.25  ppm
selenate-selenlum.  Growth of  C_.  vulgarls  1n a 4.0  ppm solution  of selenium
was  97%  of  that  observed   1n controls  and  completely  Inhibited  1n  P_.
foveolarum  (TMpathl  and  Pandey,   1985).   Selenium retarded  growth of  the
marine dlnoflagellate,  Prorocentrum mlcans.  1n  cultures  Incubated  with  100
and  1000  ppm  selenium.   Growth  was  slightly  enhanced  at  10  and 50  ppm
selenium  for  the first 15 days  of treatment but  was  Increasingly Inhibited
compared with controls  after 15 days.  Growth  of  the  marine  dlnoflagellate,
Crypthecod1n1um  cohnll.  was  severely  Inhibited   by   selenium   at  >10  ppm
selenium  within  1.5  days  of  the  Initiation  of  treatment   (Prevot  and
Soyer-Goblllard,  1986).
    Concentrations  of   selenium 1n  soil and  earthworms  ranged   from  trace
levels to  1.3  mg/kg at an  Industrial  site and from trace  quantities  to  7.6
0144d                               4-50                             04/01/89

-------
mg/kg,  respectively  (Beyer  and Cromartle,  1987).   In an experimental  field
study,  the  concentration of  selenium In  earthworms, Aporrectodea  tubercu-
lata. Aporrectodea turglda. and Lumbrlcus  rubellus. from control plots  (<0.1
mg Se/kg soil) was  16 mg/kg,  while concentrations of selenium  1n worms  from
treated plots were 36, 43, 51, 36  and  78  mg/kg.   Uptake  of  selenium by  worms
was  not  Influenced  by pH  or  other  soil  variables (organic matter  content,
phosphorous,  potassium  or  magnesium)  (Beyer  et al.,  1987).   The 8-hour
LD5Qs  for   selenlte   and  selenate   In   earthworms,  Lumbrlcus  terrestMs.
treated  Intraperltoneally  were  31  and  60  mg/kg, respectively  (Serda and
Furst, 1987).
    Exposure  of   American  coot,   FuUca   americana.  mallard,  Anas  platy-
rhynchos.  northern   pintail,  Anas  acuta,  cinnamon   teal,  Anas cyanoptera.
gadwall, Anas  strepera.  black-necked stilt, Hlmantropus mexlcanus.  American
avocet,  Recurvlrostra  americana.  and  eared  grebe, Podlceps nlgrlcolHs.  to
selenium-contaminated  Irrigation dralnwater  ponds (-300 ppb Se) resulted  In
an  Increase  In  reproductive  Impairment.   Frequency  of dead  and  abnormal
embryos  ranged  from 2.5  (ducks)  to 31.7%  (grebes) and  4.0 (ducks) to  8.8%
(coots), respectively.   Overall,  19.6% of  the  347 nests monitored  produced
at least one embryo  or chick with an  abnormality.  There were  no  abnormali-
ties  1n  embryos   of  birds from  92  nests  In an  uncomtamlnated  area over  a
2-year period.  Average  selenium concentrations 1n bird  livers  and  eggs  from
nests  1n contaminated areas  ranged  from  9.1-81.4  ppm  dry weight,  compared
with  4.1-6.1  ppm  In  Hvers  of   birds  from  an   uncontamlnated  area.   The
no-effect  level  for  dietary  selenium as sodium selenlte or  selenomethlonlne
1n  mallard  ducklings,  Anas   platyrhynchos.  1s  ~10  ppm  (Ohlendorf  et  al.,
1986).
0144d                               4-51                             06/15/89

-------
    Concentrations of  selenium 1n water  and sediment  from  two bays on  the
Gulf Coast of Texas were 0.11  and  1.44  mg/kg.   Concentrations  of  selenium In
barnacle, Balanus  aburneus.  crab,  CalUnectes  sapldus.  oyster,  Crassosstrea
v1rg1n1ca. clam, Rangla cuneata. and polychaete, Nereis  sp., from those  bays
were 0.77, 0.08, 0.14, 0.54 and 0.49 mg/kg  wet  weight,  respectively (Guthrle
et  al.,  1979).   The  level  of selenium 1n  algae,  Euglena sp., from  tailing
areas of  the  Elliot  Lake  mining district, Canada, was  2700  ng/g  dry  weight,
compared  to  a concentration of 0.2 ng/g  In world  river water  (Mann  et  al.,
1988).   Selenium  concentrations  In water,  sediment,  and  clam  valves  and
viscera of Asiatic clams, Corblcula flumlnea. from a  control  site 1n  the New
River, Virginia, were 0.11, 0.88, 0.29  and  3.90 ppm,  respectively.   Selenium
concentrations  1n  samples  from sites  receiving thermal  effluent  discharges
were 0.10, 0.60, 0.50 and 16.5 ppm,  respectively (Rodgers et  al.,  1980).
    The  concentrations  of  selenium  1n  water   and   sediment  1n  the  upper
Mississippi   River were  below  detection  (1 yg/8.  and  ~0.2  yg/g,  respec-
tively)  but  concentrations of  selenium  1n  fillets,  livers  and  kidneys  of
common  carp   ranged  from  0.161-0.356,  0.858-2.17  and  0.943-1.62  yg/g  wet
weight, respectively.  Concentrations of  selenium  In  fillets  from smallmouth
bass  and  sauger  ranged  from  0.36-0.425  and  0.284-0.369 yg/g  wet  weight,
respectively.   Concentration  factors  for  the  edible  tissue of  common  carp
ranged from 322-712 (Boyer, 1984).
    Concentrations of  selenium 1n  surface  and bottom  waters  of a  cooling
water  reservoir  ranged from  
-------
Selenium concentrations  1n shad, catfish  and  blueglll ranged  from  3.6-28.8
yg/g  wet  weight.   Selenium  concentrations  In  shad  and  blueglll  from  an
uncontamlnated site were  35-  to 50-fold lower.  Exposure of  golden  shiners,
Notemlqonus  crysoleucas.   to   5-15  yg  Se/8,  Insltu  resulted   In   tissue
levels >5  yg/g wet weight  after  20 weeks (Woock  and  Summers, 1984).   In  a
separate study,  concentrations  of  selenium  1n the  water  column of  a  power
plant cooling  reservoir were  20-30  times higher than  background  levels  with
a  mean  value  of  10  yg/a,  while  concentrations   1n  flora  and  fauna  were
10-15  times   background   levels.   Concentrations   of   selenium  In   surface
sediments   were ~4  yg  Se/g wet  weight.  Selenium was  519- and  3975-fold
higher 1n  perlphyton and fish,  respectively,  than  1n water  (Lemly,  1985a).
    Concentrations  of  selenium  In  water  from  utility wastewater  treatment
basins were  7.0,  3.0,  <2.0,  <2.0  and  <2.0 yg/SL.  Concentrations  of  sele-
nium  In black  crapple,  Pomoxls  nlgromaculatus.  pumpklnseed,  Lepomls  aurUus.
brown bullhead,  Ictalurus nebulosus. and  carp,  CypMnus  carplo, from  these
basins  ranged  from  2.7-37.6  mg/kg  (Skinner,   1985).   Concentrations  of
selenium  1n   carcasses  of  bluegllls,  Lepomls  macrochlrus.  and  largemouth
bass, Hlcropterus  salmoldes from cooling  water  reservoirs  that receive ash
pond effluent  ranged from  4-9 ppm.  Selenium concentrations  In gonads varied
between  sexes  within  reservoirs  with  significantly higher  levels  (<2-fold)
1n ovaries (<1 to -12  ppm) than  1n testes (<1 to  ~7 ppm).   Selenium concen-
trations 1n  carcasses  of  bluegllls  and  bass from a cooling  water  reservoir
that did not  receive  ash  pond effluent was  ~1  ppm.  Selenium concentrations
In  carcasses  of   bluegllls  and  bass  from a municipal water  reservoir  were
<0.5 ppm (Baumann and  Glllesple, 1986).
    Concentrations  of  selenium  1n  oysters,  Crassostrea   v1rg1n1ca.   and
sediment from  a  site  In Lake Pontchartraln,  Louisiana, were  0.013  and  0.007
0144d                               4-53                             04/01/89

-------
yg/g  dry  weight,  respectively.   Concentrations  of  selenium   In  clams,
Rangla cuneata.  and  sediment  from  a  second site were  0.032 and 0.031  yg/g
dry  weight,  respectively.   Concentrations  of  selenium 1n  clam tissue and
sediment  from  a  third  site were  0.041  and  0.05  yg/g dry  weight,  respec-
tively (Byrne and DeLeon,  1986).
    A selenium  concentration  of  14 yg/g  was  measured at  the northern end
of Lake Macquarle, New  South Wales, but the concentration decreased  rapidly
to  a plateau  level   of 4  yg/g.    Concentrations  of   selenium 1n  seagrass,
Zostera  caprlcornl.  from the northern and  southern sections  of the  lake  were
-3.9  and   ~0.6   yg/g,  respectively.   Concentrations   of  selenium  1n algae
Enteromorpha sp.  were higher  In  samples  collected  from the  northern  section
of  the  lake,   ranging  from  0.3-1.6   yg/g.   Whole  body  concentrations  of
selenium  1n  mussels,  TMchomya hlrsuta.  and cockles,  Anadara  trapezia,  from
the  northern  end of  the   lake  were  3.3  and  6.4  yg  Se/g,  respectively
(Batley, 1987).
    Phytoplankton from  Xiamen  Bay,  Fujlan  Province, China contained  1.24 ppm
selenium,   while  seaweed  concentrations   ranged  from  0.08-0.61  ppm.   The
concentration of  selenium 1n zooplankton ranged  from  2.16-6.30 ppm.   Concen-
trations  of selenium 1n other   Invertebrates  and fish  were  approximately
equal to  those  found  In zooplankton on a dry weight basis  but concentrations
were dependent on the tissue analyzed  (Liu  et al.,  1987).
    Selenium concentrations  In the top 10  cm of sediment  from 14  Ontario
lakes ranged from  <1-16  yg/g.   Concentrations of  selenium  1n tissues  of
lake  trout,  whHeflsh,  common   sucker,   yellow  perch,  northern  pike and
walleye from the study lakes  were  0.78, 0.84,  0.55,  0.38,  0.37  and  0.25  yg
wet welght/g, respectively (Johnston,  1987).
0144d                               4-54                             06/15/89

-------
    Concentration  ranges  for  selenium  1n water  and sediment  from the  San
Luis Drain, Kesterson ponds  and Volta waterways  In  California  were  0.29-0.33
and 65-100,  0.009-0.32  and  1.8-67,  and 0.0002-0.0014 and <0.2-0.5,  respec-
tively.   Concentration  ranges  for   selenium  In  algae  and  net  plankton  1n
these  systems  were  63-72 and  not   measured,  12-330 and  58-120,  and  below
detectlon-1.4  and  1.4-2.8,  respectively.  Concentration  ranges  for  selenium
1n  all  aquatic Insects and  mosqultoflsh  1n  these  systems were 170-330  and
140-370,  16-290  and  104-290,  and 0.68-3.0  and 1.2-1.4,  respectively.   The
Investigators  noted   that  selenium   concentrations  generally  Increased  from
water to sediments to plants  to animals  (Salkl and Lowe,  1987).
    Mean  selenium  concentrations  In  kidney, liver  and  muscle  of  dolphins,
Lagenorhynchus  alb1rostr1s.   were  5.85,  8.15  and   1.91   mg/kg  dry  weight,
respectively.  Selenium concentrations  In blubber,  kidney,  liver and  muscle
of  pilot  whales,  Globlcephala melaena.  from two different  collection  sites
were 0.49  and 0.59  (mg/kg wet weight),  13.0 and  11.3,  50.5  and  31.4,  and
1.22 and 2.94 mg/kg dry weight, respectively (Mu1r et al.,  1988).
    BCFs for phytoplankton,  peMphyton  and plants exposed  to selenium In  the
field  ranged  from 237-1320,  158-1070,   and  166-24,400,  respectively.   BCFs
for  zooplankton,   Insects,   annelids,  crustaceans  and   molluscs  exposed  to
combined waterborne and dietary sources  of selenium under  natural  conditions
In  the   field  ranged  from  176-2080,   371-5200,   770-1320,   420-1975  and
600-2550,  respectively.   BCFs for carnivorous,  planktlvorous  and  omnivorous
fish  1n  the  field  ranged  from  590-35,675,  445-27,000  and  364-23,000,
respectively (Lemly, 1985b).
    Addition  of  selenium  to field   mesocosms  resulted   1n a  replacement  of
chrysophytes  by  chlorophytes  In  low-dose enclosures and  cyanophytes  1n  the
0144d                               4-55                             04/01/89

-------
high-dose  enclosure.   There  were no  significant  differences  between  zoo-
plankton  communities  among  1,  10  and 100  yg selenlte-selenlum  treatments
(Salkl et a!., 1985).
0144d                               4-56                             04/01/89

-------
                             5.   PHARMACOKINETICS
5.1.   ABSORPTION
    Rhoads and  Sanders  (1985)  found rapid  clearance  of  75Se-selenate (0.35
yg)  and   75Se-selen1te  (0.21  yg)   from  the  lungs  of  female  rats  treated
by  Intratracheal  Instillation.   For both  compounds,  -50J4  of  the  dose  was
cleared 1n 44 minutes.
    In a  study  by  Welssman  et  al.   (1983),  10  beagle  dogs  of  each  sex
(sedated   with  sodium phenobarbHal and  trlflupromazlne  hydrochloMde)  were
exposed  (nose-only)  for  10-40  minutes  to  aerosols  of  75Se-selen1ous  add
or  75Se-selen1um.   Particle  sizes  were  .0.5+^2.4 ym  for  selenlous  add  and
0.7+K5 ym  for  selenium.  Two  dogs were  sacrificed  2  hours and 2,  4,  8,
16, 32, 64,  128, 194 and  256  days after exposure.   The  Initial  amounts  of
selenium   deposited  were  28+12  yg/kg  following  exposure to  selenlous  add
and  22^9  yg/kg  following exposure  to  selenium.   Determination  of  radio-
activity   1n the  tissues of dogs sacrificed at 2 hours Indicated that -97 and
80% of the  deposited selenium was  absorbed  following exposure to selenlous
add and  selenium,  respectively.   Welssman  et al.  (1983)  also exposed dogs
to  radlolabeled  selenlous  acid or  selenium (2-30 yg Se/kg)  by  Intravenous
Injection, nasal Instillation, gavage  or  1n  the diet.  The dogs were sacri-
ficed  4  days  after  dosing and the  amount  of selenium  1n  muscle following
oral or  nasal  Instillation compared with  Intravenous exposure was  used  to
estimate   absorption.   Absorption  was  estimated  at  73+13.7,  95.8+16.4  and
49.5+9.9X of  the administered  dose following nasal Instillation,  gavage and
feeding of selenlous add, respectively, and 52.U59, 72.7*04.2 and 9.5+1.7X
following nasal  Instillation,  gavage and  feeding of  selenium,  respectively.
0145d                               5-1                              06/15/89

-------
    Medlnsky et  al.  (1981) studied  the  absorption  of  selenium  In  anesthe-
tized  male  and  female  F344  rats  exposed  for  10  minutes  to  resplrable
particles   of   7SSe-selen1um   (1.3^0.7   yg/rat)   or   75Se-selen1ous   add
(3.6^2.4  yg Se/rat)  by nose-only  Inhalation.   Additional  groups   of  rats
were  treated  with   75Se-selen1ous  add  or  75Se-selen1um by  gavage  (<20
yg/rat)  or  nasal   Instillation  (<20  yg/rat).   The  dermal  absorption  of
selenlous add was  also determined.   The Investigators estimated absorption
by comparing tissue  levels of  selenium  following  Inhalation,  oral,  nasal  or
dermal exposure to tissue  levels of selenium following  Intravenous exposure.
The  nasal  absorption  of   selenium was  determined  In  anesthetized   rats  1n
which  the  stomach   was  surgically   exposed   and  Ugated  at  the  pyloHc
sphincter to trap material  cleared Into the stomach.  The calculated absorp-
tion of selenium following exposure to selenlous add or selenium by various
routes  Is  shown  In  Table  5-1.   The  absorption of  selenium  from selenlous
add and  selenium metal was  similar  following  nasal  exposure,  but  alveolar
and  gastrointestinal  absorption  of  selenlous add  was  greater  than  the
absorption of selenium metal.
    Based on studies of acute  Inhalation  exposure of rats  (Medlnsky et al.,
1981)  and  dogs  (Welssman  et  al.,  1983)  to  selenium and selenlous  add,
Medlnsky  et al.   (1985) developed a  model  to project  steady-state tissue
selenium  concentrations  1n humans exposed  by  Inhalation  to  selenium.   The
results, shown In Table 5-2,  Indicate that exposure to  selenium In urban air
(1 ng  Se/m3)  for a  lifetime would contribute  little to the selenium burden
1n human  tissues, while exposure at  the TLV  (0.2 mg Se/m3),  8 hours/day,  5
days/week for a lifetime would greatly Increase the lung and liver burden of
selenium compared with  the range of selenium levels generally found  In these
tissues.
0145d                               5-2                             06/15/89

-------
                                   TABLE  5-1
           Calculated Internal Absorption of 75Se 1n Rats Expressed
                       as  Fraction  of  Administered  Dose3
Total Absorption^
Chemical
Form Nasalc
Selenlous 0.18+0.092
acid
Elemental 0.16+0.037
selenium
Gastro- Alveolar6
1ntest1nald
0.87+0.094 0.94+0.029

0.504-0.28 0.57+0.17

Cutaneous^
0. QUO. 01

NR

aSource:  Medlnsky et al., 1981
bMean +. SD
cRats were sacrificed 4 hours after dosing.
dRats were sacrificed 24 hours after dosing.
eFract1on absorbed per day for 9 days
NR * Not reported
0145d                               5-3                              04/01/89

-------
                                  TABLE 5-2
          Projected Steady-State Concentrations of Selenium In Organs
      and  Tissues of People Exposed to Selenium 1n Urban Atmosphere or In
    the Workplace and  Measured Selenium  Concentrations  1n  Human  Tissues3*^
Organ
Lung
Liver
Blood
Other tissues
Inhaled
Atmosphere0
0.4
0.02
0.01
0.002
Selenium
In the Workplace at
TLV Concentrations'1
22,000
1,200
440
78
Selenium In
Human Tissues6
100-300
70-680
100-340f
60-370
aSource:  Medlnsky et al..  1985
bng Se/g blood-free tissue  wet weight
cAssumes 23 mVday Inhaled, 1  ng Se/m3
^Assumes 0.2 mg Se/m3 Inhaled  8 hours/day,  5  days/week
eCasey  et  al.  (1982).   Measured concentrations  for  numbers Indicate  range
 of values reported.
fAllaway et al. (1968)
0145d                               5-4                              04/01/89

-------
    Venugopal and  Luckey (1978)  stated  that gastrointestinal absorption of
selenium  1n  the  form of selenltes,  selenates  and selenlferous compounds 1n
mammals 1s  rapid  and efficient,  but  metal  selenldes and elemental  selenium
are absorbed poorly.
    Thomson  and   Stewart (1974)  treated  three  fasting  women  orally  with
7SSe-selen1te at  a  dose of  <10  yg  Se.   Absorption  of  selenium was  calcu-
lated to be 70, 64  and  44%  1n  the three  Individuals, Indicating  variation In
the gastrointestinal absorption of selenlte-selenlum.
    Young  et  al.  (1982) fed  volunteers  chicken meat Intrinsically  enriched
with   the   74Se   Isotope    (chickens   were    treated   by   gavage    with
74Se032~),  and  a   simultaneous  dose  of   76Se   selenlte.   Examination  of
fecal  excretion   of  74Se   and  76Se   Indicated   a  lower  absorption  of
selenium from selenlte than  from selenium Incorporated In  chicken  meat.
    Selenium 1s an  essential  element  for  humans.   The  element 1s a part of
the  enzyme  glutathlone peroxldase,   which  Is  Important  1n  the  cellular
defense  against   oxldatlve   damage.    Because   selenium   Is   essential,  the
concept  of bloavallabllHy  Is  discussed often.   Factors  that  affect the
bloavallabllHy of  selenium,  as  reviewed by  Young et al.   (1982),  Include
Intake levels of  the element  and Us chemical  form;  the  presence  or absence
of  promoters  or  Inhibitors  such  as  ascorbic  acid,  phytate,  fiber,  sugars,
fats  and  proteins;  mineral-mineral  Interactions;  type  and  degree  of  food
processing; and  the 1ngest1on  of certain drugs.   Young  et  al.   (1982)  also
suggested  that nutritional  state, physiological  states  (such as growth and
pregnancy) and pathological  states can Influence selenium  utilization.
    Levander et al.  (1983) studied the bloavallabllHy of selenium  In  groups
of  10 Finnish men  provided with selenium  supplementation  (200  yg  Se/day)
1n  the form of Se-r1ch  wheat,  Se-r1ch  yeast  or  sodium selenate for  11  weeks,
0145d                               5-5                              04/01/89

-------
followed  by  a  10-week  posttreatment  period.    Twenty   men   treated  wHh
placebos  served  as controls.   All  subjects  In  the  study  had low  prestudy
plasma  selenium  levels  (-70  ng/mi).   Selenium b1oava1labH1ty was  assessed
by  the  determination  of  platelet glutathlone peroxldase  activity,  and  the
determination  of  plasma  and  erythrocyte  selenium  levels.   Treatment with
selenate  resulted  1n  a  43 ng/ma Increase  1n plasma  selenium at 11 weeks,
when plasma  concentration  appeared  to plateau. Supplementation with  Se-Mch
wheat or  yeast  resulted  1n plasma selenium  Increases of  102 and 97  ng/mi
at  11   weeks,  respectively,  and these  levels appeared  to  be  Increasing.
Erythrocyte  selenium  levels did  not  appear  to Increase following  treatment
with selenate, while treatment with wheat and  yeast  more  than  doubled eryth-
rocyte  selenium  levels  without   signs  of  a  plateau.  Platelet  glutathlone
peroxldase  activity  following  treatment  with  selenate  or  Se-r1ch  wheat
Increased  to 170  and  162%  of  control activity,  respectively,  following  4
weeks of  treatment,  and appeared to  plateau.   Treatment  with  Se-rlch  yeast
resulted  In  a  slow Increase  In platelet  glutathlone peroxldase  activity,
approaching  the  activity of  the  other  treatment groups  by  week 11.   During a
10-week  posttreatment   period,  plasma  selenium  levels and  platelet  gluta-
thlone  peroxldase  activity  returned to  control levels 1n  the  selenate  group,
but  remained elevated  1n  the  Se-r1ch  wheat  and yeast  groups.  The  response
of  plasma glutathlone  peroxldase activity  to selenium supplementation  (all
forms)  was   found  to  be  very poor.    The  Investigators stated that  seleno-
methlonlne (SeMet)  Is the  major  form  of  selenium  In  Se-r1ch wheat,  and  H Is
also thought to be a major component  of yeast selenium.
     In  a  similar   study   1n  New Zealand,  where  selenium levels  are  low,
Thomson et al.  (1982)  treated 16 volunteers with  dally supplements  of  SeMet
or  sodium selenlte (100  yg  Se) for  at   least  11  weeks.    Treatment with

0145d                               5-6                              06/15/89

-------
SeMet resulted  1n  a  greater  Increase In whole blood,  erythrocyte  and  plasma
selenium compared  with  treatment  with  selenlte.   Increases  In whole  blood
glutathlone  peroxldase  were  similar  following   treatment  with  the  two
compounds.
5.2.   DISTRIBUTION
    Following  nose-only  Inhalation  exposure  of  beagle  dogs  to aerosols  of
7SSe-selen1ous  add  or  selenium,  most  of  the  selenium  was  found  1n  the
lung, liver, kidneys,  spleen,  heart and blood (Welssman et al.,  1983).  The
percent of the  Initially deposited  selenium retained  1n  these tissues  during
the  first  90  days  postdoslng  ranged  from  -0.1-0.2% In  the liver  to  0.0002-
0.001%  1n  the  blood.   Concentrations of  7SSe  1n   the  liver  peaked 2  hours
after exposure, while concentrations  1n other tissues  did  not peak until  2-4
days   after   exposure.   Elimination   half-lives   of   selenium,   estimated
graphically, were -42-46 days for the liver and  30  days for other  organs.
    Two weeks  after Fisher  rats were  given  Intratracheal Instillations  of
7sSe-selenate   (0.35  vq)   or   75Se-selen1te   (0.21   vg),   ~20X  of   the
administered activity was found  In the  carcass, 5% of  the dose was found 1n
the  liver  and  5%  of the dose  1n  the  skeleton  (Rhoads and  Sanders,  1985).
Results for the two compounds were similar.
    Thomson  and Stewart (1974)  treated  three fasting  women orally  with
7SSe-selen1te  at  a  dose  of  <10  yg   Se.    Based  on  urinary   and  fecal
excretion,  the  Investigators  estimated  that  53, 51 and 35%  of the dose  was
retained by the three subjects 14 days after  dosing.
    McAdam  and Levander  (1987)  fed  purified Torula  yeast-based  diets con-
taining 2.5  or 5.0  yg Se/g as  D-SeMet,  L-SeMet,   sodium  selenlte or  sodium
selenate to groups  of  eight male weanling Sprague-Dawley  rats  for 6 weeks.
0145d                               5-7                              06/15/89

-------
Although Inorganic selenium appeared to be more toxic based on  mortality and
body weight  gain  (Section  6.1.2.1.),  skeletal muscle  selenium levels were
higher  following  exposure  to  organic selenium compounds.  Plasma, red  blood
cell,  liver,  heart  and  skeletal  muscle  selenium  levels are  presented  In
Table  5-3,  which  Indicates  that  higher  muscle  selenium  levels  were not
reflected consistently  by  proportionate  Increases  1n  plasma  or  red  blood
cell  selenium;   therefore,  monitoring  of   whole  blood  selenium  may not
adequately reflect selenium body burden.
    Using light  and  electron  microscopic  visualization, a series  of  studies
Indicated   that    selenium   can   accumulate   In   the   anterior   pituitary
(Thorladus-Usslng and Danscher, 1985), adrenal  gland  (Thorladus-Usslng and
Rasmussen, 1986) and oocyte {Thorladus-Usslng et al.,  1986)  of rats  treated
with sodium  selenlte orally  or  by 1ntraper1toneal Injection.   The  selenium
found was reported  to  be In the form  of  selenlte or  a selenlde bound  to  an
endogenous  metal  (possibly  zinc).    In   the  adrenal   gland  and  anterior
pituitary, selenium  located  1n  secretory  cells  and  In the  ovary;  selenium
accumulated  1n   secondary   and  mature  Graaflan  follicles.   The  amount  of
selenium located  1n  the  tissues of concern  In  relation to  administered dose
was not provided.
    Gregus and Klaassen  (1986)  examined tissue  selenium levels  4 hours  after
4-6  urethane-anesthetlzed,  bile  duct-cannulated male Sprague-Dawley  rats
were  given  an  Intravenous  Injection  of   selenium   (sodium  selenlte and
[75Se]selen1ous  add)  at  doses  of  0.03,   0.1,  0.3  or  1.0  mg/kg.   The
distribution of  selenium was  found to  be  Independent  of dose  within the dose
range  examined.    Selenium levels  were   highest  In  the  liver and  kidney,
accounting for 7.42-12.6  and  4.34-7.52% of   the dose/10 g of  tissue,  respec-
tively.  Blood  levels  of selenium accounted  for  0.621-1.78%  of the  dose per
10 ml.

0145d                               5-8                              06/15/89

-------
tn
o.
                                                     TABLE 5-3


       Concentrations of  Selenium  1n Tissues  from  Rats  Fed  Diets Containing Added  D-Selenomethlonlne (SeMet),
                                          L-SeMet,  Selenlte or Selenatea»b
tn
i
us
Dietary Se
Level
(vg/g)
Control
2.5
2.5
2.5
2.5
5.0
5.0
5.0
5.0
Form


D-SeMet
L -SeMet
selenlte
selenate
D-SeMet
L -SeMet
selenlte
selenate
Plasma
(vg Se/ml)


0.6H0.03b
0.56+0.06b
0.43+0.04b
0.42+0. Olb
0.8U0.043
0.58+0.08b
0.64+0.04b
0.56+0.05b
RBC
(yg Se/g)


2.05+0.12b
1.6H0.10b
1.07^0.04C
0.85±0.12C
3.07*0.11a
3.02+0.38a
1.92f0.34b
1.64+0.21b
Liver
(vg Se/g)


2.83+0.20b
1.86+0.12d
1.35f0.07e
1.34±0.14e
3.40+0.223
3.49+0.20a
2.39+0.15c
1.8H0.15d
Heart
fug Se/g)


2.20+0.20a
1.15+0.23b
0.28+0.02C
0.47+0.10°
1.94+0.05a
2.0H0.10a
0.50+0.06C
0.44*0.03C
Skeletal Muscle
(yg Se/g)


1.27+0.31b
0.84+0.06C
0.12+0.01d
0.14±0.01d
1.57+0.113
1.69+0.073
0.17*0. 03d
0.17+0.01d
    aSource:   McAdam and  Levander,  1987


    bMean  +  SEM,  n=5  except  In  5.0 vg/g  selenlte
     common  superscript letter  differ  p<0.05.
group  In  which n=3.   Means  In  a  column not  sharing a
CD

-------
5.3.   METABOLISM
    Reviews  of   the  metabolism  of  selenium  (Mushak,  1983;  Ganther,  1986)
Indicate that reduction  1s  a  major  pathway of selenium  blotransformatlon  In
mammals.   The  reduction  pathway  of  selenium metabolism  1s  shown In  Figure
5-1.  In  the presence of ATP,  sulfurylase enzyme and  glutathlone,  selenate
(+6)  1s  reduced  to selenlte  (+4).  Selenlte  Is  further  reduced  by  reducing
agents,   Including  thlols  and  ascorbic add, to form  selenotMsulfldes.  The
reaction of  selenlte  with  glutathlone,  the major  cellular thlol,  results  1n
the formation of  GSSeSG,  which  1s further reduced to  hydrogen  selenlde  (-2)
by  the  glutathlone reductase  pathway.   The  formation of hydrogen  selenlde
occurs 1n the liver and the erythrocytes (Ganther,  1986).
    Elemental selenium can be formed  by the spontaneous  decomposition  of the
selenopersulf1de Intermediate In  the  glutathlone reductase pathway or  by the
spontaneous  oxidation  of  hydrogen  selenlde  (Ganther,   1986).   Elemental
selenium formed  1n the tissues may  be biologically active;  H  Is  hydrophoblc
and can  bind tightly to proteins, and  may  bind to  nucleic adds  and Uplds.
    The   major  elimination  pathway  of  selenium  1s   by  the  methylatlon  of
hydrogen selenlde  to  dlmethylselenlde or  tr1methylselenon1um.   The formation
of  tr1methylselenon1um,  which 1s excreted In the  urine, predominates  when
selenium levels  are low, while at higher  levels dlmethylselenlde  accumulates
and 1s exhaled through the lungs  (Mushak,  1983).
    Hydrogen selenlde  1s  converted  to  methylated selenium by  at least two
methyl  transferases:  the  most active transferase,  found  1n  the  cytosol  of
the liver, and a  transferase  found  1n liver mlcrosomes (Mushak,  1983).   The
methyl  donor  for both enzymes  1s S-adenosylmeth1on1ne.   Increased  selenium
Intake results In the enhancement of mlcrosomal  transferase activity.
0145d                               5-10                             04/01/89

-------
               Se04*~  •»  RTP  	»    Rdenosine  Selenophosphate  +  PPt  	
              (•tl«n*ti>       tulfurylui, Hje*
               SSSeOj*'  *  RHP  	» SeO,*'   +  6SSG
                                 ISH   (Illtnltl)
                4 RSH (69.  6SH)  t> H,SeOj 	> 6SSeS6  +  3H20
                                            ((•l«nep*riul'
               GSSeSS 	»•        6SSeH  	»•   HzSe
                       MPOPH, flutcthlen*          KSDPM, (lutathlen*  
                            nductu*                 p«dutt»n
                HzSe 	»> CH3SeH  	» CH3-Se-CH3  	•> (CHj),  Se*
                                         (dl*«thyli«loniun>     (triitthylstlonlun)
                                          FIGURE  5-1
                                   Metabolism of Selenium
                          Source:   Muschak,  1983; Ganther, 1986
0145d                                       5-11                                     04/01/89

-------
    Hydrogen  selenlde  may  also  react  with  metal   Ions   to   form   metal
selenldes, or with  protein and metal  Ions  to form a  protein-metal  selenlde
complex (Garberg and  Hogberg,  1986).   These pathways  may help prevent  toxic
effects of heavy metals (e.g.,  cadmium, mercury).
    Selenocyanate has  been  Identified  as  a  minor  urinary metabolite 1n  rats
treated with selenium  at  relatively  high  doses (Vadhanav1k1t et  al.,  1987).
A  study  of  the  metabolism of  selenocyanate 1n  rats  (VadhanavIkH et  al.,
1987) Indicates  that  H  also  1s methylated, with 61%  of  a  subcutaneous  dose
of 75SeCN recovered as tr1methylselenon1um In the  urine of rats.
5.4.    EXCRETION
    Welssman  et  al.   (1983)  reported  that  fallowing  nose-only  Inhalation
exposure   of   beagle  dogs   to  7SSe-selen1ous  add   (28^9  ygSe/kg)   or
7sSe-selen1um  (22+.9  pg  Se/kg), urinary  excretion  accounted  for 70-80%  of
the  excreted  dose, presumably  during  the  first  32 days  after   dosing,  the
period  during  which  urine and  feces  were collected.   Exhalation  of  73Se
accounted for only 0.6%  of the excreted dose during the  first 10 days  after
exposure; the remainder of the  excreted dose appeared  In the feces.
    Determination of urinary and  fecal excretion  of selenium after  male  and
female F344 rats were  treated  by  Intravenous  Injection,  nose-only Inhalation
exposure  or  gavage with  low  doses  (<20  yg  Se/rat)  of  7SSe-selen1ous  add
or 75Se-selen1um Indicated similar  distributions of  Se 1n  the   excreta  24
hours after  dosing  (Medlnsky  et al.,  1981).  Following  Intravenous,  Inhala-
tion and  gavage  treatment, 27-52, 20-28  and 14-45% of the dose  appeared  1n
the  urine, respectively,  while 5-11,  6-8  and 12-28% of  the dose  appeared  In
the feces, respectively.
    During  a  2-week  period  following  an  Intratracheal   Instillation  of
7SSe-selenate   (0.35   jig)  or   75Se-selen1te  (0.21   pg),  rats   excreted
-40% of the dose 1n the urine and feces (Rhoads and  Sanders, 1985).

0145d                               5-12                             04/01/89

-------
    Thomson  and  Stewart   (1974)  treated  three  fasting  women  orally  with
7SSe-selen1te  at  a  dose  of  not  more  than  10 jag  Se.   Fourteen days  after
treatment ~14,  9  and 7% of the dose  was  excreted In the urine, with  33,  40
and 58% of the dose excreted 1n the feces.
    Gregus and  Klaassen (1986) reported that  4  days  following an Intravenous
dose of  selenium  given to  male  Sprague-Dawley rats (0.3 mg  Se/kg  as  sodium
selenlte  and  [75Se]selen1ous  add),  17.6+3.6 and  4.34^0.34%  of  the  dose
was excreted 1n the urine and feces, respectively.
    McConnell  and  Roth  (1966)  studied the  excretion of  selenium In  young
male rats  given  subcutaneous  Injections  of  varying doses of  7SSe-selen1te
or  75Se-L-SeMet.   Following treatment  with  selenlte at  a  dose of 0.005  or
0.913 mg Se/kg, 0.2  and 10.6% of  the dose was exhaled In  24 hours,  with 32.9
and 23.2%  of the  dose  excreted  1n the urine,  respectively.   Treatment with
selenlte at  doses >2.146 mg Se/kg  resulted  1n the exhalation  of at least 41%
of  the  dose, with 2-20% of the  dose excreted  1n the urine.   L-SeMet  treat-
ment resulted  1n  the  exhalation  of  1.3  and  6.1%  of  the dose  1n  24  hours,
with 27.1 and  35.8%  of the dose excreted In  the  urine at  doses  of  0.001 and
1.065 mg Se/kg, respectively.  At  doses of  L-SeMet  >2.45  mg Se/kg,  >17.1% of
the dose  was exhaled, while  -7-22% of the dose  was excreted In  the  urine.
Following treatment  with  either  selenlte or  L-SeMet, exhalation of selenium
was greatest during  the  first 6  hours  after  treatment.    The  Investigators
noted that larger amounts  of  selenium were  excreted following treatment with
selenlte compared to treatment with L-SeMet.
    Calabrese  (1985)  stated that the  renal excretion  of  tr1methylselenon1um
1s  slower In male  rats  compared  with female rats.  This  difference 1n secre-
tion was eliminated  1n  castrated male rats,  Indicating  that the excretion of
tr1methylselenon1um   In  male   rats  Is  androgen-dependent.    Selenium  and


0145d                               5-13                             06/15/89

-------
methylated selenium compounds  have  been  shown to be more toxic  to male  rats
than  to  female  rats.   This  Increased  toxldty   to  male  rats   was   not
eliminated  1n  castrated  male rats,  suggesting an  additional  mechanism  of
Increased sensitivity of male rats to selenium.
5.5.   SUMMARY
    Studies in dogs (Welssman  et  al.,  1983)  and  rats  (Medlnsky  et al.,  1981}
Indicate that selenium metal  and  selenlous add  are absorbed readily follow-
ing Inhalation exposure.   Gastrointestinal absorption  and bloavallabllUy  of
selenium Is  greater  with organic selenium compounds  such  as SeMet  compared
with Inorganic compounds.  Among  Inorganic selenium compounds,  selenltes  and
selenates  are  absorbed   more  readily  than   metal  selenldes  and elemental
selenium (Venugopal and Luckey, 1978).
    Following absorption,  selenium  1s distributed throughout the body,  with
higher  levels  found  1n  the  liver and  kidneys.   McAdam and  Levander  (1987)
found higher muscle  selenium levels  In rats   fed diets  with  organic  selenium
compounds (0- or L-SeMet) compared to rats fed sodium selenate or selenlte.
    Reduction  of  selenium   compounds   to   hydrogen   selenlde   followed  by
methylatlon 1s  the major  transformation  pathway leading to  the  excretion  of
selenium  (Mushak,  1983).   Methylatlon  occurs  predominantly  1n the  liver.
Trlmethylselenonlum,  excreted  1n the  urine,  Is  the major excretory product
at  low  doses of  selenium.   At higher doses  of  selenium,  larger amounts  of
dlmethylselenlde  are  produced.    Olmethylselenlde   1s   exhaled  through  the
lungs.   Metal  selenldes  and  metal-protein  selenlde  compounds  may  also  be
formed from hydrogen selenlde.
0145d                               5-14                             06/15/89

-------
                                 6.  EFFECTS
6.1.   SYSTEMIC TOXICITY
6.1.1.   Inhalation Exposure.
    6.1.1.1.   SUBCHRONIC -- Pertinent   data   regarding   the   toxldty   of
selenium  following  subchronlc  Inhalation  exposure were  not  located In  the
available literature dted In  Appendix  A.
    6.1.1.2.   CHRONIC -- Hamilton  and  Hardy  (1974)  reviewed  symptoms  that
have  been  reported 1n  chronic  Industrial  selenium poisoning.   The  symptoms
reported most  frequently  Include a strong  garlic  odor 1n the breath,  sweat
and urine, acute sore  throats and  cold-Uke symptoms,  gastrointestinal  Irri-
tation, lacrlmatlon, and a metallic taste  1n  the mouth.   Exposure  concentra-
tions  resulting  In  these effects  were not  provided.   The  development  of
garlic  breath  and  cold-Uke   symptoms  Is  thought  to be  a  result  of  the
hepatic production  of dlmethylselenlde, which  1s  exhaled through the  lungs
(Dlskln et al., 1979).
    Dlskln et  al.  (1979)  described a  case of a 71-year-old man who  had  been
employed  1n  selenium  refining  for 50  years.   The man was  admitted to  the
hospital  with  chest   pain  and  ECG evidence of  cardiac  Infarction.    The
patient  had reddish   orange  hair   and  red  fingernails,   a  result  of   high
selenium exposure.  The patient  developed  cardlogenlc  shock and died 8  days
after  admission.   At  autopsy,  sections  of the lungs  showed  numerous  perl-
vascular  noncaseatlng  granulomas  that  varied In  age.   Analysis  of  tissues
for selenium revealed  abnormally high  levels In  the lungs (109 ppm, normal
levels 0.15-0.21 ppm),  perlbronchlal  lymph nodes  (26 ppm, normal  levels  0.1
ppm),  and  hair  (213  ppm, normal  levels   0.36-0.74  ppm), but  not In  other
tissues,  Including  the trachea  (0.33  ppm, normal  levels not  reported)  and
0146d                               6-1                               04/03/89

-------
liver  (0.47  ppm,  normal   levels  0.43-0.53 ppm).   The  Investigators  stated
that  the  perlvascular  location of  the granulomas  suggested a  blood-borne
toxin, presumably d1methylselen1de.
    Glover  (1967)  measured  urinary  selenium concentration  of workers  at  a
selenium rectifier plant over  a 5-year  period.   A1r  concentrations,  measured
sporadically,  appeared  to  correlate with  urinary  selenium  concentrations.
Glover  (1967)  recommended  a maximum allowable  urinary  concentration  of 0.1
mg  Se/i,  which he  believed  corresponded to an  air  concentration of  0.1  mg
Se/m3.  Glover  (1967)  did  not  believe  that  garlic odor of the breath  was  a
reliable  guide to  selenium absorption.   Garlic  breath was  absent  In  men
excreting as  much as  0.1  mg  Se/a,  1n  the urine,  while  1t  was  present  In
workers  with  urinary  selenium  concentrations  of  0.5-1.0  mg/i.    Glover
(1967) reported that the cause  of  death of  17  selenium workers was  not found
to  differ from expected  values.  The duration of  exposure  for these workers
was not provided.
6.1.2.   Oral Exposure.
    6.1.2.1.   SUBCHRONIC -- Halverson  et a'i.  (1966) fed groups  of  8  or  10
young  male   Sprague-Dawley  rats  sodium selenlte  or  selenlferous   wheat  at
dietary levels  of  1.6,  3.2, 4.8,  6.4,  8.0  or  9.6 ppm  selenium for  6 weeks.
Assuming rats  eat  food  equivalent  to 5% of  their  body weight/day (U.S. EPA,
1986a), these  diets  provided  selenium  doses  of 0.08,  0.16,  0.24,  0.32, 0.4
or  0.48 mg/kg/day.   An additional group  of rats  was  fed  selenlferous wheat
at  a  dietary level of 11.2 ppm selenium (0.56  mg/kg/day).   Control  rats were
fed  the basal  diet, which  contained 0.7 ppm Se  (0.04  mg/kg/day).   All rats
treated with  selenlferous  wheat  at  11.2  ppm  died, while  5/8 and  1/8 rats
treated with  selenlferous  wheat at  9.6 and 8.0 ppm died.   In rats treated
with  sodium  selenlte,  1/10 and  1/8  died at 9.6 and 8.0 ppm, respectively.

0146d                                6-2                              04/03/89

-------
All deaths  occurred  after the  fourth  week  of the study.  Treatment  of  rats
with either  form of  selenium  at >6.4  ppm  resulted In  significant  (p<0.01)
growth  depression.   A  small  growth  depression was  also observed   1n  rats
treated with selenlte (but not  selenlferous wheat)  at  4.8 ppm.   Rats treated
with selenlferous wheat  at >6.4 ppm  Se had  significantly (p<0.05)  enlarged
spleens, while  at >8.0  ppm  a  significant  (p<0.05)  reduction  In  hemoglobin
and an  enlargement  of   the  pancreas  was  observed.    Liver  weights   of  rats
treated at 6.4-11.2 ppm  Se were reduced significantly  (p<0.01)  and livers of
rats treated  at >6.4 ppm Se showed  varying  degrees  of  mottling,  roughness
and discoloration.   It   Is not  clear  1f the  liver  effects were observed 1n
selenlferous wheat or selenlte  treated  rats,  or  In  both  groups  of  rats.   The
dietary level of  3.2  ppm Se  (0.16 mg/kg/day)  1s considered a NOEL  (U.S.  EPA,
1984a).
    NCI/NTP  (1980a)  treated  groups of  10  male  and  10 female  F344  rats  and
B6C3F1  mice by  gavage with selenium sulflde  1n  CMC  7 days/week  for 13 weeks,
followed by 1 week of observation.  Rats were treated  at doses  of  0  (vehicle
control),  3.2,  5.6,   10,  17.8  or 31.8  mg/kg/day,  and mice  were  treated at
doses of 0  (vehicle  control),   21.6,  46.4,  100,  216  or 464 mg/kg/day.  Doses
were not  calculated  1n   terms of selenium  because the substance under study
was not defined  clearly;  analysis Indicated that the substance  was primarily
selenium  monosulflde with  selenium   dlsulflde, selenium and sulfur  also
present.  All rats survived  the treatment,  and  no  effect on  growth  rate was
noted.   The only hlstologlcal  change  noted  In rats  was focal  coagulation
necrosis  with  Infiltration  by  Inflammatory  cells  1n  the   livers   of  rats
treated  at  31.8  mg/kg/day.   In mice,  464 mg/kg/day  was clearly an effect
level;  four  females  and  one  male died, body weight gain was  suppressed 1n
females, and  from week   3  to  the end  of the  study,  both  sexes  appeared  thin

0146d                               6-3                              06/15/89

-------
or  showed  hunched  posture on  both.   An Increased Incidence and  severity  of
microscopic Interstitial  nephritis was  also observed 1n mice  treated  at  464
mg/kg/day.   Except  for  the  death   of one  female  mouse  treated  at  216
mg/kg/day, compound-related effects were not observed 1n mice at lower  doses.
    Beems  and  van  Beek  (1985)  fed  groups of  eight male  and eight  female
weanling  Syrian  hamsters  sodium  selenlte  1n the  diet  at  concentrations  of
0.1 (unsupplemented  control),  1, 5,  10 or  20  ppm Se for 42 days.   Based  on
food Intake data,  the  Investigators  calculated that the  selenium  Intake  was
0.007,  0.1, 0.31, 0.61 or  1.21 mg/kg/day for  males and  0.007,  0.1, 0.3, 0.63
and 1.26  mg/kg/day  for  female  hamsters  fed at  0.1,   1,  5,  10  or 20  ppm,
respectively.   No  effects  on survival  or  behavior were noted.   Body  weight
gain  was  reduced  1n  males  at  all  doses  compared   with  unsupplemented
controls,  but  the  reduction was  found  to  correspond to  differences  1n food
Intake.  No changes  1n  body weights  were  noted  In females.   At necropsy,  no
gross   treatment-related  abnormalities  were   noted.    The  only  microscopic
changes observed  were lesions  In the  liver  of  most hamsters  treated  at  20
ppm, Including necrosis,   oval-cell  proliferation  1n  the  perlportal  areas,
enlargement of  hepatocytes,  and  hepatocellular nuclei  mainly  In  the centrl-
lobular areas, and accumulation  of brown pigment  1n  macrophages In the perl-
portal  region.   This study  Identifies  a  subchronlc  LOAEL  of  1.21  mg/kg/day
1n hamsters, and a NOEL of 0.63 mg/kg/day.
    In a  60-day study, McAdam and  Levander (1987) fed  groups  of  eight male
weanling  Sprague-Dawley  rats  Torula yeast-based purified  diets  (<0.02  yg
Se/g)   with 2.5,  5  or  10  vg/g   (ppm)  added  selenium  as  D-SeMet,  L-SeMet,
sodium selenlte or  sodium  selenate.   Assuming  rats eat  food equivalent to 5%
of  their  body  weight/day, these  diets  provided selenium at  doses  of  0.125,
0.25 or  0.5 mg/kg/day.   All  rats  provided  with diets  containing selenium
(all  forms) at  10  yg/g  died within  29  days.   At 5.0 yg/g,  5  selenlte-,

0146d                               6-4                              06/15/89

-------
3  L-SeMet-,  1 D-SeMET- and  1  selenate-fed  rat  died.  At  2.5  yg/g all  rats
survived  except   two  rats  fed  selenate.   Body  weight  gain  was  decreased
significantly  1n  rats  fed  Inorganic  selenium at 5.0  jig/g,  but not  In  rats
fed organic selenium.  Little  difference  1n  body  weight  gain  was  observed 1n
rats fed  2.5 yg  Se/g.   Determination of muscle  selenium  levels  at  the end
of the  study  Indicated  a  greater accumulation of selenium following  organic
selenium exposure compared with  exposure  to  Inorganic  compounds  (see  Section
5.2.).   Hlstopathologlcal  examinations were not completed.
    In  a  study  by  Keller  et al.  (1986),  the   Immune  responses of  female
Sprague-Dawley  rats  (12/group)   provided  with   drinking   water   containing
sodium  selenlte  at  0,  0.5,  2.0  or  5.0  ppm  Se  for  10 weeks were  examined.
Based  on  an  estimated  water  consumption  of  0.02  a/day  for a  0.25  kg  rat
(U.S.  EPA,  1980b),  the rats were  treated with additional  selenium at  doses
of ~0,  0.04,  0.16 or  0.4 mg/kg/day.  Determination  of NK  activity Indicated
significantly enhanced  cytoxlclty  against YAC-1   tumor cells  at 0.5  and 2.0
ppm,  with  NK activity  similar   to  controls  at  5  ppm.   Humoral antibody
production  against  keyhole  limpet  hemocyanln was  reduced In all  selenium-
treated rats,  with  a  statistically  significant  (p<0.05)  decrease observed
only at 5.0 ppm.  Selenium  treatment  significantly  (p<0.01)  reduced delayed-
type hypersensltWHy  response at  all doses.  Additional  effects  noted  were
a  significant  (p<0.05)  decrease 1n  prostaglandln synthesis,  an  Increase 1n
relative  thymus  weight  at 5.0 ppm,  and  a dose-related  Increase  1n relative
liver  weights.    The  only  microscopic   lesions  noted were  mild  hepatocyte
hypertrophy and mild congestion  of  renal cortical veins.  The  dose at  which
these  effects  occurred was  not  stated.   From  their  results,  the  Investi-
gators  suggested  that the effect of  selenium on  the  Immune system may be the
underlying mechanism of  the antlcarclnogenlc  and carcinogenic  properties of
selenium.

0146d                               6-5                             04/03/89

-------
    6.1.2.2.   CHRONIC -- Smith  et  al.  (1936)  conducted  an  epidemlologic
study  of  the  chronic  toxlclty of  selenium  1n  humans  on  farms  1n  South
Dakota,  Wyoming  and  Nebraska where  animals  were known to  be  suffering from
alkali  disease.   Clinical  signs  1n  farm workers  that  were attributed  to
selenium  Included  bad  teeth,  jaundice,  chloasma,  vertigo, chronic  gastro-
intestinal  disease,  dermatitis,  nail  changes,  arthritis,  edema,  lassitude
and fatigue.   Analyses  of food  resulted  In  an  estimation  of  dally  selenium
Intake  of  0.1-0.2  mg/kg.   Of  the urine  samples analyzed  for  selenium,  45X
contained  0.2-1.33  yg/mi,.   The  observed   symptoms  appeared  to  correlate
with a urinary selenium concentration >0.2 yg/mi.
    U.S. EPA (1980a) summarized  a  number of  studies  that  show an association
between  high selenium Intake and an  Increase 1n  dental  carles 1n humans.   No
exposure  data  were  provided.    Increased  Incidences  1n  dental carles  have
also  been  produced  In  rats fed  high levels  of dietary selenium  (Buttner,
1963),  and  In  monkeys,  Hacaca  1rus.  provided with  selenium  In  the  drinking
water at 1 ppm (Bowen, 1972).
    Beath  (1962) reported symptoms of lassitude,  alopecia,  discoloration of
the skin and loss  of  fingernails  In  persons  exposed  to well water  containing
9  mg/ft,  selenium.   Assuming  a  70  kg person   consumes  2   t  of  water/day,
this water  concentration  corresponds to  a dose  of -0.26  mg/kg/day.   Discon-
tinuation of the use  of selenium-contaminated water resulted  1n  regrowth of
hair and nails and Increased mental alertness.
    Yang  et al.  (1983)  reported outbreaks  of  selenium  Intoxication  In  the
Hubel  province  of  the  People's  Republic  of  China.  Dietary  selenium Intake
1n  this region  was  compared with a  region  of  high selenium  Intake  without
selenium Intoxication.   In  five villages  In  the Hubel  Province,  severe out-
breaks  of  selenosls  occurred,   affecting an  average  of 49.2X  of  the  popula-
tion,  with  85.5% of  the population  affected  1n  one  village.  Evacuation and
0146d                               6-6                              06/15/89

-------
change  1n  the  diet resulted  1n  recovery.   The Investigators  Indicated  that
the selenium source  appeared to be  high  levels of selenium  1n  surface  coal
that  leached   Into  the  soil and  became  available  to  uptake by  vegetable
crops.  Apparently,  the  outbreak was caused  by an Increase  In  the consump-
tion  of  selenium-accumulating vegetables,  Including corn and  turnip  greens,
as a  result of a drought that led to failure of the rice crop.   The  contri-
bution  of  selenium  In  drinking  water  was  considered  negligible,  compared
with dietary selenium.
    The symptoms  reported  Included bMttleness of  nails,  loss of  nails  and
hair,   pruritus of the  scalp, dermatitis characterized  by  hyperemla,  edema
and eruptive blistering, and  nervous  symptoms  such  as  peripheral  anesthesia,
acroparesthesla  and  pain In  the  limbs.   In  more  severe cases,  exaggerated
tendon  reflexes,  numbness,  convulsions,  motor dysfunction  progressing  to
paralysis,   and  hemlplegla   developed.   Mottled  teeth  and  Increased  tooth
decay  were  also found,  but  because  fluoride levels were  high 1n  the  area,
these  observations are  difficult  to Interpret.   The  above symptoms  were
associated with  dally  selenium  Intake ranging from 3.2-6.69 mg/day,  with  an
average  Intake of 4.99  mg/day,  estimated  from  analysis of  components  from
the diets  of   six  persons.   Analysis  of  the diets  of  three persons  from  an
area  with  high  selenium without  selenosls,  showed a  dally  selenium Intake
ranging from 0.24-1.51 mg/day with an average  Intake  of 0.75 mg/day.  Hair,
blood  and urine  selenium levels  found 1n high  selenium  areas  with  and  with-
out selenosls  and 1n areas  with  adequate  selenium are  presented In Table 6-1.
    Yang et al.  (1983)  also reported the case  of a male worker  who consumed
one tablet  of   sodium selenlte   dally  (-0.9  mg  Se) for  >2  years.   Effects
observed were  garlic odor  In his dermal excretions  and thickened,  fragile
nails.   After   stopping  selenium  Intake,  his  nails   became  smooth.    The


0146d                               6-7                              04/03/89

-------
CD
                                                             TABLE  6-1


                  Selenium Levels 1n Hair, Blood and Urine of Residents of High- and Adequate-Se Areas of China3
Area Sampled
H1gh-Se area where chronic
selenosls was observed
H1gh-Se area where chronic
selenosls was observed
H1gh-Se area where chronic
selenosls was not observed
Se-adequate area
Hair
n vq Se/g
1745 (1.9-8.2)
65 32.2
(4.1-100)b
14 3.7
(1.9-8.2)
1745 0.36
(±0.17)
Blood
n pg Se/ma.
Ill (0.35-0.58)
72 3.2
(1.3-7.5)
14 0.44
(0.35-0.58)
111 0.095
(±0.091)
Urine
n pg Se/mH
19 (0.88-6.63)
17 2.68
(0.88-6.63)
14 0.14
(0.04-0.33)
19 0.026
(±0.012)
           aSource: Young et al., 1982

           bF1gures 1n parentheses are range or standard deviation.
CD
IO

-------
authors suggested  that  a  dally Intake of 1.0 mg  selenium  over  a  long period
may  be  toxic  to  humans.   Blood  (0.179  ppm) and  hair  selenium  (0.828  ppm)
concentrations were  well  below those associated  with  selenosls  1n  the  high
selenium  area.  The  authors  suggested   that  tissue  levels  alone  are  not
sufficient  to  predict selenium Intoxication;  the form of  selenium  Ingested
must also  be  known because organic selenium accumulates In the tissues  to a
greater extent than Inorganic selenium.
    Longnecker et al. (1987) found no evidence of selenium toxldty  among 77
South Dakotans  with high  dietary  selenium  Intake  (up  to  590  yg/day).   The
ranges of  selenium levels 1n  plasma, whole  blood and toenalls  were 118-295
ng/ml, 176-521 ng/ma and 0.67-2.00 yg/g,  respectively.
    Chronic exposure  of  livestock to high  levels of selenium  can result In
alkali disease.  The  symptoms  observed Include emaciation,  lack of  vitality,
loss  of  hair,  separation of  the hoof,  atrophy  and  decompensation of  the
heart, cirrhosis of the liver, anemia, glomerulonephrltis  and erosion 1n the
joints of the long bones (U.S.  EPA, 1980a).
    In the  chronic  NCI/NTP (1980a) gavage study,  groups of 50  F344  rats/sex
and  50  B6C3F1  mice/sex were treated  by  gavage with selenium sulflde  1n CMC
(see  Section  6.1.2.1. for description  of the  test  material)   dally  for  103
weeks at dosages  of 3 or   15 mg/kg/day (rats) or  20  or  100 mg/kg/day (mice).
Equal  sized  groups  of  untreated and  vehicle-treated  controls  were  main-
tained.  High-dose  rats  of both  sexes had  reduced rate of body  weight  gain
that became evident  after  -16  weeks  of treatment  and  resulted  1n lower  body
weights  at termination.   Exposure to  selenium  had  no apparent effect  on
survival of rats.   Nonneoplastlc  hlstopathologlc lesions  In   rats  Included
focal hepatocellular  changes 1n high-dose males  and  Increased  lung  pigmenta-
tion  In  high-dose  males   and  In  low-  and  high-dose  females.    Exposure to

0146d                               6-9                              06/15/89

-------
selenium had  no  effects on body weights  or  survival  of mice.  There was no
mention of nonneoplastlc hlstopathologk  lesions  1n  mice.
    FHzhugh  et  al.  (1944)  found  decreased  growth rate,  a  restriction of
food Intake and  liver lesions  In  female  Osborne-Mendel rats (18/group)  fed
selenium In the diet  In  the form of  selenlferous  wheat  or  corn  at  5,  7  or 10
ppm [0.25, 0.35 or  0.5 mg/kg/day assuming a  rat  food  Intake equivalent  to 5%
of body weight (U.S.  EPA, 1986a)]  for  up  to  24 months.   Additional  groups of
rats were  fed  Inorganic  selenlde  (ammonium  potassium selenlde)  at  10,  20 or
40 ppm  Se  (0.5,  1  or  2 mg/kg/day).   The basal  diet  used 1n this  study was
suboptlmal 1n  the  level of protein.  Mortality was  Increased In rats fed at
>10 ppm  selenium.   Cirrhosis  of  the liver,  found  at  all  treatment  levels,
was  observed   1n  71/100  treated  rats surviving  for  >3  months.    Further
Information concerning  the  protocol  and  carcinogenic   effects  observed In
this study (described 1n Nelson  et  al., 1943) are presented 1n Section 6.2.2.
    Schroeder  and Mltchener (1971a)  treated  groups  of -50  male  and  50 female
Long-Evans BLU:LE rats with sodium selenlte  or sodium selenate  1n  the drink-
Ing water  at  a  concentration  of  2-3  mg  Se/l [0.28-0.42  mg/kg/day  assuming
a  0.35  kg rat  body  weight  and a  0.049  a/day  drinking  water  Intake  (U.S.
EPA, 1986a)].  Similar  groups  of rats provided  with  drinking water  with no
added selenium served  as  controls.   The  experiment  was  Intended to  continue
throughout  the  Hfespan  of   the   treated   animals  as   a  carclnogenlcHy
bloassay.  Male rats  were  especially sensitive  to  sodium  selenlte, with  50%
dying within  2 months.   Male  rats  surviving selenlte-treatment  were  given
sodium  selenate  for  the  remainder  of  the   experiment.   Similar  effects on
survival were not observed In female sodium  selenlte-treated  rats.   No  other
parameters of toxldty were reported.
0146d                               6-10                             04/03/89

-------
6.1.3.   Other  Relevant  Information.   Selenium,   an  essential  element  for
humans,  Is  a  part  of  glutathlone  peroxldase.   Additional  proteins  that
contain  selenium  Include selenoproteln of  muscle and  Se-transport  protein.
Other selenoprotelns and selenoamlno acids have been  found  but  have  not been
defined  further (Koller  and  Exon,  1986).   The dally  safe and  adequate level
of  selenium  Intake for  adults  1s  considered  to  be  50-200 yg (NAS,  1980).
This  level,  along with  safe  and adequate  dally  selenium  Intake  levels  for
Infants and children,  Is  presented In Table 6-2.
    Selenium  deficiency  1n  humans 1n China  has  been  associated with  Keshan
disease, a  cardlomyopathlc  condition  characterized  by cardiac  enlargement,
ECG  abnormalities,  heart failure,  gallop rhythm  and cardiac   shock  (Koller
and  Exon,  1986).   In  persons  In low selenium areas  In New Zealand,  stress
seems  to result  In  the  expression  of  selenium-deficiency  syndromes,  which
Include  Intermittent  muscle  tenderness and  pain  and  white fingernail  beds.
Epidemiology  studies  1n  low  selenium areas 1n the United States and  Finland
appear  to   show  an   association  between  low  selenium  and   cardiovascular
disease  (Koller ana Exon, 1986).
    Carter  (1966)  reported a  fatal  case of selenium  Intoxication  1n  a child
who  drank  gun-bluing  compound  containing  1.8% selenlous  add. The  clinical
signs of toxldty observed  Included peripheral vascular  collapse, pulmonary
edema and coma.  Autopsy  revealed pulmonary congestion,  hemorrhage  and edema.
    Oral I.DCQ values   for  selenium  compounds  1n  animals are   summarized  In
Table  6-3.   The   lowest LD5Q  was  1.8  mg  Se/kg for  sodium  selenate  1n
rabbits  (Muehlberger  and Schrenk, 1928),  while the  highest oral  LD5Q,  6700
mg Se/kg 1n rats,  Is for  selenium metal  (Cummins and  Klmura, 1971).
    Interactions  of selenium  have been reviewed  by  Dlplock (1976) and  U.S.
EPA  (1980a).   Selenium  has  been  shown   to  be  antagonistic   to  the  toxic
0146d                               6-11                             04/03/89

-------
                                  TABLE 6-2
              Safe and Adequate Ranges  of  Dally  Selenium  Intake*
Group
Infants
Children
Adults
Age
(years)
0.0-0.5
0.5-1
1-3
4-6
7 +
NA
Dally Selenium Intake
(yg)
10-40
20-60
20-80
30-120
50-200
50-200
*Source:  U.S. EPA, 1980a;  NAS,  1980
NA = Not applicable
0146d                               6-12                             04/03/89

-------
0
-p.
o.


s



o
-p.
0
oo



Species/Strain Sex
Hlce/CD-1 F
Rats/Sprague- M
Dawley
Rats/Sprague- M
Dawley
Rats/Sprague- M
Dawley
Rats/Sprague- M
Dawley
Rabbi ts/NR NR
RabbHs/NR NR
NR = Not reported



TABLE 6-3
Acute Oral Toxlclty of Selenium Compounds
Compound LD5Q 10 ^Q
(mg/kg compound) (mg/kg selenium)
sodium selenlte 8.4 3.8
sodium selenlte 7 3.2
selenourea 50 32.1
selenium sulflde 138 76.2
elemental selenium 6700 6700
sodium selenlte NR 2.85
sodium selenate NR 1.8






Reference
Plasterer
et al., 1985
Cummins and
Klmura. 1971
Cummins and
Klmura. 1971
Cummins and
Klmura, 1971
Cummins and
Kumura, 1971
Muehlberger and
Schrenck, 1928
Muehlberger and
Schrenck. 1928




-------
effects of  silver,  cadmium,  Inorganic  mercury  and thallium.  Rats  provided
with  a selenium  supplemented  diet  were  protected  from paraquat  toxlclty
compared with selenium deficient rats (Omaye et  al., 1978).
    Sodium arsenlte  1n  drinking water  (5  mg/8.)  has been shown  to  alleviate
chronic and  acute toxlclty  1n  rats  fed  selenlferous  grains  (Moxon,  1938).
Arsenic appears   to  alleviate  the   toxlclty  of  selenium by  enhancing  the
biliary excretion of selenium.
    Dietary sulfate partially restores  the growth  of  rats treated with  sele-
nlte or selenate  1n  the diet  (Halverson and  Monty,  1960).   Sodium sulfate In
the diet protected against  liver necrosis  1n  rats  fed  selenate (10  mg Se/kg)
but not selenlte or selenlferous wheat (Halverson et al.,  1962).
6.2.   CARCINOGENICITY
6.2.1.   Inhalation.   Pertinent  data   regarding  the  carclnogenlclty   of
selenium  following   Inhalation  exposure were  not  located  1n the  available
literature dted 1n Appendix A.
6.2.2.   Oral.  Nelson  et  al.  (1943) (see  Section 6.1.2.2.  for  noncarclno-
genlc  effects  In this  study that were reported by  FHzhugh et al.,  1944)
treated  groups  of   18  female   Osborne-Mendel   rats  with  diets  containing
selenium at  5,  7  or 10 ppm  In  the  form of selenlferous corn  or  wheat  for 2
years.  An  additional  group of  rats was  fed  selenium at  10 ppm  In  the diet
1n  the form  of  a  solution of  ammonium  potassium  selenlde and  ammonium
potassium  sulflde.   Eighteen rats  were maintained  as controls.  The  basal
diet  used  1n this study was  suboptlmal 1n the  level  of  protein.   Mortality
was  high  and  tended  to  be proportional  to  selenium Intake;  only  four rats
treated with  selenlferous  corn  and  wheat and six  rats  treated with selenlde
survived  for  2 years.  At  all  doses,  cirrhosis of the liver  was observed.
0146d                               6-14                             06/15/89

-------
Cirrhosis was  found  1n rats that died  or  were  killed as early as  3  months,
and  1n  43/53  rats  surviving 2  years.   Liver tumors  were  observed  only  In
rats with cirrhosis.   Of the 53  rats  surviving  2  years,  11  developed  hepato-
cellular  "adenomas  or  low  grade carcinomas."   The  14 control  rats  that
survived to 18 months  showed no  evidence  of  neoplasla.  No  discussion of the
statistical significance of these findings was presented.
    Schroeder  and Mltchener  (1971a)  reported  a  statistically  significant
Increase  In  the  number  of tumors 1n  rats  treated  with sodium selenate  In
drinking  water  at  2-3 ppm  Se.   IARC  (1975)  noted  that  "an evaluation  of
these  results   was  not  possible  because  not  all  autopsled  animals  were
examined  hlstologlcally  and  because  treated,  animals   lived  longer  than
controls."  Further  details concerning methods and  effects on survival  are
presented 1n Section 6.1.2.2.
    Harr et al. (1967) and  Tlnsley et  al.  (1967)  did not find a  carcinogenic
effect  1n male  and  female Wlstar  rats  (1437   rats  used  In the  studies)
provided  with  sodium  selenlte  or sodium  selenate  at 0.5-16  ppm  Se  In  the
diet  throughout   their   lifetimes.    Liver  effects   (accentuated  lobular
pattern, hyperemla, cellular degeneration,  mildly prol1ferat.1ve  hepatocytes,
blnucleatlon and multiple nucleoll) were  observed  1n  rats  treated  at  4,  6, 8
or 16 vg/g (4,  6,  8 or 16 ppm)  1n the diet.
    Schroeder and MHchener  (1972)  treated groups of  -50 male and 50 female
Swiss mice with selenate or  selenlte  1n  double  delonlzed drinking  water  at a
dose of  0  or  3 ppm.   Treatment  began at  weaning  and  was continued until  all
mice  were  dead.    Under   the  conditions  of  the  study,   selenium  did  not
significantly Increase the Incidence of tumors.
    In  an  NCI/NTP  (1980a)  bloassay,  groups  of  50 male  and  50 female  F344
rats  and  B6C3F1  mice were  treated   by  gavage  7  days/week  with  selenium

0146d                               6-15                             04/03/89

-------
sulflde  1n  CMC.   Rats were  treated  at doses of  3 or 15  mg/kg/day  and mice
were  treated  at  doses of  20 or 100  mg/kg/day.   Similar groups of  rats and
mice  treated  with  the vehicle  or  left untreated  served  as  controls.   Doses
were  not  calculated  1n terms  of  selenium because  the  substance  under study
was not defined clearly; analysis  Indicated  that  the substance was primarily
selenium  monosulflde  with  selenium dlsulflde, elemental  selenium  and  sulfur
also  present.  Survival   of  rats  and mice  was  not  affected  by  selenium
sulflde treatment.   Tumor  Incidence data, presented  In  Table  6-4, Indicated
that  primary  liver tumors  were  Increased  significantly  (p<0.001)  1n  male and
female rats and female mice  (p<0.001)  at  the  high dose, and lung tumors were
Increased significantly (p<0.001)  In  female mice.  NCI/NTP (1980a) concluded
that  under  the conditions  of  the  study,  selenium  sulflde was carcinogenic 1n
male  and  female  F344 rats and  female B6C3F1  mice,  but  was  not  carcinogenic
1n  male   B6C3F1 mice, although  male  mice  may  have been  able to  tolerate
higher doses.
6.2.3.   Other  Relevant  Information.   The  NCI/NTP  has  conducted  dermal
cardnogenlclty studies  of  selenium  sulflde (NCI/NTP,  1980b)  and  Selsun*
(antldandruff  shampoo  containing  2.5% selenium sulflde)  (NCI/NTP,  1980c) In
ICR  Swiss  mice.   In  these  studies,  groups  of   50 mice/sex  were  treated
dermally  3  times/week  for  86 weeks with  selenium sulflde  at 0.5 or  1  mg, or
with  0.05 mi. of a 25 or  50% Selsun*  solution.   Similar groups  of  vehicle
controls  were maintained.   The studies were  terminated at 86 weeks  because
of  poor   survival.   Amyloldosls,  a  known  cause of  death 1n  Swiss  mice,
contributed to the deaths  of most mice after  1 year.  Dermal  application of
selenium  sulflde  or   Selsun* did  not  result  1n  treatment-related Increases
In  tumor  Incidences.   Both   studies  were  limited by  the relatively  short
llfespan of ICR mice.
0146d                               6-16                             09/15/89

-------
                                                    TABLE 6-4


                        Incidence  of  Tumors  In F344 Rats and B6C3F1  Mice Treated by Gavage
                                       with  Selenium Sulflde for 103 Weeks8
Species/Strain Sex Dose
(mg/kg/day)
Rat/F344 H 0 (vehicle)
3
15
F 0 (vehicle)
3
15
Mouse/B6C3Fl N 0 (vehicle)
20
100
F 0 (vehicle)
20
100
M 0 (vehicle)
20
100
F 0 (vehicle)
20
100
Duration
of Study
(weeks)
104
104
104-105
104
104
104-105
104
104-105
104-105
104
104-105
104-105
104
104-105
104-105
104
104-105
104-105
Tumor Type
hepatocellular carcinoma
or neoplastlc nodule
hepatocellular carcinoma
or neoplastlc nodule
lung alveolar/bronchlolar
carcinoma or ademona
lung alevolar/bronchlolar
carcinoma or adenoma
hepatocellular carcinoma
or adenoma
hepatocellular carcinoma
or adenoma
Tumor Incidence
(p value)
1/50 (p<0.001)b
0/50 (NS)
24/49 (p<0.001)c
1/50 (p<0.001)°
0/50 (NS)
37/50 (p<0.001)c
4/40 (p=0.022)b
10/50 (NS)
13/50 (p=0.016)c
0/49 (p<0.001)b
3/50 (NS)
12/49 (p<0.001)c
15/50 (p<0.026)b
14/50 (NS)
23/50 (NS)
0/49 (p<0.001)b
2/50 (NS)
25/49 (p<0.001)c
o
CO
OD
IO

-------
                                                 TABLE 6-4 (cont.)
.*»
en
a.
                                                Quality of Evidence

                        The  compound  was  administered  to adequate  numbers  of  animals  of  both  sexes  by  a
                        relevant  route of exposure at  two  dose levels.

                        The  substance  under  study  was  not  clearly  defined  -  analysis  Indicated  that  the
                        substance  was  primarily  selenium  monosulflde  with  selenium   dlsulflde.  elemental
                        selenium  and  sulfur  also present.   The  amounts of  the Individual  compounds were  not
                        determined.  Hale mice may not  have been  treated at  the maximum  tolerated dose.
Strengths of Study:
Weakness of Study:
   Overall  Adequacy:     Adequate
   aSource:   NCI/NTP.  1980a

?] bCochran-Arm1tage Test
co
   cF1sher  Exact

   NS =  Not  significant
o
CO
oo
lO

-------
    Further evaluation of the data, however, failed  to  Identify  the  specific
selenium  salt  associated  with  neoplastlc  lesions; therefore  the  NCI/NTP
(1980) study 1s considered Inconclusive.
    Selenium, an  essential  element,  has  been extolled  as  an  antlcardnogen.
Several  epidemiology  studies reviewed  by Wlllett and  Stampfer  (1986)  have
reported  lower  blood  selenium   levels   1n  cancer  patients  compared  with
cancer-free  subjects.  Wlllett  and  Stampfer   (1986)   Indicated  that  these
studies  have  limited  significance because  the effect  of cancer  on  blood
selenium level  was  not distinguished from  an  effect of selenium on  cancer.
Several  more  recent  studies  (Wlllett et  al.,  1983; Salonen  et al.,  1984,
1985; Kok  et  al.,  1987)  examined selenium levels of serum samples  collected
before the  diagnosis  of  cancer.    These  studies have also  shown a  relation-
ship  between  low  serum  selenium  levels and  cancer.   Kok  et  al.  (1987)
observed a significant association between cancer and low  selenium  levels  In
men  but  not  women,  and  Indicated  that   similar  results  were  found 1n  the
studies by Wlllett et al.  (1983)  and Salonen et  al.  (1985).
    Numerous  studies  1n  animals  have also  reported  antlcardnogenlc  effects
of  selenium  against carcinogen-Induced  carcinomas.   Animal studies  reviewed
by  Shamberger  (1985)  have  found  that selenium  treatment  reduces  the  Inci-
dence of  skin tumors  Induced by 3-methylcholanthrene and  benzopyrene,  liver
tumors Induced  by 3-methyl-4-d1methyl-am1noazobenzene,  2-acetylam1nofluorene
and  aflatoxln   B,,  tracheal   tumors   Induced   by   1-methyl-l-nltrosourea,
mammary  gland  tumors  Induced by  7,12-dlmethyl-benzanthracene, anthracene and
N-methyl-N-nltrosurea, tumors of the colon and  large and small bowel  Induced
by  1,2-d1methylhydraz1ne  and  azoxymethane, and  lung  and large  and  small
bowel tumors  Induced  by b1s(2-oxopropyl)n1trosam1ne.  Mice, rats or  hamsters
were treated  In  these  studies with  selenium  In  the diet at 0.1-6 ppm,  1n the
0146d                               6-19                             09/15/89

-------
drinking water at  1-4  ppm,  or  dermally with a 0.005X selenium solution.  The
observation  that  selenium  protects  against  fat-sensitive  types of  cancers
may  Indicate  that  selenium  protects  against  cellular  damage  caused  by
peroxldatlon of fat (Shamberger, 1985).
    In  a  study  of the  effect  of  selenium  on  the  Immune  system  In  rats,
Koller  et  al.  (1986)  found that  selenium  In  the  diet at  0.5 and  5.0  ppm
enhanced  NK  activity,  while  treatment  at  10  ppm  resulted  In NK  activity
similar  to  controls   (see  Section  6.1.2.1.).   From  these  results,  the
Investigators postulated  that  enhancement  of NK activity may be a  mechanism
of the  antlcarclnogenlc activity of  selenium,  suggesting that neoplasms that
are NK sensitive may be prevented or are responsive to selenium therapy.
6.3.   MUTAGENICITY
    Hutagenlclty  data  for  selenium  compounds are  summarized 1n Table 6-5.
The results  were  mixed,  with  both positive  and negative results  reported.
The  most  frequently  tested  compounds  were  sodium  selenlte  and  sodium
selenate.    Shamberger   (1985)  reviewed  mutagen1c1ty data  for   selenium  and
concluded  that   1n general,   sodium  selenlte  (+4)  was more  mutagenlc  1n
mutation  systems   than  sodium   selenate  (+6).   The only  study of  selenium
metal   found  that  the  metal  was more active  1n  Inducing sister  chromatld
exchange  1n  human  whole  blood  cultures  than  selenium  dioxide,  sodium
selenlde or sodium selenlte (Ray and Altenburg, 1980).
6.4.   TERATOGENICITY
    Robertson (1970)  reported   that  during a  5-year  period,  all but one  of
four  certain  pregnancies  and  one   probable  pregnancy among  eight  women
exposed  to  selenium  1n  the  preparation  of medium  for Salmonella  culture
ended    In   miscarriage   (clinical   factors   may  have  accounted   for  two
miscarriages).  In  the single  pregnancy that went  to  term,  the  Infant  had

0146d                                6-20                             09/15/89

-------
                                                                                IflDLC  O-D


                                                               Hutagenlclty Testing of Selenium Compounds
^ 	
°- Assay
Reverse
mutation


Reverse
mutation


Rec assay
DNA damage



Rec assay
, ONA damage
rsj
Chromosome
aberrations


Chromosome
aberrations



Chromosome
aberrations
Indicator/
Organism
Salmonella
typhlmurlum
TA98. TA100
and 1A1537
S. typhlmurlum
(specific
strains not
stated)
Bacillus
subtllls
17A (rec».
45T (rec-)

B. subtllls
17A (rec»)
457 (rec-)
rat lymphocytes
In vitro


Chinese hamster
ovary cells



human lympho-
cytes Jin vitro
Compound
Na2Se03.
Na2Se04


selentte.
selenate


sodium
selenlte
(Na2Se03).
selentous acid
(H2Se03) (M)
selenlc acid
(H2Se04) (»6)

Na2Se03



Na2Se03.
Na2Se04



Na2Se03.
H2Se03.
Application
plate
Incorporation


plate
Incorporation


spot test




spot test


added to
cultures


added to
cultures



added to
cultures
Concentration Activating Response
or Dose System
0.12 |iN none «•



NR none - selenlte
«• selenate


0.8-10 mg none «•




1 or 10 mg none


1. 5, 7.5, 1 none »
and 2.5xlO"«
H
-
10~* to » glutathlone »
10~« N thlone »



6.5. 13. 26 none all » except
or 53xlO~» N Na2Se04
Comment
Heakly * only In TA100



Selenate resulted In an
Increase In base-pair
substitutions

None




NaOH was added to raise
the pH

Significant Increase of
abnormal metaphases at
1xlO~» N (p<0.01) and
2.5xlO~» N (p<0.001)
Selenate was more active
In Inducing aberrations
without glutathlone;
glutathlone enhanced the
activity of selenlte
Efficiency of chromosome -
breaking activity was In
Reference
Noda et al..
1979


Lofroth and
Ames. 1978


Nakamuro
et al.. 1976



Nakamuro
et al.. 1976

Newton and
Lilly. 1986


Whiting
et al.. 1980



Nakamuro
et al.. 1976
                                       H2Se04,
                                       sodium selenate
                                       (Na2Se04).
                                       selenium oxide
                                       (Se02)
the following order:
H2Se03 > Na2Se03 > Se02 »
H2Se04 > Na2Se04; activity
of the M compounds was
greater than the +6
compounds
00
IO

-------
                                                                            TABLE 6-5 (cont.)
o<
I
rsj
Assay
Cloning
efficiency
Nick transla-
tion assay -
DNA damage
(strand breaks)

Sister
chromatld
exchange
Indicator/
Organism
human dlplold
flbroblasts
human dlplold
flbroblasts

human whole
blood cultures
Compound Application
Na2Se03 added to
cultures
Na?Se03 added to
cultures

sodium selenlde added to
(Se^-); cultures
selenium
Concentration Activating Response
or Dose System
0.1-1.0 raM «• glutathlone *
1 S-9 «•
50-500 tiN _f glutathlone <•

1.12x10"* to none «•
8.0xlO~> N
Comment Reference
S-9 and reduced glutathlone Snyder, 1987
Induced similar enhancement
of cloning ability
Reduced glutathlone In- Snyder. 1987
creased DNA damage; HO-
scavenger, mannltol and
peroxide scavenger, catalase
had no effect on strand
break formation, suggesting
that the breaks most likely
do not arise via free oxygen
radical formation
SCE Inducing abilities of Ray and
Se In decreasing order of Altenburg.
effectiveness; selenium > 1980
dioxide (Se4*);
selenium (Se°).
sodium selenlte
(Se**)
selenium dioxide > sodium
selenlde > sodium selenlte
Sister
chromatld
exchange
Crossing over





Chromosome
aberrations



human whole
blood cultures

Drosophlla
aelanoqaster




rat lymphocytes
and bone marrow
cells


sodium
selenate

selenocystlne.
seleno-
oethlonlne



Na2Se03




added to
cultures

added to
media




In vivo expo-
sure of rats
(Intravenous)


1.12x10'* to none
8.00xlO's H

1.63. 5.22. NA
16.3. 81.6
iiH seleno-
cystlne; 1.63,
5.22. 16.3 vH
selenomethlonlne
1.5. 2, 2.5, NA
6. 5 mg/kg for
2 treatments
or 6 tag/kg for
1 treatment
None


reduction In None
recombination
combination



- lymphocytes; Doses of 5 or 6 mg/kg were
<• bone marrow near lethal doses
at 5 or 6 mg/kg
for 2 treatments

Ray and
Altenburg,
1986
Ting and
Walker, 1969




Newton and
Lilly. 1986



      NR = Not reported; NA = not applicable
 u>
 CO

-------
bilateral club-foot.  Urinary  selenium  levels  of these women toward  the  end
of the 5-year period were similar to controls living 1n the same  area.
    In a review of  the  reproductive  effects  of  selenium,  Barlow  and  Sullivan
(1982)  Indicated  that hoof abnormalities  have been  noted 1n foals,  calves
and lambs  born  to animals grazing  on  selenlferous pastures.  Rosenfeld  and
Beath  (1947,  1964)  described  other  malformations  In  lambs born  1n  selenlf-
erous areas.  The majority (85%) of  lambs  born  In  the area had abnormal  eyes
and  limbs  and died  at   birth  or shortly  after birth.   In survivors,  limb
deformities  Impaired  locomotion.    Eye  deformities   Included   large  cysts
protruding  through   the  Hds,  sclera  fused  to  bone  preventing  rotation,
colobomas,  and  abnormal  or  absent  lens,  cornea or  Iris.   Some lambs  were
mkrophthalmlc with  multiple  ocular  cysts.   Hlstologlc  examination of  the
eyes  revealed bizarre structural arrangement resulting from  arrested  growth
and poor  differentiation.  Hypoplasla  of reproductive organs was also  noted
1n  lambs  that  survived.  Estimates  of selenium  doses  resulting  1n  these
effects were not provided.
    Numerous  studies  reviewed by  U.S.   EPA  (1980a)  Indicate  that  In  hens,
selenHe In  the diet  at 8-10 mg/kg  can  result  1n  fetal deaths and  extensive
malformations of  the  central  nervous system, extremities  and beak.   Because
these studies are not useful  for risk assessment,  they will not  be  discussed
further.
    Fetotoxlc effects of  selenium 1n rodents appear to be less dramatic  than
those  1n domestic animals  (Barlow and Sullivan,  1982),  predominantly  result-
Ing  In a  decrease  In  fetal  body  weight.   Nobunaga   et  al.  (1979)  treated
groups  of   14  female  IVCS mice  with  selenium as  sodium  selenHe  In  the
drinking water at 0,  3  or 6 ppm Se  for  30 days  before mating with  untreated
males.   Selenium  treatment was  continued until  gestation day  18  when  the
mice  were  sacrificed and  the  fetuses  were  examined.   No  effects  on  the

0146d                               6-23                             09/15/89

-------
number  of  Implants/dam,  the  number  of dead  fetuses,  the number of  resorp-
tlons, or the number of malformed  fetuses  were  noted.   Fetal  body weight was
decreased significantly (p<0.05) at 6 ppm,  but not at 3 ppm.
    Plasterer  et  al.  (1985)  treated  groups  of  45  pregnant  CD-I  mice  by
gavage with  distilled  water  or  with sodium selenlte  1n  distilled water  at a
dose of 7 mg/kg/day on gestation days  7-14.   Shortly after  birth,  Individual
pups  were  examined  for  gross  structural  abnormalities, and  Utters  were
weighed on both  the day of  birth  and  48 hours  later.   Under  the  conditions
of  the study,  sodium  selenlte did  not  result  In  any maternal  or  fetal
effects.
    Holmberg  and   Perm  (1969)  did  not  find  any  teratogenk  effects  In
hamsters given  an  Intravenous  Injection of  sodium selenlte  at  2 mg/kg  on
gestation day  8.    Pretreatment with  sodium selenlte  markedly reduced  the
teratogenlc   effects of  Intravenous Injections of  cadmium sulfate and  sodium
arsenate.
    Lee et al. (1979)  studied the  effects  of  selenium treatment and  simulta-
neous  selenium and  methylmercury  treatment  on the  Induction  of  cleft  palate
In mice.  Pregnant  ICR mice  were treated with single subcutaneous  Injections
of  selenium  or  selenium and  methylmercury  on gestation day  10,  or  multiple
Injections on  gestation days  9-12 or  7-12.   Treatment  with selenium  at  2
mg/kg/day had  no  effect  on mean Utter  size,  mean  fetal  weight  or  the
percent of  mice with  cleft  palate.   At  3.5 mg  Se/kg/day,  the only  effect
noted  was a  significant decrease  In mean  fetal body weight.  Treatment  of
mice with methylmercury  and  selenium  simultaneously appeared  to  potentiate
the  ability  of methylmercury  to  Induce cleft  palates  when treatment  was
administered on gestation days 9-12.
OH6d                               6-24                             09/15/89

-------
6.5.   OTHER REPRODUCTIVE EFFECTS
    Schroeder  and  MHchener  (1971b)  completed a  multi-generation study  1n
which  five  pairs of  CD-I  mice were  fed  a purified diet  containing  minimal
(0.056  ppm)  selenium,  and  were  provided  with   drinking water  containing
selenate at  a  concentration  of 3 ppm Se  [0.57 mg/kg/day,  assuming a  0.03 kg
mouse  body  weight  and  a  0.0057   l/day  drinking  water   Intake   (U.S.  EPA,
1986a)].   Treatment  was  Initiated  at weaning.   Control  mice were  provided
with the same  diet.   Pairs of mice were  allowed  to breed up to 6 months  of
age.   The   second  generation  was   bred   from  Fl   ,  F,.   and  F,    offspring,
                                                 a    ID       1C
and  the third  generation  was obtained  from F.   and F.  offspring.   The
parameters   of  toxlclty  evaluated  Included Intervals between Utters,  age  at
which  first  litter  was  produced,  number  of  runts,   number  of  stillborn,
failure to  breed successfully, congenital abnormalities  and maternal  death.
By  these  criteria,  selenium  at 3  ppm 1n  the drinking  water  was toxic  to
mice.   The  strain began  to  die  out  at  the  third  generation;  seven  pairs
failed  to  breed,  Including 16 runts among 23  mice born to three  pairs.   For
all  generations,  selenium treatment  resulted  1n   93  runts of 389 live-born
offspring,  1 stillborn Utter, the  death  of  23 offspring  shortly  after birth
and 1 maternal death.  The  Increase In the number  of runts was  statistically
significant  for  all  generations, while  deaths of  offspring  were only  sig-
nificant In the F, generation.  Congenital malformations were not  reported.
    In  a study concerning the effects of selenium on  the  testes,  Nebbla  et
al.  (1987)  treated  groups of  7-12  male  Wlstar rats with  sodium  selenlte  In
the drinking water at a concentration of  4,  8  or  16 ppm (1.8,  3.7  or  7.3 ppm
Se)  for 240  days.   Assuming  that  a  0.35  kg rat  drinks 0.049 l/day  (U.S.
EPA, 1986a),  the  1.8,  3.7 and 7.3  ppm Se concentrations  correspond  to doses
of  0.25, 0.52 and 1.0 mg/kg/day,  respectively.   Control   rats were  provided
with tap water.   Body weights of rats treated at  7.3  ppm were  significantly
0146d                               6-25                              09/15/89

-------
lower than  controls,  and  testes  weight and testes-to-body weight  ratio  were
Increased significantly compared  with controls.  Microscopic  examination  of
the  testes  revealed  damaged tubules  1n  rats  treated at  7.3  ppm,  exhibiting
ol1gosperm1a  and  vacuollzatlon of  spermatlds  without multlnucleated  cells.
Moderate  Intertubular edema  was  also observed.   Similar but  less  severe
effects were  noted  at 3.7  ppm;  no  effects  were noted at  1.8  ppm.   Measure-
ment  of  testlcular  enzyme  activities compared  with controls  Indicated  a
significant  decrease  1n  SDH  activity at  3.7 and  7.3  ppm,  a  significant
Increase  1n  GGT  activity  at 7.3  ppm, significant decreases  1n  B-G activity
at  1.8  and  3.7 ppm  and a significant Increase  1n  B-G activity at  7.3  ppm.
Testlcular LDH activity was  found to be Increased significantly at  all  dose
levels.   The  significance of the changes  1n  enzyme activity  are  not  clear.
The reproductive capacity  of these rats was  not determined.
6.6.   SUMMARY
    Selenium  Is an  essential element  that  Is  a part of  glutathlone peroxl-
dase.   The  dally  safe and  adequate  level  of  selenium  Intake for  adults  1s
considered  to be  50-200  v9  (0.05-0.2 mg)  (NAS,  1980).   Selenium 1s  also
quite toxic,  with  selenosls (brlttleness of nails,  loss of nails  and  hair,
dermatitis,   nervous  symptoms)  reported  In  humans  following chronic  dietary
Intake of 3.2-6.69 mg/day  (Yang et al., 1983).
    Adverse  effects  reported  In  animals treated orally with selenium  com-
pounds  Include effects on growth,  reduced survival,  hlstopathologlc  changes
In  the  liver and  testes  and  Immune  system  effects.  Subchronlc  NOELs  that
have been  Identified In   laboratory animals are 0.16 mg  Se/kg/day  of  sodium
selenlte  or  selenlferous  wheat  In rats (Halverson et al., 1966) and  0.63  mg
Se/kg/day  of  sodium  selenlte  1n  hamsters   (Beems  and  van  Beek,  1985).
Limited chronic studies of selenium compounds  do not  Identify  NOELs.
0146d                               6-26                             09/15/89

-------
    Selenium toxldty  following  Inhalation  exposure  has  not been studied In
animals.   Symptoms  reported  1n  humans  following occupational  exposure to
selenium Include a  strong  garlic  odor  1n the breath, sweat and urine, acute
sore  throats  and cold-like  symptoms,  gastrointestinal  Irritation,  lacrlma-
tlon,  and  a metallic  taste  In the  mouth  (Hamilton and  Hardy,  1974).  The
development of  garlic breath  and  cold-Uke  symptoms  1s  thought  to   result
from  the  hepatic production  of  dlmethylselenlde, which  Is exhaled through
the lungs (D1sk1n et al.,  1979).
    Carclnogenlclty   studies   of  selenium  are   not  conclusive.   Selenium
sulflde has tested  positive  for  cardnogenldty  1n  rats  and  female mice 1n
an NCI/NTP  (1980a)  bloassay.   Increased  tumor Incidences were also reported
by Nelson  et  al. (1943)  In  rats treated orally  with  selenlferous  corn and
wheat  (diets were  suboptlmal  1n  protein),  and In  rats  treated  with   sodium
selenate (Schroeder  and  MHchener,   1971a).   Negative  results were reported
1n rats treated with sodium selenate or  selenlte  (Harr et  al., 1967; Tlnsley
et  al.,  1967)  and  In   mice  treated  with   sodium selenate  or  selenlte
(Schroeder  and  MHchener,  1972).
    Epidemiology studies  (WHlett and  Stampfer,  1986;  Salonen et al., 1984,
1985; Kok et al., 1987) and studies  using animals  (Shamberger, 1985) suggest
that selenium has antlcarclnogenlc activity.   This  effect may be due  to the
ability of  selenium to protect  against  cellular damage  by peroxldatlon of
fat  (Shamberger, 1985), or 1t may be  a  result  of the  effect  of selenium on
the Immune  system (Koller  et  al.,  1986).
    The  available   evidence   suggests  that   selenium   Is  fetotoxlc  and
teratogenlc.  A  study by  Roberts  (1970) weakly  associates miscarriage and
birth  of  Infants  with  bilateral  club  foot  to  occupational   exposure to
selenium.
0146d                               6-27                             09/15/89

-------
    Selenium  1s  teratogenlc  to  livestock (Barlow  and  Sullivan, 1982),  but
developmental  effects   reported   1n  laboratory   rodents  are   limited   to
decreased  fetal  body  weight  (Nobunaga   et  al.,  1979).   A  limited  multi-
generation  study  (Schroeder and MHchener,  1971b) found  that  by the  third
generation,  CD-I mice  treated with  selenate In  the  drinking water at  0.57
mg/kg/day  failed  to  breed,  or   produced  a  large   proportion  of  runts.
Testlcular  degeneration  and ol1gosperm1a  were reported 1n rats  treated  with
selenium at 1 mg/kg/day 1n drinking water  for 240 days (Nebbla et al.,  1987).
0146d                               6-28                             09/15/89

-------
                     7.   EXISTING  GUIDELINES AND  STANDARDS
7.1.   HUMAN
    The WHO  (1984)  maximum guideline for selenium 1n  drinking  water  1s  0.01
mg/l.
    The AWQC  for selenium 1s  0.01  mg/st (U.S.  EPA,  1980a,  1982), which  Is
also the current  drinking  water  standard (see below).  The  value  was  chosen
with the consideration  that  only  5-10% of dally  exposure  to selenium  should
come  from  water-related   sources,  for   example  drinking  water  and aquatic
organisms.
    U.S. EPA (19855) recommended  10-day  health advisories  of 0.041  and 0.144
mg/8. for a  10 kg child  and  a 70  kg adult,  respectively.   These  values  are
based  on  the  subchronlc  rat  study  by  Halverson  et  al.  (1966).   The  AADI
(U.S. EPA, 1985b) for  a 70 kg adult  1s  0.106  mg/i based on  the human study
by Yang et al. (1983).
    U.S. EPA  (1985a)  has  proposed  an  RMCL  of  0.045 mg/i.   The  RMCL  1s  a
nonenforceable  health  goal  that   would  result  1n no  known or  anticipated
health  effects.   The current  enforceable MCL for  selenium,  based  on  the WHO
(1984) guidelines, 1s 0.01  mg/il (U.S. EPA, 1985c).
    The verified  oral  RfDs for selenlous  acid and selenourea  are  0.003 and
0.005 mg/kg/day  (U.S.  EPA, 1985c,d),  respectively.   Both  values  are based on
the  human   ep1dem1olog1cal study  by  Yang  et  al.   (1983),  which  reported
selenosls 1n persons exposed to selenium at 0.046 mg/kg/day.
    Final  RQs  for   selenium   compounds  are   1000 for selenourea,  100  for
selenium  [(a  release   of  selenium  does  not  need  to  be  reported   If  the
diameter of  the metal  Is  >100 ym  (0.004  Inches)] and sodium  selenlte,  and
10 for  selenlous  add  and  selenium oxide (U.S.  EPA,  1986b).   The proposed RQ
for selenium dlsulflde Is  10 pounds (U.S. EPA,  1987b).

0147d                               7-1                              06/15/89

-------
    The OSHA  PEL  for selenium  compounds  (as  Se) 1s 0.2 mg/m3  (OSHA,  1985).
ACGIH  (1987)  has also  recommended a  TLV-TWA  of 0.2  mg Se/m3  for  selenium
compounds to protect against systemic toxlclty and Irritation (ACGIH, 1986).
    U.S. FDA  (1987)  has  ruled  that  feed  for  major food producing animals may
be  supplemented  with sodium  selenlte  or selenate  at  a level  not  to  exceed
0.3 ppm  Se.   Feed supplements  for  limited  feeding may  also be supplemented
with selenium at a rate  not  to exceed  an Intake of 0.7 mg/head/day for sheep
and 3 mg/head/day for beef cattle.
7.2.   AQUATIC
    Adams and Johnson (1981) proposed  a  criterion for  selenium 1n freshwater
systems  of   52   yg/a   based  on  data  demonstrating   that  the  amphlpod,
Hyallela azteca.  was the most  sensitive  species tested.  The  criterion  was
calculated  from  the product  of  the  14-day   LC5Q  of  0.07  mg/l  for  this
species and an  average  application factor of 0.74  derived  from the ratio of
time-Independent  LCrQ  values  to  the  geometric  means  of  MATCs  for  rainbow
trout and Daphnla magna.
    Lemly  (1985b)  recommended  a maximum  permissible  level  of  selenium In
warm water  lakes  and reservoirs  of 5  yg Se/l  as  a  long-term average  (>30
days).    The  author  suggested   that  10  yg  Se/8.  was  a  reasonable  criterion
for  selenium  In   rivers  and   streams  containing  centrarchlds  and  50  yg
Se/a.  1n rivers  containing  only  salmonlds.   A  criterion   for  selenium  1n
cold water  lakes  and  reservoirs  of  50  yg Se/i  was  recommended 1f  only
salmonlds were  present,  but  a  criterion  of 5  yg Se/8,  was  recommended If
centrarchlds were present.   Lemly (1985b) apparently  based  these criteria on
data collected  from  field studies rather than  from  calculations or extrapo-
lation from laboratory-derived toxlclty data.
0147d                               7-2                              04/03/89

-------
    U.S.  EPA   (1987a)   reported   that   "The   procedures   described   1n  the
'Guidelines for  Deriving Numerical National Water Quality  Criteria1  for  the
Protection  of  Aquatic  Organisms  and  Their  Uses   Indicate  that,  except
possibly  where  a  locally  Important  species  1s  very  sensitive,  freshwater
aquatic organisms and  their uses  should not  be affected  unacceptably 1f  the
4-day  average  concentration  of  selenium  does  not  exceed  5.0  yg/t  more
than once  every 3  years on  the  average and  If  the  1-hour  average concen-
tration  does  not   exceed   20  yg/l  more  than once  every  3  years  on  the
average."
    U.S.  EPA  (1987a)  also  reported  that  "The procedures  described 1n  the
'Guidelines for  Deriving Numerical  National  Water  Quality Criteria  for  the
Protection  of  Aquatic  Organisms  and   Their  Uses'   Indicate  that,  except
possibly  where  a   locally  Important species  1s  very sensitive,  saltwater
aquatic organisms and  their uses  should not  be affected  unacceptably 1f  the
4-day  average  concentration   of   selenium does  not   exceed  71  ^g/i  more
than once  every 3  years on  the  average and  1f  the  1-hour  average concen-
tration  does  not   exceed  300  vq/i more  than once  every  3  years on  the
average."   Despite   these  criteria for  marine  systems,  U.S.  EPA  (1987a)
recommended that the status of  marine  fish communities be monitored  whenever
the concentration of selenium 1n saltwater was >5.0
0147d                               7-3                              04/03/89

-------
                              8.   RISK  ASSESSMENT

    Statements  concerning available  literature  1n  this  document  refer  to
published, quotable  sources  and  are In no way  meant  to  Imply  that  confiden-
tial business  Information  (CBI),  which this  document could not  address,  are
not  In  existence.   From examination of  the  bibliographies of the  CBI  data,
however,  1t  was determined  that  CBI  data that would alter the approach  to
risk assessment or the risk assessment  values presented herein  do not  exist.
8.1.   CARCINOGENICITY
8.1.1.   Inhalation.   Pertinent   data   regarding  the  carclnogenldty   of
selenium  compounds   following  Inhalation  exposure  were  not  located  1n  the
available literature cited In Appendix  A.
8.1.2.   Oral.   Oral  carclnogenldty  studies  of  selenium compounds   have
been Inconclusive.   Nelson et al.  (1943)  found  Increased  tumor  Incidences  1n
rats fed  selenlferous  corn or wheat In a basal diet that  was suboptlmal  1n
protein.  Schroeder  and MHchener  (1971a) reported a  significant  Increase  1n
tumor Incidences  In  rats  treated with  selenate 1n the drinking  water at 2-3
ppm  Se.   An  evaluation of this  study  1s  not possible, however, because not
all  autopsled  rats  were  examined h1stolog1cally,  and  treated  rats  lived
longer  than  controls (IARC,  1975).  Positive  results were also reported  1n
male and female rats and  female  mice treated by gavage with selenium  sulflde
(NCI/NTP,  1980a).    However,  the  Identity  of  the  compound associated  with
neoplastlc   evidence  Is  unclear;   therefore   the   study  Is   considered
Inconclusive.
    In contrast, Harr et al.  (1967) and "Mnsley et  al.  (1967)  did  not find a
carcinogenic  effect  In  male and  female Wlstar  rats provided  with sodium
selenlte or  sodium  selenate  1n  the diet at  0.5-16  yg Se/g diet  throughout

0148d                               8-1                             09/15/89

-------
their lifetimes.  Negative  results  were  also reported  In  mice  treated  with
selenate  or  selenlte  In   the   drinking  water  at  3  ppm  (Schroeder   and
MHchener, 1972).
    Numerous studies,  reviewed  by Shamberger (1985),  1n  which  animals  were
treated with  a known  carcinogen and  selenium  In  the  diet  (0.1-6  ppm)  or
drinking  water  (1-4  ppm),   Indicate  that   selenium   has  antlcardnogenlc
activity.   In addition, epidemiology studies reviewed by Wlllet and Stampfer
(1986) have reported  an association  between  lower  selenium blood  levels and
cancer.    It  has  been  suggested  that selenium may protect  against  cellular
damage caused by peroxldatlon of  fat (Shamberger,  1985).   From results  of a
study concerning the effect of selenium on the Immune system In rats, Koller
et  al.  (1986)  suggested  that   selenium  may act  as  an   antlcardnogen  by
enhancing  NK activity.
8.1.3.   Other   Routes.    Dermal   carclnogenldty    bloassays   of   selenium
sulflde (NCI/NTP, 1980b)  and Selsun®  (antldandruff  shampoo  containing  2.5X
selenium  sulflde) (NCI/NTP,  1980c)   1n mice  have not  resulted  1n  Increased
tumor Incidences.
8.1.4.  Weight  of   Evidence.    After  reviewing  available  carclnogenldty
data, IARC  (1975) concluded  that animal  data were  Insufficient  to  evaluate
the  carclnogenldty  of  selenium compounds,  and  that  human  data   did  not
suggest that selenium was carcinogenic.
    U.S.  EPA   (1980a)   Indicated   that   the  carclnogenldty   of   selenium
compounds   Is   a  complex   Issue  for  the  following   reasons:   1) there  1s
evidence  that  selenates,   selenltes  and  selenldes  at  nontoxlc  levels  have
antlcardnogenlc  activity  1n   animals;    2) even   at   moderate   dietary
concentrations  (5-10  ppm)  of selenates and  selenltes,  the chronic  toxldty
1s  high,  and  this  toxldty  Interferes  with tumor  development because  of
0148d  .                             8-2                              09/15/89

-------
early  deaths;   3)  the   solubility  and,  therefore,  the  availability   of
different  selenium  compounds  Is  highly variable;  4)  low concentrations  of
selenium are essential; and  5) the chronic toxldty  studies are  difficult  to
compare because of  the large number of  different  selenium compounds  studied,
the dependence of tumor  Induction  on  changes 1n  protein  and  selenium  levels
1n  the  diet,  and Incomplete hlstopathologkal  examinations  In  a number  of
studies.
    There are  no  data regarding the  carclnogenlclty of  selenium to humans.
Because  the  animal  evidence 1s  conflicting  and  difficult  to Interpret,  1t
can be considered  Inadequate evidence of a  carcinogenic  effect.  Therefore,
based on no  human carclnogenlclty  data  and  Inadequate animal data,  selenium
can be considered,  according to U.S. EPA  classification scheme, as Group  D
carcinogen - not classifiable as to human  carclnogenlclty  (U.S.  EPA, 1986b).
However,  because  of positive  evidence  of  carclnogenlclty  1n  both  rats and
female mice  (NCI/NTP,  1980a),  selenium sulflde,  based  on  U.S.   EPA (1986b)
classification scheme, could be  considered  as  Group B2  - probably  carcino-
genic to humans.
8.1.5.   Quantitative Risk Estimates.
    8.1.5.1.   INHALATION ~ The   lack  of  data  concerning  the   carclnogen-
lclty of  selenium  following Inhalation  exposure  precludes the derivation  of
an Inhalation potency factor.
    8.1.5.2.   ORAL  — According  to  U.S.  EPA  (1986b)  guidelines,  quantita-
tive  risk  assessments  are not appropriate for  agents  that are Judged  to  be
1n Group D.   Therefore,  a potency factor  for oral  exposure to selenium will
not  be   derived.   Since  H Is  unclear regarding  the  chemical  species  of
selenium associated with carclnogenlclty In  the NCI/NTP  (1980a)  study,  H  Is
Inappropriate  to  estimate  a potency  factor  for  oral exposure   to  selenium
sulflde.

0148d                               8-3                               09/21/89

-------
8.2.   SYSTEMIC TOXICITY
8.2.1.   Inhalation Exposure.
    8.2.1.1.   LESS  THAN  LIFETIME  EXPOSURE  (SUBCHRONIC) -- Pertinent  data
regarding the  toxlclty  of selenium following subchronlc  Inhalation  exposure
were not located 1n  the  available  literature, precluding  the derivation  of a
subchronlc Inhalation RfD.
    8.2.1.2.   CHRONIC  EXPOSURE -- Data  regarding the  toxlclty of  selenium
compounds following  chronic  Inhalation  exposure are  limited  to reports  of
occupational exposure  that are not sufficient for RfD  derivation.   Symptoms
reported  following  occupational exposure  to  selenium  compounds  (concentra-
tions  not  specified)  Include  garlic  odor  In the breath,  sweat and  urine,
acute  sore   throats   and  cold-Uke  symptoms,  gastrointestinal  Irritation,
lacrlmatlon  and  a  metallic  taste  1n  the  mouth  (Hamilton and  Hardy,  1974).
Dlskln et al.  (1979)  described a  case of  a selenium worker  who had  peMvas-
cular  noncaseatlng granulomas  1n  the lungs at  autopsy that were  suggestive
of  a  blood-borne  toxin.   Dlskln  et  al.  (1979)  suggested that  the  lung
effects observed 1n  this case and other effects  of  occupational  exposure  to
selenium may  result  from  the  hepatic  production of dlmethylselenlde, which
1s exhaled through  the lungs.
    Pharmacoklnetlc  studies  using  dogs  (Welssman  et  al.,  1983)  and  rats
(Medlnsky et   al.,   1981)  Indicate  that  selenium  and  selenlous  acid  are
absorbed  readily following  Inhalation   exposure.   Based  on these  studies,
Medlnsky  et  al.  (1985)  developed  a  model  to  project steady-state  tissue
selenium  concentrations   In   humans  exposed  to selenium  by  Inhalation.
Results from the model  suggest that exposure to  selenium  1n  urban air (1  ng
Se/m3)  for  a  lifetime  would  contribute  little  to  the  selenium burden  1n
human  tissues compared with oral exposure, while  exposure  at the TLV (0.2  mg
0148d                               8-4                              09/21/89

-------
Se/m3),  8  hours/day,  5  days/week   for  a  lifetime  may  result  1n  liver
burdens  2-17  and  lung  burdens  73-220  times  greater  than  selenium  levels
normally  found  1n  these  tissues.    Although  lung  and  liver  burdens  of
selenium  associated  with  adverse effects  are not  known,  Medlnsky  et  al.
(1985)  concluded  that chronic  exposure to  selenium  near the  TLV  may  be  a
hazard to some Individuals.
8.2.2.   Oral  Exposure.
    8.2.2.1.   LESS  THAN LIFETIME  EXPOSURE  (SUBCHRONIC) — Selenium  1s an
essential element that 1s also  very  toxic.   The  safe  and adequate Intake for
adults  1s  estimated  at  50-200  yg/day  (0.71-2.9  yg/kg/day)  (NAS,   1980).
Selenosls  has  been  reported  1n  humans  chronically  exposed  to  dietary
selenium at a  dose as low as 3.2 mg/day (46 yg/kg/day) (Yang  et al., 1983).
    The  available  subchronlc  animal  studies  of  selenium  are difficult  to
evaluate because of  the various  selenium  compounds  studied. The results  of
the subchronlc studies  Indicate that death occurs In rats at  a dose  of -0.5
mg/kg/day  (MeAdam  and Levander,  1987;  Halverson  et al.,  1966).   Additional
effects  reported  1n  rats  and  hamsters treated at  >0.24 mg/kg/day  Included
hlstopathologlc changes  In  the  liver and changes In  body  weight  (MeAdam and
Levander,  1987;  Halverson  et   al.,  1966; Beems  and  van  Beek,   1985).   The
highest NOEL  below the LOAEL 1n a subchronlc animal  study  1s  0.16  mg/kg/day
1n rats reported by Halverson et al.  (1966).
    Because of the narrow range between a dose considered safe and adequate
for humans and a dose  resulting 1n effects  In  humans, greater  confidence can
be placed  1n  an  RfO  based on human  rather than animal  data.   Therefore, the
chronic  oral  RfD  of  4.6  yg/kg/day or  3.2 mg/day  for a  70 kg  human  Is
recommended as the subchronlc RfD for  selenium.   The  derivation of  this RfO,
based on  the  study  by Yang et  al. (1983), 1s  described In  Section  8.2.2.2.


0148d                               8-5                             09/21/89

-------
Confidence In this RfD  Is high  based  on  low confidence  In  the  study  and  high
confidence In the data base.
    8.2.2.2.    CHRONIC   EXPOSURE — Chronic   oral   studies    In    animals
(Fltzhugh  et  al.,  1944; Schroeder  and MHchener,  1971a,b)  have used doses
that resulted 1n Increased mortality, and did  not  Identify  LOAELs, NOAELs  or
NOELs potentially useful for risk assessment.
    The  toxldty  of  selenium  depends  on   numerous  factors  Including  the
selenium  and  nutritional   status  of  the   Individual,  b1oava1labH1ty  of
different  forms  of selenium  and Interactions with heavy metals.  Because  of
these factors, the limited  number of  human   studies reporting  selenium doses
resulting  1n  effects  are  difficult  to  compare.   An epidemiology  study  In
selenlferous  areas In South  Dakota  (Smith et al.,  1936)  reported  symptoms  of
selenosls  In  persons  with  a dietary  selenium Intake  of  0.1-0.2 mg/kg/day
(7-14 mg/day  for a 70  kg  human).  Beath (1962)  found symptoms of selenosls
1n  persons  drinking  well   water  containing  9  mg  selenium/a.  (18 mg/day
assuming a drinking water  Intake of  2  a/day).  In a single case, selenosls
was  observed  In a man  who  consumed  one sodium  selenHe  tablet dally  (0.9
mg/day) for 2 years (Yang et al., 1983).  The  verified  chronic oral  RfDs for
selenlous  add  (U.S.  EPA,  1985c) and  selenourea  (U.S.  EPA, 1985d),   and the
AIC  for  selenium (U.S.   EPA,  1984a) are  based on the  study by Yang et al.
(1983), which reported outbreaks  of selenosls  1n  China  among persons  consum-
ing  selenium In  the  diet  at   3.2-6.69  mg/day (based  on  the  diets  of  six
persons).  Selenosls  was not observed 1n  areas with selenium Intakes  averag-
ing 0.75 mg/day  (based on the diets of  three persons).   The RfDs  and  AIC are
based on the 3.2 mg/kg  LOAEL.   The  values were derived  by  dividing the LOAEL
by an uncertainty  factor of  10  to  extrapolate from a LOAEL, and  a modifying
factor of  1.5 because  of  Information suggesting  that  selenium  1n  drinking

0148d                               8-6                              09/21/89

-------
water  Is  absorbed more  efficiently  than selenium  In  food.  An  uncertainty
factor of  10  to account  for  human  variability was not  considered  necessary
because the LOAEL  was  from a large population of  humans  (248  Inhabitants  of
the  5  most heavily affected  villages,  although  estimates of dietary  Intake
were based on  the  diets  of 6 persons, and the LOAEL, 3.2 mg/day,  1s  a  value
for  1  person).   The RfDs, 0.21  mg/day  for  selenlous  acid  (U.S.  EPA,  1985d)
and  0.35   for  selenourea  (U.S.  EPA,  1985c),  and  the  RfD  and  AIC of  0.21
mg/day for selenium are  at the  upper  end of  the  range  of a  safe and adequate
dally Intake of 50-200 vq Se/day (0.05-0.2 mg/day) (NAS,  1980).
    Confidence  In  the  Yang et  al.  (1983)  study  1s low because  the  range  of
doses  resulting  1n effects was  based  on the analysis of  the diets of  only
six  people.    Confidence  In  the  data  base  Is   considered  high  based  on
additional supporting  epidemiology studies  and  reviews  of  the  toxldty  of
selenium resulting  1n  long-term selenium Intake  levels similar  to  or  higher
than the RfO.  Reflecting  confidence  1n  the  data  base, confidence 1n the RfD
of 0.32 mg/day (0.05 mg/kg/day)  1s high.
0148d                               8-7                              09/21/89

-------
                          9.  REPORTABLE QUANTITIES
9.1.   BASED ON SYSTEMIC TOXICITY
    The  toxlclty  of  selenium was  discussed 1n  Chapter  6.   Data  considered
for RQ derivation are presented  In Table 9-1.   Data  from the  NCI/NTP  (1980a)
gavage study  using  rats and  mice  are not  Included  because the Identity  of
the  test material  was  not   reported  sufficiently  to   allow  estimation  of
dosages  of  selenium.   In  the  study by Beems and  van  Beek  (1985),  hlstologk
changes  1n  the  liver  were  observed 1n hamsters  treated  with  sodium selenlte
1n  the diet at a dose  of  1.21  mg Se/kg/day  for 42 days.  Halverson et al.
(1966) reported  deaths  of  rats  fed  selenlferous  wheat  at 0.56 mg Se/kg/day
for 6 weeks, while liver effects were  noted 1n  rats  at  0.32 mg  Se/kg/day;  1t
1s  possible that  deaths  from  exposure to selenium 1n this  study occurred  at
lower dosages  as  well.   Death occurred 1n  rats  treated  with  sodium selenlte
1n  the diet at a dose of 0.5 mg Se/kg/day  for  29  days  (MeAdam  and Levander,
1987), and  In  rats  treated with  sodium selenlte  In  the diet 0.28 mg  Se/kg
day  for  2 months (Schroeder  and MHchener, 1971a).  In a  chronic study  In
rats  (FHzhugh et al., 1944), Increased mortality  anc! cirrhosis of the  liver
were  observed  1n rats treated with  ammonium potassium   selenlde 1n the  diet
at  a dose  of  0.5  mg  Se/kg/day.   Reduced food  Intake and  growth  and the
presence  of liver lesions were  observed at  0.25 mg/kg/day.   Schroeder and
MHchener (1971b) reported  Increased mortality of offspring and an Increase
1n  the  number  of  runts  from  mice  treated  with  sodium  selenate  In the
drinking  water at a  dose of  0.57  mg Se/kg/day  for 3 generations.   In  an
epidemiology study  (Yang et  al.,  1983),  selenosls  was observed   In  persons
consuming diets  containing  selenium  that  provided doses of  3.2-6.69 mg/day
(based on  analysis  of  the  diets  of  six  persons).  The  symptoms reported
Included  brlttleness   of  nails,  loss of  nails  and  hair,  prurltls  of the
scalp,                dermatitis                characterized                by
0149d                               9-1                             06/15/89

-------
                  TABLE 9-1



Toxlctty Summary for Oral  Exposure to Selenium
VO
o.
Species/Strain Sex
Hamsters/ N
Syrian
Rats/Sprague- N
Dawley
Rats/Sprague- N
Dawley
Rats/Sprague- N
Dawley
i
*«» Rats /Os borne- F
Nendel
Rats/Osborne- F
Nendel
Rats/Long- N.F
Evans RLV:LE

Nice/CD-I N.F

o
o
00
Ifi
Average
No. at Body
Start Height
(kg)
8 0.14b
8-10 0.35b
8-10 0.35b
8 0.35b
18 0.35b
18 0.35b
50/sex 0.35b

5/sex 0.03b




Vehicle Compound
diet sodium selentte
diet selenlferous
wheat
diet selenlferous
wheat or sodium
selenlte
diet D-SeNet. L-Se
Net. selenlte
or selenate
diet ammonium potas-
sium selenlde
diet selenlferous
grain
drinking sodium selenlte
water

drinking sodium selenate
water




Exposure
20 ppm sodium
selenlte In
the diet for
42 days
11.2 ppro Se
for 6 weeks
6.4 ppm Se
for 6 weeks
10 v9/g Se In
the diet for
29 days
10 ppm Se for
up to 24 months
5 ppm selenium
In the diet
for 24 months
2 mg Se/l for
2 months

3 ppm for 3
generations,
parental gen-
eration treated
for 6 months



Transformed Equivalent
Animal Dose Human Dosea
(mg/kg/day) (ng/kg/day)
1.21C 0.15
0.56d 0.10
0.32d 0.055
0.5d 0.09
0.5d 0.09
0.25d 0.043
0.28* 0.048

0.57f 0.043




Response
Hlstopathologtc changes In
the liver (oval cell pro-
liferation, necrosis of
cells and foci of cells)
Death after 4 weeks of
treatment
Reduced liver weights.
mottling, roughness and
discoloration of the liver
Death
Increased mortality.
cirrhosis of the liver
Reduced feed Intake and
growth; liver lesions
Increased mortality of
offspring before weaning.
an Increase In the number
of runts
Increased mortality at
2 months




Reference
Beems and van
Beek. 1985
Halverson
et al.. 1966
Halverson
et al.. 1966
NcAdam and
Levander.
1987
Fltzhugh
et al.. 1944
Fltzhugh
et al.. 1944
Schroeder and
Nltchener,
1971a

Schroeder and
Nltchener,
1971a





-------
                                                                        TABLE 9-1 (cent.)

No. at
Species/Strain Sex Start

Humans N.F NR



Average
Body
Height Vehicle
(kg)
70b diet





Compound

dietary
selenium



Transformed
Exposure Animal Dose
(mg/kg/day)
3.21 mg/day NA
average sele-
nium Intake


Equivalent
Human Dose8
(mg/kg/day)
0.046





Response Reference

Selenosls: brlttleness and Vang et al..
loss of nails, loss of 1983
hair, dermatitis, nervous
symptoms
   Calculated by multiplying the animal transformed dose by the cube  root  of  the ratio of the animal body weight to the human body weight
   bReference body weights (0.14 kg hamster; 0.35 kg rat; 0.03 kg mouse;  70 kg human)  (U.S. EPA. 1986c)
   C0ose  provided by Investigators
   Calculated assuming a rat eats food equivalent to 5X of Its body weight/day (U.S.  EPA. 1986c)
   Calculated assuming a 0.35 kg rat drinks 0.049 ft of water/day (U.S. EPA. 1986c)
vo 'Calculated assuming a 0.03 kg mouse drinks 0.0057 ft of water/day (U.S.  EPA.  1986c)
i
" NR = Not  reported; NA = not applicable
o
CO

-------
hyperemla, edema and eruptive blistering, and nervous system effects Includ-
ing peripheral anesthesia, acroparesthesla,  pain In the limbs, and numbness,
convulsions and  motor  dysfunction  progressing  to paralysis  1n  more severe
cases.
    The derivations  of candidate  CS  and RQ  values  are  presented  1n  Table
9-2.  Candidate CSs were not computed for data from the studies by Beems and
van Beek  (1985), Halverson  et  al.  (1966),  McAdam and Levander  (1987)  and
Schroeder   and  MHchener   (1971a)  because  adequate chronic  data  were avail-
able.    In  the  epidemiology  study (Yang  et al.,  1983),  the  syndrome  of
selenosls  was observed  In persons consuming selenium 1n  the  diet  at a dose
as  low  as  3.2 mg/day.   This dose  corresponds  to an RVd  of 4.7.   It 1s not
clear  what  specific  effects were  observed  at the reported dose.   The con-
servative  approach 1s to assume that the most severe effects,  nervous system
effects Including convulsions and paralysis, occurred at  the  lowest  reported
dose.    The  effects  on   the  nervous  system correspond  to  an  RV   of  9.
Multiplying the RVg by  the RVrf  yields a  CS of  42.3.
    Mortality,  corresponding  to  an RV   of  10,  was  reported In  both  the
Fltzhugh  et  al.  (1944)  chronic rat study  and   the  Schroeder and MHchener
(1971b) 3-generat1on study  using  mice.   Mortality occurred In the FHhzhugh
et al.  (1944) rat study at  a human  MED  of 6.3 mg/kg, which  corresponds to an
RVd  of 4.3,  while  In  the Schroeder  and  MHchener   (1971b)  3-generat1on
study,  Increased mortality  of offspring  occurred at  a  human  MED of  2.8
mg/kg,  corresponding  to  an   RV.   of  4.8.    Multiplying  the  RV  s  by  the
RV s  yields  CSs  of 43 and  48  for  the  FHzhugh  et  al.  (1944) and Schroeder
and MHchener (1971b) studies,  respectively.
    U.S.  EPA  (1983)  also  derived  an RQ of 10 based on a  CS of 49  calculated
from  the   Schroeder  and  MHchener  (1971b)  3-generat1on  study  using  mice.
0149d                               9-4                             06/15/89

-------
VO
o.
                                                           TABLE  9-2

                                                 Composite  Scores for  Selenium
10
1
en

Species
Rats
Mice
Animal Dose
(mg/kg/day)
0.5
0.57
Chronic
Human MED
(mg/day)
5.9
3.01
RVd
4.3
6.3
Effect
Increased mortality,
cirrhosis of the liver
Increased mortality
RVe
10
10
CS
43
62.8
RQ
10
10
Reference
Mtzhugh et al .,
1944
Schroeder and
         Humans
NA
3.21
       of offspring before
       weaning

4.7    Selenosls (effects
       on the nervous system,
       Including convulsions
       and paralysis)
                                                                                                    MHchener, 1971b
42.6    10    Yang et al.,
              1983
         NA = Not applicable
CO
10

-------
The CS  reported  In U.S. EPA  (1983)  differs  slightly  from the CS derived  1n
Table 9-2  because  different  assumptions  were  used to  estimate  dosage  from
the 3 ppm drinking water concentration.  U.S. EPA  (1983)  calculated  the  dose
assuming  mice  drink  water  equivalent  to   17X  of  their  body  weight/day,
resulting  In  a  dose  of  0.51  mg/kg/day, while   the  CS derived  above was
calculated assuming  a  0.03  kg mouse  drinks 0.0057  l  water/day (U.S.  EPA,
1986c),  resulting In a dose of 0.57  mg/kg/day.
    The  CSs computed  for mortality  1n  rats   (FHzhugh et  al.,  1944), mortal-
ity In offspring of mice (Schroeder  and HHchener,  1971b) and  severe nervous
effects  1n humans  (Yang et al., 1983)  range  from 42.3-48, and  all correspond
to an RQ  of  10.    The  CS of  42.3  and RQ: of  10.computed  for the  Yang et al.
(1983)  study  1s  chosen  over  either  of  the slightly  higher  CSs calculated
from  the  animal  data  because  the  uncertainty  associated with extrapolation
from animals  to humans  1s eliminated.  This  CS and RQ are presented  1n Table
9-3.
9.2.   BASED  ON CARCINOGENICITY
    Oral  studies  of  the  carclnogenlcHy of selenium  compounds In animals
have  not   been  conclusive.   Nelson  et   al. (1943)  found  Increased   tumor
Incidences In  rats treated with selenlferous corn or wheat.   This  study  Is
limited  because the  basal  diet was  suboptlmal  1n  protein.  Positive results
were also reported In  male and  female  rats and female mice  treated  by  gavage
with  sodium  selenlde  (NCI/NTP, 1980a).   This  study  Is  limited  because the
compound  under  study  was  not  defined   clearly.   Schroeder  and  MHchener
(1971a)  reported a significant Increase  1n   the Incidences  of  tumors 1n  rats
treated  with selenate  1n the  drinking  water  at 2-3 ppm Se.  An evaluation  of
these data 1s  not  possible  because  not all  autopsled  rats  were  examined
hlstologlcally and treated  rats  lived longer than controls   (IARC, 1975).

0149d                               9-6                              06/15/89

-------
                                  TABLE 9-3
                            Selenium and Compounds
           Minimum  Effective  Dose  (MED)  and Reportable Quantity (RQ)

Route:                  oral
Dose*:                  3.2 mg Se/day
Effect:                 severe nervous symptoms:  convulsions,  paralysis
RVd:                    4.7
RVe:                    9
Composite Score:        42.6
RQ:                     10
Reference:              Yang et al., 1983
*Equ1valent human dose
0149d                               9-7                              06/15/89

-------
Harr  et  al.  (1967) and  Tlnsley et al.  (1967)  did not  find  a carcinogenic
effect  In  male  and  female  Wlstar rats  provided with  sodium  selenlte  or
sodium selenate  1n the diet  at 0.5-16 yg Se/g diet  throughout  their  life-
times.  Negative results were also reported In mice treated with selenate or
selenlte at 3 ppm (Schroeder  and Kitchener, 1972).
    Determining  the  carcinogenic potential of  selenium 1s  further  compli-
cated by the  essentiality  of selenium and numerous  studies  reporting  anti-
carcinogenic  activity.   Human  epidemiology  studies  also  tend  to  show  an
Inverse relationship between  blood selenium levels and cancer.  Based on the
lack  of  human  cardnogenldty  data  and Inadequate animal evidence, selenium
can  be   considered,  according  to  EPA classification  scheme,  as  Group  0
substance  -  not  classifiable as  to human  cardnogenlclty.   Substances  with
an EPA Group D  classification  are not given  a hazard ranking; therefore,  an
RQ based on  cardnogenlclty  cannot  be assigned.   Evaluation  of  the  NCI/NTP
(1980a)  study failed to demonstrate the association of a chemical species  of
selenium with  cardnogenlclty.   Therefore,  1t  Is  Inappropriate  to estimate
an RQ based on cardnogenldty of selenium sulflde.  However, because of the
evidence of  cardnogenldty   In  both   rats  and female  mice   1n  this  study,
selenium sulflde may be classified  as  a  Group D substance  based on the U.S.
EPA (1986b) classification  scheme.
0149d                               9-8                              09/21/89

-------
                                10.  REFERENCES

ACGIH  (American  Conference  of  Governmental  Industrial  Hyg1en1sts).   1986.
Documentation of Threshold Limit Values and  Biological Exposure  Indices,  5th
ed.  Cincinnati,  OH.   p.  517-518.

ACGIH  (American  Conference  of  Governmental  Industrial  Hyglenlsts).   1987.
Threshold  Limit   Values   and  Biological   Exposure   Indices   for   1987-1988.
Cincinnati, OH.   p.  32.

Adams, W.J.  and  B.B.  Heldolph.   1985.  Short-Cut Chronic Toxldty  Estimates
Using  Daphnla magna.  Aquatic  Toxicology  and  Hazard  Assessment.   Seventh
Symposium,  ASTMST  854,   R.D.   Cardwell,   R.  Purdy  and  R.C.   Bahner,   Ed.
American Society  for  Testing and Materials,  Philadelphia, PA.   p.  87-103.

Adams, W.J.  and  H.E. Johnson.   1981.  Selenium: A hazard  assessment and a
water  quality criterion  calculation,  p. 124-137.    Jji:  Aquatic  Toxicology
and  Hazard  Assessment:  Fourth  Conference,  American Society  for Testing  and
Materials, O.R.  Branson  and K.L. Dlckson,  Ed.  ASTM  STP 737.

Ahsanullah,  M and G.W.   Brand.   1985.  Effect  of  selenlte  and  selenlferous
fly ash leachate on growth and  viability  of  the marine amphlpod Allorchestes
compressa.  Mar.  Blol.  89(3):  245-248.

Allaway, W.H., J. Kuboto,  F. Losee  and  M.  Roth.  1968.  Selenium,  molybdenum
and  vanadium 1n  human blood.   Arch.  Environ.  Health.  16:  342-348.   (CHed
In Medlnsky et al.,  1985)
0150d                               10-1                             06/15/89

-------
AMmoto, R.,  R.A.  Duce, B.J.  Ray and  C.K.  Unn1.   1985.   Atmospheric trace
elements at  Enewetak  atoll:   II.  Transport  to the  ocean  by  wet  and  dry
deposition.  J.  Geophys. Res.,  D.  Atmos.  90:  2391-2308.

Atkinson,  R.  1985.   Kinetics  and mechanisms of the  gas-phase  reactions of
the  hydroxyl  radical with  organic  compounds under  atmospheric  conditions.
Chem. Rev.   85:  69,  155-157.

Barlow,  S.M.  and  P.M.  Sullivan.  1982.   Reproductive  Hazards of  Industrial
Chemicals.   An  Evaluation  of  Animal  and  Human Data.   Academic  Press,  New
York.  p. 483-500.

Barrows, M.E., S.R.  PetrocelH, K.J. Macek and  J.J. Carroll.  1980.   Blocon-
centratlon   and  elimination  of selected  water   pollutants by  the  blueglll
sunflsh  (Lepomls  macrochlrus).   Dyn.   Exp.  Haz.  Assess.  Tox.  Chem. Symp.
p. 379-392.

BarYosef,  B.   1987.   Selenium  desorptlon from  ca-kaol1n1te.   Commun.  Soil
Sc1. Plant  Anal.   18: 771-779.

Batley,  G.E.  1987.   Heavy metal  spedatlon  1n  waters,  sediments and biota
from Lake MacquaMe, New South Males.   Aust.  J.  Mar. Freshwater Res.   38(5):
591-606.

Baumann, P.C. and R.B.  G1llesp1e.   1986.   Selenium bloaccumulatlon In  gonads
of  largemouth bass  and  blueglll  from  three  power  plant cooling  reservoirs.
Environ. Toxlcol.  Chem.   57(7): 695-701.
0150d                               10-2                             04/03/89

-------
Beath,  O.A.    1962.    The  story  of  selenium  In  Wyoming.   University  of
Wyoming, Laramle.   (Cited In  U.S.  EPA,  1980a)

Beems, R.B. and L. van Beek.  1985.  Short-term (6-week) oral toxldty study
of selenium 1n Syrian hamsters.   Food Chem.  Toxlcol.   23(10): 945-957.

Bennett, B.G.   1986.  Exposure  assessment  for metals  Involved  In cardno-
genesls.  |n: IARC Sd. Publ. 71(Env1ron. Cardnog. Sel. Methods Anal.).  8:
115-127.

Bennett, W.N.,  A.S.   Brooks  and  M.E.  Boraas.   1986.   Selenium  uptake  and
transfer 1n an aquatic food  chain and  Its effects on fathead minnow larvae.
Arch. Environ. Contam. Toxlcol.   15(5):  513-517.

Bertram, P.E. and A.S. Brooks.   1986.   Kinetics  of accumulation of selenium
from  food  and water   by  fathead minnows  (Plmephales  promelas).   Water Res.
20(7): 877-884.

Beyer, W.N.  and  E.J. Cromartle.  1987.   A survey  of  Pb,  Cu,  Zn,  Cd,  Cr, As
and  Se  1n  earthworms and soil  from  diverse sites.  Environ. Monlt. Assess.
8(1): 27-36.

Beyer, W.N.,  G. Hensler  and  J.  Moore.   1987.  Relation on pH and other soil
variables  to  concentrations  of  lead,  copper,  zinc,  cadmium  and selenium In
earthworms.  Pedob1olog1ca.   30(3):  167-172.
0150d                               10-3                            06/15/89

-------
Bowen, W.H.  1972.  No title provided.   J.  Ir.  Dent.  Assoc.   18:  83.   (Cited
1n Underwood, 1977)

Boyer, H.A.  1984.   Trace elements  1n  the  water,  sediments and  fish  of  the
Upper Mississippi River, Twin cities Metropolitan  area.   Contam.  Upper Miss.
River, Proc. Annu.  Meet.  M1ss.  River  Res.  Consortium.   15th Meeting  Date.
102(4): 31702J.

Boyum, K.W.  and A.S.  Brooks.   1988.  The  effect  of  selenium  In water  and
food on Daphnla populations.   Arch. Environ. Contam.  Toxlcol.   17: 555-560.

Buttner,  W.  1963.  Action of  trace elements on the  metabolism of  fluoride.
J. Dent.  Res.  42:  453.  (Cited 1n U.S.  EPA, 1980a)

Byrne, C.J. and L.R. DeLeon.  1986.  Trace  metal residues 1n  biota  and sedi-
ments  from  Lake Pontchartraln, Louisiana.   Bull.  Environ. Contam.  Toxlcol.
37(1): 151-158.

Calabrese, E.J.  1985.   Toxic  Susceptibility:  Male/Female Differences.   John
Wiley and Sons, New York.  p. 115-117.

Callahan,  M.A.,  M.W.   Sllmak,  N.W.  Gabel,  et  al.   1979.    Mater-related
Environmental  Fate  of  129 Priority  Pollutants.  Vol.  I.   EPA-440/4-79-029a.
p. 16-1 to 16-3.

Cao,  T.  and  R. Tang.  1985.   Effect of selenium  on  synchronous  division of
cells of tetrahymena pyrlformls Bj4.  Huanjlng Kexue.  6(3): 22-24.


0150d                               10-4                             04/03/89

-------
Cardwell,  R.O.,  D.G.  Foreman,  T.R.  Payne and  O.J.  Wilbur.   1976.  Acute
toxldty of  selenium dioxide to  freshwater  fishes.   Arch. Environ.  Contam.
Toxlcol.  4(2): 129-144.

Carter,  R.F.   1966.   Acute  selenium poisoning.   Med.   J.  Aust.   1:   525.
(Cited In U.S. EPA, 1985a)

Casey, C.E., B.E GuthMe, G.M. Friend and  M.F. Robinson.   1982.   Selenium  1n
human  tissues  from  New Zealand.   Arch.  Environ.  Health.    37:   133-135.
(Cited 1n Medlnsky et al., 1985}

Chau,  Y.K.,  P.T.S.  Wong,  B.A.  Sllverberg,  P.L.   Luxon  and  C.A.   Bengert.
1976.  Methylatlon  of selenium 1n  the  aquatic  environment.   Science.   192:
1130-1131.

Cole, R.H., R.E.  Frederick, R.P.  flealy and  R.G.  Rolan.   1984.   Preliminary
findings of  the  priority  pollutant  monitoring  project  of  the nationwide
urban runoff program.  J. Water  Pollut.  Control Fed.   56:  8S8-908.

Cooke,  T.D.  and  K.W.  Bruland.    1987.    Aquatic  chemistry   of   selenium:
Evidence of blomethylatlon.   Environ.  Scl.  Technol.  21: 214-219.

Cummins, L.M.  and E.T. Klmura.  1971.  Safety evaluation  of selenium sulflde
antl-dandruff  shampoos.  Toxlcol.  Appl.  Pharmacol.   20: 89-96.

Dean, J.A., Ed.  1985.   Lange's Handbook of Chemistry, 13th ed.   McGraw-Hill
Book Co., NY.   p. 4-59, 4-104.


0150d                               10-5                              04/03/89

-------
Dlplock,  A.T.   1976.   Metabolic  aspects  of selenium  action  and  toxlclty.
CRC Cr1t. Rev.  Toxlcol.  4: 271-329.

D1sk1n, C.J., C.L. Tomasso, J.C. Alper, M.L.  Glaser  and  S.E.  Fllegel.   1979.
Long-term selenium exposure.   Arch.  Intern. Med.   139:  824-826.

Elsler, R.   1985.   Selenium  Hazards to  F1sh,  Wildlife and Invertebrates:  A
Synoptic Review.  NTIS PB86-137346.

Elkln,  E.M.   1982.    Selenium  and  selenium  compounds.   JJK   Klrk-Othmer
Encyclopedia of Chemical  Technology,  Volume 20,  3rd ed.,  M.  Grayson and  D.
Eckroth, Ed.  John Wiley & Sons, Inc.,  New York.   p.  575-601.

Etnler,  E.L.,  R.E.  Meyer, E.B.  Lewis and  L.C.  Folmar.   1987.   Update  of
Acute and Chronic Toxlclty Data for Heavy  Metals  and Organic  Chemicals  Found
at Hazardous Waste Sites.  Prepared  by Oak  Ridge  National Laboratory,  Oak
Ridge, TN.  PC  A10.  p. 1-5,  11. 131-207.

Flnley,  K.A.   1985.   Observations  of bluegllls  fed  selenium-contaminated
Hexagenla nymphs collected from Belews  Lake,  North Carolina.   Bull.  Environ.
Contam. Toxlcol.  35(6): 816-825.

FHzhugh,  O.G.,  A.A.  Nelson   and  C.I.   Bliss.   1944.   The  chronic   oral
toxlclty of selenium.  J. Pharm. Exp.  Therap.  80: 289-299.

Ganther,  H.E.   1986.   Pathways of selenium  metabolism Including  respiratory
excretory products.  J. Am. Coll.  Toxlcol.  5(1):  1-5.

0150d                               10-6                             04/03/89

-------
Garberg, P. and J. Hogberg.  1986.  The  role  of  selenium-oxygen  Interactions
In selenium metabolism.   Amblo.   15(6):  354-355.

Gartrell,  M.J.,  J.C.  Craun, D.S.  Podrebarac  and  E.L.  Gunderson.   1986a.
Pesticides,  selected elements   and  other  chemicals  In  adult  total  diet
samples.   October,  1980  - March,  1982.  J.  Assoc. Off.  Anal.  Chem.   69:
146-161.

Gartrell,  M.J.,  J.C.  Craun, D.S.  Podrebarac  and  E.L.  Gunderson.   1986b.
Pesticides,  selected  elements  and other  chemicals  1n  Infant  and  toddler
total  diet  samples.  October,   1980  -  March, .1982.  0.  Assoc.  Off. Anal.
Chem.  69:  123-145.

GennHy, J.M.,  N.R.  Bottlno,  R.A Zlngaro,  A.E.  Wheeler  and  K.3.  Irgollc.
1985.   A  selenlte-lnduced  decrease  1n  the  I1p1d  content  of a  red alga.
Phytochemlstry.   24(2):  2823-2830.

Gibson,  T.L.,  P.E.  Korsog and  G.T.  Wolff.   1986.   Evidence for the  trans-
formation  of  polycycllc  organic  matter  In  the atmosphere.  Atmos.  Environ.
20: 1575-1578.

G1llesp1e,  R.B. and  P.C.  Baumann.  1986.  Effects of high  tissue  concentra-
tions  of  selenium  on  reproduction  by  bluegllls.   Trans.  Am.  Fish.  Soc.
115(2): 208-213.

Glover,  J.R.  1967.  Selenium  1n human  urine: A  tentative  maximum  allowable
concentration for  Industrial  and rural  populations.  Ann. Occup.  Hyg.   10:
3-14.

0150d                               10-7                             04/03/89

-------
Gregus Z.  and  C.G.  Klaassen.  1986.   Disposition  of  metals  1n rats: A com-
parative study of fecal,  urinary  and  biliary excretion and tissue distribu-
tion of eighteen  metals.   Toxlcol.  Appl.  Pharmacol.  85(1): 24-38.

Gruebel, K.A.,  J.O.  Leckle and J.A. Davis.   1987.  Effect of  changes  1n oxi-
dation-state on  mobility  of  selenium.   Abstracts  of  papers  of the  American
Chemical Society, Vol.  193.   Stanford University,  Stanford, CA.  p. 210.

Guthrle, R.K.,  E.M. Davis, D.S.  Cherry and H.E. Murray.  1979.   Blomagnlfl-
catlon of  heavy  metals  by organisms In  a marine microcosm.   Bull.  Environ.
Contam. Toxlcol.   21(1-2): 53-61.

Halverson,   A.W.  and K.J.  Monty.   1960.   An  effect  of dietary  sulfate  on
selenium poisoning  In the  rat.   J.  Nutr.   70:  100-102.  (CHed 1n U.S. EPA,
1985a)

Halverson,  A.M.,  P.L.  Guss and O.E. Olson.   1962.  Effect of  sulfur  salts on
selenium poisoning  In the  rat.   J.  Nutr.   77:  459-464.  (CHed 1n U.S. EPA,
1985a)

Halverson,  A.W.,  I.S. Palmer  and  P.L.  Guss.   1966.  Tox1c1ty of  selenium to
post-weanling rats.   Toxlcol. Appl.  Pharmacol.   9:  477-484.

Hamilton  A. and H.L.  Hardy.   1974.   Industrial  Toxicology.    Publishing
Sciences Group, Inc.,  Acton,  MA.   p.  167-170.
0150d                               10-8                             04/03/89

-------
Hansen, T.B.  and  B. Anderson.   1986.   Ozone and  other  air pollutants  from
photocopying machines.   Am.  Indust.  Hyg.  Assoc.  J.   47: 659-665.

Harr,  J.R.,  J.F.  Bone,  I.J. Tlnsley,  P.M.  Weswlg  and R.S. Yamamoto.   1967.
Selenium  toxldty  In rats.   II.  Hlstopathology.   In.: Selenium 1n  B1omed1-
dne,  O.H.  Muth,  J.E.  Oldfleld, P.M.  Weswlg, Ed.   Proc.  of 1st Int.  Symp.,
Oregon  State   University,   1966.    AVI   Publishing  Co.,   Westport,  CT.
p. 153-178.  (Cited 1n  U.S.  EPA, 1985a)

Hawley,  G.G.   1981.    The  Condensed  Chemical   Dictionary,  10th   ed.   Van
Nostrand Relnhold  Co.,  New York.  p.  914,  951.

Heinz, G.H. and L.G. Gold.  1987.  Behavior of  mallard ducklings from  adults
exposed to selenium.  Environ.  Toxlcol. Chem.  6(11):  863-865.

Heinz, G.H., D.J.  Hoffman,  A.J. Krynltsky and D.M.G. Weller.   1987.   Repro-
duction 1n mallards fed selen'.um.   Environ.  Toxlcol. Chem.   6(6): 423-433.

Heinz,  G.H.,  D.J.   Hoffman  and L.G.  Gold.   1988.    Toxldty  of organic and
Inorganic  selenium  to  mallard  ducklings.  Arch.  Environ.  Contam.  Toxlcol.
17: 561-568.

Hlraoka, Y., S. Ishlzawa,  T.  Kamada and  H. Okuda.   1985.   Acute toxldty  of
14 different  kinds  of  metals affecting  medaka  fry.  Hiroshima J. Med.  Sd.
34(3): 327-330.
0150d                               10-9                             04/03/89

-------
Hodson, P.V.,  J.W.  Hilton  and S.J. SUnger.   1986.   Accumulation  of water-
borne selenium by rainbow  trout  (Salmo  galrdnerl).  eggs,  fry and juveniles.
F1sh Physlol. Blochem.   1(4):  187-196.

Holllbaugh,  J.T., D.L.R.  Selbert and W.H.  Thomas.   1980.   A  comparison of
the  acute  toxUHIes  of   ten  heavy  metals  to  phytoplankton   from  Saanlch
Inlet, B.C. Canada.   Estuarlne Coastal Mar. Sd.  10(1): 93-105.

Holmberg,   R.E.,  Jr.  and V.H.  Perm.   1969.   Interrelationships  of  selenium,
cadmium and arsenic  In mammalian teratogenesls.  Arch. Environ.  Health.  18:
873-877.

Hunn,  J.B.,  S.J.  Hamilton and  O.R.   Buckler.   1987.   Toxldty of  sodium
selenlte to rainbow  trout  fry.   Water  Res.  21(2): 233-238.

IARC  (International  Agency for  Research on  Cancer).  1975.   Selenium and
selenium compounds.   Iri:   Evaluation  of  Carcinogenic  Risk  of  Chemicals to
Man.  Monograph No.  9.   Lyon,  France,  p.  245-260.

Jenkins, R.A.   1986.   Occurrence  of  selected  metals  In  cigarette  tobaccos
and smoke.   In; IARC Sd.   Publ.  71 (Environ.  Cardno. Sel. Methods Anal).  8:
129-138.

Johnson,  M.G.    1987.   Trace  element  loadings  to  sediments   of  fourteen
Ontario lakes  and  correlations  with  concentration  In fish.  Can.  J.  F1sh.
Aquat. Scl.  44(1):  3-13.
0150d                               10-10                            04/03/89

-------
Johnston, P.A.  1987.  Acute toxldty of Inorganic selenium to Daphnla magna
(Straus) and  the effect  of  sub-acute exposure upon growth and reproduction.
Aqua. Toxlcol.  10(5-6):  335-352.

Klauda,  R.J.   1986.   Acute  and Chronic Effects  of Waterborne  Arsenic  and
Selenium on the Early Life  Stages  of Striped Bass (Horone saxatnis).  NTIS
PC A10/MF A01.

Klaverkamp, J.F.,  D.A   Hodglns and  A.  Lutz.  1983.   SelenHe  toxldty  and
mercury  selenium   Interactions  In   juvenile  fish.   Arch.   Envlrn.  Contam.
Toxlcol.  12:  405-413.

Klelnow, K.M.  and  A.S.  Brooks.  1986.  Selenium compounds   1n  the fathead
minnow  (Plmephales  promelas):  I.  Uptake,  distribution  and  elimination  of
orally  administered  selenate,  selenlte  and  Ievo-selenometh1on1ne.   Comp.
Blochem. Physlol.  C.  Comp.  Pharmacol. Toxlcol.  83(1):  61-70.

Kok,  F.J.,  A.M. De  Bru1j1n, A.  Hofman,  R.  Vermeeren and  H.A.  Valkenburg.
1987.   Is  serum selenium  a  risk  factor  for  cancer  1n  men only?   Am.  J.
Epldemlol.   125(1): 12-16.

Koller, L.D. and J.H. Exon.  1986.   The two  faces  of  selenium-deficiency and
toxldty are similar  In  animals and man.  Can.  J.  Vet.  Res.   50(3):  297-306.

Koller,  L.D.,  J.H.  Exon,  P.A. Talcott, C.A.  Osborne and  G.M.  Hennlngsen.
1986.   Immune  responses  In  rats  supplemented  with  selenium.   Cl1n.  Exp.
Immunol.  63(3): 570-576.
0150d                               10-11                             04/03/89

-------
Krlshnaja, A.P., M.S.  Rege  and A.G. Joshl.  1987.  Toxic  effects  of  certain
heavy metals  (mercury,  cadmium, leads, arsenic  and  selenium) on  the  Inter-
tidal crab Scylla serrata.  Mar. Environ.  Res.   21(2):  109-119.

LeBlanc,  G.A.   1980.   Acute  toxlclty  of  priority pollutants to  water  flea
(Daphnla maqna).  Bull. Environ. Contam. Toxlcol.  24(5):  684-691.

Lee,  M.,  K.K.  Chan,  E.  Sa1renj1  and  T.  N11kun1.  1979.   Effect  of  sodium
selenlte  on  methylmercury-lnduced  cleft palate  In the mouse.   Environ.  Res.
19: 39-48.

Lemly,  A.D.    1985a.   Toxicology  of  selenium  In  a  freshwater  reservoir:
Implications   for  environmental  hazard evaluation and safety.  Ecotoxlcol.
Environ. Saf.  10(3):  314-338.

Lemly,  A.D.   1985b.   Ecological basis  for regulating  aquatic emissions  from
the power Industry: The case  with  selenium.  Reg. Toxlcol.  Pharmacol.   5(4):
465-486.

Levander, O.A., G.  Alfthan,  H. Arv1lomm1, et al.  1983.   B1oava1lab1l1ty  of
selenium  to   Finnish  men  as  assessed by  platelet  glutathlone  peroxldase
activity and  other  blood parameters.  Am.  J.  Clln.  Nutr.   37:  887-897.

L1u,  O.L., Y.P.  Yang, M.H. Hu,  P.J.  Harrison  and N.M. Price.   1987.   Sele-
nium  content  of marine food  chain  organisms from the coast  of  China.   Mar.
Environ. Res.  22(2):  151-165.
0150d                               10-12                            04/03/89

-------
Lofroth, G.  and B.N. Ames.   1978.   Mutagenlclty of  Inorganic  compounds  In
Salmonella   typhlmurlum:  Arsenic,  chromium  and  selenium.   Mutat.  Res.   53:
65-66.

Longnecker, M.P., P.R. Taylor, O.A. Levander, et al.  1987.  Tissue selenium
levels and  Indices of  selenium  exposure  1n a selenlferous area.  Fed. Proc.
46: 587.

Mann, H.,  W.S.  Fyfe and  R.  KerMch.   1988.  The chemical  content  of algae
and waters: B1oconcentrat1on.  Toxic.  Assess.  3(1):  1-16.

Martin,  M., K.E. Osborn,  P.  Bllllg and N. GUcksteln.  1981.  Toxlcltles of
ten  metals  to  Crassostrea  qlqas  and HytHus  edulls  embryos   and  cancer
maglster larvae.  Mar.  Pollut. Bull.   12(9):  305-308.

Mayer, F.L. and M.R. Ellersleck.  1986.  Manual  of Acute Toxldty: Interpre-
tation and  Database  for  410  Chemicals and 66 Species of Freshwater Animals.
U.S. Dept.  Int.  Fish and H1ldl.  Serv.   Resource  Publ.  160, Washington, DC.

McAdam,  P.A.  and D.A.  Levander.   1987.   Chronic toxldty  and  retention  of
dietary  selenium  fed   to rats  as  D- or  L-selenometh1on1ne,   selenlte  or
selenate.   Nutr. Res.  7(6):  601-610.

McCarty, M.F.   1984.   A practical  prescription  for  cancer prevention syner-
glstlc use of chemopreventlve agents.   Med.  Hypoth.   14: 213-0225.
0150d                               10-13                           04/03/89

-------
McConnell,  K.P.  and  D.M.  Roth.   1966.   Respiratory  excretion of  selenium.
Proc. Soc. B1ol.  Med.  123: 919-921.

Medlnsky, M.A., R.G.  Cuddlhy and  R.O.  McClellan.   1981.   Systemic  absorption
of  selenlous  add  and elemental  selenium aerosols  1n  rats.   J.  Toxlcol.
Environ. Health.   8: 917-928.

Medinsky,  M.A.,   R.G.  Cuddlhy,  W.C.  Griffith,  S.H.   Welssman   and   R.O.
McClellan.  1985.   Projected  uptake  and toxlclty of selenium  compounds  from
the environment.   Environ.  Res.   36(1): 181-192.

Moxon,  A.L.   1938.   The effect  of  arsenic on  the toxlclty of  selenlferous
grains.  Science.   88: 81.   (Cited In U.S.  EPA,  1985a)

Muehlberger,  C.W.   and  H.H.  Schrenk.   1928.    The effect  of  the  state  of
oxidation on  the toxlclty  of certain  elements.   J.  Pharmacol.  Exp. Ther.
33: 270.  (CHed  In  Olson,  1986)

Mu1r, D.C.G., R. Wagemann,  N.P.  Grift, R.J. Norstrom, M. Simon and  J. L1en.
1988.   OrganochloMne  chemical  and  heavy metal contaminants In  white-beaked
dolphins  (Laqenorhynchus   alblrostrls)  and   pilot   whales   (Globlcephala
melaena)  from the  coast  of  Newfoundland,  Canada.   Arch.  Environ.  Contam.
Toxlcol.  17: 613-629.

Mushak,  P.   1983.    Mammalian blotransformatlon  processes Involving  various
toxic metalolds and metals.  In.: Chemical Toxicology and Clinical  Chemistry
of  Metals,  S.S.   Brown and  J.   Savory,  Ed.  Academic  Press,  Inc.,  London.
p. 230-233.

0150d                               10-14                            04/03/89

-------
Nakamuro, K.,  K.  Yoshlkawa,  Y.  Sayoto, et  al.   1976.   Studies  on selenium-
related  compounds.   5.  Cytogenetk effect  and  reactivity  with  DNA.   Mutat.
Res.  40: 177-183.

NAS  (National  Academy  of Sciences).   1976.   Selenium.   National  Academy of
Sciences, Washington, DC.   p.  1-13,  20-27,  41-50.

NAS  (National  Academy of  Sciences).   1980.  National  Academy  of Sciences,
Food  and Nutrition  Board.   Recommended  Dietary Allowances, 9th  Rev.  Natl.
Res. Council, Natl. Acad. Sd.,  Washington, DC.  p. 162-164.  (Cited 1n U.S.
EPA, 1985a)

NCI/NTP  (National  Cancer  Institute/National  Toxicology  Program).   1980a.
Bloassay  of  selenium  sulflde (gavage)  for  possible  cardnogenldty.   NCI
Technical Report Series  No.  194.   NTP  No.  80-17.

NCI/NTP  (National  Cancer  Institute/National  Toxicology  Program).   1980b.
Bloassay  of  selenium sulflde  (dermal  study)  for  possible cardnogenldty.
NCI Technical Report Series  No.  197.   NTP  No.  80-18.

NCI/NTP  (National  Cancer  Institute/National  Toxicology  Program).   1980c.
Bloassay  of  Selsun* for  possible cardnogenldty.   NCI  Technical  Report
Series No. 199.  NTP No.  80-19.

Nebbla,  C.,  C. Brando,  E.  Burdlno, et al.  1987.  Effects of  the chronic
administration  of sodium selenlte on rat  testes.   Res. Commun. Chem. Pathol.
Pharmacol.  58(2): 183-197.
0150d                               10-15                            04/03/89

-------
Nelson, A.A., O.G. FHzhugh and H.O. Calvery.  1943.  Liver tumors following
cirrhosis caused  by  selenium  1n rats.  Cancer Res.   3:  230-236.   (Cited 1n
U.S. EPA, 1980a; IARC,  1975)

Nelson, D.A., 3.E.  MUler  and A. Calabrese.   1988.   Effect  of heavy metals
on  bay  scallops,  surf  clams  and  blue  mussels  In acute   and  long-term
exposures.  Arch. Environ.  Contam. Toxlcol.   17:  595-600.

Newton, M.F.  and  L.J.  Lilly.   1986.  Tissue-specific  clastogenk  effects of
chromium and selenium salts 1n vivo.   Mutat.  Res.  169(1-2): 61-69.

N11m1,  A.J.  and  Q.N.  Laham.    1975.   Selenium  toxldty  on  the  early  life
stages of zebraflsh.   J.  F1sh. Res.  Board  Can.  32(6): 803-806.

N11m1,  A.J.   and  Q.N.   Laham.   1976.   Relative  toxldty of organic  and
Inorganic  compounds  of  selenium  to  newly  hatched  zebraflsh (Brachydanlo
rerlo).  Can. J. Zool.   54(4): 501-509.

Nobunaga, T., H.  Satoh and T. Suzuki.   1979.  Effects of sodium selenlte on
methylmercury embryotoxlclty  and teratogenldcty  1n  mice.   Toxlcol.  Appl.
Pharmacol.  47:  79-88.

Noda, M., T.  Takano and H. Sakural.   1979.   Mutagenk  activity of selenium
compounds.  Mutat. Res.   66: 175-179.

Ohlendorf,  H.M.,   D.J.  Hoffman,   M.K.   Salkl  and  T.W.  Aldrlch.    1986.
Embryonic mortality  and  abnormalities  of  aquatic birds:  Apparent  Impacts of
selenium from Irrigation  dralnwater.  Scl.  Total  Environ.   52(1-2): 49-63.
0150d                               10-16                            04/03/89

-------
Olson, O.E.   1986.   Selenium toxldty 1n  animals  with emphasis on man.   J.
Am. Coll! Toxlcol.  5(1): 45-70.

Olson,  G.J.,  W.P.   Iverson  and  F.E.  Brlnckman.   1983.   B1odeter1orat1on
Potential of  Standard Reference Materials  (SRMS).  In:  B1odeter1orat1on,  5th
ed., T.A. Oxley and S.  Barry, Ed.   p.  434-444.

Omaye, S.T.,  K.A. Reddy and C.E.  Cross.   1978.   Enhanced lung toxldty  of
paraquat  In   selenium-deficient   rats.    Toxlcol.  Appl.   Pharmacol.    43:
237-247.  (Cited In U.S. EPA, 1985a)

OSHA  (Occupational   Safety  and  Health  Administration).   1985.   Occupation
Standards, Permissible  Exposure Limits.   29 CFR 1910.1000.

Owsley,  J.A.  and  D.E.  McCauley.    1986.   Effect   of  extended   sublethal
exposure to  sodium selenlte  on Cer1odaphn1a afflnls.   Bull.  Environ.  Contam.
Toxlcol.  36(6): 876-880.

Pagano,  G.,   M.  dpollaro,  G.  Corsale,  et   al.   1986.   The  sea  urchin:
Bloassay for  the  assessment  of damage from environmental  contaminants.  ASTM
Spec. Tech.  Publ. Vol.  920,  Community  Toxic. Test.  p.  66-92.

Page, G.W.   1981.  Comparison  of  groundwater  and surface water  for  patterns
and  levels  of  contamination by  toxic  substances.   Environ.  Sc1.  Techno!.
15: 1475-1481.
0150d                               10-17                            04/03/89

-------
Palawskl, 0., J.B. Hunn and  F.J.  Dwyer.   1985.   Sensitivity  of  young striped
bass  to  organic  and  Inorganic  contaminants  1n  fresh and  saline  waters.
Trans. Am. F1sh. Soc.  114(5): 748-753.

Pelletler,  E.    1986.   Variations  of  the  bloaccumulatlon  of  selenium  by
Hytllus  edulls  In the  presence  of  organic or  Inorganic  mercury.   Can.  J.
F1sh. Aquat. Sd.  43(1):  203-210.

Plasterer,  M.R.,  W.S.  Bradshaw,  G.M.  Booth, M.W.  Carter,  R.L. Schuler  and
B.D.  Hardln.   1985.   Developmental   toxldty  of  nine  selected  compounds
following prenatal exposure  1n  the mouse:  Naphthalene,  p-n1trophenol, sodium
selenlte, dimethyl phthalate, ethylenethlourea and  four  glycol  ether deriva-
tives.  J. Toxlcol.  Environ.  Health.   15(1): 25-38.

Prevot, P.  and  M.O.  Soyer-Goblllard.   1986.  Combined action of cadmium  and
selenium  on  two marine dlnoflagellates  In culture, Prorocentrum  mlcans  and
Crypthecodlnlum cohnll.   J. Protozool.   33(1):  42-47.

Ray,  J.H. and  L.C.  Altenburg.   1980.  Dependence  of   the  sister-chromatld
exchange-Inducing abilities  of  Inorganic  selenium  compounds  on the valence
state of selenium.  Mutat.  Res.   78(3): 261-266.

Reading,  J.T. and A.L.  Bulkema.   1983.   Chronic  effects  of  selenlte-selenlum
on Daphnla pulex.  Arch. Environ.  Contam. Toxlcol.  12:  399-404.

Reamer, D.C. and  W.H.  Zoller.  1980.  Selenium blomethylatlon  products  from
soil and sewage sludge.   Science.   208: 500-502.


0150d                               10-18                            04/03/89

-------
Rhoads, K. and C.L. Sanders.  1985.  Lung  clearance,  translocatlon  and  acute
toxlclty of arsenic, beryllium, cadmium, cobalt,  lead  selenium,  vanadium  and
ytterbium  oxides  following deposition  In  rat  lung.   Environ.   Res.   36(2):
359-378.

Robertson  D.S.F.    1970.   Selenium  -  A  possible  teratogen?   Lancet.   1:
518-519.

Robin, J.P.  1984.  Safety and health aspects  of  selenium  and  tellurium In a
copper refinery.  Proc. Int. Symp. Ind.  Uses  Selenium Tellurium.   3:  28-56.
Selenium-Tellurium Dev. Assoc. DaMen,  Conn.   >04(20):  173620.   [CA 104(20):
173620w]

Rodgers,  J.H.,  Jr.,  O.S.  Cherry,  R.L.  Graney,  K.L.  Dlckson and J.  Cairns.
1980.  Comparison of heavy metal  Interactions  In  acute and artificial  stream
bloassay  techniques for  the  Asiatic  clam  (Corblcula  flumlnea).   JJK  Aquatic
Toxicology, ASTMSTP707,  J.G.  Eaton,  P.R.  Parrlsh  and  A.C. Hendrlcks,  Ed.
American Society for Testing and Materials, Philadelphia, PA.  p. 266-280.

Rosenfeld,  I.  and O.A.  Beath.   1947.   Congenital  malformations of eyes  of
sheep.  0. Agrlc.  Res.   75: 93-103.  (Cited 1n Barlow and Sullivan,  1982)

Rosenfeld, I.  and O.A. Beath.  1964.  The  effect  of selenium on reproduction
and   congenital   malformations.    In:   Selenium   Geobotany,   Biochemistry,
Toxlclty  and  Nutrition.   Academic Press, London,    p.  198-208.   (Cited  In
Barlow and Sullivan, 1982)
0150d                               10-19                            06/15/89

-------
Sa1k1, M.K.  and  T.P.  Lowe.  1987.  Selenium  In  aquatic  organisms from sub-
surface agricultural drainage  water,  San Joaquln Valley, California.  Arch.
Environ.  Contam.  Toxlcol.   16(6):  657-670.

Sakural,   H.  and  K.  Tsuchlya.   1975.   A  tentative  recommendation  for  the
maximum dally Intake of selenium.   Environ.  Physlol. Blochem.  5:  107-118.

Salkl, A.,   M.  Turner, K.  Patalas,  3.  Rudd  and D.  Flndlay.    1985.   The
Influence of  fIsh-zooplankton-phytoplankton  Interactions on  the  results  of
selenium toxldty experiments within large enclosures.   Can. J.  Fish. Aquat.
Sc1.  42(6):  1132-1143.

Salonen,  J.T., G.  Alfthan,  J.K.  Huttunen and  P.  Puska.   1984.    Association
between serum  selenium and the risk of  cancer.   Am.  J.  Ep1.  120:  342-349.
(CHed 1n Wlllett and Stampfer, 1986)

Salonen,  J.T., R.  Salonen,  R.  Lappetlalnen,  P.H.  Maenpaa,  G. Alfthan and P.
Puska.   1985.    Risk   of  cancer  In  relation  to serum concentrations  of
selenium and vitamins  A and  E: Hatched case-control  analysis of  prospective
data.  Br. Med. 0.  290:  417-420.   (CHed 1n Hlllett and  Stampfer, 1986)

Schroeder, H.A.  and  M. MUchener.  1971a.   Selenium and tellurium 1n rats:
Effect on growth, survival  and  tumors.  J. Nutr.   101:1531-1540.

Schroeder, H.A.  and  M. MHchener.  1971b.   Toxic  effects  of trace  elements
on the reproduction of  mice and rats.   Arch.  Environ. Health.  23: 102-106.
0150d                               10-20                            04/03/89

-------
Schroeder, H.A.  and M. Mltchener.   1972.   Selenium and  tellurium In mice:
Effects on growth, survival  and tumors.   Arch.  Environ.  Health.   24: 66-71.

Schroeder, W.H., H.  Dobson,  D.M.  Kane and N.D. Johnson.  1987.  Toxic trace
elements associated with airborne partlculate matter: A review.  Air  Pollut.
Control. Assoc.  11: 1267-1287.

Serda,  S.  and A.  Furst.   1987.   Acute  toxkUy  of selenium to earthworms.
Proc. West. Pharmacol.  Soc.   30:  227-228.

Shamberger, R.J.  1985.  The genotoxldty of selenium.  Mutat. Res.   154(1):
29-48.

Skinner, W.F.   1985.   Trace element  concentrations 1n wastewater  treatment
basin-reared  fishes: Results of a pilot  study.  Proc. Pa. Acad.  Sc1.  59(2):
155-161.

Smith,  M.I.,   et  al.   1936.   The  selenium problem  In  relation  to  public
health.   A preliminary  survey  to  determine   the  possibility  of selenium
Intoxication   1n  the  rural   population   living  on  selenlferous  soil.   U.S.
Publ. Health  Rep.  51:  1496.  (Cited In  U.S.  EPA,  1980b)

Smith, R.A.,  R.B. Alexander and M.G.  Wo1man.   1987.  Water-quality  trends  In
the Nation's  rivers.  Science.   235: 1607-1615.

Snyder,  R.D.    1987.   Effects   of   sodium  selenlte  on  DNA  and carcinogen-
Induced DNA repair 1n human  deplold flbroblasts.  Cancer Lett.   34(1): 73-81.


0150d                               10-21                           04/03/89

-------
Sorensen, E.M.B., P.M. Cumble, T.L. Bauer, J.S. Bell and C.H.  Harlan.  1984.
H1stopatholog1cal, hematologlcal,  condition-factor  and organ weight changes
associated  with  selenium accumulation   In  fish  from  Belews  Lake,  North
Carolina.  Arch. Environ.  Contam.  Toxlcol.  13(2):  153-162.

SMvastava,  O.K. and R.K.  Tyagl.   1985.   Toxicology of selenium  and  vanadium
to  the  striped  gouraml,  Collsa  fasclatus.   Acta.  Hydroblol.   25-26(3-4):
481-486.

Stahl,  Q.R.   1969.   Preliminary  A1r  Pollution Survey of  Selenium  and  Its
Compounds.  A  Literature  Review.   Prepared  under  Contract No.   PH22-68-25
for the U.S. Department of Health, Education  and Welfare,  p.  38.

Tang,  R., J.  Sh1, Y.  Liu, M.  Van, T. Cao  and  Y.  Chen.   1985.   Toxldty of
selenium  on the  growth  of  Tetrahymena   pyrlformls.   Beijing Daxue Xuebao,
Z1ran Kexueban.  1:  58-64.

Thomson,  C.D.   and   R.O.H.   Stewart.    1974,    The  metabolism   of   [7SSe]
selenlte 1n  young women.   Br. J.  Nutr.   32:  47-57.

Thomson, C.O., M.F.  Robinson,  D.R. Campbell  and  H.M. Rea.  1982.  Effect of
prolonged  supplementation with  dally  supplements  of  selenomethlonlne  and
sodium  selenlte  on  glutathlone peroxldase activity  In blood  of  New Zealand
residents.  Am. J. CUn.  Nutr.   36: 24-31.

Thorlac1us-Uss1ng,  D.  and  G.   Danscher.   1985.   Selenium  1n  the  anterior
pituitary of rats exposed to sodium selenlte: Light  and  electron microscopic
localization.   Toxlcol. Appl. Pharmacol.   81(1):  67-74.

0150d                               10-22                            04/03/89

-------
Thorlaclus-Usslng, 0. and  B.L.  Rasmussen.   1986.   Light and electron micro-
scopical visualization  of  selenium 1n  adrenal  glands  from  rats  exposed to
sodium selenlte.  Exp.  Mol. Pathol.   45(1):  59-67.

Thorladus-Usslng, 0.,  J.  Rungby,  B.  Moeller-Madsen and G.  Danscher.   1986.
Selenium  accumulation   1n  the  oocyte  after  exposure  to  sodium  selenlte.
Acta. Pharmacol. Toxlcol.   59(7):  139-141.

Ting, K.P.  and G.W.R.  Walker.   1969.   The distributive  effects  of  seleno-
amlno add  treatment on corsslng-over  1n DrosophUa  melanogaster.   Genetics.
61: 141-155.  (Cited In U.S.  EPA,  1985b)

Tlnsley, I.J.,  J.R.  Harr,  G.F. Bone,  P.M.  Weswlg  and  R.S.  Yamamoto.   1967.
Selenium toxlclty  1n  rats.  I. Growth  and  longevity.  Jji:  Selenium  In  Bio-
medicine, O.H. Muth, O.E.  Oldfleld and  P.H. Weswlg,  Ed.   First  International
Symposium,  Oregon  State   University.   AVI  Publishing  Co.,  Westport,   CT.
p. 141-152.  (Cited 1n U.S. EPA, 1985a)

Trlpathl,  A.K.  and S.N. Pandey.   1985.  Toxlclty  of selenium to  Chlorella
vulgarls and Phormldlum foveolarum.   Natl.  Acad.  Sd. Lett.   8(10):  307-309.

Underwood,  E.J.   1977.   Trace  Elements In Human  and Animal Nutrition,  4th
ed.  Academic Press, New York.   p.  310-346.

Unsal,  M.   1987.   Evaluation  of  the  synerglstlc  effect  of selenium on  the
acute  toxlclty  of  mercury  1n  fish  larvae.    Rev.  Int.   Oceanogr.   Med.
p. 87-88, 125-135.

0150d                               10-23                            04/03/89

-------
USOI  (U.S.  Department of  Interior).   1982.   Minerals  Yearbook.   Centennial
Edition, 1981.  Volume I. Metals and Minerals.   Bureau  of  Mines,  Washington,
DC.  p. 948-951.

USDI (U.S. Department of Interior).  1988.   Preprint  from  the  1987 Bureau of
Mines  Mineral  Yearbook.   Other  Metals.   U.S.   Department  of  the  Interior,
Bureau of Mines, Washington, DC.

U.S.  EPA.   1977.   Computer  print-out  of non-confidential  production  data
from TSCA Inventory.  U.S.  EPA, OPTS,  CIO, Washington, DC.

U.S.  EPA.   1980a.   Ambient  Water  Quality Criteria  for Selenium.   Environ-
mental  Criteria  and  Assessment   Office,  U.S.  EPA,  Cincinnati,  OH.   EPA
440-80-070.  p. C-l to C-7.

U.S.  EPA.   1980b.   Guidelines  and Methodology  Used 1n  the  Preparation  of
Health  Effect  Assessment  Chapters  of  the   Consent Decree  Water  Criteria
Documents.  Federal Register.  45(231):  79347-79357.

U.S.  EPA.   1982.    Errata   for  Ambient  Water   Quality  Criteria  Documents.
(Updated: February 23, 1982).  Prepared by the  Office of  Health  and Environ-
mental Assessment,  Environmental Criteria and Assessment  Office,  Cincinnati,
OH for the Office of Water  Regulations and Standards, Washington,  DC.

U.S. EPA.  1983.   Reportable Quantity Document  for  Selenium (and  compounds).
Prepared by the Office  of  Health and  Environmental  Assessment, Environmental
Criteria and  Assessment  Office,  Cincinnati,   OH  for  the Office of Emergency
and Remedial Response, Washington,  DC.

0150d                               10-24                            04/03/89

-------
U.S. EPA.  1984a.  Methodology and Guidelines  for  Ranking  Chemicals  Based  on
Chronic Toxlclty  Data.   Prepared by  the  Office of Health  and  Environmental
Assessment, Environmental Criteria and  Assessment  Office,  Cincinnati,  OH for
the Office of Emergency and Remedial  Response,  Washington,  DC.

U.S. EPA.   1984b.  Health  Effects Assessment  for Selenium  (and  compounds).
Prepared by the Office of Health and  Environmental Assessment,  Environmental
Criteria and  Assessment  Office,  Cincinnati, OH  for  the Office of  Emergency
and Remedial  Response, Washington, DC.  EPA/540/1-86-058.

U.S. EPA.   1985a.  National  Primary Drinking  Water  Regulations:  Synthetic
Organic Chemicals,  Inorganic  Chemicals  and Micro-organisms.  Proposed Rule.
Federal Register.   50(219):  46936-47025.

U.S. EPA.  1985b.   Drinking Water Crltera Document for Selenium.   Office  of
Drinking Water, U.S. EPA, Washington,  DC.   NTIS PB86-118098.

U.S. EPA.   1985c.   Integrated  Risk  Information  System   (IRIS).   Reference
Dose (RfD)  for  Oral  Exposure  for Selenlous  Add.   Online.   (Verification
date 08/19/85.)   Office  of  Health  and  Environmental  Assessment,  Environ-
mental  Criteria and Assessment Office, Cincinnati, OH.

U.S. EPA.   1985d.   Integrated  Risk  Information  System   (IRIS).   Reference
Dose (RfD)  for Oral  Exposure for  Selenourea.  Online.   (Verification  date
08/19/85.)   Office of  Health  and  Environmental  Assessment,  Environmental
Criteria and Assessment Office, Cincinnati, OH.
0150d                               10-25                            04/03/89

-------
U.S.  EPA.   1986a.   Reference  Values  for Risk  Assessment.   Prepared by  the
Office  of  Health and  Environmental Assessment,  Environmental  Criteria  and
Assessment Office,  Cincinnati, OH for  the Office  of  Solid  Waste,  Washington,
DC.

U.S.  EPA.   19865.    Guidelines   for   Carcinogen  Risk  Assessment.   Federal
Register.  51(185):  33992-34003.

U.S.  EPA.   1986c.    Superfund  Programs;  Reportable  Quantity  Adjustments;
Final Rule.  Federal Register.  51(188):  34534-34549.

U.S.  EPA.   1987a.    Ambient   Water  Quality  Criteria  for  Selenium,  1987.
Office  of  Water  Regulations  and  Standards,  Criteria and Standards  Division,
Washington, DC.  EPA/440-5-87-006.

U.S.  EPA.   1987b.    Hazardous Substances;  Reportable Quantity  Adjustments;
Proposed Rules.  Federal Register.  52(50):  8140-8186.

U.S.  FDA.   1987.   Food  additives  permitted  1n  feed  and  drinking  water  or
animals; selenium.   Federal Register.   52(65): 10887-10888.

Vadhanav1k1t, $., R.J. Kraus  and  H.E.  Ganther.   1987.  Metabolism of seleno-
cyanate 1n the rat.   Arch. Blochem.  Blophys.   258(1): 1-6.

Venugopal,  B.  and   T.D.   Luckey.   1978.   Metal  Toxldty  1n  Mammal.   2.
Selenium.  Plenum Press, NY.   p.  235-245.
0150d                               10-26                            04/03/89

-------
Wang, W.  1986.  ToxIcHy  tests  of  aquatic  pollutants by using common duck-
weed.  Environ. Pollut.  Ser.  B.  Chem.  Phys.   11(1): 1-14.

Watenpaugh,  D.E. and  T.L.  6e1t1nger.   1985a.  Absence of selenate avoidance
by fathead minnows  (Plmephales  promelas).  Water  Res.   19(7): 923-926.

Watenpaugh,  D.E. and  T.L.  BeUlnger.   1985b.  Oxygen consumption In fathead
minnows (Plmephales promelas).   Comp. Blochem. Physlol., C: Comp. Pharmacol.
Toxlcol.  80C(2): 253-2256.

Weast,  R.C.,  Ed.  1985.   CRC  Handbook of  Chemistry and  Physics,  66th  ed.
CRC Press, Inc., Boca  Raton,  FL.   p.  B-136,  137,  145.

Weir, P.A.  and C.H.  H1ne.   1970.  Effects  of  various metals  on behavior of
conditioned goldfish.   Arch.  Environ.  Health.  20(1):  45-51.

Welssman,  S.H.,  R.G.  Cuddlny and H.A. Medlnsky.   1983.  Absorption, distri-
bution and  retention  of Inhaled selenlous  add  and  selenium  metal  aerosols
In beagle dogs.  Toxlcol. Appl.  Pharmacol.   67: 331-337.

Whiting, R.F.,  L.  Wei  and  H.F.  Stlch.  1980.  Unscheduled DNA  synthesis and
chromosome  aberrations  Induced  by Inorganic  and organic selenium compounds
In the  presence of glutathlone.  Mutat. Res.  78:  159-169.   (CUed In U.S.
EPA,  1985b)

WHO  (World  Health Organization).    1984.   Guidelines  for   Drinking  Water
Quality, Vol. 1.  Geneva,   p.  53, 57-58.
0150d                               10-27                            04/03/89

-------
Wlersema,  J.M.,  L.  Wright,  B.  Rogers,  et al.   1984.   Human  Exposure  to
Potentially Toxic  Elements  through  Ambient  Air 1n  Texas.   Proc.  77th  APCA
Annu. Meet., Austin, TX.  84-1.2.   p. 15.

Wlllett, W.C.  and  M.J.  Stampfer.   1986.   Selenium  and human cancer.   Acta
Pharmacol.  Toxlcol. Suppl.  59(7):  240-247.

Wlllett, W.C.,  B.F.  Polk,  J.S.  Morris, et  al.   1983.  Pred1agnost1c  serum
selenium and  risk  of  cancer.  Lancet.   2:130-134.    (Cited  1n  Wlllett  and
Stampfer, 1986}

Wlndholz, M.,  Ed.   1983.   The  Merck Index, 10th  ed.   Merck 8. Co.,  Rahway,
NJ.  p. 4709,  8271, 8286.

W1tz,  S.,  J.A. Wood  and M.W.  Wadley.   1986.   Toxic  metal and  hydrocarbon
concentrations 1n  automobile  Interiors  during  freeway  transit.  JJK  ACs  D1v.
Environ. Chem.  192nd Natl.  Mtg.   26: 302-305.

Wolfenberger,   V.   1986.   Survival  of the hermit crab,  Cllbanarlus  vlttatus.
exposed  to  selenium and  other environmental  factors.   Bull.  Environ.  Contam.
Toxlcol.  37(3): 369-374.

Wolfenberger,   V.A.   1987.    Influence  of  environmental   factors  on  oxygen
consumption of Cllbanarlus  vlttatus (striped  hermit   crab).   Tex.  J.  Sd.
39(1): 37-47.
0150d                               10-28                            04/03/89

-------
Woock,  S.E.  and  P.B.  Summers,   Jr.   1984.   Selenium  monitoring   1n  Hyco
Reservoir (NC)  waters (1977-1981) and biota (1977-1980).  Electr. Power Res.
Inst.  EPRI   EA.   EPRI  EA-3329.   Workshop  Proc.   Eff.   Trace  Elem.  Aquat.
Ecosyst.  102(24): 208969e.

Woock,  S.E., W.R. Garrett,  W.E.  Partln  and W.T.  Bryson.   1987.  Decreased
survival and teratogenesls during laboratory selenium exposures  to blueglll,
Lepomls macrochlrus.   Bull.  Environ.  Contam. Toxlcol.  39(6): 998-1005.

Yang, G., S. Wang, R. Zhou and S. Sun.   1983.  Endemic  selenium  Intoxication
of humans 1n China.   Am.  J.  Clln.  Nutr.   37P: 872-881.

Young,  V.R., A.   Nahapetlan  and  M.   Janghorbanl.   1982.   Selenium bloavall-
abllHy  with  reference  to   human  nutrition.   Am.  J.  CUn.  Nutr.   35(5):
1076-1088.

Z1eve, R. and P.J. Peterson.   1984.   Volatilization  of  selenium from plants
and soils.  Sd.  Total Environ.   32:  197-202.
0150d                               10-29                            04/03/89

-------
                                  APPENDIX A

                              LITERATURE  SEARCHED



    This  HEED  1s  based  on  data  Identified  by  computerized  literature

searches of the following:

              CHEMLINE
              TSCATS
              CASR online (U.S. EPA Chemical Activities Status Report)
              TOXLINE
              TOXLIT
              TOXLIT 65
              RTECS
              OHM TADS
              STORET
              SRC Environmental Fate Data Bases
              SANSS
              AQUIRE
              TSCAPP
              NTIS
              Federal Register
              CAS ONLINE (Chemistry and Aquatic)
              HSDB
              SCISEARCH
              Federal Research 1n Progress


These  searches  were  conducted  In  May,   1988,  and  the following  secondary

sources were reviewed:
    ACGIH  (American  Conference of Governmental  Industrial  Hyglenlsts).
    1986.   Documentation  of the  Threshold  Limit Values  and  Biological
    Exposure Indices, 5th ed.  Cincinnati, OH.

    ACGIH  (American  Conference of Governmental  Industrial  Hyglenlsts).
    1987.   TLVs:  Threshold  Limit Values for  Chemical  Substances  1n  the
    Work  Environment  adopted   by   ACGIH   with   Intended  Changes  for
    1987-1988.  Cincinnati,  OH.  114 p.

    Clayton,  G.D.  and  F.E.  Clayton,  Ed.    1981.   Patty's  Industrial
    Hygiene  and  Toxicology,  3rd rev.  ed.,  Vol.  2A.   John  Wiley  and
    Sons, NY.  2878 p.

    Clayton,  G.D.  and  F.E.  Clayton,  Ed.    1981.   Patty's  Industrial
    Hygiene  and  Toxicology,  3rd rev.  ed.,  Vol.  2B.   John  Wiley  and
    Sons, NY.  p. 2879-3816.
0151d                               A-l                              04/03/89

-------
    Clayton,   G.D.  and  F.E.  Clayton,   Ed.   1982.   Patty's   Industrial
    Hygiene and  Toxicology,  3rd  rev.   ed.,  Vol.  2C.   John  Wiley and
    Sons,  NY.   p.  3817-5112.

    Grayson,  M. and  D.  Eckroth,  Ed.   1978-1984.  Klrk-Othmer  Encyclo-
    pedia  of  Chemical Technology,  3rd ed.  John Wiley and  Sons,  NY.   23
    Volumes.

    Hamilton,  A.  and H.L.  Hardy.  1974.  Industrial Toxicology,  3rd ed.
    Publishing Sciences  Group,  Inc.,  Littleton,  MA.   575 p.

    IARC  (International  Agency for  Research  on Cancer).   IARC  Mono-
    graphs on  the  Evaluation  of Carcinogenic  Risk  of  Chemicals   to
    Humans.  IARC,  WHO,  Lyons,  France.

    Jaber, H.M.,  W.R.  Mabey,  A.T.  L1eu,  T.W.  Chou and  H.L.  Johnson.
    1984.    Data  acquisition  for  environmental  transport  and  fate
    screening  for  compounds  of  Interest to  the Office  of Solid Waste.
    EPA  600/6-84-010.   NTIS  PB84-243906.   SRI   International,   Menlo
    Park,  CA.

    NTP (National  Toxicology  Program).   1987.   Toxicology Research and
    Testing  Program.   Chemicals   on   Standard  Protocol.    Management
    Status.

    Ouellette,  R.P.  and  J.A.  King.    1977.    Chemical  Week Pesticide
    Register.   McGraw-Hill  Book  Co.,  NY.

    Sax, I.N.   1984.  Dangerous Properties of  Industrial  Materials, 6th
    ed.  Van  Nostrand Relnhold Co., NY.

    SRI (Stanford  Research  Institute).   1987.   Directory of  Chemical
    Producers.   Menlo Park,  CA.

    U.S.  EPA.   1986.  Report  on  Status  Report  1n  the  Special Review
    Program,   Registration   Standards   Program   and   the   Data  Call   In
    Programs.   Registration  Standards  and  the Data Call  1n Programs.
    Office of  Pesticide  Programs,  Washington, DC.

    USITC   (U.S.   International  Trade   Commission).    1986.   Synthetic
    Organic Chemicals.  U.S. Production  and Sales, 1985,  USITC  Publ.
    1892,  Washington, DC.

    Verschueren,  K.   1983.   Handbook of  Environmental  Data  on  Organic
    Chemicals, 2nd ed.   Van Nostrand  Relnhold Co., NY.

    Wlndholz,  M.,  Ed.  1983.  The Merck  Index,  10th  ed.   Merck  and Co.,
    Inc.,  Rahway,  NJ.

    Worthing,  C.R.  and  S.B. Walker, Ed.   1983.  The Pesticide  Manual.
    British Crop Protection Council.  695 p.
0151d                               A-2                              04/03/89

-------
    In  addition,  approximately 30  compendia of aquatic  toxUUy data were

reviewed, Including the following:


    Battelle's  Columbus  Laboratories.   1971.   Water  Quality  Criteria
    Data  Book.   Volume  3.   Effects  of  Chemicals   on   Aquatic   Life.
    Selected  Data  from the Literature  through  1968.   Prepared for  the
    U.S. EPA under Contract No. 68-01-0007.   Washington,  DC.

    Johnson,  W.W.  and  M.T. Flnley.   1980.   Handbook  of  Acute  Toxldty
    of  Chemicals  to  F1sh and  Aquatic  Invertebrates.   Summaries  of
    Toxldty  Tests  Conducted  at  Columbia  National  Fisheries  Research
    Laboratory.   1965-1978.    U.S.  Dept.  Interior,  Fish  and  Wildlife
    Serv. Res. Publ. 137,  Washington, DC.

    McKee, J.E.  and  H.W.  Wolf.  1963.  Water Quality Criteria, 2nd  ed.
    Prepared  for  the  Resources   Agency  of  California,  State   Water
    Quality Control Board.  Publ.  No. 3-A.

    Plmental, D.   1971.   Ecological  Effects  of -Pesticides  on  Non-Target
    Species.  Prepared for the U.S.  EPA, Washington,  DC.   PB-269605.

    Schneider, B.A.  1979.  Toxicology  Handbook.   Mammalian and Aquatic
    Data.  Book  1:  Toxicology  Data.   Office  of  Pesticide  Programs,  U.S.
    EPA, Washington, DC.  EPA 540/9-79-003.   NTIS  PB  80-196876.
0151d                               A-3                              04/03/89

-------
                                                           APPENDIX  B



                                            Summary Table  for  Selenium and Compounds
oo
CT>
00
Species
Inhalation Exposure
Subchronlc ID
Chronic ID
Carc1nogen1c1ty ID
Oral Exposure
Subchronlc human
Chronic human
Carc1nogen1c1ty ID
REPORTABLE QUANTITIES
Based on Chronic Toxlclty:
Based on Carc1nogen1c1ty:
Exposure Effect RfD or qi* Reference

ID ID ID NA
ID ID ID NA
ID ID ID NA

3.2 mg/day selenosls 0.21 mg/day Yang et al., 1983
3.2 mg/day selenosls 0.21 mg/day Yang et al., 1983
ID ID ID NA

10 Yang et al., 1983
ID
          ID  =  Insufficient Data

-------