United States
Environmental Protection
Agency
Environmental Monitoring Systems
Laboratory
Las Vegas, NV89114
Research and Development
EPA/600/M-85/022 Sept. 1985
Jp
ENVIRONMENTAL
RESEARCH BRIEF
Instream Tests for Toxicity Persistence from Heavy Metals
John R. Baker1, Barry P. Baldigo1, and Wesley L. Kinney2
^.Abstract
Instream toxicity tests used the larval fathead minnow
Pimephales promelas and the cladoceran Ceriodaphnia in
Prickly Pear Creek and Spring Creek, Montana, waters to
examine toxicity persistence in a stream receiving metal
inputs. The toxicity source was Spring Creek, a tributary of
Prickly Pear Creek. Tailing and settling ponds, related to
gold mining in the Spring Creek drainage, release zinc,
copper, and cadmium to Prickly Pear Creek. Stream surveys
characterized flow regimes, water quality, and biotic
conditions in conjunction with toxicity testing. The study
objectives were to: (1) develop a data base for validation of a
toxicity persistence model; (2) assess the applicability of
data from the Prickly Pear Creek study relative to model
assumptions; and (3) assess field techniques for acquiring
model input data.
Toxicity to test organisms was primarily due to zinc and
copper in Spring Creek waters. Changes in Prickly Pear
Creek toxicity downstream from the Spring Creek conflu-
ence were primarily due to dilution and hence were
consistent with model assumptions. However, Spring Creek
was not the sole source of toxicity in Prickly Pear Creek
waters as unidentified toxicants were present in other
tributaries. Ceriodaphnia was highly sensitive to toxicity in
Spring Creek waters and provided useful model input data.
P. promelas had a higher tolerance, and bioassay data from
these organisms could not be used for model input. In the
field, nutritional problems were encountered with test
organisms using procedures described in bioassay proto-
cols for each, suggesting either a quantitative food regime
should be developed or a nonfeeding test be used in the
future.
'Lockheed Engineering and Management Services Company, Inc , P O Box
15027, Las Vegas, NV 89114
2Advanced Monitoring Systems Division, Environmental Monitoring Sys-
tems Laboratory, Las Vegas, NV 89114
Introduction
The U.S. Environmental Protection Agency's (EPA) Office of
Water Regulations and Standards, Monitoring and Data
Support Division (MDSD), is examining persistence and
degradation rates of toxic wastes in streams. MDSD is
seeking to identify methods most suitable for assessing
instream persistence of whole effluent toxicity m receiving
waters. Specifically, methods are required for site-specific
assessment of effluent toxicities, both acute and chronic,
prior to discharge, at the discharge point, and at points
downstream where dilution, degradation, and partitioning
to other compartments result in reduced toxicant concen-
trations. Particular interest centers on validation of toxicity
models designed to predict instream toxicity persistence
and validation of methods for acquiring input data for these
models.
Instream toxicity testing has recently been conducted at
several sites by EPA's Environmental Monitoring Systems
Laboratory in Las Vegas, Nevada (EMSL-LV), and by the
Environmental Research Laboratory in Duluth, Minnesota
(ERL-D). Validation of a stream dilution model will be based
on results from these investigations. Assumptions for the
model presently being assessed are: (1) toxic chemicals and
toxicity itself follow a conservative (not enhanced or
degraded) mixing behavior; (2) physical, chemical, and
biological interactions do not substantially alter toxicity at
the point of complete mixing; and (3) variations in effluent
toxicity are reflected in the varying toxicity of receiving
waters and can be described by mass balance relationships.
To provide instream toxicity persistence data to MDSD,
EMSL-LV conducted a stream toxicity study in the fall of
1983 at Prickly Pear Creek, Montana. The objectives of this
study were to: (1) develop a data base to be used for model
validation; (2) assess the applicability of data from Prickly
Pear Creek relative to model assumptions; and (3) assess
-------
field techniques for acquiring model input data. The study
consisted of short-term acute and chronic toxicity tests
using two test organisms and stream survey characteriza-
tion of flow regimes, water quality, and biotic conditions.
Prickly Pear Creek flows north from its headwaters in the
ElkhornMountainsfor approximately 64 km before entering
Lake Helena andthe Missouri River(Figure 1). Gold mining
in the Corbin and Spring Creek drainage basin (draining into
Prickly Pear Creek) began in the early 1860's. Tailing and
settling ponds remain as prominent features within these
drainages and release high concentrations of zinc, copper,
and cadmium which are carried into Prickly Pear Creek via
Spring Creek. Areas along Prickly Pear Creek were also
subjected to extensive mining operations in the early
1900's. Over 75 percent of Prickly Pear Creek was subjected
to stream bed modifications and dredging during the mining
process.
Prickly Pear Creek, Montana
Figure 1. Station locations on Prickly Pear Creek and Spring
Creek, Montana.
Procedures
Spring Creek toxicity and instream toxicity persistence in
Prickly Pear Creek were determined using static renewal
bioassays designed to measure both acute and chronic
toxicity. Test organisms were the cladoceran Ceriodaphnia*
and the larval fathead minnow Pimephales promelas.
Toxicity tests were conducted with 24-hour composite
stream and effluent water collected from September 30
through October 9, 1 983. Ceriodaphnia toxicity testing
actually began on October 1. Sampling stations are
described in Table 1
Water quality and hydrological parameters were also
measured as part of the study. For toxicity tests, grab
samples of control waters were collected each day from the
upstream Prickly Pear Creek station 011. These waters
were diluted with Spring Creek water (the toxicity source) to
obtain dilution test volumes with varying metals concen-
trationsfor comparison to ambient Prickly Pear Creek water
toxicity.
Table 1.
Location ol'Stations on Prickly Pear Creek and Spring
Creek, Montana, 1983
Station No.
Description
'Taxonomy uncertain; may be C affm/s or C reticulata x C aff/nis From
Ceriodaphnia Workshop (U S EPA Region VIM) in Fort Collins, Colorado,
March 6-7,1984, personal communication Dr Dorothy Berner, Museum of
Comparative Zoology, Harvard University, Cambridge, Massachusetts
011 Prickly Pear Creek, 1.1 km upstream from
Spring Creek confluence
Spring Creek Spring Creek, 100 m upstream from Spring
Creek confluence
013 Prickly Pear Creek, 300 m downstream from
Spring Creek confluence
014 Prickly Pear Creek, 3.8 km downstream from
Spring Creek confluence, 100 m downstream
from Dutchman Creek confluence
018 Prickly Pear Creek, 12 km downstream from
Spring Creek confluence, 3 km downstream
from Lump Gulch confluence
Results
Metal Concentrations
Spring Creek metal contributions caused significant in-
creases in metals in Prickly Pear Creek water. However,
approximately a two-fold decrease occurred between
stations 013 and 018, due primarily to tributary dilution
(Table 2). Dissolved metal concentrations were generally
well below acute criteria in all downstream tributaries.
Total recoverable cadmium, zinc, and copper concentrations
in Spring Creek and Prickly Pear Creek samples consistently
exceeded U.S. EPA acute criteria for aquatic life during the
toxicity testing period (Table 2). Concentrations of arsenic
and lead were below the aquatic life criteria at all stations.
Silver exceeded the acute criterion on October 6 at station
013, but was well below the acute criterion for all other
dates and stations. Although cadmium exceeded the acute
criterion, concentrations were below reported toxic levels
for Ceriodaphnia and larval fathead minnows. Toxicity in
test organisms was attributed to zinc and/or copper, but
Ceriodaphnia bioassays indicated that another unidentified
toxicant was present. Zinc and copper concentrations in
Spring Creek were variable over the 9-day testing period
with peak total recoverable concentrations on test days 1
and 5 (test numbers 1 through 9 refer to dates, October 1 -9)
(Figure 2).
-------
Table 2. Total Recoverable Concentrations of Selected Metals in Spring Creek and Prickly Pear, andU.S. EPA Calculated Acute Criteria
for Aquatic Life
Mean Values are 10-day averages (September 30-October 9, 1983).
Number of days criteria were exceeded are given in parentheses.
Station
Total Metals (fjg/l)
Cadmium*
Lead
Zinc*
Copper*
Silver
Arsenic
X
Flange
Criterion flange
X
flange
Criterion Range
X
Range
Criterion Flange
X
Range
Criterion Range
X
Range
Criterion Flange
X
Range
Criterion Range
011
2(6)
1-3
1.5-1.8
13(0)
7-22
74-100
100(10)
49-183
180-224
12(2)
6-13
12-15
0.6(0)
<0.2-0.9
1.2-1.9
2(0)
<0.5-11
440
Spring
Creek
7.6(9)
6-12
4.7-6.1
72(0)
44-238
291-389
2119(10)
1260-3625
464-562
84(10)
37-220
33-41
1.6(0)
0.2-3. 1
8.5-12.8
27(0)
1.5-84
440
013
5(10)
2-9
2.0-2.7
30(0)
20-54
108-155
580(10)
481-656
238-303
28(9)
12-47
16-20
1.9(1)
<0.2-11.2
2. 1 -3.5
6(0)
3-10
440
014
4(5)
1-6
1.9-2.8
19(0)
11-26
103-160
236(10)
261-372
230-308
14(3)
<6-22
15-21
0.2(0)
<0.2-0.5
2.0-3.6
4(0)
3-7
440
018
3(6)
2-9
2.2-3.2
15(0)
8-28
121-183
203(0)
169-232
255-338
12(0)
7-15
17-23
0.1(0)
-------
Table 3. Mean Number of Neonates Produced and 95-Percent Confidence Limits, Ceriodaphnia Tests 1 Through 9
Chronic effect concentrations are noted for individual tests.
Comparisons were not made between tests.
Dilution
Treatment
Spring Creek
(%)
Control x
(95% C. L.)
1%~x
195% C. L.)
2.5% 7
(95% C. L.)
5% 7
(95% C L.)
70% T
(95% C. L.)
20%~x
(95% C. L.)
Culture x
Water
(95% C. L.)
1
13.0
(11.7-14.4)
10.6
(8.4-12.8)
10.3
(8.6-12.0)
10.1
(7.3-13.0)
1.0*
(-0.5-2.5)
0
23.3
(20.1-26.8)
2
4.2
(1.9-6.5)
3.7
(2.5-4.9)
6.0
(2.5-9.4)
7.2
(5.6-8.8)
7.0
(4.3-9.5)
0*
26.5
(22.4-30.6)
3
17.6
(15.5-19.7)
20.9
(15.1-26.5)
25.9
(23.4-28.4)
25.3
(22 0-28.5)
17.7
(15.6-19.8)
0*
28.6
(26.8-30.5)
4
27.3
(21.8-32.7)
24.1
(21.2-27.0)
27.4
(26.6-28.2)
21.8
(18.3-25.4)
3.8"
0
38.5
(32.6-44.3)
lest
5
28.5
(26.2-30.8)
22 5
(16.7-28.4)
25.6
(22.6-28.6)
9.9*
(5.2-14.4)
13.8
(96-18.0)
0
37.9
(35.1-40.7)
6
33.7
(31.8-35.9)
296
(26.5-32.9)
34.2
(33.2-35.21
21.2*
(14.4-27.7)
0
0
35.8
(34 0-37.6 J
7
33.8
(22.8-28.7)
No
Data
28.7
(27.0-30.3)
22.9*
(20.1-25.9)
10
0
30.9
(28.7-33.0)
8
25.7
(11.7-14.4)
25.6
(23.2-28.1)
22.8
(17.5-28.2)
18.4
(12.2-24.3)
15.0*
(12.6-17.4)
0
25.3
(22.6-28.1)
9
28.0
(1 5.5- J 9.7)
18.8
(14.0-23.9)
23.8
(20.9-26.6)
18.9
(17.4-20.4)
13.5
(10.4-16.6)
0*
25.2
(21.8-28.6)
*Significantly different from control treatment, based on 95 percent confidence limits, indicating chronic effect level.
relationship between toxicity and metal concentrations
was poor, due primarily to the occurrence of the unidentified
toxicant.
Control water toxicity was evident in tests 1 through 3 with
significantly lower neonate production in the control tests
relative to the culture water tests (Table 3). Bioassays
conducted on water collected from the tributary streams on
October 16 revealed a potential source of control water
toxicity due to Copper Creek inflow, located 100m upstream
from control station 011. Significant difference in neonate
production (control vs. culture water) was also found in test
5, but this was probably due to nutritional differences in the
two waters. The culture water supported high concentra-
tions of algae (Closterium) and bacteria, and provided a
greater food supply for Ceriodaphnia than was available in
stream waters.
Instream and Dilution Test Comparisons
Toxicity in Spring Creek dilution tests and instream Prickly
Pear Creek tests was compared to determine if instream
changes in toxicity were due strictly to inflow of Spring
Creek water (Figures 4-7). Dilution volumes of Spring Creek
water at instream stations 013, 014, and 018 were 17.3,
7.2, and 2.4 percent, respectively, and approximated
dilution volumes of Spring Creek water used in the
Ceriodaphnia dilution tests (20, 10, and 2.5 percent).
Mortality in dilution and instream tests having comparable
Spring Creek dilution volumes showed a high degree of
Ceriodaphnia
LC-50 (48 hr.)
25
Figure 4.
34 56789
Test Number
Percentage Spring Creek water resulting in 48-hour
LC-50s and 95-percent confidence limits, Cerio-
daphnia tests. Confidence limits could not be
determined for tests 3 and 4 because mortality was
100 percent in the 10 percent and 20 percent
dilution treatments.
similarity. However, higher mortality in some instream
tests suggested an additional instream toxicant similar in
nature to the unidentified toxicant in Spring Creek. Neonate
production in dilution and instream test comparisons also
showed no significant difference in a majority of the tests
(Table 4).
-------
Acute Toxicity Cenodaphnia (48 Hr.j
Pimephales promelas
roo-
~
80-
b(J~
_
40-
-
20-
r>-
020% Spring fj 01 3 Prickly
Creek Pear
$
:;::
1
s
;;:
|
:';
p
1
1
?
yi
:-::
1
.f;
I
1
1
~
£
:':
j^j
!:|:
|
':J:
:
^
yj
;.-:
1
1
g
:|::-
1
:l
^
S'
:|.'
|
1
|
|
I
:|;
ii
?
J;
:;
Si
X'
3
!?
i
i
?"
:.
f
'':
P
?;
:;
:
|
.j{:
|
«
1
:;
':_
I
|
A
;;
;:;!
;.:
:s
.;:
S
1
1
£
$
y
:?
;-.
V1
1
ij;
i
i
F71
;!':;
;
£
;.-:
*:
£
3
1
S
3 14151 6 I/! 8
Test Number
Figure 5. Percent mortality in 20-percent Spring Creek water
and Prickly Pear Creek station 013 treatments,
Ceriodaphnia tests.
100-
80-
t 60-
ra
O
5 40-
20-
n.
i3 10% Spring Q 01 4 Prickly
Creek Pear
?.
':
''t
ij
$
n fe
i
p.
-
'<
'.
H
cn
Test Number
Figure 6. Percent mortality in 10 percent Spring Creek water
and Prickly Pear Creek station 014 treatments.
Ceriodaphnia tests.
Larval fathead minnows were less sensitive to Spring Creek
toxicity than were Ceriodaphnia. Estimated LC-50s for
fathead minnows were at dilution volumes greater than
25-percent Spring Creek water, which was greater than
dilution volumes found for instream Prickly Pear Creek
stations (Figure 8). This lower sensitivity was also reflected
in the instream station tests which showed little or no
mortality.
Toxicity was highly variable for fathead minnows. Minimal
mortality occurred in tests 2, 8, and 9. A significant decline
in toxicity in tests 6 and 7 indicated that the unidentified
substance toxic to Ceriodaphnia was not toxic to fathead
minnows. Higher toxicity in tests 1 and 5 corresponded to
elevated total recoverable concentrations of zinc and
copper (Figures 2 and 7); however, a strong relationship for
these metal concentrations and toxicity was not clearly
evident in fathead minnow data.
High control mortality occurred after the third or fourth day
and at test termination mortality was greater than 30
percent in six of the 10 tests (0, 1, 2, 4, 7, and 8). High
control mortalities are usually indicative of procedural
problems. However, mortality declined in the lower dilution
treatments with little or no mortality at either 12.5 or 25
percent suggesting that Spring Creek water was amelio-
rating conditions in the control water. This mortality decline
may have been due either to dilution of control water
toxicity or to the addition of some factor enhancing survival.
Demonstration of chronic effects in fathead minnows was
impossible due to highly variable growth rates. Growth was
significantly increased with increased feeding in a separate
feeding experiment, indicating test fish were probably
underfed and that a quantitative food regime should be
developed for future tests. Nevertheless, fathead minnows
raised at EMSL-LV from identical egg batches .showed
variations in length approaching 400 percent after 30 days.
This kind of growth variability highly influences test results
and a nonfeeding lethality test may be more appropriate for
field testing.
100-
^ 80-
(13
I 60-
§
85
40-
20-
o
D 2.5% Spwiff D 07S Prickly
Creek Pear
m n i
-
i
Test Number
Figure 7. Percent mortality in 2.5-percent Spring Creek water
and Prickly Pear Creek station 018 treatments,
Ceriodaphnia tests.
Stream Survey
Water Quality
Nonmetal stream water quality parameters revealed no
additional sources of toxicity as ammonia, cyanide, and
chlorine were below detection. All other water quality
parameters were typical of fall conditions for good quality
streams of the region.
Hydrology
Stream flow in Prickly Pear Creek increased from 11 cfs at
station 011 to 37 cfs at station 018, with tributary inflows
accounting for 62 percent of the increase. Flows increased
between stations 013,014, and 018 by approximately 5 and
8 cfs, respectively, as a result of ground water.
The percentage of Spring Creek water volume to the total
water volume at downstream Prickly Pear Creek stations
-------
Table 4.
Mean Number of Neonates Produced and 95-Percent Confidence Limits for Comparable Dilution and Station Treatments,
Ceriodaphnia Tests 1 Through 9
Comparable dilution and station treatments were 20 percent and station 013:10 percent and station 014: and2.5 percent and
station 018. Comparisons were not made between tests.
Treatment
Treatment
Treatment
Test
1 X
(95% C.LJ
2~x
(95% C.LJ
3~x
(95% C.L.)
4~j
(95% C.L.)
5T
(95% C.LJ
67
(95% C.L)
7~x
(95% C.LJ
8~x
(95% C.L.)
sir
#5% C.L.)
20%
0
0
0
0
0
0
0
0
0
013
0
0
0
0
0
0
0
0
0
10%
1.0*
(-0 5-2.5)
7.0*
(4.3-9.5)
17.7
(15.6-19.8)
3.8
13.8*
(9.6-18.0)
0
10
150*
(12.6-17.4)
1 3.5
(10.4-16.6)
014
6.6
(3.8-9.3)
0
19.2
(18.1-20.3)
1.0
3.5
(0.9-6. 1)
0
0
1.0
(-0.6-2.8)
5.8
(-1.3-12.9)
2.5%
10.3*
(8.6-12.0)
6.0*
(2.5-9.4)
25.9
(23.4-28.4)
27.4
(26.6-28.2)
25.6
(22.6-28.6)
34.2*
(33.2-35.2)
28.7
(27.0-30.3)
22.8
(17.5-28.2)
23.8
(20.9-26.6)
018
14.3
(12.5-16.0)
11.8
(9.9-13.7)
23.4
(18.8-26.8)
20.3
(13.9-26.6)
30.6
(28.0-33. 1
0
14
12.6
(2.9-22.3)
23.0
(14.0-32. 2)
*Significant difference in comparable dilution and station treatments based on 95-percent confidence limits.
100
Figure 8.
Test Number
Percentage Spring Creek water resulting in 96-hour
LC-50s and 95-percent confidence limits, Pime-
phales promelas tests.
013, 014, and 018 was 17.3, 7.2, and 2.4 percent,
respectively, based on concentrations of Rhodamine WT
injected into Spring Creek on September 23. Dye retention
time from the Spring Creek confluence to station 018 was
just over 11 hours.
Biota
Salmonid fishes were abundant at all Prickly Pear Creek
stations but there was a downstream shift in species
abu ndance from brook trout (Salvelinus fontinalis), to brook
and rainbow trout (Salmo gairdneri) to brook, rainbow, and
brown trout (Salmo trutta). The species shift in salmondis
was probably due to increased pool habitats downstream.
Previous investigations conducted during summer have
shown major reductions in both macroinvertebrate and
periphyton numbers and diversity in the Prickly Pear Creek
impact zone (station 013) and a gradual downstream
recovery between stations 014 and 018. A superficial
examination of macroinvertebrate communities indicated
no evident reduction in either species types or species
number in the impact zone. This lack of reduction may have
been a physiological response to cooler water temperatures.
Water temperatures during this investigation were approx-
imately 7°C compared to summer temperatures of 16 to
20°C.
Metals in Sediment and Tissue
Sediment metal concentrations in Spring Creek and at
station 013 were approximately an order of magnitude
higher than those found at the control station 011 (Table 5).
Table 5. Mean Sediment Metal Concentrations in Spring
Creek and Prickly Pear Creek, Montana, September
27-29, 1983
Sediment Metal Concentrations (mg/kg)
Station Cadmium Lead
Zinc Copper Silver
011
Spring
Creek
013
014
018
3
29
30
14
9
135
3612
3240
1243
668
502
4975
4937
2765
1680
133
1142
967
372
202
1
36
34
12
6
No arsenic analysis
-------
Concentrations decreased downstream but at station 018
were still substantially higher than concentrations found at
the control station.
Tissue metal concentrations were highest in periphyton
followed by macroinvertebrates and fish (Table 6). Peri-
phyton and macroinvertebrate tissue concentrations were
highest at station 013 and decreased downstream. Fish
tissue metal concentrations were not exceptionally high
and there was no substantial difference in tissue concen-
trations between stations. Previous investigations have
found significantly higher tissue metal concentrations in
most organs (kidneys, gills, brains, heart, and gonads) from
fish collected in the impact areas of Prickly Pear Creek.
However, muscle tissue did not have elevated metal
concentrations. In this investigation, whole fish were used
for tissue analyses and the inclusion of muscle tissue
probably masked metal concentrations in the organs.
Conclusions
Metal concentrations in Prickly Pear Creek were signif-
icantly increased downstream from Spring Creek, which
contributed elevated levels due to gold mine tailing and
settling ponds in the drainage basin. Concentrations of
cadmium, zinc, and copper measured over a 10-day period
exceeded U.S. EPA acute criteria for aquatic life at one or
more of the downstream sampling stations in Prickly Pear
Creek. Elevated metal concentrations were the only identi-
fiable water quality problems observed in Prickly Pear Creek
during this investigation.
Spring Creek toxicity to test organisms (Ceriodaphnia and P.
promelas) was primarily due to zinc and copper. Other
unidentified toxicants were present and Spring Creek was
not the only source of toxicity for Prickly Pear Creek waters;
however, changes in toxicity (persistence) in Prickly Pear
Creek were primarily due to downstream dilution of Spring
Creek water. Spring Creek toxicity exhibited a conservative
behavior in its downstream distribution in Prickly Pear
Creek and was consistent with toxicity model assumptions.
Sensitivity of the two test organisms to toxicity in Spring
Creek and Prickly Pear Creek was very different. Cerio-
daphnia was highly sensitive, and bioassay results were
applicable in assessing toxicity persistence in Prickly Pear
Creek. P. promelas had a higher tolerance and could not be
used in assessing toxicity persistence. Although sensitivity
of the two animals was different, both appeared to be highly
representative of toxic effects in Prickly Pear Creek native
fish and macroinvertebrate communities found in previous
studies.
Problems were encountered in the field bioassay proce-
dures used for both organisms. These problems were
related to the food regimen used in each bioassay. Cerophyl
proved to be a better food source than yeast in Ceriodaphnia
tests. Chronic toxicity was not measured in P. promelas,
apparently because of underfeeding, and either a quanti-
tative food regime should be developed for this test or a
nonfeeding test should be used in future field testing.
Table 6. Tissue Metal Concentrations in Prickly Pear Creek, Montana, September 27-29, 1983
Tissue Metal Concentrations (mg/kg)
Organism
Periphyton
Macrophyte
Macroinvertebrates
Fish
Salvelinus fontinalis
Salmo gairdneri
Salmo trutta
Coitus spp.
Station
011
013
014
014
011
013
014
018
011
011
013
014
014
014
018
011
014
018
Cadmium
1
37
9
12
1
12
4
2
<1
1
1
1
1
1
1
<1
1
<1
Lead
35
1588
175
252
18
165
47
26
3
7
10
5
8
12
10
8
6
6
Zinc
285
4640
1615
2630
326
2038
660
444
70
230
92
145
225
255
220
135
265
255
Copper
46
1190
135
330
37
276
65
37
11
20
10
14
8
16
12
10
28
7
Silver
1
19
2
4
1
2
1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
Arsenic
6
343
-
61
2
32
8
7
<1
1
<1
<1
1
<1
<1
<1
<1
<1
&U. S. GOVERNMENT PRINTING OFFICE: 1985/559-111/20673
------- |