EPA 910-R-96-019
           Toxiclty and Exposure Concerns

                    to Arsenic in Seafood'


      An Arsenic Literatim Review for Risk Assessments
                Prepared by: Christine M. Chew
                ICF Kaiser - Region X ESAT
                    Seattle, WA 98101
                   Revised: March 1996
      ICF KAISER
 Submitted in fulfillment of Region X ESAT Work Unit Document 4038 under
Technical Instruction Document 10-9601-815 as requested by Patricia drone,
          Task Monitor and (JSEPA Risk Evaluation Unit Chief.

-------
            Toxscsty and Exposure Concerns

                      to Arsenic in Seafood:


       An Arsenic Literature Review for Risk Assessments
                  Prepared by:  Christine M. Chew
                  ICF Kaiser - Region X ESAT
                     Seattle, WA 98101
                    Revised: March 1996
 Submitted in fulfillment of Region X ESA T Work Unit Document 4038 under
Technical Instruction Document 10-9601-815 as requested by Patricia drone,
          Task Monitor and USEPA Risk Evaluation Unit Chief.

-------
TABLE OF CONTENTS
INTRODUCTION	    1
      Objectives  	    1

PROPERTIES AND ENVIRONMENTAL SOURCES	    1
      Species of Arsenic	,,,.,,	    1
      Physical and Chemical Properties	    2
      Environmental Sources	, , ,	    2
      Biological Sources  	,	    4

CONSIDERATIONS FOR                            	    4
      Points of Exposure	    4
      Contributions from Environmental Media 	    4
      Seafood Exposure Contributions	    5
             Species Present  	    6
             Pacific Northwest Speciated Data	   13
             Measurement and Speciation Methods	   14
      Exposure Assessment	   18
             Sources Assessed	   18
             Consumption of Seafood 	   19
             Potential  Tools  	   25

CONSIDERATIONS FOR TOXICITY ASSESSMENTS	   26
      Metabolism	   26
             Inorganic Arsenic and Methylated Metabolites 	   26
             Organic Arsenic Species found in Seafood	   28
      Toxicity	   30
             Inorganic Arsenic and Methylated Metabolites 	   31
             Organic Arsenic and Seafood Arsenic Species	   37

RISK ASSESSMENT IMPLICATIONS	   40
      Regulatory Criteria	   40
             National  United States Standards	,	   40
             Other Standards	   42
      Applications	   43
             Friberg (1988) Study	   43

-------
             Puget Sound Seafood Risk Assessment  	   44
             Kensington Mine Risk Assessment	   45
             Lower Columbia River Bi-state Program  	   46

CONCLUSIONS	   48

REFERENCES	   50
                                         in

-------
INTRODUCTION

       Arsenic is a naturaJly occurring element found in all environmental matrices (soil, air and water)
as well as in living matter.  It is found in comparatively high levels in fish and other edible seafood.
Consequently, its  influence on human health  is of interest in health  risk assessments pertaining to
individuals or populations exposed to edible marine or freshwater organisms.  Arsenic in the environment
can be found in elemental  form as well as many other species, both inorganic and organic.  In seafood,
the organic forms tend to dominate, but inorganic forms are also present.

Objectives

       The objectives of this report are:   1) to determine a range of conservative concentrations of
inorganic arsenic in seafood (including fish, invertebrates and algae) for use in Region 10 human health
exposure assessments and  related activities; 2) to identify speciation methods which may be used by the
EPA Region 10 laboratory to study inorganic and total arsenic content  in northwest fish,  shellfish and
edible marine plants which currently do not have documented data available; and, 3) to present facts and
implications of the toxic potential of arsenic in seafood to human populations ingesting this seafood and
discuss related exposure assessment approaches.  How significant  is dietary intake of inorganic arsenic
from seafood?  What is currently known about the long-term effects of organic arsenic from seafood
consumption? These questions in particular, and others related to the stated objectives, will be discussed
in this report. This report has been prepared for use by US EPA  Region 10; consequently, references
to "the northwest"  apply to the region's four states;  Alaska, Idaho,  Oregon and Washington.
PROPERTIES AND ENVIRONMENTAL SOURCES

Speties of Arsenic

        Arsenic has two primary valence states:  trivalent (As*3) and pentavalent (As*5). Arsenic in each
of these valence states forms both inorganic and organic compounds.  The different species of arsenic
vary in reactivity, solubility, toxicity and other properties.  Of greatest interest in human health risk
assessments are those species taken up by people, and the biotransformed products that result.  Inorganic
arsenic may be taken up in trivalent or pentavalent form; associated species found in urinary excretions
are the two inorganic arsenic forms  and the monomethylated arsenic  (MMA) and dimethylated arsenic
(DMA) species.   Arsenic  species specifically  associated with seafood include a limited amount of
inorganic  arsenic, arsenobetaine (a.k.a. carboyxmethyl(tri-methyl) arsonium  bromide),  arsenocholine
(a.k.a. 2-hydroxyethyl(trimethyl)arsonium bromide) and arsenosugars (mostly found in seaweeds). Fipre
1 depicts the structure of some arsenosugars. Table 1 lists some of the more common arsenic compounds
and their structures.

-------
                                            EPA 91Q-R-96-Oly
           Toxlcity and Exposure Concerns

                    to Arsenic In Seafood:


      An Arsenic Literature Review for Risk Assessments
                Prepared by:  Christine M. Chew
                ICF Kaiser - Region X ESAT
                   Seattle, WA 98101
                   Revised: March 1996
      ICF KAISER
 Submitted in fulfillment of Region X ESA T Work Unit Document 4038 under
Technical Instruction Document 10-9601-815 as requested by Patricia drone,
          Task Monitor and USEPA Risk Evaluation Unit Chief.

-------

-------
TABLE OF CONTENTS
INTRODUCTION	   1
      Objectives	   1

PROPERTIES AND ENVIRONMENTAL SOURCES	   I
      Species of Arsenic	   1
      Physical and Chemical Properties	   2
      Environmental Sources	   2
      Biological Sources	   4

CONSIDERATIONS FOR EXPOSURE ASSESSMENTS  	   4
      Points of Exposure	   4
      Contributions from Environmental Media  	,	   4
      Seafood Exposure Contributions	   5
             Species Present  	   6
             Pacific Northwest Speciated Data	   13
             Measurement and Speciation Methods	   14
      Exposure Assessment	   18
             Sources Assessed	   18
             Consumption of Seafood 	   19
             Potential  Tools  	   25

CONSIDERATIONS FOR TOXICITY ASSESSMENTS	   26
      Metabolism  	   26
             Inorganic Arsenic and Methylated Metabolites  	   26
             Organic Arsenic Species found in Seafood	   28
      Toxieity	   30
             Inorganic Arsenic and Methylated Metabolites  	   31
             Organic Arsenic and Seafood Arsenic Species	   37

RISK ASSESSMENT IMPLICATIONS	,	   40
      Regulatory Criteria	   40
             National  United States Standards	   40
             Other Standards	   42
      Applications	   43
             Friberg (1988) Study	   43

-------

-------
            Puget Sound Seafood Risk Assessment  	  44
            Kensington Mine Risk Assessment  	  45
            Lower Columbia River Bi-state Program  	  46

CONCLUSIONS	  48

REFERENCES	  50
                                         in

-------

-------
INTRODUCTION

        Arsenic is a naturally occurring element found in all environmental matrices (soil, air and water)
as well as  in living matter.  It is found in comparatively high levels in fish and other  edible seafood.
Consequently,  its influence on human health is  of interest in health  risk assessments pertaining to
individuals or populations exposed to edible marine or freshwater organisms.  Arsenic in the environment
can be found in elemental  form as well as many other species, both inorganic and organic.  In seafood,
the organic forms tend to dominate, but inorganic forms are also present,

Objectives

        The objectives of this report are:   1) to  determine a range of conservative concentrations of
inorganic arsenic  in seafood (including fish, invertebrates and algae) for use in Region 10 human health
exposure assessments and related activities; 2) to identify speciation methods which may be used by the
EPA Region  10 laboratory to study inorganic and total arsenic content  in northwest fish, shellfish and
edible marine plants which currently do not have documented data available; and, 3) to present facts and
implications of the toxic potential of arsenic in seafood to human populations ingesting this seafood and
discuss  related  exposure assessment approaches.   How significant  is dietary intake of inorganic arsenic
from seafood?  What is currently known about the long-term effects of organic arsenic from seafood
consumption? These questions in particular, and others related to the stated objectives, will be discussed
in this report.  This report has been prepared for  use by US EPA  Region 10; consequently, references
to "the northwest" apply to the region's four states:  Alaska, Idaho, Oregon and Washington.
PROPERTIES AND ENVIRONMENTAL SOURCES

Species of Arsenie

       Arsenic has two primary vaJence states: trivalent (As*3) and pentavalent (As*5). Arsenic in each
of these valence states forms both inorganic and  organic compounds.  The different species of arsenic
vary in reactivity, solubility, toxicity and other properties.  Of greatest interest in human health risk
assessments are those species taken up by people,  and the biotransformed products that result.  Inorganic
arsenic may  be taken up in trivalent or pentavalent form; associated species found in urinary excretions
are the two inorganic arsenic forms  and the monomethylated arsenic  (MMA) and dimethylated arsenic
(DMA) species.   Arsenic species specifically associated with seafood include a limited amount of
inorganic  arsenic, arsenobetaine (a.k.a.  carboyxmethyl(tri-methyl) arsonium  bromide),  arsenocholine
(a.k.a. 2-hydroxyethyl(trimethyI)arsonium bromide) and arsenosugars (mostly found in seaweeds). Figure
1 depicts the structure of some arsenosugars. Table 1 lists some of the more common arsenic compounds
and their structures.

-------
                                                     S'
                         i t
                        2CHCI
                                                         s\—(a1
                                                          OH  OH
                                                     81
A-X   !

A-XI

A-XII
A-XIII
A-XIV
A-XV
                                                                 R2
                                                                -OH
                                                                              R3
                                                                         -OH
                                                                                     3.
                                                                         -OPOCH,CH SOH »CH,OH
                                                                           1   '       '
                                                                          -so
Physical and Chemical Properties

       Arsenic is a metalloid in group
V(A)  of the  Periodic Table.  Arsenic
exists  as   a   metalloid  in the  zero
oxidative state;  it also exists  in  two
other oxidative states, trivalent (+3 or -
3) and  pentavalent  (+5)  (Hindmarsh
and  McCurdy 1984).   In  both   the
trivalent and  pentavalent states, arsenic
is able to  form  stable   compounds.
Being   in   the   same   group    as
phosphorous,  arsenic  competes,   in
biological   environments,    for
phosphorous binding  sites.  Two arsenic
analogues  to  the phosphorous  species
phosphatidyl  choline  and acetylcholine have been identified (Hedlund et a!. 1982, Christkopoulos et al.
S988b),  Arsenic binds covalently  with many nonmetals to form a variety of organic and inorganic
compounds (Hindmarsh and McCurdy 1984).
                                                                -OH
                                                                          -so3
                                                                          -SO,
                                         SOURCE:  Shibataero/. 1993
                                                        Figure 1 Arsenosugars
       The reactivity and toxicity of these compounds varies with the nature of the compounds.  Of
primary importance is the valence state of the arsenic (Edmonds and Francesconi 1993). Trivalent arsenic
is more reactive and has demonstrated a higher toxicity than pentavalent arsenic.  Yet, when ingested by
people, pentavalent arsenic is converted by the body to trivalent arsenic, which undergoes subsequent
detoxification.

       Arsenic (III)  solubility is low in water but high in acid or alkaline solutions. In water, arsenic
is usually found as arsenate or arsenite (JonnaJagadda and Rao 1993).  Pentavalent  arsenic is  more
prevalent in well oxygenated surface waters; however, in the reducing coastal zones and  in estuaries with
high biological activity, levels of trivalent arsenic and sometimes, methylated species, meet or exceed
those of arsenic (V) (Mafaer and Butler, 1988),

Environmental Sources
       Arsenic is released from earth's crust into the environment via volcanic activity and weathering
of arsenic-containing sulfides (Phillips, 1990).  It is ubiquitous in the environment and naturally cycles
through it, traveling from  soils into  the  air and plants  through  ocean and  lake sediments and  into
groundwater.  A more in-depth discussion of this cycle can be found in the Hindmarsh and McCurdy
review (1984).

-------
    Tabft 1 Some Common Arsenic Compounds
Name                    Formula
Arsenate                 HjAsOA HAs042", As043"
Arsenite                 HjAs031", HAs032", As033'
Monomethylarsonic acid   CH3AsO|OH)2
    Oimethyiarsrnic acid
    Arsenobetains
    Arsenocholine
    Arsenous Acid
    Arsenic Acid
    Arsenic Pentoxide
    Arsenic Trioxide
    Trimethylarsine oxide
    Tetramethyiarsonium
    ion
    Dimethylarsinylethanol
                         (CH3)2AsO(OH|
                         (CH3)3Ast-CH2COOH
                         (CH3)3As*CH2CH2OH
                         H3As03
                         H,As04
                         (CH3)3AsO
                             Arsenic is also released into the
                      environment  from  human  activities.
                      Primary anthropogenic sources include
                      mining/smelting activities, generation of
                      coal   power,   pesticide
                      manufacture/application,    and   wood
                      preservative  treatment  (Leonard  and
                      Lauwerys  1980).   In  1983  Woolson
                      estimated  tliat  anthropogenic  arsenic
                      sources accounted for  more than three
                      times the natural releases of arsenic into
                      the   environment   (Hindmarsh   and
                      McCurdy  1984),   Three  mechanisms
                      influence  the  fate   of   arsenic  in
                      environmental matrices:  (1) methylation
                      and volatilization,  (2)  adsorption  and
                      precipitation,  and  (3)  oxidation  and
                      reduction  (Hindmarsh  and  McCurdy
                      1984).
SOURCES;  Hindmarsh and McCurdy (1984),
Shibata et al. (1992S
                                Tablt 2 Background Concentration Ranges of Arsenic in
                                                Environmental Media
                              Media
              Concentration     Source
                                             sO.Ol
                                             0.1-40 trig/Kg
       Arsenic
concentrations    in
environmental   media  vary
depending  on  surrounding
anthropogenic   contributors.
Table 2  lists  total  arsenic
concentrations   considered
"background*   levels.
Arsenic  is  the tenth  most
abundant element  in the sea
(JonnaJagadda and Rao 1993)
and   the   twentieth   most
abundant  element  in   the
earth's crust, with a mean concentration of 2 ppm (Leonard and Lauwerys 1980).  Associated with this,
elevated  arsenic concentrations have been found in hot springs in Central America (Lacayo et al.  1992).
Air
Soils
Freshwater

Saltwater
                                               2-5 fjg/L
                                              23x1 0-f M
Hindmarsh 4 McCurdy 1984

Hindmanh 4 MeCurdy 1984

Mahtr & Byttar 1933
Hirtdrrwrsh 4 WeCurtly 1984

Leonaitf «nd Lauweryi 1980
JonralBgadds & Rao 1993

-------
Biological Sources

       Arsenic from environmental media is also circulated through the biosphere through piant uptake
and the food chain.  As with environmental media, biological organisms host a broad range of arsenic
compounds.  Species of arsenic in aquatic environments are reflective of the food  chain's influence as
they include not only inorganic pentavalent and trivalent arsenic, but also their biological metabolites;
monomethylarsonate (MMA) and dimethylarsinite (DMA) (Maher and Butler, 1988).

       Levels of arsenic in terrestrial animals are usually less than 1 ppm-dry weight.  Within the animal
kingdom, fish contain highest total arsenic levels. Studies of Quebec wildlife and marine organisms are
consistent with this generalization (Langlois and Langis 1995). Concentration ranges include:  2.5-4.9
mg/kg in bottom feeding fish (e.g.,  cod, halibut, flounder);  1.2-10.9 mg/kg in crustaceans (including
clams, scallops, lobster and shrimp); and 0.2-0.8 mg/kg in non-bottom-feeding fish (including pickerel,
pike, smelt, whitefish, saJmon, bluefsn tuna and herring) (Hindmarsh and McCurdy 1984).  In  general,
total arsenic concentrations range from 0.1 ppm (in herring) to about 140 ppm (in sole) (Hall et at. 1978).
Arsenic  in aquatic organisms occurs  as both lipid-soluble and water-soluble species (Lunde  1977).

       Marine organisms have the ability to convert inorganic arsenic to organic arsenic species (Lunde
1977). Arsenobetaine is the major form of arsenic  found in fish and shellfish; it is suspected to be the
metabolic endpoint for arsenic in marine environments (Cuilen and Reimer, 1989).  Arsenosugars are the
primary arsenic species found  in algal organisms (Edmonds and Francesconi 1993).  Organic arsenic
compounds  isolated  from marine  organisms  include:   arsenobetaine,  methylated arsenic  acids,
arsenosugars, arsenocholine and unidentified lipid soluble compounds (Irgolic ei al.  1977, Lunde 1977
Shinagawa et al. 1983).

CONSIDERATIONS FOR EXPOSURE ASSESSMENTS

Points of Exposure

       Arsenic, in its  various forms, is present in soil,  air,  water  and biological tissue.   Significant
human exposure pathways are ingestion and inhalation. Inhalation is primarily concerned with respiratory
intake of arsenic from air (as opposed to dermal contact with arsenic in air),  Ingestion occurs from three
sources:   incidental ingestion of soil, ingestion of drinking water, and  ingestion  of foods containing
arsenic.   Human exposure involves both inorganic and organic arsenic compounds; bioavailability is
dependent upon the nature of the specific compound.  For example, arsenic (III)  and arsenic (V) are
absorbed efficiently across the gastrointestinal tract, but to a lesser extent than arsenobetaine.

Contributions from Environmental  Media

       Concentrations  of arsenic have generally been measured as total arsenic and represent a  range of

-------
organic and inorganic arsenic contaminants.  As it has become increasingly apparent that the particular
species of arsenic in an exposure defines its toxic potential and course of action, reported concentrations
are beginning to be detailed by species.

        According to the World Health Organization, normal daily intake of inorganic arsenic is generally
less than 50 jig (Vahter et al,  1983). Total        intake ranges from 0,5-4.2 rag/day, and is
on diet. Two of the most significant contributors to arsenic  ingesdon  are drinking water and seafood.
Soil  arsenic content is  of greater concern in residential  scenarios involving incidental ingestion of
contaminated soil by children. Inhalation exposure to  arsenic usually contributes less than  1 jig/day to
arsenic  intake, but may be of greater concern  in an occupation setting (Leonard and Lauwerys  1980).
                            ...__.                      As noted in table 2,
      Table 3 Percentage of Inorganic Arsenic Compared to
                 Total Arsenic in Selected Foods                      a11    environmental
                                             _                      contain  some  background
                                             Psrcentagt                                  °
    Food                                  inorganic Arsenic          concentrations   of
   ~"*"    '  — — — —   '     j   _-___—__—   ,  — ,                which  may  contribute  to
    Milk and dairy products                       75                                          ,
                                                                    exposure.    However,  the
    Meat- beef and pork                          75                 concentrations   presented
    Poultry                                       65                 indicate total arsenic content
    Fish-- saltwater                               0                  and   do   not  differentiate
                                                                             different   arsenic
      -freshwater                               10
                                                                    compounds.    Due  to  the
    Cereals                                       65                         ,
                                                                    extensive   information
      e                                                             available, a general consensus
    Vegetables                                   0.5                 exists that inorganic arsenic is
    Potatoes                                     10                 harmful   to   people.
    Frujts                                        10                 Therefore, inorganic arsenic
                                                                    content   is   of   particular
                                                                    interest    Table 3 displays
                                                                                of
                                                                    that   inorganic   arsenic
          for various          of foods.  Values       for fish inorganic arsenic content are not widely
accepted; a discussion is          below.

Seafood Exposure Contributions

       Arsenic, because it is ubiquitous in the environment, has been identified in various forms in most
freshwater and saltwater fish, shellfish and seaweed.  Winger et al, (1990), in a study of 102 fish and
fiddler crabs from the lower Savannah River, measured total arsenic concentrations in these organisms.

-------
They found that arsenic does not biomagnify and it does not accumulate in fish to the same extent as it
does in lower trophic organisms. They also found that arsenic concentrations in planktivorous fishes were
elevated over predators and omnivores. Suedel et al. (1994) reiterate Penrose's conclusions that little
evidence exists for food-chain biomagnification of arsenic in marine ecosystems; however, arsenic may
biomagnify in tertiary consumers who do not have the same abilities to convert it to less harmful organic
arsenic species.

       In other studies, Zook et al,  (1976) found that total  arsenic concentrations in different seafood
species varied more than other metal contents and were comparatively elevated above other metals in
some species.  Consistent with these findings, the Group of Experts on Scientific Aspects of Marine
Pollution (GESAMP) data indicate that the dwelling area of seafood will  influence  its arsenic content
(Friberg  1988).  In experiments with flounder and cod from Norway, Staveland (1993) found  that total
arsenic concentrations in fish are essentially independent of age, sex and, for the most part, season.

Species Present

       Species of arsenic present in seafood vary among organisms. Both inorganic and organic arsenic
compounds exist in seafood. In fmfish and shellfish, the principal arsenic species identified  has been
arsenobetaine;  findings of lesser amounts of arsenocholine have also been reported (Edmonds  and
Francesconi 1993, Vahter 1994). Inorganic arsenic has been  found in fish and shellfish at concentrations
ranging from 0-9.5 percent of total arsenic (see tables 4 & 5).  Trace amounts of other species have also
been identified.  Arsenic  species reported in  seaweed are  primarily arsenosugars  (see figure 1)  and
inorganic arsenic, at concentrations ranging from 0-60 percent of total arsenic (see table 6),
   Fish/Species
    Table 4 Speciated Arsenic Concentrations in Finfish
Location     Total     Inorganic    Percent
   of       Arsenic     Arsenic    Inorganic
 Sample    frrtg/kgS*    (mg/kgl*    Arsenic
Source
Anchovy
Barracuda
Chrysophrys
major
Conger
Eel
Flounder
Garfish
Japan
Japan
Japan

Japan
Japan
Japan
Australia
2.33
0.88
1.21

2.88
0.15
7.2
1.3
0
0
0

0.12
0
0
0.01
0
0
0

4.2
0
0
0.8
Kaise et al, 1 988
Kaise et al. 1988
Kaise &t al. 1 988

Mohri et al. 1 990
Mohri et aL 1 990
Shinagawa et al.
Mahsr and Butier






1983
1988

-------
Fish/Species
Goby

Haddock
Hairtail
Halibut .
Herring


Mackerel


N&odiirsma
ransonneti
Op/ggnathus
fasciatus
Panpris tipoma
trillineatum
Pneumatopttorus
japonicus
Prionurus
micmlspido tii3
Salmon
Sardine

Saury
Shark
Siganus
Skate
Sole
Table 4 Speciated Arsenic Concentrations in Finfish
Location Total Inorganic Percent
of Arsenic Arsenic inorganic
Sample (mg/kgl* (mg/kg)' Arsenic Sourc®
Australia
Japan
U.K.
Japan
Japan
Australia
Japan
U.K.
Japan
Japan
Japan
Japan

Japan

Japan

Japan

Japan

Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
0.5
0.74
2.6
1.4
1.91
1.1
1.3
1.1
5.1
1.36
0.32
0.72

9.38

0.62

1.1

0.1

0.7
3.5
4.51
1.1
2.1
0.47
64.05
2.3
0
0
0.02
0.05
0
0.01
0
0.04
0.01
0
0
0

0

0

0

0

0.01
0.06
0
0.04
0.2
0
0
0
0
0
0.8
3.6
0
0.9
0
3.6
0.2
0
0
0

0

0

0

0

1.4
1.7
0
3.6
9.5
0
0
0
Maher and Butler 1 988
Kaise ef al. 1 988
Brooke and Evans 1981
Mohri ef al. 1 990
Kaise ef al. 1988
Maher and Butler 1988
Kaise ef al. 1 988
Brooke and Evans 1 981
Shinagawa ef al. 1 983
Kaise ef al. 1 988
Kaise ef al, 1 988
Kaise ef al. 1 988

Kaise ef al. 1 988

Kaise ef al. 1 988

Shinagawa ef al. 1 983

Kaise ef al. 1 988

Mohri ef al. 1 990
Shinagawa ef al. 1 983
Kaise ef al. 1 988
Shinagawa ef at. 1 983
Yasui 1978
Kaise ef al. 1 988
Mohri ef al. 1 990
Kaise et al. 1988

-------
Tabls 4 Speciated Arsenic Concentrations
Fish/Species
Stephanolepids
cirrhifsr
Stingfish
Stingray
Whiting
Yellowtail
Location
of
Sample
Japan
Japan
Japan
Australia
Japan
Total
Arsenic
(mi/kg)'
4,35
2,9
17,08
2.2
1
Inorganic
Arsenic
(mg/kgl*
0
0
0
0.01
0.03
" Maasursmsnts ar» primarily of muscle tissue (not whole body}
although this is not always plainly noted in the source materials
Percent
Inorganic
Arsenic
0
0
0
0.5
3.0
and assumed to
in Finfish
Source
Kaise et al. 1 988
Mohri et al, 1 990
Mohri ef al. 1 990
Maher and Butler 1
Shinagawa ef al. 1
t>» givan in rng/kg W«t Weight,





988
983



Fish/Species
Abaione
Barnes dilatsts
Clam
Cockle
Crab
Lobster
Table 5
Location
of
Sample
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Speciated
Total
Arsenic
Cmg/kgJ*
2.6
0.2
3,5
6.8
1.72
6.82
2,8
1.91
1.95
3.17
4.6
4.2
29.6
12.4
Arsenic Concentrations
inorganic
Arsenic
(mg/kgl*
0.01
0.19
0.01
0.03
0.05
0
0.07
0
0.05
0.07
0.08
0.06
0.28
0.09
Percent
Inorganic
Arsenic
0.38
95
0.29
0.44
2.91
0
2.5
0
2.56
2,21
1.74
1.43
0.95
0.73
in Shellfish
Source
Shiorni 1984
Mohri 1990
Shinagawa 1983
Shiomi 1984
Mohri 1 990
Kaise er si. 1 988
Shiorni 1984
Kaise ef si. 1 988
Ranjak 1 984
Flanjak 1 984
Maher and Butler 1
Maher and Sutler 1
Ranjak 1 984
Maher and Butler 1






988
988
988

-------
Table S
Location
of
Fish/Species Sampl®
Mollusc



















Mussel


Oyster

Prawn





Scallop


Hon@Kong
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Japan
Australia
Japan
Japan
Japan
Japan
Australia
Australia
Australia
Australia
Japan
U.K.
Japan
Japan
Japan
Speeiated Arsenic Concentrations in Shellfish
Total Inorganic Percent
Arsenic Arsenic Inorganic
Imglkgl* Img/kg}* Arsenic Source
19.4
3
5,1
8.9
33.9
16.8
44,2
36,9
52.2
67,9 ,.
1,02
1.31
1,08
1.64
3.53
17.28
126.92
38.73
123.79
61.61
3
2,4
4.36
4.2
9.S5
6,6
3.17
4.2
8.3
3.6
14
7.2
1.1
1.93
0.3
0
0.07
0.1
0.34
0.02
0.18
0.3
0.57
0.27
0
0
0
0
0
0
0
0
0
0
0,04
0.01
0
0.08
0
0.08
0.07
0.04
0
0,02
0,04
0.04
0.01
0
1,55
0
1.37
1.12
1
0.12
0.41
0.81
1.09
0.4
0
0
0
0
0
0
0
0
0
0
1.33
0.42
0
1.43
0
1.21
2.21
0,95
0
0.56
0,29
0.56
0.91
0
Phillips & Depledge 86
Shlnagawa 1983
Shiomi 1984
Shiomi 1984
Shiomi 1984
Shiomi 1984
Shiomi 1984
Shiomi 1384
Shiomi 1984
Shiomi 1984
Kaise et al. 1 988
et al. 1 988
Kaise et al. 1 988
et a/. 1 988
Kaise et at, 1 988
Kaise et al. 1 988
et at. 1 988
et at, 1 988
Kaise et al. 1 988
Kaise et al. 1 988
Maher and Butler 1988
Shiomi 1984
Kaise et al. 1 988
Shiomi 1984
et al. 1988
Flanjak 1 982
Flanjak 1983
1984
Shinagawa 1 983
Mahur and Butter 1 988
Brook® and Evans 1981
Maher and Butler 1 988
Shiomi 1984
Kaise ef al. 1 988

-------
Table 5 Speciated
Location Total

Fish/Species
Shrimp

'Measurements are
materials.
Of
Samp!®
Japan
Japan
assumed to be

Arsenic
{mg/kg!*
1.9
1.17
Arsenic Concentrations
inorganic Percent
Arsenic
(mg/kg)*
0.03
0
given in mg/kg Wet Weight,


inorganic
Arsenic
1.58
0
although this is

in Shellfish

Source
Maher and Butler 1 988
Mohri 1 990
not always plainly noted in the source




Seaweed/
Species
Brown,
unspecified
Cystoprion
manitoformus
Estnia bicyrlis
Elkonia radiate


Hinkie fusiforme





Laminariu
jttponica


Nmnaeystus
decipiens
Red, unspecified

Table 6
Location
of
Sample
Japan

Australia

Japarv

Australia
Japan

Japan
Japan
Japan
Japan
Japan

Japan
Japan
Japan
Japan

Japan
Japan
Speciated
Total
Arsenic
{mg/kg!*
19.6

11,4

8,9

9
6,1

1.2
41.31
9.3
8.2
6.1

31.21
49.76
2.5
0.14

21
16.5
Arsenic Concentrations
Inorganic
Arsenic
(mg/kg)'
1.9

0.31

0,2

0,36
0.6

0,48
1.47
5
2.3
3,7

0.23
0
0.08
0.08

0.7
4.7
Psreent
Inorganic
Arsenic
9.69

2,72

3,39

4
9.84

40
3.56
53.76
28.05
60.66

0.74
0
3,2
57,14

3.33
2S.4S
in Seaweed


Source
Le et al, 1 994b

Maher and Butler 1988

Yasui 1978

Maher and Butler 1 988
Yasui 1978

Mohri et at. 1 990
Kaise era/. 1988
Yasui 1978
Yasui 1978
Shtnagawa et al. 1 983

Mohri et al. 1 990
Kaise et al. 1 988
Shinagawa et al. 1 983
Mohri era/. 1990

L« et al. 1 994b
Mohri ef al. 1 990
10

-------
TabS® 8 Speciated Arsenic Concentrations in Seaweed

Seaweed/
Species
Sargassum
bractaolosam
Undaria
pinnatifida


Maesuremartts
materials.
Location
of
Sample
Australia


Japan
Japan
Japan
are assumed to b#

Total
Arsenic
Img/kg}"
7.9


1.78
1.6
0.8
given in ing/kg

Inorganic
Arsenic

-------
findings, revealed inorganic arsenic percentages of approximately one percent for organisms with Sow
arsenic concentrations to approximately half a percent for organisms with total arsenic levels approaching
20 mg/kg,  A second linear regression including the Lunde data yielded similar results, none of which
concur with the GESAMP analysis  (Edmonds and Francesconi 1993).

       Currently, specific data for species in question will provide the best quantification of inorganic
arsenic present.  In  general, data in table 4 show inorganic arsenic concentrations in fish to  range from
zero to one percent of total arsenic concentrations, which is in keeping with Edmond's and Francesconi's
analyses.  However, the exceptions of 3.6% inorganic arsenic in saury (Shinagawa et al. 1983, herring
(Brooke and Evans 1981) and hairtail (Mohri et al. 1990), 4.2% in conger (Mohri el al. 1990) and 9.5%
in shark (Yasui et  al. 1978) indicate that Friberg's interpretation of 2-10%  inorganic arsenic is not
entirely incorrect.  Shellfish, because of their bottom-dwelling location and their filter-feeding revealed
slightly higher percentages of inorganic  arsenic.   Data in table 5 range from 0-2.91%, without any
apparent outliers.

       Further investigation into inorganic arsenic  content in shellfish  has been conducted  by Kalman
(1987) who examined human urine following shellfish ingestion.  For three weeks, the entire protein
portion of the diet of human volunteers was replaced with either crab, shrimp, oysters, mussels, clams
or squid. While the laboratory conducting this investigation is located in Seattle, Washington, the source
of this seafood in unspecified; furthermore, how and when urine samples were taken is unspecified. Total
and inorganic urinary arsenic concentrations were analyzed; data were presented from subjects consuming
oysters, mussels, crab and clams.

       Kalman  (1987) calculated percentages inorganic arsenic in the urine, finding a range from 4%
(clam consumption) to 77%  (mussel consumption).   In the control  (no seafood replacement in diet),
inorganic arsenic accounted for over 75%, however, the total urinary arsenic concentration for the control
was over an order  of magnitude lower  than the experimental  subjects.  Kalman's inorganic arsenic
measurements represent a total of Arsenic (III),  arsenic (V), MMA  and DMA (those species detected
from hydride generation preparations).   Since a significant amount  of organic seafood arsenic is not
expected to be converted to  inorganic arsenic via human metabolic  pathways (see  later discussion on
metabolism), these data seem to point at highly elevated levels of  inorganic arsenic in seafood.

       As it has been found that most  ingested organic seafood arsenic  (primarily arsenobetaine) is
excreted in human urine unchanged (see later discussion on metabolism), Kalman's data seem to imply
that a significant portion of the arsenic ingested was  inorganic. However, seafood arsenic content is not
reported  for comparison. Whether or not seafood came from a contaminated source      may also have
influenced type and level of seafood  arsenic content.  Furthermore, other portions of the subjects' diets
may have influenced the amount and  type of ingested arsenic.  Kalman's data confirm a need for further
studies to determine how reflective human urinary excretion data are of seafood arsenic content. These
data are presented in table 7, along with speciated data from Kaise « al, (1988) for comparison purposes.
                                               12

-------

inorganic

Species
(Source
Medium)*
Oyster-
oyster tissue
human urine
Mussel
mussel tissui
human urine
Crab
human urine
Clam
clam tissue**

human urine
Table 7
Arsenic

% A$lMS<
(As** &
As**}

0
—

0
—

..

0
3
«
Comparison of
Percentage of
Total Arsenic
in Shellfish and Human Post-Shellfish-lngestion

i

% MWA

0
--

0
--

-

0
0
"



% DMA

5
--

10
„

--

24
17
-•
% Hydrid®
Generating
As I As,W8,
DMA.MMA!

5
57

10
77

4

24
20
48

Urine
%Tri-
methylated
organic As
species

95
43

90
23

96

76
80
52
    -Indicates no measurement was made.
     'Human urine data from Kalman (1987); seafood tissue data from Kaise et al, (19881.
    "Second set of seafood tissue data for clam from Mohri et a/. (199Q),
Pacific Northwest Specwted Data

       Data on speciation of arsenic in Pacific Northwest seafood has been collected in conjunction with
regionaJ risk assessment reports. Table 8 lists inorganic arsenic content in six free-swimming fish of the
Lower Columbia River,  Percentages ranged  from ,45 to  8,5%, making it difficult to generalize among
species (Tetra Tech 1995),   Data representing organic arsenic content of seafood connected with the
ASARCO Tacoraa Smelter Site (Parametrix 1993, Parametrix 1995) is presented in Table 9, Since these
data were collected for risk assessment purposes in areas subject to possible contamination, it may be
inappropriate to look at these concentrations  in light  of determining  "background" and/or  "safe*
concentrations of inorganic arsenic in fish; however,      data are particularly valuable in advancing an
overall understanding of seafood arsenic content within region  10. The context in which these data were
collected is further discussed in a subsequent section of this report.
                                              13

-------
Measurement and Speciation
Methods

       Background
Information   from
Laboratories

       Fish tissue  can  be
analyzed  for arsenic content
either  via  concentration  of
tissue  sample  to  a  dried
powder or digestion of tissue
sample.   Analysis  of dried
powder   using   x-ray
fluorescence  is  a   multi-
element technique not often
Table
Fish
Species
coho
Chinook
sturgeon
sucker
carp
steelhead
SOURCE:
8 Percentages Asinof9 in Lower Columbia River Fish
Mean AsltW5
Concentration
(00/fl)
0.003
0.013
0.039
0.014
0.001
0.007
Tetra Tech 1 995
Mean As,ot
Concentration
U*S/fl!
0.373
0.960
0.577
0.148
0.221
0.711


Percent Aslrax,
.80
1.4
6,9
8.5
.45
.94

Table
Seafood
Species
striped
seapgreh
(EmbiotocB
lateraiis)
rock sole
mussels
(Mytilus
sp.)
sea
lettuce
9 Arsenic
As+3
Cone.
o.oiu.
O.Q02U-
0.077
0.02-0.05
0.01U-0.04
Concentrations
As+s
Cone,
0.01U
0.002U-
0.082
0.15-0.31
0.19-1.78
in Seafood
MMA
Cone.
0.02U
0.002U-
0.002
0.01 U-
0.02U
0.02U-
0.04U
at ASARCO
DMA
Cone.
(0S/S)'
0.02U
2.2-6.6
0.01 U-
0.02U
0.02U-
0.04U
Tacoma Smalter Site
"Si^g "Swt
Cone. % Cone.
I0g/gl" Aslno,8" 108/81*
0.01 1.9
0,6-8.6
0,17-0.3$ 9.4
0.2-1. 82
      'SOURCE:  Weston (1996}
      "SOURCE: TsujH1993!
employed for measurement of trace contaminants due to its 1-2 ppm detection limit, but it may sometimes
be appropriate for arsenic measurements (personal communication with Dr. Eric Crecelius,  Battelle
Northwest Laboratories, 03/01/96).  Most frequently, however, arsenic speciation in fish tissue is begun
with a digestion.  How aggressive the digestion is will effect the extent to which arsenic compounds will
                                             S4

-------
be  available for  detection.   Total  arsenic measurements mandate a more aggressive digestion than
inorganic arsenic only measurements.  Conditions related to the arsenic-containing medium can effect the
digestion efficiency.   For example,  in order to measure total arsenic in some slags, a particularly
aggressive  digestion was employed  in order to insure release of the arsenic from a silica rind (personal
communication with Dr. J. Lowry,  NEIC lab, 02/27/96).

        Standards can help to assess the recovery efficiency of a particular digestion system.  However,
standards represent optimal  analytical bias (personal communication with J.Lowry, 02/27/96) and for
many species standards are not available. Dr. Lowry suggested that total arsenic standards are available
in an albacore tuna medium and a new arsenobetaine standard in a shellfish medium may also be available
(personal communication, 02/27/96).  Dr. Lowry also discussed some of the limits of hydride generation,
noting that-the arsenic-organic  bond is not necessarily broken, and for detection purposes,  it'must  be
cleaved. Furthermore, he noted that a microwave system is not necessarily an optimaJ approach to total
arsenic detection because the  pressure build-up of a closed  system becomes the rate-limiting step;
however, the newer, focused microwave system allows for settings which  do  not volatilize the arsenic,
thereby enabling operation of an open system (personal communication with J.Lowry, 02/27/96).

        Samples

        Most experiments addressing detection methods have measured arsenic from laboratory distilled
water and other liquid samples, including seawater, freshwater, urine and blood.  Detection of arsenic
from fish tissue has relied upon digestion techniques,  For hydride generation techniques, which do not
speciate among  various  organic   arsenic  compounds,  digestion  techniques  involving  virtually
indistinguishable  breakdown of  organic  compounds  are sufficient.   However,  for more specific
chromatographic techniques, more sophisticated extraction procedures must  be used. As discussed below,
investigations  into more sensitive detection procedures is ongoing; such  extraction methods will depend
on the  needs of specific detection techniques. The importance of developing such techniques is discussed
briefly by Vela et al. (1993), who suggest that supercritical fluid extraction (SFE) may show promise.
Seafood  may also be dried into a powdered form  which can  be analyzed  for total arsenic using x-ray
fluorescence; however, this is a multi-element technique not often  applied to environmental samples due
to its high detection limit, on the order of 1-2 ppm (conversation with E.Crecelius 03/01/96).
       Defection and Spedation

       Historically, arsenic measurement in biota and environmental media have reflected total arsenic
content.  Arsenic from inorganic species has been grouped with arsenic from organic species; arsenic in
water soluble structures  has been grouped with arsenic in iipophilic structures.   Differences in the
physiological effects of various arsenic compounds were just beginning to be articulated by Coulson and
in the earlier  half of the twentieth century (Lunde 1977).  It was not until the nineteen eighties and
                                               15

-------
nineties that reasonable laboratory speciation techniques allowed for such detailed measurements; current
methods, however, are still in need of refinement,

       The most efficient approach to arsenic speciation/detection should be one in which a  single
procedure allows for speciation and identification of each individual arsenic species, both organic and
inorganic.  In keeping with toxicity and regulatory exposure data, detection ability would be in the range
of parts per billion (ppb).  Furthermore, such a method could be employed with environmental samples,
animal tissue samples, and blood and urine samples.  Recent developments by Le et al. (1994b), Ataliah
and Kaiman (1991)  and  Momplaisir et al. (1994) have advanced this line of thinking.  The following
paragraphs discuss the development of their procedures as well as current limitations.

       One of the most common approaches to measuring inorganic arsenic and its human metabolites
is the hydride generation (HG) preparation and  subsequent  detection, usually by atomic absorption
spectrophotornetry (AAS).  Hydride generating methods, while able to detect  arsenic  in the parts per
billion range, are limited.  Two drawbacks have been discussed: (1) HG can cause an  underestimation
of arsenic present because of DMA's slower reaction rate. (2)  HG has a tendency to produce non-linear
standard  curves, leading to overestimated results at low concentrations when linear regression analysis
is used to estimate the calibration equation (Murer et al, 1992b).

       Le et al. (1993) found that addition of cysteine to the reaction will stabilize the DMA alleviating
the  need to significantly  lower pH, which would create non-optimal conditions for the other species.
Murer and colleagues (1992b) suggested  that the use of standard additions and flow injection analysis
(FIA)-AAS appears capable of monitoring long-term, low-dose exposures.  This method has two primary
limitations:  a 20% reduced recovery of arsenate, and inability to address organic arsenic species found
in seafood. Murer et al. (1992b) further recommended use of Larsen's recently optimized Zeeman-AAS
methods.  While detection limits for these techniques are reasonable, they do not speciate and measure
organic arsenic species commonly found in seafood (KaJman 1988, Hanna et al. 1993).

       In order to account for these trimethylated arsenic species (i.e., arsenobetaine and arsenocholine)
a strong  acid digestion,  or other manipulation,  is required prior to hydride generation and subsequent
detection (Le et al.  1993, Le a al.  1994b, Buchet et al. 1994, Mohri et al, 1990,  Kaise et al,  1988).
A common digestion is conducted using nitric, sulfurie and perchloric acids (Fanner and Johnson 1990).
In 1992, Navarro et  al.  introduced a  microwave dissolution in a closed teflon bomb; this technique
accelerated digestion. More recently, Le et al.  (1993)  attempted a microwave oven digestion aided by
potassium persulfate and sodium hydroxide;  this method convened organoarsemeals to arsenate, but
recovery was poor.  Experiments in microwave digestion continue (Sheppard et al,  1994)

       Following digestion of the trimethylated arsenic species, HG preparation may be  used and
concentration may be measured utilizing AAS or atomic emission spectrometry  (AES)  techniques.
Concentration of seafood organic arsenic is then calculated as the difference  in arsenic concentrations
                                               16

-------
detected  in samples with and without the strong acid digestion step.  This method does not allow for
speciation of the trirnethylated compounds.

       A chromatographic system can be employed to separate out arsenic species,  Nixon and Moyer
(1992) found that silica-based cation-exchange cartridges allowed for a faster fractionation over HPLC
columns.  A chromatographic system utilizing inductively coupled plasma (ICP) and AES can detect all
arsenic species without a need for chemical alteration (e.g., hydride generation).  However, the sensitivity'
of such a method is approximately 35 times less than that of the HG-AAS system (Murer et al. 1992a).
Replacing the AES with mass spectrometry (MS) improves sensitivity significantly (Larsen et al. 1993a);
however, this technique has not been widely tested and MS equipment costs may be prohibitive.

       While chromatography  can  assist  in separation  of arsenic compounds, its application  is not
entirely straightforward.  Arsenic (III), arsenic (V), MMA and  DMA can be  separated with an anion
exchange column; but arsenobetaine and  arsenocholine require a cation  exchange column.   Le et al,
(1994b) implemented such a two-column system using one column to separate inorganic species, their
metabolites and arsenosugars, and the second column to separate arsenobetaine and arsenocholine.  Murer
et al. (1992a) found that  running  the two  exchange  columns in series produced arsenic  (V) and
arsenobetaine as  co-eiutants,  Le et  al.  (1994a) avoided  this  problem by using a polymer-based anion
exchange column. Improvement to  HPLC-ICP-MS techniques has been  achieved with the elimination
of chloride interference (Story et al.  1992). Desemay et al. (1994) were able to employ HPLC-ICP-MS
techniques with detection limits of arsenic species ranging from ten to thirty picograms.

       An alternative system has been developed dependent  upon  HPLC separation of arsenic species
followed by conversion of all  arsenic  species to arsenate using  an on-line photo-oxidation system;
subsequent detection is conducted with AAS (Atallah and Kalman 1991).  Atallah and Kalman's method
has only  been shown to work with aqueous solutions  and may be impeded by ammonium ion and urea
interference with photo-oxidation.  Capillary zone  electrophoresis  (CZE) also shows promise for
separation of species (Vela et al, 1993), but it has yet only  been used to separate arsenic III, arsenic V,
MMA and DMA  (Li and Li 1995, Lin et al.  1995).

       Applications of gas  chromatography  (GC)  are discussed by  Kalman (1987) and Kaise et  al.
(1988).   Kalman (1987)  incorporated  a GC  separation  into  the HG-AAS approach,  but it was only
applicable to inorganic arsenic plus  its metabolites.  Kaise et al. (1988)  used GC-MS for detection of
arsenic compounds following alkaline digestion and  subsequent sodium borohydride reduction.  Their
system demonstrated successful separation and detection.  One limitation was that arsenocholine was not
converted to trimethylarsine (TMA) as other trimethylated arsenic species were.   Also, this method
would not be able to detect separate  trimethylated arsenic species as they should all have been digested
and reduced to a single compound, TMA.  Another potential  drawback to this  detection method is that
arsenosugars are converted  to DMA, and  therefore could not be distinguished (Le et al.  1994b,  Mohri
1990).
                                              17

-------
       Momplaisir el a/., (1994) have tried yet another approach to arsenic speciation:  on-line hydride
generation. Using a specially designed silica T-tube, the authors were able to interface HPLC with AAS
detection.   Species were first separated  on a cyanopropyl chromatographic  column  prior to on-line
therrnochemical  hydride generation,  which  enabled  conversion  of all  arsenic species,  including
trimethylated  compounds, to their  respective hydrides.   Both  inorganic arsenic species,  and their
monomethyl   and   dimethyl   metabolites,   in    addition   to   arsenobetaine,   arsenocholine,
phosphorylarsenochoiine and tetramethylarsonium ions, were  completely separated from each other.
Further  investigation and possible refinements to this method  must be executed; however, it appears
promising.

       Current Detecnon/Speciation in the Northwest

       Arsenic speciation data for  both  the Lower Columbia  River study (Tetra Tech 1995) and the
ASARCO Tacoma Smelter Sediment Site (Parametrix 1993, Parametrix 1995) were determined at Battelle
Pacific Northwest Laboratories (conversation with Battelle PI, E.Crecelius 03/01/96).  The method that
Battelle is employing is a difference  method, determining inorganic arsenic content separately from total
arsenic content; organic arsenic is calculated from the difference  in totals.  Total arsenic measurements
are performed using a nitric acid digestion followed by ICP-MS analysis.  Inorganic arsenic detection
requires a milder digestion, and has been accomplished using hot dilute acid (e.g.  2M  HC1 solution) or
base leach (e.g. 2M NaOH solution). This digestion is followed  by the hydride generation-AAS detection
approach which speciates among trivalent and pentavalent inorganic arsenic as well as MMA and DMA
(Battelle 1986).

       The relative  advantages and disadvantages of these methods have been discussed  above.
Depending on the outcome of studies into toxicity of organic arsenic found in seafood and determination
of northwest  seafood  arsenic species (particularly regarding arsenocholine), a method able to  speciate
trimethylated organic arsenic compounds may not be  necessary.  Furthermore, Battelle's method does
allow for  separate  speciation  of  DMA,  the carcinogenic  potential of  which  is  currently under
investigation.  Hence, this method may be sufficient for regional needs.

Exposure Assessment

Sources Assessed
       Table 10 outlines the exposure concerns generally assessed in arsenic exposure assessments. As
indicated by McKone and Daniels (1991), drinking water ingestion and food ingestion are the primary
exposure pathways of concern.  Drinking water contributions to arsenic intake  have been of concern in
locations with elevated arsenic  concentrations in the drinking  water supplies;  this arsenic tends to be
mostly inorganic.  A review of selected studies was conducted by Brown and  Chen (1994).  A  survey
                                               18

-------
of  dietary arsenic
contributions   of     *"ma™ms**^
                             Tabit 10 Human Arsenic Exposure Points Commonly Assessed*
various   food
groupings   was                          Air     Gfound'    Surface      Soi!       Food
                                                      water     Water                Chain
COndUCted  in  SiX     =a=ss=s=ai	i^~Jaii~-J^r^-Jiaaa—r;	an-=,   'is.,, •  -=	SI-TIT.	-	•	aassag-  	.....———-
Canadian cities. Of      inhalation           0,P        :»
dietary intake,  fish      ingestion             ---         P         —        P"         p
contributes    the      Derma! Contact
greatest percentage     -=-=-^--=-.-^==*=- ===:=:==	—^s==^ =, .=:t=        . _	
  ..          ,,  .         0 indicates specific to occupational exposures; P indicates general
of arsenic with the      papulatim exposures.
meat  and  poultry      "Soil ingest/on is most commonly associat&d with childho&d exposure in
grouping  a  distant     ^rMOfn^semarto^^^^^^^^^^^^^^^^^^^^^^
second;   mean
concentration   of
total  arsenic   was
1662 ng/g (an average of 64%) for fish and 24.3 ng/g for mean and  poultry (Dabeka et a/. 1993).
Marine fish contained  more total arsenic than freshwater fish and shellfish arsenic content fell between
the two.

       Consistent results are reported by the USEPA (1988) and Leonard and Lauwerys (1980). USEPA
determined, from FDA studies, that average dietary intake of inorganic arsenic is approximately 50 jig,
approximately  80% of which is accounted for  by the meat, fish and poultry  component  of the diet.
Leonard and Lauwerys report that average daily  oral intake  of total arsenic ranges from 0,5-4.2 mg/day,
depending on diet.  They also  note that  inhalation exposure to  arsenic  has been measured at under 1
fig/day for nonoccupational exposures. Hence,  ingested contributions for the general  population are of
greater concern. Bennett (1981) completed his own study to  determine representative dietary total arsenic
intake from terrestrial foods was about 40 pg/day and from seafood was about 80 ^g/day. Bennet tallied
one year of human ingested arsenic to amount to  a total of 15 mg from terrestrial foods, 33 mg from
aquatic foods and 0.5 mg from drinking water, or 48.5 mg-As/year. This translates to an average of 133
|ig-As/day.

Consumption  of Seafood

       Default Ingestion Rates
       The US EPA draft Exposure Factors Handbook (1995) breaks down fish consumption rates into
four categories:  general  population,  recreational marine anglers, recreational freshwater anglers and
Native American freshwater anglers.  The EPA lists several intake rates for each category, rather than
calculating a combined mean rate because of differences among individual surveys.  Table 11 outlines
                                              19

-------
       Table 11 United States Fish Intake Rates
                           Average*
                            fi/day)
                 95*
              Percent!!®
               !g/day»
general population
 - per capita
general population
 - consumers only
recreational marine
anglers
recreational freshwater
anglers
Native American
freshwater anglers
 11-17"
117-124
37-50*"
  7-24
 63-305
   42

   284

 146-339
 107""
  15-94

170-913 *
the intake rates presented in the Draft
Exposure Factors  Handbook.   It is
noted that recreationai marine anglers
consume more fish than  recreational
freshwater anglers.

       The EPA  recommends  that
whenever possible, consumption rates
for the population(s) involved  in a
given risk assessment be specifically
investigated because data will  vary
with location,  climate,  season  and
ethnicity  of  the  angler/consumer
population^).  If such a  study is not
possible, data from EPA should not
be averaged; data from a study with
similar  characteristics   to   the
population  and location  of concern
should be chosen from those in the
recommendations  section   of  the
Exposure Factors Handbook  (EPA
1995), The EPA also recognizes that     ««»™"^^
consumption of fish internal organs
may  increase  exposure to  certain
contaminants.  Two attributes of fish concentration data should be carefully noted: (1) Is concentration
given in  units of contaminant per gram of fish bodyweight or in units of contaminant per gram of fish
fat content? (2)  Is  the  concentration given in wet weight or dry weight terms? It is important that
consistent units be applied for comparison purposes.  A Canadian study reports that for both freshwater
and marine species, cooked fish contains a higher concentration of arsenic per gram of fish bodyweight,
However, this increase in concentration agreed closely with the resultant decrease in weight from the
cooking process (Dabeka et d.  S993),

       Northwest Seafood and Other Fish

       Seafood and fish consumption  in the  northwestern  United  States  includes a diverse set of
organisms.  Both marine and freshwater systems  provide food to surrounding populations; some diets
include fish from both sources,  some  from only one.  There are  numerous  subpopulations in the
northwest,  each of which consumes  a distinct set of these organisms.  Subpopulations are divided by
ethnicity, profession, and location.  Unfortunately, specific, detailed studies regarding types of seafood
consumed have not  been carried out  for these  northwest populations.  Table  12 lists  marine and
  Average presented is tn arithmetic mean,
 "These valuta nfltet consumption of fish from »H sources
including stars-boughs, canned, seff-csught, etc.
 '"Ths study with t msmn of SO yielded a median of 21,
""The 107 is t 9Q* percent/I*.
  * These percent/las were listed as "upper percsnti/as * and were
not funhw specifi&ti.

SOURCE:  USEPA 1995
       20

-------
freshwater organisms that comprise part of the diet of at least one northwest subpopulation. Local studies
have been  conducted,  around the lower Columbia River and  in  Puget Sound,  to  determine  fish
consumption rates.  Results of these studies are presented in Table 13.
Tabl® 12 Seafood Consumed
Fish/Species
FINFiSH
Albacore tuna
Atlantic cod iGadus morhua]
Bass, striped
Bluefin tuna
Buffalo fish
Cabezon
Carp
Catfish, channel
Chinook salmon
Coho salmon
Eel
Eel, conger
Eulachon jcandlefish)
Filefish (Stephanotepic cirrhifisr]
Flounder
Greenling
Grtenling-- ling cod
Hairtail Ibsitfishl
Halibut
Herring
Japanese striped knifejaw (Oplegnathus
fascfaws)
Largtscale sucker
Pacific cod (Gadus macrocepitalus)
Pacific flounder (Platichthys stellars]
Pacific hake fwhiting)
in the Northwest
Aqueous
Environment

marine
marine
freshwater
marine
freshwater
marine
freshwater
marine
freshwater
freshwater
freshwater
marine
marine
marine
marine
marine
marin®
marine
marine
marine
marina
freshwater
marin®
marine
marine
Source"

local
imported
imported
imported
imported
local
local
local0
local
focal
imported
imported
local
imported
imported,
local
local
local
imported
local
local
imported
local
local
local
loca!
                                              21

-------
Table 12 Seafood Consumed
Fish/Species
Pacific mackerel IScontier japonicus)
Pacific smelt
Pile Perch (Rhacocheifus vacca)
Pmumatophortts jsponicus
Pomfret
Rabbit fish (Sfganus fuscescens)
Rainbow trout (onchorhynchus mykiss)
Rock fish
Sablefish fblack cod)
Salmon
Saury
Shark
Sinx Peron (Embiotocidae]
Skate
Skipjack tuna
Smelt
Sole
Squawfish (PtychocheiSus}
Steelhead trout
Stingray
Suckers (Catastomus)
Tai/Sgabream IChrysophrys major)
Ulapia
White sturgeon
Yellowfin tuna
in the Northwest
Aqueous
Environment
marine
marine
marine
marine
marine
marine
freshwater
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
freshwater
freshwater
marine
freshwater
marine
freshwater
freshwater
marine

Soures*
local
local
local
local
imported
imported
local0
local
local
local0
locai
local,
imported
local
local,
imported
imported
local
local,
imported
local
local
imported
local
imported
localc
local
imported
SHELLFISH/OTHER
Abalone
Clam/cockie
marina
marine
imported0,
local
local,
imported
22

-------
Table 12 Seafood Consumed
Fish/Species
Conch
Crab, blue
Crab, Pacific dungeness
Crab, Alaskan king
Crab, red rock
Crab, tanner
Crayfish
Freshwater prawn {Macrobrachium)
Goose barnacles
Herring roe
Lobster
Mantis shrimp (PseudosQuillal
Moon snail
Mussel
Octopus
Oyster
Saltwater prawn
Scallop
S®a cucumber
Sea urchin
Sea urchin roe
Shrimp
Squid
in the Northwest
Aqueous
Environment
marine
marine
marine
marine
marine
marine
marine,
freshwater
freshwater
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
marine
Source*
imported
imported
local
local
local
imported
local
imported
local
local
imported
imported
local
local,
imported
local,
imported
local,
imported0
local,
imported
local,
imported0
local,
imported
local
local
local,
imported
local,
imported
23

-------
           Fish/Species
Table 12 Seafood Consumed in the Northwest

                               Aqueous
                             Environment
           Source"
            'Local includes sny organisms harvested for human consumption from
            Alaska, Idaho, Oregon or Washington.  Imported implies from another
            state or nation,
            e'Indicates a cultured species

            SOURCE: Consultation  with Dr, C.Michael Watson, USEPA Region 10
            Toxicologist.  November 1995.
Table 13 Northwest Fish Consumption Rates
Study Group
Shore and boat
Arithmetic
Mean
Ig/dsy)
39
50th
Percentile
(g/day)
10
iOth
Percent!!®
Ig/dayJ
78
Source
Pierce st »/. 1981
      anglers - some
      shellfish, no salmon

      Shore anglers - no          55
      shellfish

      Boat Anglers               15

      Squaxin population         54

      Squaxin - shellfish          13
      only

      Lower Columbia           63.2
      River population
                      25


                       5

                      33

                       5


                     38.9
157        Tens Tech 1988


 45         TatraTeoh 1988

161       Toy et a/. 1995 draft

 41       Toy si st. 1395 draft



97.2         CRITFC  1994
       Seaweed Consumption


       Seaweed consumption in the northwest is not weS! defined. Little to no local seaweed is available
at the locaJ markets in the Seattle area (personal visits); hence, most local seaweed of concern is gathered
independently and non-cornmerciaJSy. Currently, an investigation into the eating habits of Asian-Pacific
Islander populations in  the northwest is underway (personal communication from  Dr. R.Lorenzana,
USEPA Region 10 Toxicologist, 11/95).  Results from this study may help to identify northwest seaweed
species of concern. Common forms of marine algae ingested by Japanese populations include Japanese
nori, wakame, konbu and hijiki; scientific classification of these species, respectively, are;  Porphyru
spp., Undaria pinnatifida, Laminaria spp. (mostly Laminariajaponica) and Hizitia-Jusiforme (Edmonds
                                             24

-------
 and Francesconi 1993).  As indicated above, seaweed consumption is of concern because of the significant
 percentage of Inorganic arsenic which it contains (see table 6),  However, this percentage varies greatly
 with species, and therefore, arsenic content of particular northwest species of concern should be assessed.

        In  Washington state, for example, while  absolutely no  commercial  seaweed harvesting is
 permitted,  recreational harvesting  to  a maximum of ten  pounds wet  weight per day  is  allowed.
 Prevention of commercial harvesting is strongly enforced by the state, but enforcement of recreational
 limits is somewhat loose.  Hence, small scale resaJe ventures may occasionally occur (conversation with
 T, Mumford, WDNR, 03/04/96). Tom  Mumford of Washington state's Department of Natural Resources
 (conversation 03/04/96) provided some  information about seaweed species gathered in Washington. Four
 species commonly  gathered and  ingested  are Sargassam muplcum, Alaria  marginata,  Laminaria
 groenlandica and nereo cyspis (also called bulk kelp, a floating variety with a long hollow stalk ending
 in a bulb with blades- blades are ingested portion). The first of these species, the Sargassam, is closely
 related to the Japanese Hizilda, which is known for high arsenic concentrations.  Korean populations in
 Washington are known to gather the Sargassam and pickle it into a spicy, kimchi-like dish, hence further
 knowledge of its speciated arsenic content is of great interest (conversation with T. Mumford,  WDNR,
 03/04/96).

 Potential Tools

        The Exposure Commitment Method was developed by the Monitoring and Assessment Research
 Centre (MARC) to (1) allow for comparisons between exposure pathway contributors, and (2) to estimate
 equilibrium concentrations which will be established based on continued release of a contaminant. Bennet
 (1981) illustrates the application of this method to arsenic.  The following exposure analysis,  based on
 a unit exposure of 1  mg/year, is described:  ingestion of one mg-terrestriaJ foods/year contributes 0.28
 jig-total-arsenic/kg-BW to a person; and ingestion of one mg-seafood/year contributes to  a concentration
 of 0.14 /ig-organic arsenic/year. This exposure commitment method is designed as "a time-independent
 approach to  pollutant assessment"  and measures  both intensity  and  duration of a  contaminant's
 environmental presence.

        Bennett used the model to discover the following.   He found that representative dietary total
 arsenic intake was about 40 |*g/day from terrestrial foods and about 80 |ig/day from seafood.  Greater
than  80% of dissolved Inorganic arsenic is absorbed across the gastrointestinal tract and  organic arsenic
 is "readily" absorbed.  Organic forms of arsenic are found to be retained in the body only about half as
 long  as inorganic forms.  In his analysis,  Bennett assumes  that aquatic intake sources provide organic
 arsenic only, and that a 90% absorption of this arsenic across  the  gastrointestinal tract is expected.
 Inorganic arsenic was determined to have an 8-day residence time and organic arsenic to have a  4-day
residence time.  This exposure commitment method is not readily found in subsequent arsenic literature,
however, it may be a useful tool for risk assessments.
                                               25

-------
       Pathway Exposure Factors (PEFs) were introduced by McKone and Daniels (1991)  to describe
the relationship between environmental concentrations of contaminants and human exposures to those
contaminants,  A  PEF is given in as a lifetime-equivalent chronic daiiy  intake in units of mg/kg-day;
these values are both medium-specific and exposure route-specific (e.g., one PEF may be for ingestion
of soil and another for dermal uptake from water). Arsenic is used as an example.  The authors assume
that arsenic's movement in the environment is governed by its solubility and  its  attachment to mobile
particles.  Major pathways for arsenic exposure are ingestion associated  with  soil-based  pathways and
ingestion connected  with water-based pathways (i.e.,  ingestion of fruits, vegetables, grains, drinking
water and fish).  No distinct values  are provided, but again, this may be a useful tool for arsenic risk
assessments.

CONSIDERATIONS FOR TOXICITY ASSESSMENTS

       The toxicity of arsenic is highly dependent upon its species and its valence state (Leonard and
Lauwerys 1980, Vahter 1994, Edmonds and  Francesconi 1993)  In general, inorganic arsenic has greater
toxicity than organic arsenic; trivalent arsenic is more toxic than pentavalent arsenic.  However, there
are certainly exceptions to this generalization, as well as potential areas for research which could further
refute it.  Arsenic species associated with seafood are fairly representative of the variety, though not the
distribution, present in the overall environment.  Finfish and shellfish primarily  contain trimethylated
arsenic species (i.e., arsenobetaine and arsenocholine), while seaweed tends to concentrate organosugars
(Vahter 1994).  Inorganic arsenic is also present is some seafood, but at  significantly lower levels than
organic arsenic species (Edmonds and Francesconi 1993). Seaweed can be an exception to this as some
forms contain up to 60%  of total arsenic as inorganic arsenic (Shinagawa et al. 1983). How each arsenic
compound is treated by  the body is dependent on many factors.  For  example,  absorption of either
inorganic or organic arsenic species  from the gastrointestinal tract  is partially dependent on the water
solubility of the given compound (Hindmarsh and McCurdy 1984).

Metabolism
Inorganic Arsenic and Methylated Metab<tes

       Inorganic arsenic is absorbed across the gastrointestinal tract, although not as efficiently as many
water soluble organic arsenic species. Inorganic arsenic is excreted  by humans in feces in small, well-
defined amounts (Tarn et al, 1982).  Fecal excretion of ingested soluble forms  of arsenic have been
measured at levels as low as 3.5% for trivalent and 6% for pentavalent arsenicals; however, insoluble
arsenicals have low absorption (Hindmarsh and McCurdy 1984).  Recovery rates of arsenic from human
urine range from 48% to 86% (Johnson and Farmer 1991, Buchet et al.  1994,  Crecelius 1977).

       Johnson and Farmer (199 i) studied  one volunteer consuming single doses of 66 ^g-arsenate/day
for ten consecutive days  and two volunteers each consuming a single dose of 220 p.% arsenate.  For the
tea day study, only 48.2% arsenic was recovered in urine after eighteen days; for  the single dose study,
                                              26

-------
 66.4% arsenic was recovered from urine within seven days.  Interpretation of data should consider that
 different measurement techniques were used to determine the amount of arsenic ingested and  amount
 excreted.  However, this study does raise the question of the validity of single dose studies for estimating
 percentage of dose retained in the body.

        Once absorbed into the bloodstream, arsenate (As(V)) cars be reduced to arsenite (Asflll)), which
 tends to exist in protonated form at  physiological  pH (Vahter 1994).  Arsenic (III) is  methylated to
 monomethy! arsenic (MMA) and dimethyl arsenic (DMA) (Thomas 1994).  Buchet and Lauwerys (1994)
 found that it takes several hours following ingestion of inorganic arsenic for the methylated derivatives
 to be the primary  arsenicals excreted  in the urine.  It has  been suggested that  at a given  exposure dose
 the methylation capacity may become saturated (Thomas 1994, Smith et al. 1992); Buchet and colleagues,
 in 1981, proposed the saturation point to  be at  500 pg/day (Carlson-Lynch et al. 1994).  (Methylation
 saturation is an ongoing debate and is discussed in more detail below.)

        Methylation has been considered the detoxification pathway for inorganic arsenic; however, this
 is currently under  debate (see discussion on DMA toxicity below).  Both methylated and unmethylated
 arsenic  forms are excreted in the urine.  For example,  Hindmarsh and  McCurdy  (1984)  report an
 experiment with humans who had ingested wine rich in arsenite:  50% of the arsenic was excreted as
 dimethylarsinic acid, 14%  was methylarsonic  acid and 8% was  excreted  as inorganic arsenate and
 arsenite.  The mono-  and  di-methylated  forms are excreted more efficiently than the unmetabolized
 inorganic  arsenic (Vahter 1994).   It has also been  proposed that inorganic arsenic detoxification may
 occur via protein binding (Snow,  1992), but extensive discussion of this possibility is  not available.

        In their review, Hindmarsh and McCurdy (1984) discuss the body's retention of inorganic arsenic
 species. Inorganic arsenic, particularly from trivalent exposures, is retained by the body with increasing
 percentages as dose increases.  This may be due to  decreasing methylation efficiency.  At lower doses,
 inorganic arsenic retained by the human body is deposited  in hair, skin and nails.  Studies in mice show
 that  arsenic retained by the body  subsequent  to inorganic arsenic exposure is retained  in  the skin,
 epithelium of the upper gastrointestinal tract, epididymis and the stomach wall.  Trivalent and pentavalent
 forms show similar distribution patterns;  in general, pentavalent arsenic  experiences shorter retention
 times than trivalent arsenic.

       Dimethylarsenic

       In general, DMA is very stable to chemical  degradation and is not decomposed by hydrochloric
or nitric acids, even with  heating; however, microbial demethyiation has  been demonstrated. DMA
 administered to mice, rabbits, and humans has been recovered in urine.  In rats,  significant demethyiation
of administered  DMA has  not been  observed  (Vahter  et  al. 1984).   Vahter  and  colleagues (1984)
performed a study  on the metabolism  of DMA,  as administered orally to mice  and rats.  Eighty percent
of the oral dose of  DMA  was absorbed  from  the  gastrointestinal  tract.  After three days,  mice had
                                               27

-------
eliminated over 99%-of the dose, and rats, about 50%. The rats tended to accumulate the DMA in red
blood cells.   For mice, highest  initial  concentrations were in  the kidneys,  lungs, intestinal raucosa,
stomach and testes.  The organs, in mice, with the longest DMA retention times were lungs, thyroid,
intestinal walls and lens.  A previous study with cows showed DMA accumulation in liver, spleen and
pancreas.  In the Vahter study, no demethylation of the DMA  was observed, although some DMA in
tissues was in a complexed form.

       As an Essential Element

       Various animal studies have suggested that arsenic may be an essential element to the human diet
(Neilsen 1990).  Walkiw and Douglas (1975), Hindmarsh and McCurdy (1984) and Shibata ei al. (1992)
have reviewed many such studies  to conclude that insufficient evidence exists to confirm this hypothesis.
In an experiment with rats, a group without supplemental arsenic in its feed was shown not to grow to
the extent that a group with supplemental arsenic did. However, Hindmarsh and McCurdy caution that
simply because an agent has a positive  effect on growth does not make it essential; the mechanism of
arsenic action as a growth promoter is not yet adequately understood to make such a determination.
Furthermore, in order to be an essential element, the agent must play a unique role, and none  has yet
been  demonstrated  for  arsenic  (Hindmarsh and  McCurdy  1984).   EPA's risk assessment  forum
acknowledges that information to establish this is weak though incomplete incomplete and recommends
that the likelihood  be "weighed qualitatively along with risk assessment information for carcinogenic
effects" (USEPA 1988).

Organic Arsenic Species found in Seafood

       Organic arsenic found  in fish and shellfish has been shown to be efficiently absorbed from the
gastrointestinal tract and to be rapidly excreted in unchanged form (Lunde 1977, Tarn et al. 1982). Yet,
it was not until  1977 that Edmonds el al, isolated and identified the predominant seafood organic arsenic
compound, arsenobetaine,  from rock lobster (Vahter et al, 1983). Coulson's early experiments with rats
in the 1930s demonstrated that a greater percentage of inorganic seafood arsenic (18%) than organic
seafood arsenic (0.7%) was retained over time, three months  in this case (Lunde 1977).

       More recent experiments have confirmed the low retention time of organic arsenic in seafood.
Freeman et al,  (1979) conducted a study on six male volunteers who, in a single day, consumed two
meals containing a total of 5 pg of "seafood arsenic".  After eight to nine days, little arsenic was still
being cleared from the urine of five subjects; and a mean of 77% of the total "seafood arsenic" ingested
had been cleared. In 1982, Tarn et al. calculated similar results: 15 men consumed a single meal  offish
containing 10.1  mg of "seafood arsenic" and, after eight days,  had renally excreted 76% of the total.
Generally, urinary excretion of seafood arsenic ranges from 60-70% in humans (Tarn et al.  1982, Vahter
1994), with some exceptions; Jongen et al. (1985) cite several  reports of at least 80% of arsenic ingested
via human fish consumption being eliminated in urine within a few days.
                                              28

-------
        More extensive studies on the distribution, retention and biotransformation of arsenobetatne and
 arsenochoiine have been conducted by  Vahter,  Marafante and  their colleagues  (Valuer et al.  1983,
 Marafante et a/,  1984)  and  by Christkopoulos et al. (1988a).  Research from both groups supports the
 theory that, in marine  environments, arsenochoiine is a precursor of arsenobetaine.  The Vahter and
 Marafante studies were  performed by exposing mice, rats and rabbits to radio-labeled "As arsenobetaine
 or arsenochoiine, while the  Christkopouios study was performed  with rat liver hepatocytes.  Marafante
 et al, (1984) explore two proposals regarding the metabolic fate of arsenochoiine:  (i) Arsenochoiine is
 incorporated into the phospholipids as choline is,  (2) Arsenochoiine is oxidized to arsenobetaine,

        Rodent exposures to arsenobetaine (Vahter et al. 1983).   BAs arsenobetaine was administered
 both orally and intravenously to mice, rats and rabbits.  As with humans, the primary route of excretion
 was urine.  A very smaJl percentage of the dose was  excreted  in the feces of intravenously exposed
 rodents; and only a smaJl amount was found in the feces of orally exposed mice. Nearly all arsenobetaine
 was excreted by rats and mice within three days, while 74% had been excreted by  rabbits. The latter is
 comparable to human excretion of seafood arsenic (Freeman et at. 1979, Tam et al. 1982). In all cases,
 greater than 99% of the radio-labeled urinary arsenic was in the form  of arsenobetaine; no radio-labeled
 inorganic  or methylated  arsenic acids were found in urine. Clearance of arsenobetaine from most tissues
 in mice appeared to be  described by first order kinetics.  Due to the low retention times, reasons for
 retention,  such as a specific binding, could not be established. Tissues with comparably lengthy retention
 times included testes, epididymis, seminal ducts and seminal  vesicles for ait three rodents,  and also
 muscle in  the rabbits.  These are obviously different retention locations than those for inorganic arsenic.
 The rabbits demonstrated a significantly  longer tissue retention time than the mice.

        Rodent exposures to  arsenochoiine (Marafante et al.  1984). ^As arsenochoiine was administered
 both orally and  intravenously to mice, rats and rabbits.  Nearly all of the arsenocholine administered
 orally was absorbed across  the gastrointestinal tract.   As  with arsenobetaine, the primary  route of
 excretion  was urine; in  all three species, 70-80% of the total  dose was excreted  in  urine within three
days.  Unlike with arsenobetaine, the unmetabolized form of arsenochoiine was only present in urine in
significant amounts on the first day and accounted for only about ten percent of the total dose. Retention
times for  arsenochoiine, as observed over four weeks of study, regardless of the exposure route, were
significantly higher than for  arsenobetaine. Rabbit tissue levels averaged five times,  and rat tissue three
times, that of mice.  After three days, concentration of radio-labeled arsenic in the brain, though low
compared  to other tissues, was three to four times that in the blood. Extractions from mice showed over
90% of water-soluble arsenic to be in the form of arsenobetaine. The tissue distribution of arsenochoiine
was similar to that of arsenobetaine.  The elimination of arsenochoiine appeared dependent on the rate
of oxidation of arsenochoiine to arsenobetaine in tissues; elimination from most tissues appeared to follow
first order kinetics.  Arsenic distribution  throughout the body following uptake of arsenochoiine is likely
reflective  of the distributions of both arsenobetaine and phosphatidy! arsenochoiine (synthesized via
choline kinase and incorporation into phospholipids).
                                                29

-------
       Biotransformation of arsenocholine  (Christkopoulos et al.  1988a).  Rat liver hepatocytes were
incubated separately with arsenobetaine and arsenocholine.   In the  incubation of arsenocholine, the
biotransformation to arsenobetaine occurred via the formation of arsenobetaine aldehyde; and the volatile
compound  liberated  during  arsenocholine oxidation  was  found  to  be  trimethylarsine  (TMA).
Trimethylarsine oxide (TMAO) was formed in a side reaction, presumably from arsenobetaine aldehyde,
during the  oxidation process.   Despite formation of  TMA  and TMAO,  most  arsenochoiine was
metabolized to arsenobetaine; however, experiments with arsenobetaine have not revealed formation of
TMA or TMAO. In humans, arsenobetaine is excreted in the urine; however, trimethylarsine oxide may
be reduced  in vivo to trimethylarsine, one of the more toxic forms of arsenic.

       In Seaweed

       Inorganic arsenic found in marine algae is primarily in the form of arsenosugars comprised of
a pentavalent arsenic atom bonded to two methyl groups, an oxygen atom and a carbon atom of a ribose
sugar (Edmonds and Francesconi 1993). Such dimethylarsenosugars are suspected to undergo degradation
to DMA, but this is still uncertain (Vahter,  1994).   However, there  is a great deal of interindividual
variation regarding the metabolism of arsenosugars, and this metabolism  itself is not well understood.

       Le  et al, (1994b) conducted a single-exposure study of one arsenosugar administered to human
volunteers  who had not ingested any seafood during the  three days prior to  the experiment.  The
arsenosugar was administered during a meal in a serving of nori made from a red algae containing the
arsenosugar.  While urinary excretion of total arsenic  was elevated in seven of the nine volunteers, peak
urinary excretion rates ranged from ten to sixty hours, and two volunteers exhibited no elevation in
urinary total arsenic levels. To confound the matter, four of the volunteers were from the same family,
sharing similar diets and activity levels, yet their results varied; one family member did not even register
a change in urinary arsenic output.

       Algae  have been shown to  synthesize lip id-  and  water-soluble  organic  arsenic species  from
inorganic arsenic in a growth medium (Lunde 1977).  Marine algae also contain a significant percentage
of inorganic arsenic.  In HiziMafusiforme, a Japanese seafood, approximately half of the total arsenic is
inorganic arsenic; other edible forms of seaweed  contain lower amounts of inorganic arsenic (Edmonds
and  Francesconi 1993).   (Discussion of inorganic  arsenic metabolic fate and  toxicity is presented
elsewhere throughout the toxicity assessment section of this document.) Unfortunately, little information
is available regarding arsenic and  marine algae.

Toxiciiy

       Hindmarsh and McCurdy (1984) present a general ordering of the toxicity of various arsenicals:
arsines > arsenites (inorganic trivalent  compounds)  >  arsenoxides (organic trivalent compounds) >
arsenates (inorganic pentavalent compounds)  >  arsonium compounds > metallic arsenic.  This  ordering
                                              30

-------
is in support of the argument put forth by Edmonds and Francesconi (1993):  The valence of arsenic in
a  given compound has a greater bearing on that compound's toxicity than whether it is  organic or
inorganic.

        Clearly, the trivalent arsenic species, including organic compounds, are of greater toxicity than
the pentavalent species. Also apparent in this ranking is the fact that trimethylated arsonium compounds
found  in  seafood (arsenobetaine and arsenocholine)  are  less toxic than inorganic trivalent arsenic
compounds.  Thomas (1994) adds the extent of methylation to the list of general toxicity-gauging criteria.

Inorganic Arsenic and Methylated Metabolites

        Employment of arsenic as a poison predates chemistry textbooks; and the first case linking arsenic
exposure with cancer occurred in the 1880's (Jongen et al.  1985),  Hence, it has been wel!  established
that inorganic arsenic, under  various conditions, is toxic to people.  Several epidemiological studies
linking arsenic with a range of toxic effects have been conducted on populations with high arsenic content
in their drinking water supplies.  Arsenic is associated with a variety of both acute and chronic effects.
Some evidence suggests that diet may  play a role in arsenic toxicity and metabolism.  Nutritional factors
such as intake  of nutrients involved  in methylation may  effect arsenic toxicity (Carlson-Lynch  et al.
1994).

        The trivalent form of inorganic arsenic is commonly associated with  arsenic's toxic effects.
Known suifhydryl reagents, trivaJent arsenicals inhibit several  thiol-dependent enzyme systems in many
different tissues.  In their review, Leonard and Lauwerys (1980) discuss some of the specific actions of
trivalent arsenic.   At  low (on the order of micromolar) concentrations, it inhibits decarboxylation of an
acid essential to the  Krebs cycle.  It can block metabolism at levels not high  enough to impact cell
division. At higher concentrations it can cause mitotic arrest; it effects DNA synthesis and repair.  Since
pentavalent arsenic has a lower affinity for hydroxy and thiol groups, it inhibits fewer enzyme systems.
Consequently, it has a lower toxicity than trivalent arsenic, but it is not non-toxic.  Pentavalent arsenic
inhibits ATP synthesis via uncoupling of oxidative phosphorylation to replace the phosphoryl group
(Leonard and Lauwerys 1980).  The  following paragraphs discuss the toxic effects induced by various
inorganic arsenic compounds.

       Acme Effects

        In a review, Philipp (1985) summarizes the acute effects of arsenic poisoning in three progressive
stages:  (!) nausea, vomiting, diarrhea, inflammation and ulceration of mucous membranes,  and kidney
damage; (2) bloody diarrhea, abdominal pains, thirst, dizziness, dehydration, muscle cramps, cyanosis,
delirium and  convulsions; (3) marked weakness, shock, muscle paralysis,  liver and kidney damage and
death due to circulatory failure.
                                               31

-------
       Chronic Effects:  General

       Arsenic is  also responsible for a great  number of chronic effects in exposed persons.  In his
review,  Philipp (1985)   lists  the  following chronic effects due to  long-term,  low-dose exposure to
inorganic  arsenic:    loss of appetite/weight, diarrhea alternating with  constipation, gastrointestinal
disturbance, cirrhosis of the liver, peripheral neuritis, conjunctivitis, hyperkeratosis and melanosis of the
skin. Gherardi el al. (1990) concur with the link  to peripheral  neuropathy  and also  note that arsenic
appears to cross the blood-brain barrier with ease.

       Hartmann and Speit (1994)  found that sodium arsenate induces DNA damage in white blood cells.
This damage was detected as DNA migration in the single cell gel (SCO) assay. The authors  further
explain that while arsenic has induced DNA strand breaks, sister chromatid exchanges and chromosomal
aberrations, it has not induced detectable gene mutations at specific gene loci.  These  results suggest that
arsenic toxicity is manifested as interference with the replication and/or repair-dependent processes. Such
effects  may be mediated through  a  reaction with  sulfhydryl  groups of tissue proteins and enzymes
(Jonnalagadda  and  Rao 1993).

       Chronic Effects:  Cancer

       Cancer is another chronic  manifestation of arsenic exposure.  There are two primary exposure
sources  linked with  arsenic-induced cancers:   ingestion of contaminated waters and exposure in an
occupational setting (e.g., pesticide manufacture/application and  smelting activities).  Inorganic arsenic
exposure has been  shown to increase human risk of lung cancer (via  inhalation of trivalent arsenic) and
skin cancers (via ingestion of inorganic arsenic  in drinking water) (Fowler et al.  1993).

       In a 1994 review, Brown and Chen summarize the epidemiological studies of regions in Chile,
Argentina, Mexico, and Taiwan to conclude that the occurrence  of dermatological lesions symptomatic
of arsenic exposure occurred  at arsenic concentrations  in drinking water of at  least 0.3 to 0.4 ppm.
Further, no physical manifestations of arsenic toxicity have been reported at drinking water concentrations
of less than 0.2 ppm.

       Evidence  in support of the increased Sung  cancer risk include a recent study of gold miners.
Kabir and  Bilgi  (1993)  identified  a statistically significant increase  in risk  for primary cancer of the
trachea, bronchus  or lung for gold  miners  with  a minimum  of five years of "dusty gold  mining
experience".  Additionally, a fifteen year latency period was found.

       Arsenic is  thought to be  a  co-carcinogen or promoter,  as opposed to an initiator (Fowler 1993,
Hindmarsh and McCurdy, 1984).  The mechanisms of arsenic in careinogenesis are discussed in detail
by both Stohrer (1991) and Snow (1992), Stohrer considers that  arsenic has caused chromosome breaks
in the absence of point mutations and that it has induced gene amplification.  Inorganic arsenic has  also
                                               32

-------
been shown to induce genes in heat shock/stress response; and it has activated herpes simplex and herpes
zoster viruses.  Hence, Stohrer reiterates McDonnell et al.'s 1989 speculation that arsenic may cause
cancer in humans via activation of an oncogenic virus such as the  human papiiloma virus.  Stohrer
supports the potential for  this argument, noting that both  arsenic  and  human papiiloma virus are
associated with rare precancerous skin lesions known as Bowen's disease as well as with cancer of the
epithelial tissues.

        In her review, Snow (1992) summarizes that arsenic exhibits comutagenic, cocytogenic and gene
amplification/expression alteration effects.  Inorganic  arsenic increases  expression of the multi-drug
resistance gene.  Snow attributes arsenic carcinogenesis to two actions: (1) inhibition of essential proteins
leading to alterations in cellular metabolism, and (2) gene amplification  which may cause the expression
of several known cancer-related genes. Arsenic itself is not mutagenic, but it does produce chromosomal
damage in mammalian cells resulting in an increase in  mutagenic response to other agents.  Inorganic
arsenic has also  been found to inhibit enzymes required for oxidative phosphorylation (Snow 1992).  It
has been demonstrated both in vitro and in vivo that trivalent arsenic  exposure is responsible for the
induction of stress proteins, including herne oxygenase  (referenced by Fowler et al. 1993). Of further
note is the fact that arsenic is used as a growth promotor/inducer  for chickens and swine (Vahter 1994).

       The ability of arsenic to induce the mechanistic activities discussed above is partially dependent
on the length of time it  is retained in  a given location  within the body.   Further considering that
detoxification of arsenate involves oxidation to  arsenite followed by two methylation steps,  exposure
duration of internal tissues to various arsenic species is dependent  not only on rate of clearance of a given
species from that tissue, but also on the rate of oxidation/reduction effecting the existence of that species
(Thomas 1994).  The availability of a methyl donor is also critical  to arsenic methyiation; and subsequent
effects on availability of methy! groups for other reactions has not been extensively explored.  Via such
action pathways, arsenic may alter gene expression, leading to pronounced toxic or carcinogenic effects.
Also, saturation  at the step  of conversion of monornethyl arsenic to dimethyl arsenic is supported with
evidence, but not proven (Thomas 1994).  (A more in depth discussion of the debate over methylation
saturation is provided later in this report.)

       Animal Studies

       The  toxicity of various arsenic  compounds  has also been tested on different animal  species.
Shukla and Pandey (1985)  review investigations of the  effects of heavy metals on freshwater  fish: (1)
Arsenic is precipitated in the mucous film formed on  the body of the fish- thus  death  is caused by
suffocation.   (2)  Arsenic in fish gills is at concentrations seven  to eight times the concentration in the
whole fish.  (3)  Breathing distress in fish is caused by clogging of the gills with precipitated mucous and
from heavy metal  ions which lead to anoxia and collapse of blood vessels.  Buhl and Hamilton (1990)
conducted two studies, in clear waters, of placer mining  contaminants, including arsenic. They  found
that arsenic was  less toxic to salmonids than copper, zinc  ajid lead; juveniles were more sensitive than
                                                33

-------
aievins or swim-up fry.  Buhl and Hamilton (1991) also found arsenic to be comparatively  (to other
common placer mining contaminants) moderate in toxicity to arctic grayling, eoho salmon and rainbow
trout.

       Cockell et al, (1991) investigated the toxicity of dietary disodium arsenate to juvenile rainbow
trout. The most prevalent indicator of chronic toxicity was inflammation of the sub-epithelial tissues of
the gallbladder wall.  Other indicators included: decreased  growth rate, mild to moderate anemia,  and
active feed refusal (for higher exposure  levels).  In  comparing the results of this study to those from a
similar study, conducted by Oladimeji el  al., with sodium arsenite, Cockell el al. suggest that pentavalent
arsenate may be as toxic as trivalent arsenite in dietary exposure to juvenile rainbow trout.

       The effects of sodium arsenate on ducklings have been studied by Hoffman et al. (1992). Diets
of 200 ppm sodium arsenate given to day-old mallard ducklings caused some reduction in growth.  Other
ducklings were fed diets with selenium and selenium & arsenic together. Arsenic was found to at least
partially alleviate some of the toxic effects of selenium, supporting the idea that arsenic has antagonistic
effects on the toxicity of selenium.

       Weis et al.  (1992)  investigated the effects of arsenic  contained  in the wood  preservative
chromated-copper-arsenic  (CCA), which is often used  in the construction of pilings and bulkheads in
estuarine environments. A sea urchin fertilization test was conducted; and wood treated with CCA was
found to reduce fertilization by 90% and  to  completely inhibit the development of fertilized larvae.
However, a smaller piece of wood (0.4 cm2)  did not significantly effect fertilization or development.
Snails exposed to CCA treated wood experienced 100% mortality within one week; and in that same time
period, exposed algae underwent chlorosis.

       The toxicity and excretion of inorganic arsenrc  were studied in C57 BL/6J mice  fed 0.5 mg-
sodium  arsenite/kg-BW (Morel  et al.  1995).   Urinary excretion  was measured and interindividual
differences were found with inorganic arsenic and MMA, but not with DMA.  Urea, SAM and creatine
excretions were also measured.  The authors claim that their results imply that interindividual variability
in total  arsenic  metabolite excretion is  reflective of the GSH-dependent redox state. Further, these
experiments support the theory that the intracellular GSH-dependent redox state may be involved in the
first, but not second methylation step.   Reduction from pentavalent arsenic to trivalent arsenic may be
enzymatically catalyzed; and, two different enzyme systems may  be  involved  in the methylation of
inorganic arsenic in mammals.  Hence, the authors propose that the second methylation is not regulated
by the GSH-mediated pathway, so individual differences in.GSH do not effect the production of DMA,
once the MMA is produced.  [These conclusions may be considered in light of the methylation pathway
saturation debate.]
                                               34

-------
       Dimethylarsenic

       Animal data has also been used to suggest an alternative hypothesis for inorganic arsenic toxicity:
that inorganic  arsenic itself is not the toxin.  Methylation may actually create a more toxic compound,
rather than "a detoxified one.  While this theory lacks compelling supporting evidence, brief discussion
is provide by  Thomas (1994),   In his review,  he identifies three toxic effects  attributed to dimethyl
arsenic:  induction of teratologic effects in rats, acute Sung injury in mice and single-strand DNA breaks •
(perhaps  via peroxyl  radical formation).

       Recent findings  suggest that dimethylarsinic  acid  (DMA), the  most common mammalian
metabolite of ingested inorganic arsenic, may be carcinogenic, or at least a  promoter of carcinogenesis.
Due to its more rapid elimination and  lower tissue retention, DMA  has  been thought to be a detoxified
product of inorganic arsenic.  However, in oral ingestion experiments  with rats,  it has  demonstrated
carcinogenic promotional activity in the bladder,  kidneys,  liver and thyroid gland (Yamamoto et al,
1995). As early as  1984, Johansen et al. suggested that DMA may be a promoter, although seemingly
a weaker one than As (III).

       Johansen et al., in unpublished results, found DMA to be as toxic as As(ffl) and As(V) in toxicity
tests, and went on to investigate DMA's carcinogenic  potential (Johansen et al. 1984). Utilizing a two-
stage  liver model  with  rats,  they induced carcinogenesis through administration  of  the  initiator,
diethylnitrosamine(DENA)and subsequently administered DMA to a subgroup of the DEN A-treated rats.
From the 11  DEN A-treated rats, two tumors  were  identified; three tumors were found in the seven
DENA/DMA  exposed rats.  The difference in the  number of tumors formed was not statistically
significant and the authors recommended testing with lower initial DENA doses, so that non-promoted
tumor production might be low enough to observe a significant difference. They did observe an increase
in the number  of liver lesions in the DENA/DMA exposed animals, furthering the  suspicion that DMA
may be a promoter.

       Following the work  of Johansen et al., several studies linking DMA with various types of cellular
damage were published (Endo el al.  1992, Dong et al. 1993, Yamanaka et al. 1991, Murai et al, 1993).
In previous studies,  Yamanaka et al. (1991) had shown that DMA administered  orally to mice induces
lung-specific strand  scissions in  mouse DNA.   In 1991, through study of cellular response to DMA-
induced oxidative damage in the lung, they were able to attribute pulmonary DNA damage to radicals
formed as a result of the metabolism of DMA (Yamanaka et al.  1991).

       In 1992 Endo and colleagues investigated the genotoxicity of nine different organic and inorganic
arsenic compounds.   DMA  alone was found to  induce tetraploids and mitotic arrest.  Many structurally
unrelated compounds,  which all induce  mitotic arrest, are also known to cause  aneuploid and cell
transformation. Since DMA was found to produce mitotic arrest, the authors suggest that it may also
cause  cell transformation.  Furthermore, Endo  et al. postulate that DMA-inhibition of lymphocyte ceil
                                              35

-------
division is likely a direct action.  They also proposed that DMA may be the ultimate arsenic carcinogen,
particularly  when considering effects from inorganic arsenic exposure; and they suggest that tetraploid
induction coyfd be one of the stages responsible for DMA's carcinogenieity,

       Murai and others (1993) investigated the effects of oral administration of DMA to 4-week-old
rats.  DMA exerted dose dependent effects on weight and mortality.  General signs of toxicity included
kidney lesions, papillary necrosis, hyperplasia of the covering epithelium of renal papillae, and proximal
tubular necrosis.  Also  considering that the kidney is a major route of elimination  of inorganic arsenic
and its methylated metabolites, Murai et al. suspected the major cause of death to be renal failure.  Rat
death from  DMA exposure was  accompanied by marked  weight loss and piloerection,  with symptoms
showing up as early as  eight weeks in females and twenty weeks in males.
                                                   3QWks
                                                    J
 I :DEN
• :DMA
                 f :MNU   f :OMH
                 ® : Siscrrtica
                                      ' BBN
!:DHPN
        Drawing upon these and
other results suggesting that DMA
possesses  both  clastogenic  and
genotoxic  properties,  Yamamoto
et d. (1995) investigated DMA's
carcinogenic potential in male rats.
Five  groups of twenty rats were
treated, over four  weeks, with a
DMBDD   multi-organ
cardnogenesis  bioassay (see figure
2) and given no further treatment
for two weeks.  Two groups of
twelve were untreated for these six
weeks.   At week six, four of the
first  groups  were administered
DMA at various concentrations (5,
100,  200, 400  ppm) in  their'
drinking water; the second two groups, at week six, were also administered different concentrations (100
& 400 ppm) of DMA in their drinking water,  DMA was found to significantly enhance tumor induction
in urinary bladder,  kidneys, liver and thyroid gland, in the DMBDD-initiated rats.  Bladder cancers were
markedly enhanced at all concentrations of DMA. Kidney and thyroid cancers were moderately enhanced
in a dose-dependent fashion.  Liver carcinogenesis was strongly enhanced in groups administered  DMA
at concentrations exceeding 200 ppm. Lung and skin carcinogenesis did not appear to be enhanced.

        Yamamoto and colleagues (1995) point out that DMA has clearly demonstrated promoting ability,
yet it is still unclear whether or not it has any initiating activity. In the kidneys of rats treated with DMA
only  at a concentration of  100 ppm, biochemical data indicated significant eel!  proliferation, but the
mechanism for this was not readily apparent.  In the past, Cohen and Ellwien suggested that increased
        Experimental protocol of muiUotgEB cardnogesesis biowwy (DMBDD mod-
 el). Aniraiij were sequentially treated wild DEN (anw. 100 mg/kg body weight, i.p^
 single dose), MNU (f, 20 mg/kg body weight i.p.. 4 limes, on d»y» 5. 8, 11, »nd 14),
 DMH (V, 40 rug/kg body weight. s.e.. 4 times, on dsyi 18, 22, 26. and 30), BBN (•.
 UWr in the drinking water, during weeks 1 ind 2), and DHPN (HI 0.1% in the drinking
 « aicr. dunng weeks 3 and 4), After a 2-week interval, groups 2-5 were given 50. 100.
 ZW. or 400 ppm DMA. respectively, in the drinking water (0), Al! survivors were killed
 3t »eek 30 (»).

SOURCE:  Yamamoto  et al, 1995
   Figure 2 Multiorgan Carcinogenesis Bioassay Protocol
                                                36

-------
 cell proliferation may account for carcinogenicity of nongenotoxic compounds.  Hence, it is possible that
 in  longer-term, lower dose exposures, DMA  may be carcinogenic, that  is,  it may demonstrate both
 initiating and promoting activities,

        Microbiological Experiments

        De Vincente et al, (1990) investigated the effects of heavy  metals on  resistance to antibiotics.
 Marine environments with little fecal pollution were found to harbor the highest frequencies of resistance
 to arsenic; these same environments also had the highest incidence of multi-resistant strains.  Rani and
 Mahadevan (1993) found the  Pseudomonas strain MR1 to have resistance to arsenic, possibly because
 of  its energy-dependent arsenate efflux system.   Of six heavy metals  tested,  the organism was  most
 resistant to arsenic.

 Organic Arsenic and Seafood Arsenic Species

        Organic arsenic species, particularly arsenobetaine, found in  fmfish and shellfish have, since the
 earlier  half  of this  century, been  generally accepted  as  being  nontoxic  (Jongen  et al.  1985).
 Arsenobetaine, while it is efficiently absorbed  across the  gastrointestinal tract, is excreted rapidly and
 unchanged (Brown et al. 1990, Vahter,  1994).  Consequently, little motivation has existed to  investigate
 the toxic potential of humaji exposure to inorganic seafood arsenic; long-term seafood exposure studies
 are not  readily  identifiable.

        At the onset of their group's investigations into the distribution, retention and biotransformation
 of arsenobetaine (Vahter et al. 1983) and  arsenocholine (Marafante et  al. 1984), Vahter et al. (S983)
 caution  that "Although it seems clear that the arsenic compounds in seafood have less acute toxicity than
 inorganic arsenic or methylated arsenicacids...the metabolism and possible effects following long-term
 exposure in mammals are largely unknown."

        Arsenobetaine

        The rodent studies of arsenobetaine yielded little evidence indicative of toxicity. In three days,
over 70% of arsenic dose was excreted unchanged in the urine as arsenobetaine (Vahter et al. 1983).
Toxicity studies conducted with mouse  embryocytes (Sabbioni et al,  199!)  and  rat hepatocytes
(Christkopoulos et al,  1988a)  also did not display  any toxic effects of arsenobetaine.

        Jongen et al. (1985) investigated the toxic  properties of arsenobetaine in vitro. Specifically, the
promoting and initiating capabilities of arsenobetaine were studied; and synergistic/agonistic effects were
 investigated.  At 1 mg-arsenobetaine/plate, no mutagenic effects were found in a bacterial assay with S.
typhimurium, in the sister chromatid exchange  (SCE) assay, or in the assay of forward mutation on the
HGPRT gene.  Differences in cytotoxicity with above assays were surprising; however, differences in
                                               37

-------
exposure may be the cause.   In the HGPRT  forward mutation assay, cells were exposed to the test
compound during pre-incubation, and arsenobetaine appeared to demonstrate a  higher sensitivity to this
pre-incubation than other test compounds..  Initiating and promoting potential were found to be negligible
in the systems tested. No synergistic and/or antagonistic effects were observed in various test systems.
Jongen  et al, (1985)  also note  that other experiments  have confirmed  the low acute  toxicity of
arsenobetaine and arsenochoiine in rats and chicks.

        In 1988 experiments were conducted on rat embryos to investigate the embryotoxicity of both
arsenobetaine and arsenochoiine (Irvin and Irgolic  1988),  Arsenobetaine, in and out of presence of rat
liver homogenate (S-9), demonstrated no embryotoxicity or embryolethality. Results implied that from
several  micrograms to several hundred micrograms of arsenic per gram no subacute or acute prenatal
toxic effects are pronounced.  Neither arsenobetaine nor arsenochoiine was found to impair embryonic
growth  in the absence  of S-9.  No toxic effects  with arsenochoiine  or arsenobetaine occurred with
approximately 20 ng (S.Sfig-As/cm3 for arsenobetaine  bromide and 6.1  ^g-As/cm3 for arsenochoiine
bromide).  Irvin and  Irgolic (1988)  recommend  that  determination of whether or  not microsomal
homogenates  from tissue of pregnant rats metabolize arsenobetaine and arsenochoiine should be  made.

        Arsenochoiine

        Arsenochoiine studies gave more cause for suspicion, but elucidated no overt signs of toxicity.
ArsenochoSine was  retained  longer than arsenobetaine in rodent body  tissues;  it was  metabolically
transformed to arsenobetaine and it was potentially incorporated into phospholipids in the cells.  For these
reasons, Marafante et al. suggest that despite arsenocholine's small contribution to total seafood arsenic,
it may be of greater  toxic concern than arsenobetaine (Marafante ei al, 1984).

        Work by Christkopoulos el al,  (19€8a)  further supports Marafante et al. 's conclusions.  Although
rat  hepatocytes  incubated with arsenochoiine yielded no indications  of cytotoxicity,  or loss  of cell
viability, arsenochoiine metabolism did  produce  some side-products.   As noted under  metabolism
(above), Christkopoulos et al. (1988a) found that during its metabolic transformation to arsenobetaine,
arsenochoiine may form trimethylarsine oxide, which may in turn be reduced in vivo to trimethylarsine,
one of the more toxic forms of arsenic.  Christkopoulos et al. (1988a) then suggest that understanding
the fate of arsenochoiine ingested by man via  seafood consumption may be important to understanding
the potential toxicity of seafood arsenic to humans.

        Francesconi  et al,  (1989) have also  explored  the  biotransformation  of arsenochoiine .to
arsenobetaine. They fed three groups offish  raw cubes of beef dosed with arsenobetaine, beef dosed with
arsenochoiine and undosed beef,  respectively.   Fish dosed with either arsenobetaine or arsenochoiine
retained muscle concentrations of arsenic forty times those in the control group.  37%  of arsenic from
ingested arsenobetaine was retained; while 39% from the  arsenochoiine was retained.   Arsenic  species
were extracted from fish in the arsenochoiine group; and less than 0.2% of the arsenic present was found
                                               38

-------
 in the form of arsenocholine.  The major arsenic fraction of from the extraction was determined to be
 arsenobetaine.

        In Seaweed

        Two medical     studies offer evidence of arsenic poisonings from ingestion of kelp-based health
 supplements (Walkiw and  Douglas 1975),  One patient admitted to  the hospital (for an unconnected
 reason) was found to have elevated urinary arsenic levels, which were eventually attributed to kelp-basal
 health' supplements.  These supplements were found to have a total arsenic content of 27,8 ng/g; one
 caplet contained 20 jig.  Twenty-six days subsequent to  the patient's  discontinuation of the kelp-based
 supplements, urinary arsenic levels no  were no longer elevated.   A  second patient showing signs of
 peripheral neuropathy was found to have elevated urinary arsenic concentrations, also attributed to kelp-
 based health supplements.  Thirty days  following discontinuation of supplements, the patient's urinary
 arsenic levels were one      of the        concentrations, and after ninety days, urinary arsenic levels
 had          to "normal" (Walkiw and Douglas 1975).  Considering the       lengthy
 times, particularly in the         case, and noting the occurrence of  peripheral neuropathy,  it
 likely that a significant percentage of the                 from the  kelp             was inorganic.
 Measurements made were limited to total arsenic, and speciation is not discussed.

        Other Organic Arsenicals

        Philipp (1985) reports that trtmethylarsine, a highly volatile and poisonous form of arsenic, has
 been found to be produced via methylation of inorganic or organic arsenic by fungi.  Inhalation of this
 compound has ended in death.

        Leonard and Lauwerys  (1980) affirm that although organic arsenicals are found to be stored in
the human placenta, they have not been  found to cross the plaeental wall in humans, cat or rabbit.

        Cfaong et al,  (1989) conducted  a study to identify binding sites of organic arsenicals in the
erythrocyte.  Organic           containing a single lipophiiic     or ary! group, such  as lewsite or
                    (PDA), are known to                                   and          the skin,
        the                     they are        primarily by erythrocytes and are readily
to other tissues.  For PDA, glutathione (GSH) was identified as the principal' binding site in erythrocytes.

        Jonnalagadda and Rao (1993) determined that organic          are first metabolized to a trivalent
          before          toxic        and that         toxicity can  be            by dithiols.
                                               39

-------
RISK ASSESSMENT IMPLICATIONS

Regulatory Criteria

National United States Standards
       Numerical Criteria

       The US EPA's Integrated Risk Information System (IRIS) (USEPA 1995) presents exposure
limiting factors under the heading of "inorganic arsenic" considering both arsenic's hazardous properties
and its carcinogenic  properties. For hazardous effects, a critical oral dose of 0,0008 mg/kg-day (derived
from Tseng's NOAEL of 0.009 mg-As/L-water), with an uncertainty factor of 3 is listed with a reference
dose of 3E-4 mg/kg-day. These data are presented with "medium confidence". The inhalation reference
dose has been removed from the database. Arsenic is listed as a group A human carcinogen with an oral
slope factor of 1.5/(mg/kg-day)  and a unit risk of 5E-5/(/tg/L),  Its inhalation  slope factor is 4E-
3/(jtig/rn3).  One in a mil! ion risk levels of 2E-2 jig/L under oral and 2E-4 ^g/m3 under inhalation are
listed in IRIS (USEPA 1995).  A maximum contaminant level (MCL) for drinking water is listed as 0.05
mg/L. Most of these values are being criticized by various parties; criticisms are discussed below. No
regulatory limits are presented for seafood content.

        Criticisms

        Perhaps the  most.widely criticized of the United States regulatory criteria is the drinking  water
MCL. The MCL is accused of by some as being too restrictive and by others as not being adequately
restrictive.  This value of 50 ^ig/L was set by the US Federal Water Quality Administration around  1942,
prior to the existence of the USEPA (Stohrer 1991). This level  is currently under court mandate to be
re-evaluated by the  USEPA.  Stohrer (1991) noted that the US MCL is half that recommended by the
Royal  Commission on Arsenic, and, is four times less than what the Royal Commission views as the
threshold for arsenic toxicity.  Based on Taiwanese epidemiological data, Stohrer placed the threshold
from 100-150 ng/L; a single threshold dose for all toxic effects of arsenic is consistent with the theory
of arsenic being an indirect, gene-inducing carcinogen.

        Conversely, the  MCL Is farther criticized by Smith el al. (1992) who presented results of brief
calculations which claimed that if arsenic in drinking water were at a concentration of 50 Mg/L, and
individuals  consumed one liter  of contaminated water  per  day, based  on linear  extrapolation of
epidemiological data from Taiwan populations, the individual cancer risk may be as high  as 13 in  1000.
They further claimed  that  at the  MCL drinking water  could contribute up  to 100 jig-arsenic per day;
USEPA currently claims that drinking water contributes approximately 5 |*g-arsenie/day.  Such drinking
water supplies could cause arsenic to account for approximately 90% of daily arsenic intake (or 80% for
consumption of only one liter of contaminated water). Smith et al. asserted  that risk of developing lung
and liver cancers based on arsenic exposure at the MCL would increase.
                                              40

-------
       Other scientists have been critical of EPA's unit risk for airborne arsenic.  Viren and Silvers
(1994) evaluated three cohort studies to conclude that the present unit risk may overestimate effects of
airborne arsenic by nearly three times.  The studies evaluated occupational arsenic exposures for smelter
workers.  Results from a northwest study based in Tacoma,  Washington were evaluated, along with new
Swedish data and a previous Montana study.  Unit risk was established based on methods analogous to
those used by  EPA to derive  the previous value  of 4,29xKJ3,   Viren and Silvers advised  that the
relationship between urinary arsenic and airborne arsenic should be clarified in order to better assess the
relationship of this arsenic and various cancers,

       Carlson-Lynch et al. (1994) questioned the USEPA's arsenic oral cancer slope factor which was
established in 1988,  Their primary reservation about the  slope factor was the assumption of a linear
dose-response relationship with data from the Taiwan epidemiological study relating skin cancer to arsenic
ingestion. Such an assumption does not allow for the possibility of a threshold or sublinear dose-response
relationship for low doses of arsenic,   Carlson-Lynch et al,  further questioned USEPA's  use  of the
Taiwan epidemiologicai study out of context.  They pointed out that dietary habits,  including nutritional
deficiencies (such as reduced protein and methionine intake) that may compromise methylation capacity,
were  not  taken into account.  Carlson-Lynch et al, also  claimed that the volume  of drinking water
consumed is not wel! supported; and toxic and carcinogenic potential of other pollutants in the drinking
water, such as fluorescent humic acids  which  are associated  with bladder cancers, was not adequately
addressed.

       An ongoing debate discusses the likelihood of saturation of the arsenic methylation pathway in
humans and its effect on exposure assessment data.  Current USEPA cancer  potency factors do not
assume that a threshold exists in arsenic response for cancer.   Several researchers in the scientific
community exploring this problem have concurred that  such a threshold appears to exist and that a
probable explanation for this threshold is a saturation of the metabolic methylation step (Carlson-Lynch
et al.  1994, Buchet et al. 1994, Vahter 1994, Beck et al.  1995, Thomas 1994).  However, there  also
exists a faction which challenges the methylation saturation theory  (Hopenhayn-Rich  et al. 1993, Smith
et al.  1992, Mushak  and Crocetti 1995).  Based on the analysis of several arsenic ingestion/excretion
studies, Hopenhayn-Rich et  al, (1993) revealed that regardless of dose, the proportion of unmetabolized
inorganic  arsenic  to the sum of unmetabolized inorganic arsenic plus mono- and di-mettiylated arsenic
remains relatively constant across the groups.

       While this conclusion itself does not disprove the methylation saturation theory, it does challenge
supporters of the  theory to  better characterize its mechanisms.  (Other explanations may exist for the
constant ratio,  such  as alternate routes of excretion for unmethylated arsenic.)  In response to  this
challenge, Carlson-Lynch et al. (1994) raised several points.   First, they noted that the average arsenic
exposures in studies used by Hopenhayn-Rich (1993) were below the suggested saturation levels; Buchet
et al,  previously suggested  complete saturation  at doses  exceeding 500 Mg^day,  and the beginnings of
saturation from 200-250 Mg/day.  Carlson-Lynch and colleagues further  contended that use of grab
                                               41

-------
samples (as  in studies utilized by Hopenhayn-Rich et al.  1993)  could  not accurately  represent renal
elimination of the arsenic dose as urinary excretion demonstrated considerable inter individual variation;
a 24 or 48 hour total was recommended.  Beck et al, (1995) submitted the approach of comparing the
ratio of monomethyl arsenic to dimethyl arsenic; they note three studies which considered this ratio and
found it to increase with increasing dose,   Del  Razo et al. (1994) have used this ratio to imply that
metnylation saturation is likely to occur at the second step rather than the first.

       The methylation debate is ongoing (Beck et al.  1995, Smith er d. 1995, Mushak and Crocetti
1995, Thomas  1994), and its implications bear heavily on arsenic risk assessment.  Furthermore, as more
information becomes available on the carcinogenic potential of DMA, it will impact this debate.  Should
DMA itself be found to be carcinogenic,  the methylation pathway  will no  longer be  considered a
detoxification pathway,
                            Tabli 14 National Regulatory Criteria for Seafood Arsenic Content
                        Country              Maximum Permissibl«              Source
                                                Arsenic Content
Australia
Hong Kong
1 mg-AsjIW8/kg-fish
6 mg-Aslot()|/ks-fish
Edmonds & Francesconi 1993
Edmonds & Francosconi 1993
                        Malaysia
                        New Zealand
                        Norway
  10 mg-As,Ma(/kg-shellfish
   1 //g-As^/g-fish-WW
      2 mg-As
4.0 mg-Ast<,!8,/kg-edible-fish-
           WW
        Mat 1994

Edmonds 81 Franoesconi 1933

    Sakulic era/. 1993
Other Standards

       Some
countries   do
provide  regulatory
limits  for  seafood
arsenic   content.
Among  these
countries   are
Australia,   Hong
Kong,    Malaysia,
New Zealand, and     	______	
Norway.   The
maximum
permissible   limits
established by these countries, as depicted in table 14, are provided in different units.  Some countries,
like Australia and New Zealand, focus strictly on inorganic arsenic content, while others present limits
in terms of total arsenic.  Most countries specify limits as per kg of organism, but Norway specifies its
limit in terms of edible fish tissue. Such a restriction may be difficult to establish in the northwest, as
various subpopulations consume different parts of the total organism. Also of note is that Hong Kong's
limits are given separately for fish and shellfish.

       Other agencies have set non-seafood-specific limits.  For  example, die Royal Commission on
Arsenic has set a maximum daily  intake factor for total arsenic of 450 ng/d%y (or 225 j*g/L) based on
consumption of two liters of water  per day.  Additionally, many states within the USA have set their own
drinking  water standards.  These standards range from 7.2 jig/L in Virginia to 190 j*g/L in New York
(Stohrer  1991).
                                              42

-------
Applications

        Currently, seafood-arsenic risk assessments focus on inorganic arsenic  content.  For the most
part, the organic  fraction has been labeled  "non-toxic" and consequently disregarded.   The following
paragraphs illustrate various approaches to risk assessments of human arsenic  exposure from seafood
ingestion.

Fnberg (1988) Study

        Friberg (1988) evaluated the human health effects,  particularly the carcinogenic potential of a
number of metals, including arsenic.  He acknowledged that seafood-mediated arsenic exposure may be
substantial,-but conceded that most seafood arsenic is comprised, of arsenobetaine, a relatively atoxic
compound.  However,  Friberg further  contended that extreme seafood consumption could result  in
significant exposures to inorganic arsenic, and he quantified this exposure to be up to several hundred
micrograms of arsenic per day.  Such an exposure may be connected with a lifetime increased chance of
developing skin cancer.

        Data used by Friberg (1988) were taken from the GESAMP study discussed earlier in this report.
Exposure to inorganic arsenic from seafood ingestion is dependent on fish content  as well as eating habits.
Friberg calculated ingestion rates of arsenic depending on concentrations in fish.  He assumed that
seafood contains two to ten percent inorganic arsenic as percentage of total arsenic.  This assumption was
taken from a World Health Organization 1981 report, Environmental Health Criteria 18 (Arsenic).  (A
challenge to the 2-10% assumption and a discussion of it is presented earlier in this report.) Four seafood
consumption rates corresponding to approximately one meaJ per week, three meals per week, one meal
per day, and an "extreme consumer" were examined in comparison to two different fish inorganic arsenic
concentrations: ten percent of 1.0 pglg in "most commercially important fish species" and five percent
of 10 p,g/g in  bottom feeding fish species.  The  investigation yielded consumption rates of inorganic
arsenic  from fish from  2-500 Mg/day;  specific  results are  shown  in table 15.   Friberg assumed that
bioavailability  of  inorganic arsenic from fish was equivalent to that of drinking water.

       Friberg (198S) compared  the risk  from  arsenic exposure  via seafood  consumption to cancer
responses presented in the WHO document, which estimated a five percent increase in arsenic-induced
skin cancer from  10 g-As/lifetime, or a daily intake of 0.4 mg inorganic arsenic  over a lifetime, or a
daily intake of one milligram inorganic arsenic over 25 years.  Hence, except for the extreme consumer,
Friberg concludes that fish consumption alone should not pose a significant increase in cancer risk; but
it may impact significantly upon total arsenic ingestion, and consequently contribute to raised cancer risk.
Even considering  the possibility that Friberg's values were overestimated, there exists a strong possibility
that for extreme consumers, seafood arsenic  exposure may contribute to an incremental increase in cancer
risk.
                                               43

-------
         Table 15 Friberg's Daily Inorganic Arsenic Intake from Seafood for Four Consumer Levels
                                                           As,    Consumption  Rates (jig/day?
Mean
Seafood Aswt-
Concentration
U/Q/Q)
most commercially 1
important fish
bottom-feeding 1 0
fish/crustaceans
Percentage Seafood Consumption Rates (g/day)
Aslnor8 in
Fish 20 60 150 1000
10 2 6 15 100

5 10 30 75 500

SOURCE: Fribsrg 1988
Puget Sound Seafood Risk Assessment (Tetra Tech 1988)
Tabli 16 Puget Sound Saafoad Consumption Rates
Organism
fish
shellfish
kelp
nori
SOURCE: Tetra
Average
Consumption
{g/day}
12.3
1,1
0.006
4.1
Tech 1 988
High-end
Consumption
!g/day!
95.1
21,5
0.023
16.2

       In  1988   a   human   health  risk
assessment was conducted in  consideration
of chemical contamination found  in Puget
Sound  seafood (Tetra Tech  1988).   The
assessment focused soleiy on  contributions
to  risk   from   ingestion   of   seafood.
Consumption rates, displayed  in table  16,
were  derived from  various  local  data.
Consumption rates for  fish were  derived
from studies  on Puget Sound recreational
anglers' catch and consumption  practices,
Median shellfish  rate was taken  from a
regional study conducted  by Nash in 1971; high' exposure to shellfish was based on an estimate of one
meal of 151 grams per week. Algal consumption rates were based on marketing information from a local
marieulture facility that specialized in nori.  Median consumption values were based on United States
habits and high end values were based on Japanese consumptions.

       Arsenic was evaluated in terms of carcinogenic risk.   The assessment considered exposures over
a 70 year lifetime using a cancer potency factor of 1.5, the interim recommendation of the USEPA Risk
Assessment Forum. Total arsenic concentrations were measured, and the assumption was made that one
                                              44

-------
Table 17 Cancer
Organism
fish
shellfish
kelp
nori
SOURCE: Tetra
Risk from ingestion
Total Arsenic
Concentration
(pg/kg-WW)
3441
3252
2559
2279
Tech 1988
of As from Puget
Average
Exposure
Cancer Bisk
9E-6
8E-7
1E-8
3E-9

Sound Seafood
High Exposure
Cancer Risk
8E-5
2E-5
1E-5
3E-6

                                                                          percent of total arsenic
                                                                          was in  the  inorganic
                                                                          form.  No citation was
                                                                          made regarding the one
                                                                          percent   assumption,
                                                                          Risk   results   are
                                                                          presented in  table 17, •
                                                                          High  end  exposures
                                                                          indicated  carcinogenic
                                                                          risks  exceeded   the
                                                                          standard  one  in  a
                                                                          million risk.  Risks for
                                                                          seaweed exposure may
                                                                          be  underestimated due
to the assumption of one percent inorganic .arsenic. The differences in inorganic arsenic content among
fish, shellfish and macroalgae is not extensively discussed.

Kensington Mine Risk Assessment (PTI, 1995)

        Potential human health  risks associated with arsenic releases from  the mining outfall  were
evaluated by the mine's owners in conjunction with the permitting process for a gold mining project at
Kensington Mine in Alaska.  Since the intended outfall location was located in seawater, the only human
arsenic exposure pathway of concern  was that of seafood ingestion. Best case assumptions regarding the
pre-discharge treatment of outfall predicted arsenic concentrations in the outfall would  be lower than
receiving seawater background concentrations (average of 1.45 /ig/L); however, worst case assumptions
were analyzed,  allowing for  an arsenic discharge concentration of up  to 3
       Standard  adult residential  exposure  parameters  were used  as  appropriate, leaving the fish
consumption and fractional intake parameters to be site-specific selections.  The amount of arsenic that
the discharge would contribute to seafood was calculated based on an estimated bioconcentration factor
(BCF) of 640 for salmon and halibut, and 1,550 for crab.  BCF values were found to vary significantly
with species and laboratory /field conditions.  Those chosen were in conservative ranges. Based on the
Edmonds and  Francesconi (1993)  evaluation of GESAMP data, and the Health  Risk Assessment of
Chemical Contamination in Puget Sound Seafood (PSEP 1988), a fraction of one percent of total arsenic
was used to estimate inorganic arsenic levels in all three species considered.

       Seafood consumption rates  were based on surveys, conducted by the Alaska Department of Fish
and Game Division of Consumption, of two  local communities which utilized part of the proposed outfall
area as fishing locations.  Total catch  was tallied for each group and the total usable pounds of fish or
crab harvested for home use by the population of the applicable community was divided by the population
                                               45

-------
of that  community.   This  average
consumption value was then doubled.      _ .,  ,„-,.„        .    n     .   „
                                        Table 18 Seafood Consumption Rates for Kensington Mine
Table 18 depicts derived  consumption                            Communities
rates.  The fraction  of  this seafood      Community            Fish lngestion     Crab |ngestjon
which would be contaminated by the                                (kg/day}           Skg/dayS
mining outfall based on the fraction                     	===:'-~—-"             -	-	-  	-1-	
                                        Haines                     0.158           0.0047
of the fishing area  impacted by the
outfall.     The  mixing  zone  was      Hoonah                    °'446           0-0085
determined to be no more than 0.002      SOURCE: PTI, 1995
percent of the fishing waters for both
the fish and the crabs.

       In choosing toxicity data for comparison with exposure data, the assessment presents a brief
review of studies supporting the threshold theory.  Following consultation with relevant scientists, the
cancer slope factor was recalculated using the exposure assumptions used by the EPA to determine the
RfD.  This reduced the slope factor from  1.75 (mg/kg-day)"1 to 1.13 (mg/kg-day)'1.  The EPA's current
oral RfD of 0.0004 mg/kg-day was used; however a recalculated RfD of 8x10"" mg/kg-day was also
applied to illustrate the impact of re-evaluating the Taiwanese epidemiological data to account for arsenic
contributions made by yams and rice. (An alternate slope factor accounting for such  a re-evaluation was
calculated to be 0.77 (mg/kg-day)"1). Both carcinogenic risks and non-carcinogenic risks  were evaluated
using the factors indicated above.  Risk contributions from arsenic outfaS! ranged from hazard indices of
0.099-0.23 and incremental cancer risks of 2xlO"n-8xlO'10.
Lower Columbia River Bi-state Program (Tetra Tech  1995 draft)

       As part of the Lower  Columbia River Bi-state Program, a report was prepared regarding the
human health risks associated with ingestion of fish from the river.  Of the many associated chemicals
of concern, arsenic was among those whose contributions to health risks  were evaluated.  Contaminant
concentrations were measured in three resident species (largescale sucker,  carp and white sturgeon) and
three non-resident species (steelhead, chinook and coho).  Risks from arsenic exposure were evaluated
for three populations: the general public, recreational  anglers and subsistence anglers.

       Arsenic concentrations were speciated into total and inorganic  concentrations. Total arsenic was
measured using ICP-MS  techniques and inorganic arsenic (plus methylated metabolites) was measured
via HG-AAS.  Samples for arsenic speciation were digested with sodium hydroxide,  which would not
decompose the organic arsenic  species, to allow for detection of inorganic arsenic. Arsenic was detected
in all six target  species.  Table 8  presents percentages of inorganic arsenic in each species.  These
percentages are elevated above Edmonds and Francesconi (1993) estimates.
                                              46

-------
       Because arsenic is a carcinogen, chemical specific chronic daily intakes were calculated over both
a 30-year and a 70-year exposure duration.  Consumption rates used were 6.5 g/day for the general
population, 54 g/day for the recreational anglers and 59 g/day for the subsistence angiers.  The first two
rates are based on EPA data and the third is based on a study conducted by the Columbia River Inter-
Tribal  Fish Commission in 1994,   Standard assessment parameters were applied for  most values,
however,  an  arithmetic average, rather than some sort of maximum value, was used  for the exposure
point concentration (in the edible fish tissue). Risk evaluations were made using EPA's oral RfD of 3E~4
and oral slope factor of 1.75.  Risk estimates are presented in table 19.  It should be noted that several
exceed a one in a million cancer risk.  For an upcoming draft of this report, risks were evaluated for
ingestion  rales ranging from nearly zero to  300 g/day; risks were found to  increase proportionally
(personal  communication with  Dana Davoii, USEPA, 03/25/96).
     Population      Typ® of
       fFish         Risk
Table 19 Risk Contribution from Arsenic in Lower Columbia River Fish

                         Arsanlc Contributions to Risk fby species)
Consumption
Rat® (fl/dayj)
General Public
(6.5)
Recreational
Anglers (54)
Subsistence
Anglers (591
General Public
(6.5)
Recreational
Anglers (54)
Subsistence
Anglers (59)
General Public
(8.5)
Recreational
Anglers (54)
Subsistence
Anglers (53)
(Exposure
DurattonJ
cancer
(30 year)
cancer
(30 year)
cancer
(30 year}
cancer
(70 year)
cancer
(70 year)
cancer
(70 year)
hazard
(30 year)
hazard
(30 year)
hazard
(30 year!
Carp
6.68E-08
5.55E-07
6.06E-07
1.83E-07
1 .3BE-OS
1 .4S1-06
3.10E-04
2.57E-03
2.81 E-O3
Chinook
8.57E-07
7.12E-06
7.78E-08
2.03E-08
1 .73E-05
1 .89E-G5
3.97E-03
3.30E-02
3. 61 £-02
Coho
1.78E-07
1 .481-06
1.626-06
4.33E-07
3.SOE-06
3.S3E-06
8.25E-04
6.86E-03
7.49E-03
Sucker
8.35E-07
6.93E-06
7.58E-06
2.03E-08
1.69E-05
1 .84i-05
3.87E-03
3.21E-02
3.51 E-Q2
Stasihaad
4.34E-07
3.61E-06
3.941-06
1.06E-06
8.78E-OS
3.59E-Oi
2.01 E-03
1.67E-02
1 .83E-02
Sturgeon
2.61 E-OS
2.17E-05
2.37E-OS
6.35E-06
5.28E-05
5.76E-05
1.21E-02
1.01E-01
1.10E-01
   Note: Risk values in bold exceed the one in a million cancer risk of the hazard index of 1.

   SOURCE:  Tetra Tech  1995
                                               47

-------
CONCLUSIONS    -

       Unquestionably, the science behind arsenic risk assessment has advanced considerably over the
past fifteen years.  The primary species present in seafood was identified as arsenobetaine.  Speciation
techniques now allow not only for quantification of organic  arsenic species, but also for speciation of
them.  The database quantifying inorganic arsenic  contributions to total seafood arsenic concentrations
has grown.  Mechanisms of human arsenic detoxification have been better characterized,  Toxicity of
seafood organic arsenic species has been further elucidated.  Yet, despite these advances, there is still
much to be accomplished and understood.

       A single value for seafood inorganic arsenic content  cannot adequately represent the variety of
inorganic arsenic concentrations present in the various seafood consumed.  Freshwater  and  saltwater
organisms vary in content; bottom-dwelling fish and fish residing higher in the water column vary in
arsenic content; free-swimming fish, shellfish and macroalgae also vary in arsenic content.  Furthermore,
broad agreement for a given type of organism has not been  established; because of inherent variations
as well as the ongoing debate surrounding inorganic arsenic content, it is difficult to determine which
percentage is  most appropriate.   Not only would such  a choice be controversial, it would require
establishment  of other dietary habits of northwesterners as well as approximation of drinking water
arsenic content, both of which influence arsenic intake.  Both vary greatly depending on the location and
ethnicity of the population(s) in question.  Such values must  be measured on a site-specific basis.

       Options exist among speciation methods for making such measurements.  Since current leanings
are towards inorganic arsenic being the only species of concern in finfish and shellfish, methods which
will separate  individual  organic  arsenic species may be unnecessary.  Such an assumption allows for
employment of methods involving destructive digestion of organic species to allow for hydride generation
and subsequent detection.  However, should arsenocholine be identified in northwest fish and shellfish,
methods  which can speciate among  organic  arsenic species will be of greater concern.    For  such
measurements, the ICP-MS detection methods when used with ionic separations, currently is the most
well-developed option.   Other methods discussed above, such  as HPLC-AAS with on-line hydride
generation also show promise, but further improvements to these methods are needed.

       More confident determinations must be made regarding the potential content of arsenocholine in
northwest seafood,  and  particularly  in  shellfish.  Once  it has  been  established  whether or  not
arsenocholine  is present, the need for its inclusion in northwest arsenic risk assessments can be assessed
and appropriate speciation and measurement techniques can be identified.  The percentage of inorganic
arsenic in northwest seafood must also be better characterized.  While a conservative percentage often
can be used to make worst case evaluations, such evaluations may inflate actual risks by a full order of
magnitude. Data from the Columbia River study indicate that even at lower percentages, inorganic arsenic
from seafood can pose and/or contributed a health threat.  Therefore, measurements should be collected.
A separate project to more specifically identify species of concern and to measure arsenic in these species
                                               48

-------
could be conducted.  However,  It may be  more appropriate to measure arsenic species of particular
concern  in conjunction with those concerns,  and to  maintain  a  growing  database of species and
concentrations.

       Resolution should be reached as to whether or not DMA is carcinogenic.  If DMA is determined
to be carcinogenic, methylation  can no longer considered to be  a detoxification pathway in humans.
Furthermore, evidence leading to the methyiation saturation theory should be re-evaluated to assess other
possible  conclusions.   Most  importantly, should DMA be determined to  be carcinogenic, speciated
measurements among inorganic  arsenic, MMA and  DMA content should be made when  measuring
seafood arsenic concentrations for risk assessments.
 * ACKNOWLEDGEMENT Credit must be extended to Michael Garry for completing the groundwork
for this report.  Before ESAT began the project,  Mr. Garry had  gathered the first set of references,
 identified most of the available speciated seafood arsenic concentrations and assembled an initial report,
 As instructed by the WUD, ESAT employed this information as a base to assist in completion of its report.
                                               49

-------
REFERENCES
Ahmed, Farid E., Dale Hattis, Richard E. Wolke and David Steirunan.  1993.  Risk Assessment and
       Management of Chemical Contaminants in Fishery Products Consumed in the USA. Journal
       of Applied Toxicology 13(6):  395-410.

Atallah, RJ. and D.A. Kalman.  1991.  On-line Photo-oxidation for the Determination of
       Organoarsenic Compounds by Atomic-Absorption Spectrometry with Continuous Arsine
       Generation,  Talama 38(2):  167-173,

Battelie, Pacific Northwest Laboratories.  1986.  Speciation of Selenium and Arsenic in Natural
       Waters and Sediments, Volume 2:  Arsenic Speciation, Final Report.  Sequim,  WA:
       Prepared for Electric Power Research Institute:  EA-4641, Vol.25  Research Project 2020-2.

Beck, B.D., P.D. Boardman, G.C. Hook, R.A. Rudel,  T.M. Slayton and H. Carlson-Lynch.   1995.
       Response to Smith et al.  Environmental Health Perspectives 103(1):  15-17.

Bennett, B.G.  1981.  Exposure of Man to Environmental Arsenic - An Exposure Commitment
       Assessment. The Science of the Total Environment 20:  99-107.

Brooke, P.J. and W.H, Evans,  1981. Determination of Total Inorganic Arsenic in Fish, Shellfish
       and Fish Products.  Analyst 106:  514-520.

Brown, Kenneth G, and Chien-Jen Chen.  1994. Observed Dose-Response  for Internal Cancers and
       Arsenic in Drinking Water in the Blackfoot Disease Endemic Region of Taiwan.  In Arsenic:
       Exposure and Health,  W.R.  Chappell,  C.O. Abernathy and C.R. Cothern, Eds.  Northworth:
       Science and Technology Letters, pp. 153-170.

Brown, R.M., D. Newton, CJ. Pickford and J.C. Sherlock.  1990. Human Metabolism of
       Arsenobetaine Ingested with Fish.  Hum Exp Toxicol 9: 41-6.

Buchet, J.P., J. Pauwels, R, Lauwerys.  1994.  Assessment of Exposure to  Inorganic Arsenic
       following Ingestion of Marine Organisms by Volunteers. Environ Res 66(1): 44-51.

Buchet, J.P. and R. Lauwerys.  1994. Inorganic Arsenic Metabolism in Humans.  In Arsenic:
       Exposure and Health.  W.R.  Chappell,  C.O, Abernathy and C.R. Cothern, Eds.  Northworth:
       Science and Technology Letters, pp. 181-190.

Buhl, Kevin J. and Steven J. Hamilton.  1991.  Relative Sensitivity of Early Life Stages of Arctic
       Grayling, Coho Salmon,  and  Rainbow Trout to  Nine Inorganics.  Ecoioxicology and
       Environmental Safety 22:  184-197.

Buhl, Kevin J. and Steven J. Hamilton.  1990.  Comparative Toxicity of Inorganic Contaminants
       Released by Placer Mining to Early Life Stages of Salmonids. Ecotoxicology and
       Environmental Safety 20:  325-342.

Carlson-Lynch, Heather,  Barbara D.  Beck and Pamela D. Boardman.  1994.  Arsenic Risk
       Assessment. Environmental Health Perspectives 102(4): 354-356.
                                             50

-------
Chong, Steven, Kilian Dill and Evelyn McGown. 1989,  The Interaction of Phenyidichloroarsine
       with Erythrocytes. Journal of Biochemical Toxicology 4(1):  39-45.

Christkopoulos, Alexandras, Harold Norin, Margareta Sandstrom, Hjordis Thor, Peter Moldeus and
       Ragnar Ryhage.  1988a.  Cellular Metabolism of Arsenocholine,  Journal of Applied
       Toxicology 8(2):  119-127,

Christkopoulos, A., B. Hamasur, H, Norin and I. Nordgren.  1988b,  Quantitative Determination of
       Arsenocholine and Actylarsenocholine in Aquatic Organisms Using Pyrolysis and GC/MS.
       Biomed Environ Mass Spec IS:  67-74.

Cockell, K.A., J,W. Hilton and W.J. Bettger.  1991.  Chronic Toxicity of Dietary Disodium
       Arsenate Hepatahydrate to Juvenile Rainbow Trout (Oncorhynchus mykiss). Archives of
       Environmental Contamination and Toxicology 2li  518-527,

Columbia River Inter-Tribal  Fish Commission (CRJTFC),  1994.  A Fish Consumption Survey of the
       Umatilta, Nez Perce, Yakarna and Warm Springs tribes of the Columbia River Basin.
       Technical Report 94-3.  Portland, OR:  CRITFC.

Crecelius, E.A.  1977.  Changes in the Chemical Speciation of Arsenic Following Ingestion by Man.
       Env Health Persp 19:  147-150.

Cullen, W.R.  and K.J. Reimer.  1989.  Arsenic Speciation in the Environment. Chem Rev 89:  713-
       764,

De Vicente, A., M. Aviles, J.C. Codina, J.J.  Borrego and P. Romero.  1990.  Resistance to
       Antibiotics and Heavy Metais of Pseudomonas aeruginosa Isolated from Natural Waters,
       Journal of Applied Bacteriology 68:  625-632.

Dabeka, Robert W,,  Arthur D. McKenzie, Gladys M.A, Lacroix, ChantaJ Cleroux, Susan Bowe,
       Russell A. Graham, Henry B.S. Conacher and Pam Verdier.   1993.  Survey of Arsenic in
       Total Diet Food composites and Estimation of the Dietary Intake of Arsenic by Canadian
       Adults and Children, Journal of AOAC Intemationan6(l):  14-25.

Del Razo L, Luz Maria, 1994.  Urinary Excretion of Arsenic Species in a Human Population
       Chronically Exposed to Arsenic via Drinking Water: A Pilot Study. In Arsenic: Exposure
       and Health.  W.R. Chappell, C.O. Aberaathy and C.R. Cothern, Eds. Northworth:  Science
       and Technology Letters, pp. 91-100.

Demesmay, C., M. OSSe and M. Porthault.  1994.  Arsenic Speciation by  Coupling High-performance
       Liquid Chromatography with Inductively Coupled Plasma Mass Spectrometry.  Fresnius'
       Journal of Analytical Chemistry 348:   205-210.

Dong, Jin-Tang and Xian-Mao Luo. 1993.  Arsenic-induced DNA-Strand Breaks associated with
       DNA-Protein Crosslinks in Human Fetal Lung Fibroblasts.  Mutation Research 302:  97-102.

Edmonds, John S. and Kevin A.  Francesconi.  1993.  Arsenic in Seafood:  Human Health Aspects
       and Regulations. Marine Pollution Bulletin 26(12): 665-674,
                                             51

-------
Edmonds, J.S., Y. Shibata, K.A. Francesconi, J. Yoshinaga and M. Morita.  1992. Arsenic Lipids
       in the Digestive Gland of the Western Rock Lobster Panulirus cygnus:  An Investigation by
       HPLC ICP-MS.  The Science of the Total Environment 122; 321-335.

Endo, Ginji, Koichi Kuroda, Akiyoshi Okamoto  and Shun'ichi Horiguchi.  1992.  Dimethylarsenic
       Acid Induces Tetraploids in Chinese Hamster Cells.  Bulletin of Environmental Contamination
       and Toxicology 48:  131-137.

Farmer, J.G.  and L.R. Johnson. 1990.  Assessment of Occupational Exposure to Inorganic Arsenic
       Based on Urinary Concentrations and Speciation of Arsenic. British Journal of Industrial
       Medicine 47:  342-348.

Flanjak, Johann.  1982.  Inorganic and Organic Arsenic in Some Commercial East Australian
       Crustacea. Journal of the Science of Food Agriculture 33:  579-583.

Fowler, Bruce A., Hiroshi Yamauchi, E.A. Conner and M. Akkerman.  1993.  Cancer Risks for
       Humans from Exposure to the Semiconductor Metals. Scand J Work Environ Health
       19(suppl.l):  101-103.

Francesconi, Kevin A., John S. Edmonds and Robert V. Stick.  1989. Accumulation of Arsenic in
       Yelloweye Mullet (Aldrichetta forsteri) following Oral Administration of Organoarsenic
       Compounds and Arsenate.  The Science of the Total Environment 19: 59-67.

Freeman, H.C., J.F. Uthe, R.B. Fleming, P.H.  Odense, R.G. Ackman,  G. Landry and C. Musial.
       1979. Clearance of Arsenic Ingested by Man from Arsenic Contaminated Fish.  Bulletin of
       Environmental Contamination and Toxicology 22;  224-229,

Friberg, Lars. 1988. The GESAMP Evaluation of Potentially Harmful Substances in Fish and Other
       Seafood with Special Reference to Carcinogenic Substances, Aquatic Toxicology 11: 379-
       393.

Gherardi, Remain K., Patrick  Chariot, Muriel Vanderstigel, Denis Malapert, Josette Verroust, Alain
       Astier, Christian Brun-Buisson and Annette Schaeffer. 1990. Organic Arsenic-induced
       Guillain-Barre-like Syndrome due to Melarsoprol:  A Clinical, Eleetrophysiological and
       Pathological Study.  Muscle & Nerve, pp. 637-645.

Hall, R.A., E.G. Zook, G.M. Meabura,  1978.  National  Marine Fisheries Service Survey of Trace
       Elements in the Fishery Resource.  NOAA Technical Report NMFA SSRF-721.

Hanna, C.P., J.F. Tyson and S. Mclntosh.  1993. Determination of Inorganic Arsenic and  Its
       Organic Metabolites in Urine by Flow-Injection Hydride Generation Atomic Absorption
       Speetrometry.  Clln Chem 39(8):  1662-1667.

Hartmann, Andreas and Gunter Speit, 1994, Comparative Investigations of the Genotoxic Effects of
       Metals in the Single Cell Gel (SCG) Assay and the Sister Chromatid Exchange (SCE) Test.
       Environmental and Molecular Mutagenesis 23:  299-305.

Hedlund, B.,  H. Norin, A. Christkopoulos, et al.  1982.   Acetyiarsenocholine:  A Cholinergic
       Agonist.  / Neurochem 39(3): 871-873.
                                             52

-------
 Hindmarsh, J.Thomas and Ross F, McCurdy,  1984.  CiinicaJ and Environmental Aspects of Arsenic
       Toxicity.  CRC Critical Reviews in Clinical Laboratory Sciences 23(4); 315-347.

 Hoffman, David J., Caroline J. Sanderson, Leonard J, LeCaptain, Eugene Cromartie and Grey W.
       Pendleton.   1992.  Interactive Effects of Arsenate, Selenium , and Dietary Protein on
       Survival, Growth, and Physiology in Mallard  Ducklings.  Arch, Environ. Contam. Toxicol
       22: 55-62.

 Hopenhayn-Rich, C., A.M. Smith and H.M, Goeden,   1993.  Human Studies do not Support the
       Methylation Threshold Hypothesis for the Toxicity of Inorganic Arsenic. Env Res 60'  161-
       177.

 Hughes, James P.,  Lincoln Polissar and Gerald van Belle,  1988.  Evaluation and Synthesis of Health
       Effects Studies of Communities Surrounding Arsenic Producing Industries.   International
       Journal of Epidemiology 17(2): 407-413.

 Irvin, T.Rick  and Kurt J. Irgolic.  !988.  Arsenobetaine and Arsenocholine: Two Marine Arsenic
       Compounds without Embryotoxicity, Applied Grganometallic Chemistry 2:  509-514.

Johansen, M.G., J.P. McGowan, S.H. Tu and D.Y. Shirachl  1984. Tumorigenic Effect of
       DimethySarsinic Acid in the Rat.  Proceedings of the Western Pharmacological Society 27;
       289-291.

Johnson,  L.R. and J,G. Farmer. 1991.  Use of Human Metabolic Studies and Urinary Arsenic
       Speciation in Assessing Arsenic Exposure. Bull Environ Contam Toxicol 46:  53-61.

Jongen, W.M.F., J.M. CardinaaJs and P.M.J, Bos.  1985.  Genotoxicity Testing of Arsenobetaine,
       the Predominant form of Arsenic in Marine Fishery Products.  Food and Chemical Toxicology
       23(7):  669-673.

Jonnalagadda, S.B., P.V.V. Prsada Rao.   1993.  Mini Review: Toxicity, Bioavailability and  Metal
       Speciation.   Camp, Biochem. Physiol. 106C(3):  585-595.

Kabir,  Habib  and Chandra Bilgi.  1993. Ontario Gold Miners with Lung Cancer. JOM 35(12):
       1203-1207.

Kaise, T., K.  Hanaoka, S. Tagawa, et al.  1988.  Distribution of Inorganic Arsenic and  Methylated
       Arsenic in Marine Organisms.  App Qrganomet Chem 2:  539-546,

Kalman, David A., James Hughes, Gerald van Belle,  Thomas Burbacher, Doug Bolgiano, Kimberly
       Coble, N.Karle Mottet and Lincoln Polissar.  1990,  The Effect of Variable Environmental
       Arsenic  Contamination on Urinary Concentrations of Arsenic Species.  Environmental Health
       Perspectives 89;  145-151.

KaJman, D.A.  1988.  Quantitation of Arsenic Species in Urine for Exposure Assessment Studies. J
       Res Nat Bureau Standards 93:  315-318.

Kalman, D.A.  1987.  Dietary Contributions to Arsenic Species in Urine.  In: Trace Elements in
       Human Health and Disease.  Edited by:  P. Granjean.  Environmental Health 20, WHO,
       Copenhagen.


                                             53

-------
Lacayo, Martha L., Adela Cruz, Sandra Calero, Julio Lacayo and Inge Fomsgaard,  1992. Total
       Arsenic in Water, Fish and Sediments from Lake XoSotian, Manasgua, Nicaragua,  Bull,
       Environ, Comam. ToxicoL 49;  463-470.

Langlois, Claude and Rene Langis.  1995.  Presence of Airborne Contaminants in the Wildlife of
       Northern Quebec.  The Science of the Total Environment 160/161:  391-402.

Larsen, Erik H., Gunnar Pritzl and Steen Honore Hansen.  !993a. Arsenic Speciation in Seafood
       Samples with Emphasis on Minor Constituents:  An Investigation Using High-performance
       Liquid Chromatography With Detection by Inductively Coupled Plasma Mass Spectrometry.
       Journal of Analytical Atomic Spectrometry 8:  1075-1084.

Larsen, Erik H,, Gunnar Pritzl and Steen Honore Hansen.  1993b. Speciation of Eight Arsenic
       Compounds in Human Urine by High-performance Liquid Chromatography with Inductively
       coupled Plasma Mass Spectrometric Detection Using Antimonate for Internal
       Chromatographic Standardization.  Journal of Analytical Atomic Spectrometry 8: 557-563.

Le, Xiao-Chun, William R. Cullen and Kenneth I. Reimer.  1994a, Speciation of Arsenic
       Compounds by HPLC with Hydride Generation Atomic Absorption Spectrometry and
       Inductively Coupled Plasma Mass Spectrometry Detection. Talanta 41(4):  495-502.

Le, X-C, W.R. Cullen and K.J. Reimer.  1994b.  Human Urinary Arsenic Excretion after One-time
       Ingestion of Seaweed, Crab and Shrimp.  Clin Chem 40(4):  617-624.

Le, Sean X.C., William R.Cullen and Kenneth J. Reimer.  1994c. Speciation of Arsenic Compounds
       in Some Marine Organisms. Environ. Sci. Technol.  28:  1598-1604,

Le, Xiao-Chen, William R. Cullen and Kenneth J. Reimer.  1993. Determination of Urinary Arsenic
       and Impact of Dietary Arsenic Intake.  Talanta 40(2):  185-193.

Leonard, A. and R.R. Lauwerys. 1980.  Carcinogenicity, Teratogenicity and  Mutagenicity of
       Arsenic.  Mutation Research 75: 49-62.

Li, Ke and  Sam F.Y. Li.  1995. Speciation of Selenium and Arsenic Compounds in Natural Waters
       by Capillary Zone Electrophoresis  after On-column Preconcentration with Field-amplified
       Injection.  Analyst 120: 361-366.

Lin, Li, Jiansheng Wang and Joseph Caruso.  1995.  Arsenic Speciation Using Capillary Zone
       Electrophoresis with Indirect Ultraviolet Detection,  Journal of Chromatographic Science 33;
       177-432.

Lunde, G.  1977. Occurrence and Transformation of Arsenic in the Marine Environment.
       Environmental Health Perspectives 19:  47-52.

Lunde, Gulbrand. 1973,  Separation and Analysis of Organic-bound and Inorganic Arsenic in Marine
       Organisms. Journal of the Science of Food Agriculture 24: 1021 -1027,

Lunde, Gulbrand. 1969.  Water Soluble Arseno-organlc Compounds in Marine Fishes.  Nature 224.
                                             54

-------
 Maher, W, and E, Butler.  1988. Arsenic in the Marine Environment.  App. Organometal, Chem. 2;
        191-214.

 Maher, W.A.  1985. The Presence of Arsenobetaine in Marine Animals. Comp. Biochem, Physiol.
        80C(I):  199-201.

 Marafante, Erminio, Marie Vahter and Lennart Dencker.  1984.  Metabolism of Arsenocholine in
        Mice, Rats and  Rabbits.  The Science of the Total Environment 34:  223-240.

 Mat, L 1994. Arsenic and Trace Metals in Commercially Important Bivalves, Anadara granosa and
        Paphia undulata. Bull Environ, Contain.  Toxicol.  52:  833-839.

 McKone, Thomas E. and Jeffrey I. Daniels.  1991. Estimating Human Exposure through Multiple
        Pathways from Air, Water and Soil. Regulatory Toxicology and Pharmacology 13:  36-61.

 Mohri, T., A. Hisanaga, N. Ishinishi. 1990.  Arsenic Intake and Excretion  by Japanese Adults:  A
        7-Day Duplicate Diet Study.  Fd Chem Toxic 28(7): 521-529.

 Momplaisir, G-M.  T. Lei and W.D. Marshall.  1994. Performance of a Novel Silica T-tube
        Interface for the AAS Detection of Arsenic and Selenium Compounds in HPLC Column
        Eluate.  Anal Chem 66:  3533-3539.

 Morel, G., J.L. Cluet, P. Teloahy, H.M. Yang, N. Thieffry and J. de Ceaurriz.  1995.
        Interindividual Variability in the  Urinary Excretion of Inorganic Arsenic Metabolites by C57
        BL/6J Mice: Possible Involvement of a Thiol/Disulfide Exchange Mechanism.  Toxicology
        Letters 78:   111-117,

 Murai, Takashi, Hiroyuki Iwata, Tatsuyuki Otoshi, Ginji  Endo, Shun'ichi Horiguchi and Shoji
        Fukushima.  1993.  Renal Lesions Induced in F244/DuCrj Rats by 4-weeks Oral
        Administration of Dimethylarsinic Acid. Toxicology Letters 66:  53-61.

 Murer, A.J.L, A.  Abildtrup, O.M. Poulson and J.M. Christensen.  I992a.   Effect of Seafood
        Consumption on the Urinary Level of Total Hydride-Generating Arsenic Compounds.
        Instability of Arsenobetaine and Arsenocholine. Analyst 117:  677-680.

 Murer, Arm J.L., Anne Abildtrup, Otto M. PouSsen and Jytte Molin Christensen. 1992b.  Estimation
        of the Method Evaluation Function for the Determination of Hydride-Generating Arsenic
        Compounds in Urine by Flow-Injection Atomic-Absorption Spectrometry. Talanta 39(5):
        469-474.

 Mushak, Paul and Annemarie F. Crocetti.  1995.  Risk and Revisionism in Arsenic Cancer  Risk
        Assessment. Environmental Health Perspectives 103(7-8):  684-689.

•Navarro, M., H, Lopez, M.C. Lopez  and M. Sanchez.  1992. Determination of Arsenic in Fish by
        Hydride Generation Atomic Absorption Spectrometry.  Journal of Analytical Toxicology 16:
        169-171.

 Nielsen, Forrest H.  1990.  New Essential Trace Elements for the Life Sciences.  In: Biological
        Trace Element Research.  Edited by:  G.N. Schrauzer. Publisher: The Humana Press, Inc.
                                             55

-------
                                                                                                      I

Nielson, Kirk K,, Arthur W. Mahoney and Vern C. Rogers.  1991, Occurrence of Arsenic in                  \
       Seafood from Fast Foods Analyzed by X-Ray Fluorescence, Journal of Food Composition
       and Analysis 4:  285-292.                                                                        I
Nixon, David E. and Thomas P.  Moyer.  1992.  Arsenic Analysis II:  Rapid Separation and
       Quantification of Inorganic Arsenic Plus Metabolites and Arsenobetaine from Urine.  Clin.
       Chem.  38(12):  2479-2483.
I
Parametrix, Inc.  1995.  Phase 1 Data Report and Phase 2 Sampling and Analysis Approach:
       Expanded Remedial Investigation and Feasibility Study ASARCO Sediments Superfund Site,             1
       Final. Kirkland, WA.  September 1995.

Parametrix, Inc.  1993.  Site-Specific Residue Levels in Edible Seafood at the ASARCO Tacoma                1
       Smelter Site, Draft Data Report.

PTI Environmental Services.  1995.  Risk Estimates for Consumption of Seafood Collected Near the             I
       Proposed Kensington Mine Outfall.                                                                *

Philipp, Robin. 1985.  Arsenic Exposure: Health Effects and the Risk of Cancer.  Reviews on                 I
       Environmental HealthV(1); 27-57.                                                               I

Phillips, D.J.H.  1990,  Arsenic in Aquatic Organisms: A Review, Emphasizing Chemical                     |
       Speciation. Aquatic Tox. 16: 151-186.                                                            I

Phillips, D.J.H. and M.H. Depledge.  1986.  Distribution of Inorganic and Total Arsenic in Tissues             |
       of the Marine Gastropod Hemifusus ternatanus. Mar Ecol Prog Ser 34: 261-266.                      \

Pierce, R.S., D.T. Noviello, S.H.  Rogers.  1981.  Commencement Bay Seafood Consumption                  .
       Report. Preliminary Report. Tacoma, WA:  Tacoma-Pierce County Health Department.                I

Rani, D.B. Rajini and A. Mahadevan.  1993.  Patterns of Heavy Metal Resistance in Marine                   _
       Pseudomonas MR1.  Indian Journal of Experimental Biology 31:  682-686.                            I
Sabbioni, E., M. Fiscbach, G. Pozzi, et al.  1991. Cellular Retention, Toxicity and Carcinogenic
       Potential of Seafood Arsenic. I. Lack of Cytotoxicity and Transforming Activity of
       Arsenobetaine in the BALB/3T3 Cell Line,  Carcinogenesis 12(7):  1287-1291.                         I

Sekulic, Bogdan, Jasenka Sapunar and Davorin Bazulic. 1991,  Arsenic in Norway Lobster
       (Nephrophs norvegicus L.) from Kvarneric Bay—Northeastern Adriatic.  Bull, Environ                  §
       Contam. Toxicol. 51:  460-463.                                                                   *

Sheppard, Brenda S. Douglas T. Heitkemper and Cynthia M, Gastoo.  1994.  Microwave Digestion             I
       for the Determination of Arsenic, Cadmium and Lead in Seafood Products by Inductively               I
       Coupled Plasma Atomic Emission and Mass Spectrometry,  Analyst 119: 1683-1686.

Sheppard, Brenda S,, Joseph A. Caruso, Douglas T. Heitkemper and Karen A.  Wolnik.  1992.                 I
       Arsenic Speciation by Ion Chromatography with Inductively Coupled Plasma Mass
       Spectrometric Detection. Analyst 117:  971-975.                                                    g

Shibata, Yasuyuki, Masatoshi Morita and Keiichiro Fuwa.   1992.  Selenium and Arsenic in Biology:
       Their Chemical  Forms and Biological Functions, Adv.  Biophys, 28: 31-80.                           •


                                             56

-------
 Shinagawa, A,, K. Shiorai, H. Yamanaka and T. Kikuchi. 1983.  Selective Determination of
        Inorganic Arsenic (III), (V), and Organic Arsenic in Marine Organisms,  Bull J Soc Sci Fish
        49(1): 75-78,

 Shlomi, K., A. Shinagawa, T. Igarasht, et al.  1984. Contents and Chemical  Forms of Arsenic in
        Shellfishes in Connection with their Feeding Habits. Bull J Soc Set Fish 50(2):  293-297.

 Shukla, N.P.  and G.N. Pandey,  1985,  Effect of Heavy Metals on Fish - A Review.  Reviews on
        Environmental Health V(l):  87-99.

 Smith, Allan  H,, Mary Lou Biggs, Claudia Hopenhayn-Rich and David KaJman.  1995. Arsenic Risk
        Assessment.  Environmental Health Perspectives 103(1):  15-17.

 Smith, Allan  H., Claudia Hopenhayn-Rich, Michael N.  Bates,  Helen M. Goeden, Irva Hertz-
        Picciotto, Heather M. Duggan, Rose Wood, Michael J. Kosnett and Martyn T. Smith.   1992.
        Cancer Risks from Arsenic  in Drinking Water.  Environmental Health Perspectives 97:  259-
        267.

 Snow, Elizabeth T.  1992.   Metal Carcinogenesis:  Mechanistic Implications.  Pharac, Ther. S3: 31-
        65.

 Stary, J., K. Kratzer, J. .Prasilova and T. Vrbska.  1982.  The Cummulation of Chromium and
        Arsenic Species in  Fish (Poecilia reticulata).  International Journal of Environmental
       Analytical Chemistry 12:  253-257.

 Staveland, G., II  Marthinsen, G. Norheim and D. Julshamn.  1993.  Levels of Environmental
        Pollutants in Flounder (Plaiichthys fleus L.) and Cod (Gadus morhua L.)  Caught in the
       Waterway of Glomrna, Norway.  II.  Mercury and Arsenic.  Arch. Environ,  Comam,  ToxicoL
       24: 187-193.

 Stohrer, Gerhard.  1991.  Arsenic:  Opportunity for Risk Assessment.  Archives of Toxicology 65:
       525-531.

 Story, W.Charles, Joseph A. Caruso, Douglas T, Heitkemper and Larry Perkins.  1992. Elimination
       of the Chloride Interference on the Determination of Arsenic Using Hydride Generation
       Inductively Coupled Plasma Mass Spectrometry, Journal of Chromatographic Science 30:
       427-432,

Suedel,  B.C., J.A. Boraczek, R.K.  Peddicord, P.A. Clifford and T.M. Dillon.  1994.  Trophic
       Transfer and Bioraagnification Potential of Contaminants in Aquatic Ecosystems. Reviews of
       Environmental Contamination and Toxicology 136: 21-89.

Tarn, G.K.H., S.M. Charbonneau, F. Bryce and E. Sandi. 1982.  Excretion of a Single Dose of
       Fish-Arsenic in Man.  Bulletin of Environmental Contamination and Toxicology 28:  669-673.

Tamaki, S. and W.T. Frankenberger Jr.  1992.. Environmental Biochemistry of Arsenic.  Reviews of
       Environmental Contamination and Toxicology 124: 79-110.
                                             57

-------
Tetra Tech.  1995,  Draft Report;  Assessing Human Health Risks from Chemically Contaminated
       Fish in the Lower Columbia River:  Risk Screening.  Lower Columbia River Bi-state
       Program. TC 9968.05.

Tetra Tech 1988. Health Risk Assessment of Chemical Contamination in Puget Sound Seafood.
       Puget Sound Estuary Program.

Thomas, David J.  1994. Arsenic Toxicity in Humans: Research Problems and Prospects.
       Environmental Geochemistry and Health 16(3/4):  107-111.

Toy, K.A., G.D. Gawne-Mittelstaedt, N.L.  Polissar and S. Liao. 1995. A Fish Consumption
       Survey of the Tulalip and Squaxin Island Tribes of the Puget Sound Region, DRAFT.
       Tulalip Tribes, Natural Resources Department, 7615 Totem Beach Road, Marysville, WA
       98271.

Tseng, W.P., H.M. Chu, S.W.  How, et al.  1968.  J Nad Cancer Insi 40:  453-463.

Tsuji, Joyce S, at Kle'mfelder, Inc.  1993. Memorandum dated Aupst 30,  3993, to Mr. Michael R.
       Thorp at Heller, Ehrman, White and McAuliffe.

United States Environmental Protection  Agency (USEPA),  1995. Online Integrated Risk Information
       System.

United States Environmental Protection  Agency (USEPA).  1989. Exposure Factors Handbook.
       Internal Review Draft.  National Center for Environmental Assessment.  Washington D.C.
       NCEA-W-005.

United States Environmental Protection  Agency (USEPA).  1988. Special Report on Ingested
       Inorganic Arsenic:  Skin Cancer; Nutritional Essentiality. Risk Assessment Forum:
       Washington, D.C.  EPA/625/3-87/013.

United States Environmental Protection  Agency (USEPA).  1984.  Health Assessment Document for
       Inorganic Arsenic.  Research Triangle Park, NC.

Vahter, Marie.  1994.  Review:  Species Differences in the Metabolism of Arsenic Compounds.
       Applied Organometallic Chemistry 8:  175-182.

Vahter, Marie, Erminio Marafante and Lennart Dencker.  19S4, Tissue Distribution and Retention of
       74As-Dimethylarsinic Acid in Mice and Rats. Archives of Environmental Contamination and
       Toxicology 13:  259-264.

Vahter, Marie, Erminio Marafante and Lennart Dencker.  1983. Metabolism of Arsenobetaine in
       Mice, Rats and Rabbits.  The Science of the Total Environment 30:  197-211.

Vela, Nohora P., Lisa K. Olsen and Joseph  A. Caruso. 1993, Elemental Speciation with Plasma
       Mass Spectrometry. Analytical Chemistry 65(13):  585A-597A.

Viren, John R. and Abraham Silvers.  1994. Unit Risk Estimates for Airborne Arsenic Exposure:
       An Updated View Based on Recent Data from Two Copper Smelter Cohorts.  Regulatory
       Toxicology and Pharmacology 20:  125-138.


                                            58

-------
WaJkiw, 0, and D.E. Douglas,  1975,  Health Food Supplements Prepared from Ke!p»A Source of
       Elevated Urinary Arsenic.  Clinical Toxicology 8(3):  325-331.

Weis, Peddrick, Judith S, Weis, Arthur Greenberg and Thomas J, Nosker,  1992.  Toxicity of
       Construction Materials in the Marine Environment: A Comparison of Chromated-Copper-
       Arsenate-Treated Wood and Recycled Plastic. Archives of Environmental Contamination and
       Toxicology 22: 99-106.

Weston.  1996.  ASARCO Sediments Site Human Health Risk Evaluation: Proposed Approach for
       Evaluating Seafood Ingestion.  Seattle, WA.  January 24,  1996.

Winger, Parley V,, Donald P. Schultz and W.Waynon Johnson.  1990.  Environmental Contaminant
       Concentrations in Biota from the Lower Savannah River, Georgia and South Carolina. Arch,
       Environ. Contam, ToxicoL  19:  101-117.

Yamamoto, Shinji, Yoshitsup Konishi, Tsutomu Matsuda, Takashi Murai, Masa-Aki Shibata, Isao
       Matsui-Yuasa, Shuzo Otani, Koichi Kuroda,  Ginji Endo and Shoji Fukushima.  1995. Cancer
       Induction by an Organic Arsenic Compound, Dimethylarsinie Acid (Cacodylic Acid), in
       F344/DuCrj Rats after Pretreatment with Five Carcinogens.  Cancer Research 55: 1271-
       1276.

Yamanaka, Kenzo, Akira Hasegawa, Ryoji Sawamura and Shoji Okada.  1991. Cellular Response to
       Oxidative Damage in Lung Induced by the Administration of Dimethylarsinie Acid, a Major
       Metabolite of Inorganic Arsenic, in Mice. Toxicology and Applied Pharmacology 108:  205-
       213.

Yasui, A., C. Tsutsumi and S. Toda.  1978.  Selective Determination of Inorganic Arsenic (III), (V)
       and Organic Arsenic in Biological Materials by Solvent Extraction-Atomic Absorption
       Spectrophotometry, Agric Biol Chem 42; 2139-2145.

Zook, Elizabeth G., John J. Powell, Betty M. Hackley, John A. Emerson, James R. Brooker and
       George M. Knobl, Jr.  1976.  National Marine Fisheries Service Preliminary Survey of
       Selected Seafood  for Mercury, Lead, Cadmium, Chromium and Arsenic Content.  Journal of
       Agricultural Food Chemistry 24(1): 47-53.
                                             59

-------

-------