The geochemical
other elements’
cycle of arsenic
in Lake Washington
and its relation
to
Eric A. Crecelius2
Department
of Oceanography,
University
of Washington,
Scattle
98195
Abstract
Abnormally
high arsenic concentrations
(> 200 ppm dry wt) in the smface sediments
of Lake Washington
are attributed
to atmospheric
input of partially
soluble arsenic-rich
dust from a copper smelter 35 km upwind and removal of dissolved As from the lake
water by bacteria or an inorganic reaction.
The suspended matter in the deeper lake water contains 9% iron, 8% manganese,
and 350 ppm arsenic. Cores dated by ““Pb indicate that As levels began to increase in
the sediments when the smelter began operation in 1890.
An arsenic budget for the lake shows equal supplies from the atmosphere
and from
and by accumulation
in the sediments
rivers, and removal by outflowing
water (45%)
( 55% ). A similar cycle was observed for antimony
which is released by the smelter
in much smaller amounts than arsenic.
About two-thirds
of the As in rain and lake water was in the form of arsenate;
however, in the interstitial
water 55% of the dissolved As was present as arsenite and
a few percent as dimcthylarsinic
acid.
The fate of arsenic in freshwater has received attentiou recently because of concern over its environmental effects. Ferguson and Gavis (1972) described the cycle
of As in a stratified lake. From generalized
reactions and thermodynamic
data they
showed that coprecipitation of arsenic with
ferric hydroxide would be important in
removing As from oxygenated water; in
oxygen depleted water or in sediments,
reduction of ferric ions could release As.
Kanamori (1965) showed that arsenic was
coprecipitated
from lake water by ferric
hydroxide and then released to the water
when the iron was reduced.
Sohacki
(1968) added sodium arsenite to a small
farm pond and showed that most of the
As accumulated in the fine-grained fraction of the sediments, thereby becoming
environmentally
inactive. These studies indicate the important role of both iron and
bottom sediments in the cycle of arsenic
in lakes.
Lake Washington was chosen for a study
___’ Contribution
No. 813 of the Department
of
Oceanography,
University
of Washington,
Seattle.
Financial
support was provided
by the RANN
division
of the National
Science
Foundation
( grant GI33325X).
a Present address: Battelle Northwest,
Ri&land,
Washington
99352.
LIMNOLOGY
AND OCEANOGRAPIIY
of the aquatic cycle of As because the lake,
downwind
of a large smelter, has been
intensively studied ( Edmondson and Allison 1970; Shapiro et al. 1971). Elevated
levels of lead, mercury, and copper in the
sediment
(Crecelius
and Piper
1973;
Barnes and Schcll 1973) and arsenic and
antimony in soils (Crecelius et al. 1974)
indicate that the copper smelter, near
Tacoma, contributes many trace metals to
Lake Washington.
Lake Washington borders Seattle on the
east (Fig. 1). The basin is U-shaped with
a relatively flat bottom, 40-60 m deep,
formed by sediment fill. The Cedar and
Sammamish Rivers supply nearly all the
input of water, which is discharged out of
the lake through the Lake Washington
Ship Canal. The residence time for water
in the lake is about 3 years (Edmondson
1972), long enough to allow most riverborne suspended matter to settle out.
During summer and fall the lake is stratificd with the thermocline between lo-20
m; in winter it is mixed.
To examine the cycle of As in the lake,
I measured inputs from rivers, rain, dustfall, and stormwater and removal in the
outlet water and the sediments. Water
an d suspended matter from the lake were
analyzed every 3 months. These data were
441
MAY 1975, V. 20 (3)
Crecelius
442
122” 16’
12’
- 44
used to calculate an As budget for Lake
Washington and to examine the processes
controlling As distribution in the lake.
I thank R. Carpenter for his support and
useful suggestions and W. T. Edmondson
for his advice.
Methods
I
I
122” 16’
CEDAR
I
R.
12’
Fig. 1. Concentrations
of arsenic (ppm dry
( O-l cm ) of Lake
wt ) in surface sediments
Washington:
@-locations
where water and suspended matter were collected;
A-locations
of
cores; @-locations
where surface sediments were
The copper smelter is southwest
of
collected.
the lake.
Samples were collected in PVC sampling
bottles with the least practical amount of
metal contact to minimize contamination
and stored frozen in acid-washed polyethylene bottles until analysis to minimize
chemical or biological changes. Tests in
polyethylene
containers indicated that arsenic concentrations did not change during
storage ( Robertson and Carpenter 1974).
All metals were analyzed by neutron activation, but another technique developed
by Rraman and Foreback (1973) was used
to determine the chemical form of dissolved arsenic.
Sediments were collected with a special
coring apparatus which has a hydrostatically slowed rate of penetration into the
sediments ( Pamatmat 1971) and negligible
disturbance of the water-sediment
interface.
Suspended matter was collected
by
placing a known volume of water (3-40
liters) in a PVC tank and forcing it through
a prcweighed 142-mm-diam Millipore filter
(HA, 0.45-pm pore size) by compressed
nitrogen gas at 5-atm pressure. Filters
were dried (70°C for 24 h) and weighed.
Dustfall and rainfall samples were collected together, in a 43-cm-diam polyethylene barrel, on top of a three-story
building near the outlet of the lake. A
drain in the bottom of the barrel allowed
the rain and dissolved material to pass
through a 47-mm-diam Millipore type IIA
filter into a glass flask, while the insoluble
dustfall material remained in the barrel
or on the filter. The flask was emptied a
few times a month. Once a month the
barrel was scrubbed down and the dust
was collected on the filter, dried, and
weighed.
samples
High volume air particulate
( HiVol) were collected at the same posi-
Arsenic cycle in L. Washington
tion by pulling air through a prcweighed
142-mm-diam Millipore type I-IA filter for
24-72 h. The filter was then dried and
weighed.
Stack dust samples from the Tacoma
smelter, supplied by J. Roberts, represent
dust released into the atmosphere when
the smelter is operating under normal conditions.
C. Hunter supplied sewage effluent and
sewage sludge samples from the Metro
sewage treatment plant at West Point,
Seattle.
Arsenic, antimony, manganese, iron, and
aluminum were determined by neutron acSolids were
dried,
analysis.
tivation
weighed, irradiated, and then counted on
a Ge( Li) diode detector with an energy
resolving capability of 2.56 kcV FWHM
at 1,333 keV (Crecelius et al. 1974). Water
samples were also analyzed for arsenic and
antimony by neutron activation analysis
after concentration by quantitative
coprecipitation with ferric hydroxide (cf. Robertson and Carpenter 1974).
Standard reference materials for the
metals determinations included USGS rock
standards, NBS orchard leaves and tuna
meal, and EPA standard solutions, Results
of analyses of these reference materials are
within
10% of the certified or recommended values. The standard deviations
for analysis of replicate samples are less
than 10% for each element.
Total carbon was determined on ovendried sediment (SOOC for 4 h) with a
LECO carbon analyzer, organic carbon by
the same method after treatment with IICI.
Binding of the trace metals in the sediments was checked by chemical extraction:
amorphous iron and aluminum compounds
were removed by both an oxalate (Saunders 1965) and a citrate-dithionite-bicarbonate procedure (Jackson 1969).
Sediment accumulation
rates over the
last 100 years were determined
from
several cores by E. S. Twiss by 210Pb
( Schell et al. 1973). Sediment samples
(l-5 g dry wt) were wet-ashed and the
210Po spontaneously plated, together with
208Po tracer for yield determination
onto
silver disks. The two polonium isotopes
443
were then measured by alpha spectroscopy.
The precision of the method was -I- 15%.
Results and discussion
Arsenic in sediments-Arsenic
concentrations in I4 surface sediments from Lake
Washington (O-l cm) ranged from 15-210
ppm As dry weight ( Fig. 1). These concentrations were much higher than those
reported for the surface sediments of Lake
Michigan (5-30 ppm) (Ruth et al. 1970)
or for fine-grain nonmarine shales (3-12
ppm) (Tourtelot 1964).
The lowest arsenic concentrations in the
surface sediments were found at the shallow ends of the lake, where the coarser
riverborne suspended matter settles out and
dilutes the As associated with fine-grained
particles.
Concentrations
in the surface
sediments of the deeper region of the lake
(> 30-m water depth) averaged 95 ppm
and varied by more than a factor of three.
Gould and Budinger (1958) pointed out
that convective currents associated with
winter overturn apparently erode and redeposit surface sediments.
Sediment cores from five locations in the
lake showed generally high As concentrations at the surface, decreasing with depth
to the usual background concentrations of
about 10 ppm (Fig. 2). Several details
of the arsenic profiles in the cores appeared to be unrelated from core to core
and indicated
local variations
in the
amount of As accumulated in the sediments. The high concentration at the 1718-cm depth in the core near Seward Park
was associated with an iron-rich
layer;
Shapiro et al. ( 1971) found the same layer
to be rich in iron and phosphorus.
The sedimentation rate determined for
three cores from the central region of the
lake by 210Pb was 3 mm year-l (see Fig. 2).
Schell ( 1974) found similar rates. All these
are in good agreement with the rates detcrmined by Edmondson and Allison (1970)
from a silt layer that formed when the
lake level was permanently
lowered in
1916. A study of the pollen assemblages
in two cores gave a sedimentation rate near
Madison Park of 3.1 mm ycarl and south
444
Crecelius
SAND PT.
As (ppm)
EVERGREEN
PT. BR.
As (ppm)
0
100
0
MADISON PARK
As (wm)
MERCER
I. BR.
As (ppm)
0
100
200
SEWARD PARK
As (ppm)
0
100
200
BUILT
F
6,
f
1
L
Fig. 2. Concentrations
of arsenic (ppm dry wt)
of the sediments in the core from near Evergreen
Cores, see Fig. 1).
in sediment
Point were
of Mercer Island of 3.7 mm year-l (Davis
1973).
Convective
currents
associated
with winter overturn or slumping of the
steep sides of the lake may cause local
variations in apparent sedimentation rates.
The arsenic concentration in the cores
began to increase at a depth of about 25
cm (Fig. 2). This depth is about the time
when the Tacoma smelter began operations in 1890. At this time the city of
Seattle also began to grow rapidly, possibly contributing As to the lake from coal
burning and sewage effluent. The arsenic
profiles in the cores show an increase in
As input to the lake over the last century.
The surface sediments from water depth
> 30 m are relatively rich in organic car-
bon, iron, and manganese with means and
standard deviations of 5.2% -t- 1.3%, 4.8%
-t- 1.2%, and 0.62% * 0.78% ( Table 1).
Linear correlation
coefficients
( r) were
calculated for the element in the surface
sediment collected from these depths. Organic carbon varies little and is poorly
correlated with As, r = 0.31, in contrast
to Lake Michigan where the correlation is
r = 0.72 (Ruth et al. 1970). Iron and
manganese showed a moderate range in the
surface sediments and correlated strongly
with arsenic: iron, T = 0.94; manganese,
r = 0.84. These strong correlations suggest
that As is associated with an iron or manganese phase, or both. The surface few
centimeters of lake sediments were gen-
Table 1.
ington and
Chemical
Fe
Mn
Al
c
8
%
%
%
(dry
of sediments and suspended matter in Lake Wash-
WE basis)
of suspended matter in the two rivers
Lake
As(ppm)
Sb(ppm)
composition
sediments
Surface
O-l cm
30m water
depth
Subsurface
20 cm -
cores from Lake Washington.
Ages
determined
by ““Pb (for location of
entering
Lake
Suspended
Lake surface
O-2
m
-
Washington.
matter
Lake
in lake
bottom
50 m
or
rivers
Rivers
Mean
Range
Mean
Range
Mean
Range
Mean
Range
Mean
Range
95
7.1
48-217
2.8-9.5
10
0.6
8-12
0.4-0.8
78
4.3
37-159
2.9-9.2
158
9.8
62-363
3-15
18
1.6
10-97
1.0-4.4
5.3
2.3
2.6
2.2-9.4
0.5-6.0
2.0-4.8
4.8
0.62
5.5
5.2
3.2-6.9
0.1-2.4
3.8-6.2
2.5-6.1
3.9
0.14
4.8
4.2
3.1-4.6
0.11-0.17
4.0-5.9
3.9-4.8
0.24
1.7
-
0.09-0.55
0.5-4.4
-
0.09
6.0
0.08-0.14
4.0-6.8
Arsenic cycle in L. Washington
erally brown or reddish brown throughout
the year.
Concentrations of organic carbon, iron,
and manganese were lower in the older
lake sediments ( > 20 cm deep), with
means of 4.2% -t- 0.5%, 3.9% -C 0.6%, and
0.14% * 0.4% ( Table 1). Similar increases
in iron and manganese near the surface of
two cores from Lake Washington
were
found by Wallace ( 1973).
Sediments were extracted by acid ammonium oxalatc ( AA0 ) and citrate-dithionite-bicarbona tc ( CDB ) to determine
the amount of arsenic bound to iron, manganese, and aluminum compounds. Of the
two surface sediments (O-5 cm), AA0 extracted 55% and the less rigorous CDB
procedure extracted 34% of the As. Of the
two subsurface sediments ( > 35 cm deep),
AA0 extracted 60% and CDB 30%. Thus
a significant fraction of the As appears to
be bound to the extractable iron, manganese, and aluminum compounds. Shapiro
et al. ( 1971) found that the amount of
acid-soluble phosphate in Lake Washington sediments ranged from 65-80%, suggesting that most of the phosphate was
associated with ferric compounds.
Arsenic concentrations
in cores from
three other western Washington
lakes
(Crccelius 1974) showed a slight to modcrate increase in surface sediments, but
these were much less dramatic than in
Lake Washington.
Arsenic input to Lake Washington-The
above results indicate an increasingly large
input of As to the lake during the last 80
years, or a process that now removes more
As from the lake water to the sediments
than in the past, or both.
Concentrations of dissolved ( < 0.45 pm)
and particulate As were measured for more
than a year (September 1972 to May 1974)
in rivers, rainfall, dustfall, and stormwatcr
runoff entering the lake. The two rivers,
Cedar and Sammamish, which supply 95%
of the water to the lake, contained 1.1 ppb
and 2.5 ppb total As, of which two-thirds
was dissolved. These concentrations
are
similar to those for rivers entering Puget
Sound and for rivers in Japan (Kanamori
and Sugawara 1965b ) .
445
Since the Cedar River has twice the
discharge of the Sammamish, the weighted
mean arsenic concentrations for the combined rivers were 1.6 ppb total and 1.0
ppb dissolved As. The concentrations of
total As in the lake water and in the
ship canal (which empties the lake) averaged about 1.6 ppb (range 1.3-2.0 ppb),
of which 90% was dissolved. The river
water entering Lake Washington
dilutes
the dissolved As levels in the lake.
The concentration of arsenic in the riverborne suspended matter entering the lake
varied greatly (lo-97 ppm) with river
discharge, grain size, and organic content
Concentrations
were
of the particulates.
usually higher during periods of low discharge, when grain size was minimal and
organic content maximal. The riverborne
suspended matter averaged 18 ppm (dry
wt basis ) , lower than concentrations in the
surface sediments. Assuming that the riverborne suspended matter did not take up
dissolved As from the lake water, it could
not bc the major source of the As in the
lake sediments.
The concentrations of arsenic in filtered
rain collected in Seattle near the lake outlet
averaged 16 ppb. To correct for the large
seasonal differences in rainfall, mean As
concentrations for each month were multiplied by the average monthly rainfall observed over the last 62 years ( U.S. Environ.
Data Ser. 1972). The As levels in Seattle
rain were much higher than in rain collected near the Washington coast, 100 km
upwind of the smelter ( 0.4 ppb ) and in
rain collected jn Japan (1.6 ppb) (Kanamori and Sugawara 1965a). These extremcly high As levels are likely due to
the very fine (about l-pm diam) arsenicrich stack dust from the Tacoma smelter.
The concentrations of As in the monthly
dustfall samples ranged from 215-572 ppm
with a mean of 360. Arsenic levels in dust
from other heavily industrialized
cities are
in this same range (Kanamori and Sugawara 1965a). The amount of dustfall at
this location near the lake outlet averaged
0.14 mg crnw2 month-‘-lower
than typical
city values of 0.35-3.5 (U.S. Dep. HEW
1969).
446
Crecelius
The concentrations of arsenic in dust
filtered from air when winds were blowing
from the smelter toward Seattle ranged
from 350-2,610 ppm, with a mean of 1,640;
dust collected during periods of northerly
winds contained from 46450 ppm As,
with a mean of 170. This dramatic relationship between wind direction and As
concentration
indicates that the Tacoma
smelter is the major contributor of atmospherically transported arsenic to Seattle
and Lake Washington.
The smelter’s atmospheric output of arsenic is 150,000 kg
yr-l (Crecelius et al. 1974). No other As
sources to the atmosphere appear to be
significant.
Stormwater runoff from Seattle into the
west side of Lake Washington contributes
only 1% of the water input to the lake
(Seattle Dep. Eng. Design preliminary
rcpt. ) . The stormwater averaged 15 ppb
As, of which 30% was dissolved. Runoff
from the Mercer Island floating bridge
contained about the same arsenic levels
as the stormwater.
Until 1967, treated sewage effluent was
discharged into the lake; it was then diverted to Puget Sound, The effluent now
entering Puget Sound contains 3-7 ppb
As, and the particulate
matter contains
2040 ppm As dry weight. Even in 1963,
when sewage input to the lake was maximal, the contribution of As from this source
would have been minor compared to that
from rivers, rain, and dust.
The chemical forms of As were de tcrmined in a few samples of rain, stormwater, and lake water; these are of interest
because arsenite is much more toxic than
arsenate and these ions may change in
proportion.
Braman and Foreback ( 1973)
have shown that methylatcd forms of As
as well as arscnite and arsenate are often
present in the aquatic environment.
The smelter stack dust is rich in soluble
arsenic trioxide. After this dust had been
in distilled water for 2 h, at pII 5.5 and
20°C 80% of the As dissolved was arsenite.
In five rain samples only 35% of the total
As was arsenitc. The concentration of arsenite in rain samples may be lower because some had oxidized to arsenate during
the week or two that they remained in the
rain ca tchcr.
Lake Washington water in April 1974
had about the same forms as the rain: 40%
arscnite and 60% arsenate. The interstitial
water from one surface sediment sample
was somewhat richer in arsenite (55%)
than the lake water.
Dimcthylarsinic
acid ( DMAA) , determined by the method of Braman and Foreback ( 1973)) was sometimes detected in
very low concentrations, 0.05 ppb, in lake
water. Interstitial water had 1 ppb DMAA,
but no methylarsonic
acid. These lake
sediments, which contained about 5% organic carbon but did not smell of hydrogen
sulfide, are an environment
where As
methyla tion can occur.
Arsenic removal processes-Arsenic
is
removed from Lake Washington by outflow of water through the ship canal and
by burial in the sediments. The outflow
water contains from 1.3-1.9 ppb As (mean
of 1.6 ppb) and removes a major portion
of the dissolved As entering the lake. Arsenic accumulates in the sediments both as
insoluble material and as dissolved arsenic
removed from the water by adsorption or
precipitation.
Shapiro et al. (1971) reported relatively
high concentrations of particulate iron and
phosphate in the hypolimnion
of Lake
Washington and suggested that a microorganism might be coprecipitating
phosphate with ferric hydroxide. Since arsenate
is chemically similar to phosphate, arsenic
may also be removed from the water by
precipitation
with ferric hydroxide.
The concentration of As in suspended
matter, collected at two stations in the lake
from September 1972 to May 1974, showed
a dramatic increase in the deep water
during summer and fall ( Fig. 3). In January, after the lake had been mixed by
thermal convection, suspend matter ranged
from 40-70 ppm As (dry wt). This is
much higher than the average 27 ppm As
in suspended matter from Lake Michigan
(Lcland et al. 1973). As the year progressed, As increased in the Lake Washington
suspended matter from deeper
water, In October, before overturn, sus-
Arsenic cycle in L. Washington
AUG
As (ppm 1
MAY
JAN
As (ppm)
As (PPm)
0
100
0
100
200
300
”
JAN
As (tw-n)
0
MAY
As (ppm)
100
AUG
As (ppm)
l-r
Fig. 3. Concentrations
of arsenic (ppm dry wt) in suspended matter from two locations in Lake
Washington.
Upper profiles from station near Sand Point, lower profiles from near Madison Park.
pendcd matter collected a few meters off
the bottom had > 200 ppm As.
The suspended matter collected from
deep water during summer contains as
much as 9% iron, 8% manganese, and 350
ppm arsenic (Table 1). Some of this ironand manganese-rich
suspended
matter
probably settles to the bottom during summer and fall. In winter, bottom currents
caused by convective overturn may resuspend and redeposit the semifluid surface
sediment, causing the large variability
observed in the distribution of arsenic, manganese, and iron.
The high correlation
coefficients
between arsenic, iron, and manganese in the
surface sediments and the relatively low
correlation
coefficients
between arsenic,
carbon, and aluminum indicated that the
precipitation
of iron and manganese was
the most important process removing As
to the sediments. Once the arsenic-rich
iron-manganese precipitate is buried in the
bottom sediments, the iron and manganese
may be reduced to more soluble forms,
Since the bottom water was at lcast 20%
saturated in oxygen, the surface sediments
remained oxidizing and tended to reoxidize
iron and manganese compomlds diffusing
out of the bottom.
The question of whether the iron-manganese-rich suspended matter in Lake
Washington is formed by bacteria has not
been answered.
Many of the particles
have the appearance of bacteria of the
genus Metallogenium
(J. Staley personal
communication).
However, inorganic precipitates may resemble certain bacteria.
Regardless of origin, this iron-manganeserich suspended matter apparently
concentrates As in the lake sediment.
The concentration of total As in Lake
Washing ton water from a number of
depths at two locations between Septcmber 1972 and May 1974 ranged bctwecn
l-2 ppb and showed no consistent change
with depth or time.
Arsenic budget for Lake WashingtonI have calculated an arsenic budget for
Lake Washington, to compare the major
448
Table
Crecelius
2. Arsenic
budget
for Lake Washington.
kg As yr
Input
Rainfall
on lake
Dustfall
on lake
Rivers
(dissolved
Rivers
(particulate
Storm water
runoff
As)
AS)
output
Ship canal
(total
As)
Balance
removed to
sediments
-1
1,300~520*
600,190
1,3001260
700?200
5OOL150
4,400 660
2,200?280
2,400 or
(2,600)'7OOt
4,400
The majority
of the uncertainty
is due to the
natural
variability
between
samples and not
analytical
error.
Uncertainties
in volumes
of
water
and solids
entering
the lake are not
included.
9
Calculated
from area oE lake
x sedimentation
rate
x As concentration.
*
inputs and sinks (Table 2). The error in
the input and output was based on the
standard deviation of the mean As concentration
Rainfall and dustfall accounted
for almost half of the As input to the
lake. Part of the arsenic in dust that falls
into the lake accumulates in the sediments;
however, most of the atmospheric input
of As was dissolved in rain. Rainfall and
dustfall inputs were calculated by extrapolating over the surface area of the lake
from the amount of dissolved and particulate As collcctcd by the rain and dust
catcher. The river input of dissolved As
amounted to about a third of the total input; particulate As accounted for another
16%. Stormwater runoff added the remaining 10% As to the lake, of which about
half was dissolved.
I have assumed that no volatile forms
of As are lost to the atmosphere. The only
outlet from the lake is the ship canal,
which removed 45% of the As added annually. For the arsenic budget to be balanced, about half of the input must be
removed to the sediment (Table 2). The
rate of As accumulation in the sediments
was about equal to the rate needed to
balance the budget (Table 2). The accumulation rate of As is the product of
the sediment accumulation rate, the area
of the lake accumulating sediment, and the
concentration of As in the surface sedimerits. The sediment accumulation
rate
was determined
at four locations by
weighing the sediment that had accumulated since 1916 and then dividing by the
years represented. The average sediment
accumulation rate, based on eight cores,
was 0.06 * 0.01 g dry sediment cm-2yr-1
( S. Wakeham personal communication).
The annual suspended load supplied to
Lake Washington by the two rivers was
calculated from the concentrations
measured between June 1972 and January 1974
and the mean monthly river flow (State
Wash. Dep. Conserv. 1962). Most of the
sediment transport was during winter when
the rivers are swollen. The annual input
of suspended matter was calculated to be
4 X lo7 kg yr-l, similar to 5 X lo7 reported
by the Puget Sound Task Force ( 1970).
The biogenic contribution to the sediments
was estimated to be < 30% of the annual
river input. Organic matter composes between lo-15%
of the lake sediments
(Shapiro et al. 1971). Diatom frustules
may compose roughly 15% of the sediment,
estimated by assuming that changes in dissolved silica in the photic zone of the
lake would result in an equal contribution
of SiOz to the sediments. The volume of
the photic zone was assumed to be the
area of the lake (90 km2) times 10-m
dcp th. The dissolved Si concentration
changes from a winter maximum of 4-5
ppm to a summer low of about 1 ppm
(R. Barnes personal communication).
If the annual river suspended load (4 X
lo7 kg yr-l) were spread evenly over the
relatively flat bottom of the lake (below
30-m water depth), which has an area
of 45 km2, then the sediment accumulation would bc 0.09 g cme2yr-l. This rate
is very close to the value of 0.06 g cmm2yr1
calculated from the best known sedimentation rates. This balance between river
loading and sedimentation indicates that
the lake is an effective sediment trap.
Since the relatively steep sides of the lake
accumulate little sediment, the area of the
lake below 30 m was chosen to calculate
As accumulation.
Taking the average As
concentration for surface sediment of 95
Arsenic cycle in L. Washington
ppm ( Table 1)) this area, plus a sediment
accumulation rate of 0.06 g cm-2yr-1, the
rate of As accumulation is 2,600 kg yr-l,
about the amount needed to balance the
budget.
Because of the large reservoir of dissolved As in the lake (about 5,000 kg,),
only 20% of the dissolved As must be removed annually to balance the budget.
Antimony cycle in Lake WashingtonAntimony is directly below arsenic in the
periodic table of elements and has a simiMost natural samples
lar geochemistry.
contain a tenth as much Sb as As. The
Tacoma smelter stack dust has 2-10% Sb,
giving the Seattle-Lake Washington area a
significant atmospheric input. Many samples analyzed for As were also analyzed for
Sb, but only a few water samples so that
a complete antimony budget cannot be
calculated for Lake Washington.
Concentrations of Sb in the surface sediments of Lake Washington
were much
higher than in the older sediments. Surface
sediments ranged from 2.8-9.5 ppm, with
a mean of 7.1 + 1.2, and older scdimcnts,
> 20-cm depth in cores, ranged from 0.40.8 ppm with a mean of 0.60 * 0.13 (Table
1). The profiles of Sb in cores were of
similar shape to those of As but at about a
tenth the concentrations.
Antimony
also
showed a modcrate correlation with organic carbon, iron, and aluminum in the
surface sediments. Techniques used to remove extractable
iron, manganese, and
aluminum extracted about 30% of the total
Sb from the sediments. Antimony in cores
from the three other western Washington
lakes paralleled As, but again at about a
tenth the concentration.
Concentrations of Sb in dust and rain
were also high. Dustfall samples collected
in the rain catcher near the lake outlet
ranged from 60-225 ppm Sb with a mean
of 107. HiVol
dust samples collected
during
periods of southwesterly
wind
ranged from 80-360 ppm Sb with a mean
of 250, and those during periods of northerly wind ranged from 12-94 ppm with a
mean of 35. HiVol dust samples leached by
lake water for 4 days released 40% of the
total Sb; the same percentage of Sb dis-
449
solved when smelter stack dust (which contained 8% Sb) was soaked in lake water
for the same period, suggesting that the
source of much of the particulate Sb in
Seattle air is the smelter. There are no
major sources of Sb in the area.
Seattle rain collected in the dust-rain
catcher averaged 3.5 ppb Sb, much higher
than surface water of Lake Washington,
which has a mean of 0.2 ppb. The relatively high concentrations
in both dust
and rain indicated that atmospheric input
could contribute a significant amount of
Sb to Lake Washington.
The suspended
matter in the two rivers entering the lake
contained 1.6 ppm Sb, accounting for less
than a quarter of the antimony accumulating in the sediments. Stormwater runoff contained 1.6 ppb Sb, about half of
which was dissolved.
The concentrations
of Sb in the suspended matter of Lake Washington were
usually 2-6 ppm (dry wt) except in the
near bottom water during summer and
fall, when the iron-manganese-rich
suspendcd matter contained lo-20 ppm Sb.
The same process that extracts As from
lake water appears also to remove dissolved
Sb.
Although a complete antimony budget
cannot be calculated for Lake Washington,
the known inputs appear to account for
the Sb accumulating
in the sediments.
Rainfall ( 260 kg dissolved Sb per year ) ,
dustfall (170 kg yr-l), river particulates
( 60 kg yr-l ) , and stormwater ( 50 kg yr-l)
give a total annual input (not including
dissolved Sb from rivers) of 540 kg. Calculations from the annual Sb accumulation rate in the surface sediments (> 30m water depth) indicate that about 200 kg
reach the sediment.
Conclusions
IIigh arsenic and antimony concentrations in the sediments of Lake Washington
indicate that the copper smelter at Tacoma
has contaminated a large area with these
metals. Atmospheric input (largely from
the smelter ) and rivers each contribute
about half of the As to the lake.
Dissolved arsenic is removed from the
450
Crecelius
lake water by an iron-manganese precipitate; only 20% of the As dissolved in the
lake must be removed each year to account
for the accumulation
in the sediments.
Iron and manganese are enriched in the
surface sediments of the lake, show a strong
correlation with arsenic, and about 60% of
the As in the lake sediments is extracted
with the extractable iron-manganese compounds. The presence of dimethylarsinic
acid in the interstitial
water indicates
methylation of arsenic can occur in lake
sediments.
References
BARNES, R. S., AND W. R. SWELL. 1973. Physical transport
of trace metals in the Lake
Washington
watershed, p. 45-53. In Cycling
and control of metals. Proc. Environ. Rcsour.
Conf., 1972, Columbus, Ohio. Natl. Environ.
Res. Center.
BRAMAN,
R. S., AND C. C. FOREBACK.
1973.
Methylated
forms of arsenic in the environment. Science 182 : 1247-1249.
CRECELIUS,
E. A. 1974. The geochemistry
of
arsenic and antimony
in Puget Sound and
Lake Washington,
WA. Ph.D. thesis, Univ.
Wash., Seattle.
133 p.
1974.
C. J. JOHNSON, AND G. C. IIOFER.
Cdntamination
of soils near a copper smelter
by arsenic, antimony,
and lead. Water Air
Soil Pollut. 3: 337-343.
-,
AND D. Z. PIPER.
1973. Particulate
lead
contamination
recorded in sedimentary
cores
from Lake Washington,
Seattle. Environ. Sci.
Technol. 7: 1053-1055.
DAVIS, M. B. 1973. Pollen evidence of changing
land use around the shore of Lake Washington. Northwest
Sci. 47:
133-148.
EDMONDSON,
W. T. 1972. Nutrients
and phytoplankton
in Lake Washington,
p. 179-193.
ln G. E. Likens [ea.] Nutrients
and eutrophication.
Am. Sot. Limnol. Oceanogr. Spec.
Symp. 1.
-,
AND D. E. ALLISON.
1970. Recording
densitometry
of X-radiographs
for the study
of cryptic
laminations
in the sediment
of
Lake Washington.
Limnol.
Oceanogr.
15 :
138-144.
FERCUSON, J. F., AND J, GAVIS.
1972. A review
of the arsenic cycle in natural waters. Water
Res. 6: 1259-1274.
GOCJLD, H. R., AND T. F. BIJDINGER. 1958. Control of sedimentation
and bottom configuration by convection
currents, Lake Washington, Washington.
J. Mar. Res. 17: 183-198.
JACKSON, M. L. 1969. Soil chemical
analysisadvanced
course, 2nd ed. Dep. Soil Sci.,
Univ. Wis.
S . 1965. Geochemical
study of arsenic in natural waters. 3. The significance
of ferric hydroxide precipitate
in stratification
and sedimentation
of arsenic in lake waters.
J. Earth Sci. Nagoya Univ. 13: 46-57.
-,
AND K. SUGAW~.
1965a. Geochemical
study of arsenic in natural waters.
1. Arscnic in rain and snow. J. Earth Sci. Nagoya
Univ. 13 : 23-35.
, AND -.
1965b. Geochemical
study
of arsenic in natural waters.
2. Arsenic in
river water. J. Earth Sci. Nagoya Univ. 13:
36-45.
LELAND,
H. V., S. S. SHUKLA, AND N. F. SHIMP.
1973. Factors affecting
distribution
of lead
and other trace elements in sediments of
southern Lake Michigan, p. 89-130. In P. C.
Singer [ed.], Trace metals and metal-organics
interactions
in natural
waters.
Ann Arbor
sci.
PAMATMAT, M. M. 1971. Oxygen consumption
by the seabed. 4. Shipboard and laboratory
expcrimen ts . Limnol.
Oceanogr.
16:
536550.
PUGET SOUND TASK FORCE OF THE PACIFIC NORTIIWEST
RIVER BASINS
COMMISSION.
1970.
Puget Sound and adjacent waters. Appendix
3, Hydrology
and
natural
environment.
RIBCO, Seattle. 205 p,
ROBERTSON,
D. E., AND R. CARPENTER.
1974.
Neutron
activation
techniques
for the measurement of trace metals in environmental
samples.
Nuclear
Sci. Ser. NAS-NS
3114.
79 p.
Rum, R. R., E. J. KENNEDY, AND N. F. SHIMP.
1970. Distribution
of arsenic in unconsolidated sediments from southern Lake Michigan. Ill. State Geol. Surv. Environ.
Geol.
Notes 37, p. 1-16.
SAUNDERS, W. M. H. 1965. Phosphate
retention
by New Zealand soils and its relationship
to free sesquioxides,
organic
matter,
and
other soil propertics.
N.Z. J. Agr. Res. 8:
30-57.
SMELL,
W. R. 1974. Sedimentation
rates and
mean residence times of Pb and Pb-210 in
Lake Washington,
Puget Sound estuaries, and
a coastal region.
Lab, Radiat. Ecol. Rep.
2225-T14-6.
Univ. Wash., Seattle.
T. JOKELA, AND R. EAGLE.
1973. Natural
leid-210
and polonium-210
in a marine environment,
p. 701-724.
In Radioactive
contamination
of
the
marine
environment.
IAEA/SM-158/47.
SHAPIRO, J., W. T. EDMONDSON,
AND D. E. ALLISON. 1971. Changes in the chemical
composition
of sediments of Lake Washington,
1958-1970.
Limnol.
Oceanogr.
16:
437452.
SOHACKI,
L. P. 1968. Dynamics
of arsenic in
the aquatic enviromncnt.
Ph.D. thesis, Michigan State Univ.
101 p.
KANAMORI,
451
Arsenic cycle in L. Washington
STATE OF WASHINGTON,
SERVATION.
1962.
DEPARTMENT
Water
Supply
OF
Bull.
CON-
15.
386 p.
H. A. 1964. Minor-element
composition and organic carbon content of marine
and nonmarine shales of late Cretaceous age
in the western interior of the United States.
Geochim. Cosmochim. Acta 28: 1579-1604.
TOU~TELOT,
U.S. DEPARTMENT
WELFARE.
OF HEALTH,
National
Air
EDUCATION,
Pollution
AND
Control
Air quality
criteria
211 p.
U.S. ENVIRONMENTAL
DATA
SERVICE.
1972. Cli76( 13).
matological
data, Washington.
WALLACE,
T. L. 1973. The application
of neutron
activation
analysis
to lacustrine
sediment.
M.S. thesis, Univ. Wash., Seattle. 39 p.
Administration.
for particulate
1969.
matter.
Submitted:
Accepted:
7 October 1974
20 January 1975
Announcement
International
Association of Geochemistry
and Cosmochemistry
Academy of Sciences of the USSR
Academy of Sciences of the Ukrainian SSR
An international symposium titled “Interaction between water and living matter”
will be held in Odessa, USSR, on 6-10
October 1975. The program of the symposium is divided into three parts. 1. Elementary processes : excretion into water
and uptake out of water of separate chemical elements, organic and organomincral
compounds by organisms; kinetics of these
processes and their regularities; species of
specificity of excretion and uptake of separate elements and compounds into and out
of water; structure of water and aquatic
solutions under absence of living matter action and under its influence. 2. Chemicoecological systems: interconnecting
cxcre-
tion or uptake of several different inorganic
and organic compounds; influence of some
compounds on excretion or uptake of others;
interaction between organisms through participation of compounds of different chemical nature; connection of physical, chemical,
and biological processes in experimental or
natural systems; modeling of chemicobiological systems. 3. Problems and perspectives
of aquatic chemical ecology. Geochemical
consequences of action of living matter on
aquatic environments.
Chairman of the organizing committee is
Prof. G. G. Polikarpov, Institute of Biology
of Southern Seas, Prospekt Nakhimova, 2,
Sev,astopol, 335000, USSR; the Convener is
Prof. Yu. P. Zaitsev, Odessa Branch of the
Institute of Biology of Southern Seas, N 86,
Chcrnomorskaya Doroga, P. 0. 80, Odessa,
270080, USSR.