The Fate of Mercury
in Artificial Stream Systems
Interim Report for
EPA Project Mo. R800510
April 1, 1974-
Henry J. Kania, Robert L. Knight, and Robert J. Beyers
-------
Introduction
The purpose of our research program is to determine
the fate of mercury introduced as mercuric ion into arti-
ficial stream systems on a continuous basis, and to
consider the responses of the biotic communities to this
metal. In the past two years, we have worked with water
levels of 0.01, 1.0 and 5.0 micrograms Hg per liter. A
timetable of major events in the stream research program
which includes the dosing levels and periods is given in
Table 1. At present we are not inputting mercury but
are monitoring the disappearance of-this metal from the
channels and documenting community changes.
The experimental work: is being carried out on the
United States"'Atornlc £ne'rgy~.Com!7iission's Savannah-River
Plant. The stream systems are located several miles from
the Savannah River Ecology Laboratory (SREL) of the
University of Georgia. Analytical work and sample prep-
arations are performed in a temporary laboratory building
located several hundred yards from the stream site.
The artificial stream facility consists of six
concrete block channels each three hundred feet long,
two feet wide, and one foot deep with concrete block
pools at both ends of each s-tream. The pools and channels
are lined with a 20 ml thick black polyvinyl chloride
film. Washed builders sand was distributed in the channels
to a uniform depth of two inches and water input started
in September, 1971. Flows of twenty-five gallons per
-------
-2-
rainute Into each channel, as measured by V-notch weirs,
have been maintained since that time except when water
pump malfunctions have occured. End plates at the ends
of the channels were adjusted to maintain a water depth
of eight inches over the sand. Retention times in the
streams average two hours.
Biological communities became established in the
streams from organisms blown by the wind or carried in
by animals. Mosquitofish, Gambusia affinis, were intro-
duced a month before mercury input was begun (Table l),
400 to each channel.
Water for the channels is pumped from a deep well
located near the facility. The water is treated at the
stream site by passage through limestone filled tubes in
order to increase its hardness and pH and to decrease its
free C02 content. Because of the large expenses involved,
we have not been able to follow through with plans to
Install a water treatment system which would have also
been used to increase the dissolved organic carbon
content of the water. We have also been unable to find
an economically feasible method of Insuring the avail-
ability of water at all times. Fortunately, we have had
only one extended period without the normal water input
during the experimental period. As has been discussed
in a previous interim report, the pump failure of July 4,
1972, caused a significant set back in our research
program. Any future work with the facility must include
-------
-3-
a back-up water supply and also a better means of water
treatment.
Water flowing out of the channels is passed through
beds of shredded rubber as obtained from a tire recapping
firm. We have found this to be a relatively effective
method of removing mercury from the effluent water. In
June 1972, removal was about 51$. The system was modified
and in August, 1972, removal was about 66$. These results
are based on two radiomercury uptake studies. Chemical
analyses of the water leaving the stream facility after
the level increase of August 1, 1973 indicate that about
50$ of the mercury is removed by the rubber granules.
Approximately 50$ of the mercury not removed by the
rubber granules is removed in the first 100 yards of the
effluent channel (Figure l).
Our research efforts have been divided into two
general areas: (l) to determine the fate of mercury
introduced into the channels, and (2) to determine the
responses of the stream communities to this metal. Dr.
M. C. Kerens was largely responsible for the initial
work of determining responses of the stream communities
and the research formed the basis for her Ph. D. disser-
tation, a copy of which is appended to this manuscript.
During her work the levels of mercury established were
0.01 and 1.0 ppb. Dr. Ferens decided not to stay with
the project after completing her degree, however,
-------
and there was a considerable delay before a qualified
replacement was found. This has resulted in discontin-
uities and delays in several aspects of the program.
For example, since sampling by Dr. Ferens was done only
during the low level mercury (0.01, 1.0 ppb) inputs,
we delayed increasing the levels in the 0.01 ppb channels
as long as possible. We had hoped to give her replace-
ment time to become familiar with existing communities
before the dose levels in the 0.01 ppb channels were
Increased. This was not possible because of the time
constraints on our project. The increase from 0.01 ppb
to 5.0 ppb was necessary in order to establish sufficiently
high mercury levels in the experimental channels so that
the elimination of mercury could be studied after the
inputs were stopped. In many comparisons, channels
receiving 0.01 ppb mercury were indistinguishable from
controls.
The material that follows is divided into two sections,
one of which deals with the mercury levels in the various
components of the stream system as a function of time,
the other of which deals with the responses of the bio-
logical communities to the mercury levels as established
in August of 1973. Dr. Ferens1 dissertation serves as a
summary of work done with respect to community responses
prior to the dose increase of August 1, 1973« An interim
report issued in January, 1973* for this project summarizes
-------
-5-
the data available at that time. Some data contained in
that report are not included with this document. The
final report that will appear at the end of this project
will, of course, synthesize all past work into a single
coherent document. All members of the research staff,
including Dr. Ferens and her replacement, Mr. Robert L.
Knight have been interacting so that species identification
difficulties, methodology differences and questions involv-
ing statistical handling' of the data will be resolved.
-------
Pate of Mercury
This section summarizes data regarding mercury levels
measured in various components of the stream systems. In
many cases, treatments are lumped so that the variability
between channels treated alike, the variability between
upstream and downstream locations, the variability between
the two walls of the channel, and the effects of periodic
minor catastrophes are not evident. The purpose of pre-
senting the material in this manner is to provide a
general overview of results. Methods of handling the data
in a more sophisticated manner to extract all information
are being worked out. As can be seen even in the averages,
however, variability tends to be a great problem. We do
have, of course, estimates of the variability introduced
by our sampling and analytical techniques. We have esti-
mates of the biological variability Inherent in some of
the materials sampled but these data are by no means
complete.
Our measurements of the mercury in various components
of the stream systems have been for total mercury. Since
organic forms are so important, and all possible pathways
seem to exist between mercuric ion, mercury metal, and
organic mercury, we have had samples analyzed for methyl
mercury by personnel of the EPA's Southeast Environmental
Research Laboratory at Athens, Georgia, whenever possible.
We have indications that metallic mercury may be important
in the transport of mercury between sediments, water and
-------
-7-
and air. Methyl mercury results and information implica-
ting metallic mercury as an important transport mechanism
are mentioned in the sections that follow.
General Methods and Results
All analyses have been carried out on a Coleman MAS-
50 Mercury Analyzer System modified by the addition of a
digital read-out device and fused quartz windows on the
absorption cell. We have used several different diges-
tion techniques in the 'past but presently all samples are
digested in a mixture of concentrated sulphuric and nitric
acids (5:l) and oxidized with a 6% (w/v) potassium per-
manganate solution. When large particles are involved
as with fish, invertebrates and portions of rooted
aquatic plants, the samples are allowed to stand in the
acids overnight before the permanganate is added. The
permanganate mixture is allowed to stand an additional
24 hours before being analyzed. Additional permanganate
is added, if required, as indicated by color loss. Samples
of fine particulate material, such as periphyton scraped
from the various substrates used in the channels and
removed from the sediments, are handled in a slightly
different manner in that the permanganate is added at
the same time as the acids. These procedures have been
checked against more complex methods involving refluxing,
heat treatments, and additional reagents. We have found
-------
-8-
results to be in close agreement. We do not suggest,
however, that our methods would be appropriate in more
complex systems which might be subjected to mercury
releases other than mercuric ion.
Water
The amount of water entering each channel is con-
trolled by a manual valve located Just in front of the
limestone filled treatment tubes. Flows are checked
dally against gradations on a V-notch weir, separating
the pool at the head of each channel from the channel.
Some variation in flow rate, correlated with pump opera-
tion, has been noted.
Mercury was pumped into the channels with a four
channel peristaltic tubing pump calibrated daily. Major
deviations from the desired input occurred only when the
pump failed or pumping tubes were replaced, a monthly
procedure.
Prior to August 1, 1973* only occasional mercury
analyses were made of water and then only of samples
from the 1.0 ppb channels. These analyses showed our
water concentrations were close to what was desired, at
least in these channels. After August 1, 1973, samples
were taken routinely from all channels, including controls,
Water samples were taken in the following manner
from two locations in each channel and also from two areas
-------
of the effluent leaving the stream facility. An acid
rinsed 250.ml erlenmyer flask was filled by submersion,
the contents checked for any visible particulates and a
new sample taken if particles were seen. The clear sample
was poured into an acid rinsed polyethylene bottle
containing 2.5 ml of concentrated nitric acid. Sampling
stations in the channels were located 20 feet from the
mercury input tube and two feet from the downstream end.
Samples of the effluent water leaving the rubber mercury
removal beds were taken near the ends of the channels and
300 feet downstream.
Analyses of 100 ml aliquots of both filtered (0.45A^
Gelman Type A glass fiber filters) and unfiltered water
were made. The 100 ml samples were treated with 5 ml
concentrated sulphuric acid and 1 ml '6# potassium perman-
ganate and allowed to stand at least 30 minutes before
being analyzed. Results using this technique were compared
against those obtained from a procedure which included
potassium persulphate and a heat treatment. These extra
treatments were found to be unnecessary for our samples.
The filters used contained less than 0.002/^jgs of mercury.
Filtration of standards of HgClg treated in the same
manner as the water samples showed that the filters
removed less than 1<£ of the dissolved mercury from a 1.0
ppb solution.
A summary of mercury analyses of unfiltered water
-------
-10-
collected between August 1, 1973 and January 29, 1974 is
presented in Figure 1. The data show that the input
levels were slightly higher than desired. Figure 1 also
shows that the mercury concentrations in the water at the
ends of the channels were lower than those at the head.
Uptakes based on the figures for unfiltered water were
12.8# and 22. 3# in the 1.0 ppb channels and 16.1# and
17.5/6 in the 5 PPb channels.
Although the distance between the sampling stations
in the effluent stream was about the same as between the
two channel stations, there appears to have been a greater
uptake of mercury (54.6$) from the effluent water even
though the residence time ofv the water in this section
was short. The greater uptake might be explained by the
high turbulence in this region resulting in greater
probability of contact between a mercury ion and some
portion of the stream wall. The communities in this
region are, of course, quite different from those in the
experimental channels.
Starting at noon on January 22, 197 4 and continuing
for a 24 hour period, water samples were taken every two
hours at both stations in those channels receiving a
mercury input. Since it requires two hours for a given
water masjs to traverse the distance between sampling.
stations, the difference in mercury concentrations
between a sample taken near the head of the channel at
-------
-11-
one time-and at the end of the channel two hours later
should give an estimate of the mercury uptake in the
channel during that period. These differences as a
function of time are shown in Figure 2 for the 1.0 ppb
channels and Figure 3 for the 5«0 PPb channels. Since
all differences are positive, there is no question but
that there is an uptake. However, it is difficult to
see any consistent pattern in uptake that could be
correlated with biological activity. Eased on the results'
of this intensive sampling, 12.4$ and 13-9$ of the mercury
input was removed in the 1.0 ppb channel; 15.1$ and 16.1$
of the mercury input was removed in the 5«0 ppb channels.
These figures agree with the uptakes mentioned above
calculated from several months data. There appears to
be a constant proportion of mercury removed, regardless
of the input level. This indicates that potential
binding sites for mercury are not saturated at the 5-0
ppb levels.
The percent of mercury removed by filtration was
calculated for the water samples obtained during the
January 22-23, 197^ sampling period. Results are pre-
sented in Figure 4. In all cases, the proportion of the
total mercury removed by filtration was greater at the
downstream ends of the channels than the upstream ends.
Based on the results of bacterial counts made June to
October, 1972, the results of which are summarized in
-------
-12-
Pigure 5, it Is possible that the mercury removed by
filtration is attached to these microorganisms.
At 9 AM, January 29, 197^, the mercury inputs to
the channels were turned off. Water samples were taken
Just prior to this time and every hour for the six hour
period following. The results for filtered samples along
with results for filtered samples taken three other days
during the month of February are shown in Figures 6 and
7. By the end of February, levels were at or below our
detection limit of 0.05
Air
In the initial stages of this project, we made no
plans to monitor mercury losses from the channels at
the air-water interface. Because of the acidity of the
water (pH 5-6) we believed that formation of volatile
forms would be minimal. However, mercury levels in
some components of the control streams were high enough
to indicate a transfer from the treated channels. The
possibility of leaks was remote since each channel is
covered with a single continuous sheet of PVC film, and
there is no pressure differential across the wall sepa-
rating adjacent channels.
Air sampling systems as shown in Figure 8 were
constructed and the glass boxes suspended by stainless
steel straps in all channels. Air was drawn from under
-------
-13-
the boxes"at the rate of one cubic foot per minute,
dried with magnesium perchlorate and then passed through
a gas scrubbing bottle containing 100 mis of a 10$ (v/v)
sulphuric acid, \% (w/v) potassium permanganate solution.
On several occasions, two scrubbing bottles were connected
in series. Results Indicate a 93# mercury removal in
the first bottle. The air was sampled from under the
glass collecting boxes for a period of two or three hours.
Sampling was done in the early afternoon and only on
clear days. After the sampling period, the acid perman-
ganate solutions were transferred to EOD bottles, cleared
with hydroxylamine hydrochloride crystals and analyzed,
in the usual n:?nner. Analytical. res,ults_ are summarized
in Figure 9- Results from the control channels are not
shown because they were so low. Figure 9 shows that the
releases were erratic, especially while mercury was being
input to the channels, and that the air releases continued
after the mercury inputs were stopped. There appears to
be a positive correlation between total mercury in the
sediments and air releases but the analysis of existing
data is not complete.
In an attempt to define the form of the mercury
released at the air-water interface, a trap consisting of
silver plated copper turnings was placed between the
drying tube and the gas scrubber containing acid perman-
ganate. Comparing the results from two sampling boxes located
-------
-14-
next to each other In the same channel shows the filter
effectively removes 97$ of the mercury. This type of
filter is almost 100# effective in removing metallic
mercury from an air stream. We do not as yet know its
effectiveness in removing other volatile mercury forms.
During the air sampling we noticed that the amount
of mercury collected in the acid-permanganate solutions
seemed to be related to the quantity of bubbles rising
from the bottom. Samples of this gas were collected
and quickly injected into the closed air system of our
analyzer. This resulted in a significant absorption of
the UV beam and, although we have made no attempt to
quantify the results, it suggests the presence of
metallic mercury in the bubbles. Samples of gas from
the control channels were checked with negative results.
It appears that metallic mercury in vapor form may be
released from the stream sediments with bubbles rising
to the surface. Further work is needed in this area
to better define the form of mercury released, to relate
the magnitude of the ..release to other stream properties
such as gas production, photosynthesis, respiration and
sediment concentrations.
In the intensive 24 hour water sampling mentioned in
the previous section, air samples were also taken every
two hours. Results are shown in Figures 10 and 11. The
results suggest a diurnal release pattern but more data
are needed as is a more sophisticated handling of the data.
-------
-15-
Pish
In the discussion that follows, all analytical
results are presented as means ± 2 standard errors.
Except where noted, these results are on a wet weight
basis. Because of the small size of mosquitofish
(Gambusia affinis), the entire fish was digested for
analysis. Analytical results, therefore, are not di-
rectly comparable to literature values which are usually
based on muscle analyses. In a sample of six dissected
mosquitofish, 51 ± 4# of the total mercury in the fish
was found to be in the body muscles.
The mosquitofish used in this project were removed
from a small pond formed in an abandoned asphalt parking
lot. The mercury concentration in these fish was 0.037
± .013 ppm when they were introduced into the channels
and the stainless steel cages located at both ends of
each channel. During the study, a number of fish escaped
from both the channels and cages when the screens at
the ends of the channels plugged and water levels became
excessive. The reduced numbers of fish made sampling
difficult and also greatly limited the number of samples
that could be taken.
Results of analyses of fish collected between May,
1972, and May, 1973, from the 1.0 ppb channels are shown
in Figure 12.
In April and May of 1973* all remaining mosquitofish
-------
-16-
were removed from the channels and cages and then frozen.
Most of these samples have been analyzed. Preliminary
calculations with existing data show that the fish in the
control channels contained an average of . 0?4 ± .021 ppm
mercury after one year, fish in the 0.01 ppb channels,
an average of 0.159 ± .021 ppm after one year, and fish
in the .1.0 ppb channels, an average of 5.1 ± 1.1 ppm
after one year. No difference was found between the
caged fish fed with a commercial fish food having a low
mercury concentration (0.20 ± .02 ppm Hg on a dry weight
basis) and those free living in the channels. This indi-
cates the importance of a direct uptake of mercury from
the water by mosquitofish.
Based on two sets of analyses of fish removed from
the channels after a one year exposure to water levels
of 0.01 and 1.0 ppb mercury, the portion of mercury
present as methyl mercury was 28$. The methyl mercury
analyses were made by personnel of the Environmental
Protection Agency's Southeast Research Laboratory at
Athens, Georgia. Because of the importance of organic
mercury compounds, especially methyl mercury, with
respect to human health, this project would be greatly
improved by the inclusion of investigations into the
forms of mercury that exist In the various community
components.
In May and June of 1973, new mosquitofish were
-------
-17-
released into the channels, 400 into each and 40 into
each cage. Results of analyses of fish removed periodi-
cally are presented in Figure 13 for the 1.0 ppb channels
and in Figure 14 for the 0.01 ppb channels. The 0.01
ppb channels were increased in August 1973* to 5»0 ppb.
Levels reached in the fish exposed to 1.0 ppb were about
the same after seven months as those measured the previous
year. Because this second batch of fish was introduced
into an already- contaminated <.area, the levels rose more
quickly the second year. Figure 14 shows that fish in
the channels increased to 5«0 ppb rapidly accumulated
mercury after the increase and reached an equilibrium
concentration after about four months. The equilibrium
level of about 12 ppm wet weight is double that reached
by fish in the 1.0 ppb channels.
A single set of methyl mercury analyses was made on
fish removed from the channels in mid-August, 1973«
Although total mercury levels ranged from 0.04 ppm in
the controls to 5.6 ppm in the treated channels, the
portion of mercury present as methyl mercury was on the
order of 10^. The data are extremely limited, of course,
but suggest that the proportion of total mercury present
as methyl mercury, may be a function of exposure time.
-------
-18-
Wall Communities
Shortly after the water Input to the channels was
started In September, 1971* ^0 plastic strips (10 cm &
20 cm) of the same polyvinyl chloride (PVC) material
lining the channels were suspended at upstream and
downstream locations in each channel. These strips
were removed periodically, scraped and the removed
material analyzed for mercury. Eiomass determinations
were made based on dried and ashed samples of-a blended
suspension of this material. Sampling of these plastic
strips was discontinued in October of 1973* mainly
because the large growth of a rooted aquatic plant
(Juncus diffussisimus) in the channels interferred with
the hanging strips to the extent that they could, no
longer be considered representative of the channel walls.
A method was then devised so that a known area of channel
wall could be scraped and all material collected without
any holes being cut into the plastic lining.
Results of the biomass determinations made from the
PVC strips are shown in Figure 15 and the mercury analyses
in Figures 16, 17 and 18. The variability in all these
data is primarily due to the very small amount of material
that was available for analysis and for biomass determi-
nations. It was not uncommon for the ash free dry weight
of the material removed to be less than 20 mgs.
-------
-19-
Sediments
Samples of the channel bottom have been obtained in
two ways during this project. A specially constructed
device which removes frozen cores 14 mm in diameter has
been used, and samples have also been removed from the
pyrex pans utilized by Dr. Ferens for sampling the benthic
invertebrates. The cores provided a means of determining
the vertical profile of mercury in the sediments. The
results of core analyses show that the mercury in the
stream bottom is associated exclusively with the organic
portion of the sediment which overlies the sand. The
minute amounts of mercury occasionally detected in the
lower parts of the core are probably a result of organic
matter being pushed down into the sand during sampling.
The levels of mercury measured in the organic portion
of the sediment are shown in Figure 19. The sediments in
one 0.01 ppb channel were higher in mercury concentra-
tion than the controls in all cases. The increase in
mercury in the 0.01 ppb channels in August, 1973* was
followed by an immediate increase in the sediment levels
of mercury (note scale change in Figure 19)• These
levels appeared to be still increa'sing when mercury inputs
were stopped in January of 197^- There is no way of
estimating the equilibrium levels that would have been
reached. After mercury shut down, sediment levels in all
channels appear to have declined although the data
-------
-20-
avallable at this time are inadequate for estimates of
rates of removal of mercury from the sediments.
Large core (23 cm2) samples were removed from six
positions in each channel four days after mercury input
was discontinued. The rooted plants were removed,
washed, dryed and ashed. The organic portion of the
sediments was separated from the sand, homogenized in
a measured volume of water, and subsamples taken for
biomass and mercury determinations. A summary of results
is presented in Table 2. These data show that the biomass
of the benthic communities, excluding rooted plants
were quite similar, the rooted plants were more common
in the control channels than in the treated, and the
mercury levels in channel 3 were quite different from
channel 6 although these channels were subjected to the
same treatment.
Export
We have found no satisfactory method for quantifying
the particulate matter leaving the channels and, there-
fore, the total mercury leaving the systems attached to
this material. In any attempt to balance the mercury
input with mercury output and storage, therefore, this
component of output must be estimated by subtraction.
Sampling difficulties have arisen because of the pro-
perties of the exported material, variation in these
-------
-21-
properties with season (and experimental treatment?), and
the fact that transient phenomena, such as heavy rain-
storms and high winds, seem to be responsible for
breaking loose large portions of the stream communities.
Estimates of the mercury content of the material
leaving the channels, based on samples removed from
stainless steel screens at the ends, are shown in Figures
20 and 21. Figure 21 shows a rapid increase to what
appears to be an equilibrium level in the mercury content
of the matter exported from the channels increased to
5.0 ppb mercury in August, 1973-
Rooted Aquatic Plants
In the first summer of channel operation, the bottom
communities were relatively simple with no rooted vege-
tation. By the spring of 1973* however, a great number
of small rooted plants were noted, especially in the
upstream portions of the outer channels. Although there
appeared to be two distinct species, these were later
recognized as forms of the same species, Juncus
diffusissimus Buckley. The distribution of this plant
in April of 1973 in the channels is shown in Table 3«
Quite obviously channel one received a much greater input
of seeds than the other channels. The upstream-downstream
gradient in the distribution pattern may be due to the
input of seeds to the pools above the channels. Starting
-------
-22-
in June, 1973* samples of Juneus v:ere periodically ana-
lyzed for mercury content. Plants were removed from
the channel?, the roots v;asned free of sand and separated
from the leaves. Both roots and leaves, were weighed
and placed in 100 nil volumetric flasks for digestion.
Portions of both were also weighed, oven dried at 60°C
for 24 hours., re-weighed, ashed at 400°C for 2k hours
and again weighed. Ash-free dry weight estimates were
made of the portions digested based on the records for'
the ashed samples. All mercury concentrations were
calculated on an ash-free dry '.-'-eight basis. Analytical
results for the roots are given in Figure 22, for the
leaves, in ?:'gu-re 23- Rs-sul-fcc -f-cr -the -control .channels
were very low and are not presented. Comparing .the
figures shows that the roots were much lower than the
leaves in all cases, and that the levels were probably
still increasing when the mercury inputs were stopped
on January 29, 1974.
Invertebrates
Large invertebrates have not been common in the
channels so routine samplings for mercury analyses have
not been made. The material collected in the screens at
the ends of the channels has been saved and examined for
invertebrates, however, when time was available. Animals
sorted out are identified, blotted dry, weighed, digested
and analyzed for total mercury content. The most common
-------
-23-
organisms found to date have been nymphs of the dragon
fly, Pantala hymenea, two damsel fly nymphs, Argia sp.
and Ishnura sp., and various midge larvae. Midge larvae
have been recovered in sufficient numbers on several
occassions so as to permit mercury determinations on
groups of individuals. Eased on the results of dissec-
tions of mosquitofish removed from the channels, midge
larvae form the major food item of these fish.
Levels in the midge larvae have ranged from below
detection limits in animals from the control channels
to an average of 8.6 ppm wet weight in animals from the
1.0 ppb channels to 20 ppm wet weight in animals from
the 5.,0 ppb channels,. The, data are limited and variable
but Indicate that the midge larvae have a higher level
of mercury, on a wet weight basis, than do mosquitofish
from the same channel. Analyses of these larvae, of
course, include all gut contents: Although it is not
clear what "conversion factor" would be appropriate for
converting periphyton mercury concentrations from an
ash-free dry weight basis to a wet weight basis, it
appears that in the food chain leading from the primary
producers "to mosquitofish, mercury concentration does
not occur.
-------
-24-
Future Work
Routine sampling of selected stream components will
be continued as long as differences between mercury levels
in the channels are measureable. Appropriate statistical
analyses will be made on existing data so as to extract
the maximum amount of useful information. In some cases,
it has not been possible to get sample sizes large enough
to satisfy the requirements of the usual parametric
statistical tests so that the less favored non-parametric
analyses will be necessary. Laboratory studies suggested
by the results from stream analyses, such as determina-
tion of the forms of mercury released from contaminated
sediments, will be initiated. The formulation of a mass
balance model describing the movement of mercury within
the channel systems will be attempted.
-------
Community Responses
The study of the algal and benthic insect communities
of the artificial stream systems between January, 1972
and April, 1973 was largely the effort of Dr. M. C.
Ferens and served as her doctoral research. Her methods
for sampling plexiglass plates and benthic dishes are
summarized in her dissertation which is appended to this
report. All of her algal data are presented as analyses
of responses of individual species and no total numbers.
are given. Shannon-Weaver diversities were calculated
for both the algae and insects that she counted.
Between April and October of 1973 there was no
quantification of the periphyton and benthic communities.
Study of these communities was recommenced in October
of 1973* about sixty days after the mercury concentrations
were increased to 5«0 ppb in the former low-level streams
(0.01 ppb). At that time a study of two substrates in
addition to the plexiglass plates was then started in
order to analyze time responses of periphyton exposed to
mercury. Glass microscope slides were suspended in the
channels to study the initial colonization of the peri-
phyton, and small areas of the channel walls were scraped
and the periphyton quantified to study a substrate undis-
turbed since stream construction. Also, taxonomic treat-
ment of the algal species present was continued. Dr.
Ferens listed the species she observed in Table 1 of her
dissertation (see appendix). A list of the most common
-------
-26-
species observed since October, 1973 is given in Table 4
of this report. A complete listing of algal species
including the rare forms will be included in the final
report.
Methods
Growth on Artificial Substrates Since October, 1973
Three artificial substrates were examined for com-
position of periphyton communities. Glass microscope
slides were suspended in the streams at four different
times and changes in the thickness of the algal growth
(in two cases) and the species composition (in three
.cas.es) ..we,re -studied. .Vertical jplexiglaas plates .placed
in the channels by Dr. Ferens in spring of 1973 were
examined at four different times for the density and
composition of the algal community. Also, at four times,
the algae attached to the stream -walls were quantified.
The last set of data from these artificial substrates
was collected after the input of HgClg had been ceased.
Five washed microscope slides were suspended in the'
upper part and five in the lower part of each stream on
September 29. Starting on October 9* one of these slides
was removed each three or four days for determination of
the thickness of the attached communities. This deter-
mination was made by use of a water immersion 40x lens
and a calibrated fine focus knob. The lens was focused
-------
-27-
up until the longest filament or accumulation of cells
in the field was in 'focus and the reading on the knob
noted. The lens was then focused down until the cells
attached to the slide were in focus. The difference
between the numbers on the fine focus knob could be
directly read in microns. Ten fields were measured on
each microscope slide.
A second set of washed slides was placed in the
streams on October 31 with 15 slides at each location
( 30 per stream). At three to seven day intervals one
elide was removed at each location and the thickness
of the algal community was determined as above. Then
a cover slip was placed on the lower end of the same
slide and a count was made of the algae, beginning at a
constant distance from the lower end of the slide. Enough
areaswas covered to give a count of at least 200 cells
although many counts were taken beyond 500 cells. All
cells in colonial or filamentous, forms were counted
individually. The area counted was measured by a stage
micrometer and a calibrated ocular micrometer. Numbers
were converted to cells per square millimeter. Two more
sets of eight glass slides each were placed in the streams
on December 7, 1973, and on January 30, 197*1. The second
set was suspended in the streams at a time when water
levels of mercury were less than 0.2 ppb in.all of the
streams (the mercury input was shut off on January 29*
197*0. On these two sets of slides the thickness of the
communities was not determined because the more accurate
method of directly counting the cells was used.
-------
-28-
Plexiglass plates were removed from upstream and
downstream locations on four occasions (December 12,
January 3» January 31* and February 19). These plates
were carefully scraped clean with a rubber scraper and
the material removed was diluted into a known volume of
water. This water was blended to achieve a homogeneous
mixture and subsampled for counting. Slides for count-
ing were made by placing a cover slip over a drop of
known volume and the area of the count recorded. The
number of cells counted in the given area could be readily
extrapolated to an area.of the original plexiglass plate.
On four occasions (December 14, January 8, February 5*
and March 4J samples were removed from the PVC lining of
.each s.tr.-eam «at .upper.-and lower 1-oca-tions. 'A-known'area
O
(59 cm ) was scraped into a measured volume of water in
a three-sided plexiglass box. These were blended and
subsampled for counting and the number of cells counted
was converted as above to cells per square millimeter.
Benthos
Benthic insects were collected on November 29*-1973*
and January 11, 197^* using the some method of removing
pyrex dishes as was described by Dr. Ferens (see appendix).
The insects were preserved in 80$ ethanol until they
could be sorted and counted.
Diversity
The Shannnon-Weaver diversity index was calculated
using the following formulae given by feilou (1966):
-------
-29-
H« = -(n±/N) Iog2
J =
where S is the number of species, N is the total number
of cells counted, n is the number of cells in the ith
species, H1 is the Shannon-Weaver diversity,'- andJJ is the
evenness.
Results
Glass Slides
The data for the accumulation of periphyton on the
microscope slides is summarized in Figures 2k and 25. In
the first run (Figure 24) the thickness of the algal
communities on the glass slides in the control streams
and in the low level mercury streams (l.O ppb) was similar
throughout the period of study and still increasing on
the 24^n day of colonization. This increase was due to
the two filamentous algae Oedogonium reinschii and
Gloeotila turfosa. In the high-level streams the thick-
ness leveled off after about nine days and did not in-
crease up to the 24 day.
In the second run where the thickness of the algal
communities on the slides was measured,(Figure 25), the
growth in the control streams was greater than in the
low-level streams until November 25 when both showed a
-------
-30-
decllne In thickness. The high level streams had much
less growth throughout the study. In this second run
the accumulation of cells seems to be-much less than in
the first run. This is largely due to the fact that
the filaments of Gleotila turfosa were ignored because
they were observed to be unattached and would drift
about under the water-immersion lens giving highly
variable results. The greatest thicknesses seen in the
second run are due largely to long filaments of Oedogonlum
reinschii which are attached to the slide by basal hold-
fast cells. This species was abundant in all of the
streams receiving a mercury input.
The actual counting and identification of the algae
on the slides gives a-more 'reliable picture of the changes
in the algal communities that were occuring. The results
of two sets of microscope slides that were placed in the
streams prior to the shutdown of the mercury input are
presented graphically in Figures 26 and 27. These
graphs summarize the cell totals, with each point
being the average of four counts (two slides from each of
two replicate streams). Looking only at these totals
the two experiments gave very similar results. In both
cases there was an initial period- of about 40-50 days
when the total number of colonizing algae was increasing
and was inversely proportional to the mercury input. The
difference between the controls and the low-level slides
-------
-31-
was not as pronounced as the difference between the
1.0 ppb and 5.0 ppb treatments. At the end of this
initial period all cell concentrations were approximately
equal. During the second period the trend observed
earlier reversed in that-the higher cell counts were
found in the channels receiving the 5-0 PP*> mercury input.
In the second of these experiments the slides were
examined once after mercury•input.was shut off. This last
count confused the picture because the algal population
on the slides previously exposed to the high mercury
levels declined while the other slides continued their
trend. These data indicate that the initial rate of
colonization of the glass slides was seriously affected
by the level of mercury in the water. The second trend
seen on the graphs needs further examination.
Figure 28 presents the data for total periphyton
cell densities accumulating on glass slides since the
shut down of the mercury input. In this data the initial
trend of a retardation of growth in the high-level streams
is not apparent. However, the second trend seen in
Figures 26 and 2? is apparent in this post-mercury
experiment. These data lend support to the statement
made above that the initial colonization rate on this
glass substrate is directly affected by the level of
mercury in the surrounding water. It also indicated
that the second trend observed earlier is the result of
an indirect effect of the original mercury concentrations
in the streams. This indirect effect has also been
-------
-32-
observed on the more permanent substrates as will be
discussed in the next two sections;
Plexiglass Plates
The periphyton present on plexiglass plates located
in the artificial streams have been studied since be-
fore the mercury input was begun in May, 1972. Prom May,
1972,until August, 1973jthe mercury concentrations in the
streams were 0.01 ppb and l.Or.ppb. On August 1,1973* levels
in the 0.01 ppb channels were changed to 5.0 ppb and
maintained at thatllevel until the mercury input was
shut off in January, 197^- Total cell counts are not
available for the period prior to December, 197^« However,
Ql-versl-ty and biomass values of
-------
-33-
iments where the amount of algal growth seems to be
proportional to the mercury level in the water.
Biomass data collected from these same plexiglass
plates is presented in Figure 30 and tends to confirm
the results of the direct cell counts. In all samples
except one, there was a higher biomass associated with a
higher mercury input level. This figure also shows that
the biomass was declining during these four winter months.
Figure 31 illustrates the total mercury concentrations
in these samples on a dry-weight basis. If this figure
is compared to the preceeding one, it is observed that
as the biomass decreased during this period, the mer-
cury concentration tended to increase. This observation
will be discussed later.
The calculated values for the H1^diversity are
presented in Figure 32. This figure shows that the com-
munities receiving the highest mercury levels had the low-
est diversity or uncertainty of prediction. The controls
were significantly higher than the high level samples
in three out of the four cases. This trend is consistent
with the data accumulated by Dr. Ferens during the previous
year when the mercury concentrations were much lower. The
diversity differences were smaller, yet they were signi-
ficantly lower in the 1.0 ppb streams. The low diversities
found in the 5.0 ppb samples indicate that the higher cell
densities shown in Figure 29 are the result of the in-
-------
creased dominance of a few species. The species that
was largely responsible for this dominance was Oedog-
onium reinschii, which was also the species that most
frequently dominated the glass slides and wall lining
during this study. In the previous year, Oedogonium
reinschii was found to be most abundant during the months
from October to April and to comprise an important
percentage of the algal community(about'20# of all of the
algae observed). This filamentous alga is well adapted
to the initial colonization and continued growth on a
substrate because it possesses a basal hold-fast cell
and grows rapidly.
Stream Walls
Although strips of PVC plastic Identical to the liner
of the channels had been harvested during the previous
year for biomass and mercury analysis, no quantitative
determinations of the algal communities were made until
the end of 1973 when two counts were made before the
mercury was shut off and two since then (Figure 33)• The
cell density was found to be consistent at each treatment .
level during this brief time with the high level streams
consistently having a mean 30 times or more greater than
the control streams and 10 times greater than the low
level streams. The cell densities in the low level streams
were significantly higher than the densities in the control
streams in all of the four counts.
-------
-35-
These data are another example of the trend seen In the
glass slides and on the plexiglass plates where the cell
densities were greater with higher mercury concentrations
in the water.
The H1-diversities for these samples are plotted
in Figure 3^- It was found that while the mean for
the control samples was higher than the mean for the 5-0
ppb samples in 3 of the 4 cases, it was significantly
greater in only one case, and in no case was it signi-
ficantly different from the 1.0 ppb samples. These data
indicate that although the cell density was much higher
in the high mercury level samples this was not Just the
result of one species increasing disproportionately in
abundance, out the increase in numbers of several species.
The two species that were found to be. most abundant on
the walls in these high mercury level streams were Oedogon-
ium reinBchil and Meismopedia punctata.
Benthos
The data describing the two collections of benthic
insects is summarized in Figure 35 in the form of H1-
diversity and estimated numbers of insects per square
meter. Although the mean diversity -was lowest in the
high level streams in each case, this difference was not-sign-
ificant. As far as the density of the individual insects
in each stream, it was found that the low level streams
had the most dense populations and the control and high i
-------
-36-
level streams had similar but lower densities. The re-
sults of" the work done previous to the increased input
of mercury in August, 1973* indicated no treatment affect
in upstream samples and an increased H1 for the 1.0 ppb
treatment in the downstream samples. A more intensive
benthic sampling program is needed.
Continuing Work
The intensive gathering of data up to this point,
which will continue for the next several months, has not
allowed an in-depth analysis of the algal population
dynamics. Statistical analysis of differences due to
mercury treatment will be carried out on all data. Res-
ponses of dominant.species to mercury treatment and time
of colonization will also be analyzed. It is possible
that air releases of mercury may be related to the algae
on the bottom of the streams. Total algal counts do
not take into consideration species' characteristics; there-
fore, calculations will be made of total surface areas and
volumes for the different species present. Also, it will
be possible to calculate several other diversity indices,
importance values, and density coefficients.
Several species of algae have demonstrated obvious
responses to the treatment levels and consequently need
to be investigated in greater detail. The following ob-
servations were made: l) One species of green alga, Micro-
thamnion strictissimum, was found only in channels 3 and 6
-------
-37-
which received the high levels of KgClg throughout the
experiment. Since the mercury input has been shut off,
this species seems to be declining in abundance. While
no conclusion can be supported without culturing work,
there seems to be an obvious effect here; 2) Another green
alga, Stigeoclonium elongatum, has been very abundant in
the control streams throughout the experiment and notici-
ably less abundant in the low level streams. It had been
almost nonexistent in the high level streams -until the
mercury input was turned off. It has been a common
species in all of the streams on the latest set of glass
slides. 3) One species of filamentous algae, Gloetila
turfosa, has been abundant in all of the streams during
this study. However, morphological changes have been
consistently present in the high level streams until input
of mercury was turned off. At present the misshapen cells
are much less common.
The species mentioned above as well as several other
species that have been quite abundant need to be studied
in greater detail. Therefore, laboratory culturing of
the major species that can be isolated has been started.
Once these species have been isolated into pure cultures,
the effects of varying mercury concentrations can be
observed -with a minimum of interference from other
factors. Continuing study of the natural algal populations
in the streams will include at least one more colonization
-------
-38-
experiment with glass slides, and the monthly determina-
tion of the populations on the plexiglass plates and
PVC wall lining through the spring. At least two more
sets of benthic dishes will be sorted to quantify the
insect populations during the spring months.
Other projected work includes the identification
to species of several algae that have not yet been
identified, and the identification of a large number of
organisms that have been found in the streams during
this study. Specific expertise is necessary for a re-
liable taxonomic treatment of most of these, so specimens
will be sent to experts.
A number of groups have not been considered in
..detail .but .-may be important -in -assessing .the .-impact .of
mercury on stream systems. Also, a more extensive
knowledge of the structure of the stream communities
will permit more detailed planning for future research
projects.
-------
Table 1
Timetable of major events
during the artificial streams mercury project.
September, 1971
April, 1972
May 15, 1972
July
, 1972
May, 1973
August 1, 1973
October 1, 1973
V/ater input started to the channels at 25 gpm.
400 mosquitofish (GambusJa affinls) introduced
into each channel and 40 into each cage.
Mercury input started to establish water
levels of 0.01 ppb in two channels and 1.0
ppb in two channels.
Water pump out of operation, flows greatly
reduced in channels.
Remaining mosquitofish removed from streams
and cages. Last sampling of periphyton and
benthic communities by Dr. M. C. Perens.
New fish placed in channels and cages.
Dose change, 0.01 ppb levels in two channels
raised to 5.0 ppb.1
Mr. Robert L. Knight hired to take over work
with community responses.
January 29> 197^ All mercury inputs stopped.
-------
Table 2
Summary of data obtained from six sediment
core samples from each channel taicen four
days after mercury input to the channels
was stopped.
Channel
Number
1
2
3
4
5
6
Treatment
control
1.0 ppb
0.01-5,6 ppb
control
1.0 ppb
0.01-5.0 ppb
Mercury concentration
in organic portion on
ash free-dry wt. basis
1.93 PPm
640
1462
1.57
803
763
Average biomass
in organic portion of
sediment (ash free dry wt. )
166 g/m2
178
149
170
136
165
Total
mercury in
organic portion
of sediment
0.018 g
6.36
12.1
0.014
6.08
7.02
Ash free
dry wt. of
plants
171 g/m2
71.7
52.3
96.0
48.9
72.6
-------
Table 3
Distribution of Juncus diffuslssi^us Buckley
on April 2, 1973, in tne artificial streams
Distance from
input weir in
feet
0-25
25-50
50-75
Tp-iuO
100-125
125-150
150-175
175-200
200-225
225-250
250-275
275-300
Total
CHANNEL NUMBER
1
control
"53
261
144
Do
97
124
222
70
38
26
49
236
2
1.0
ppb
112
131
30
23
13
18
31
13
6
0
17
18
3
0.01
PPb
30
8
3
•^
3
11
20
3
0
3
2
5
4
control
170
39
16
10
11
34
33
6
2
7
2
31
5
1.0
PPb
139
10
9
5
8
18
25
1
4
1
12
5
6
0.01
PPb
294
40
20
11
5
13
16
3
2
5
5
9
-------
TAELE b: Common algae of the artificial streams from
October, 1973 to March, 197^
Chlorophyta
Cholorococcum humlcola (Naegeli) Rabenhorst
Cosmarl un s s phaeros porum Nordstedt
C. laeve vnr. sop'cenirl onale V.'ille
C. viride var. minor V.'est
Eremospnaera viridJ s de Eary
Gloeotila Uuri'osa Skuja
Hormidium suDti1e (Kuetzing) Heering
Microspor? cuaclrata Hazen
Mi crotharnnion scrictissirnum Rabenhorst
Mougeotia sp.
Oedogonium reinschli
Scenedesmug acutiformis Schroeder
Spondylos iurn planu"n V. a G. S. West
Stigeocloniutn elonga turn (Hassall) Kuetzing
Chrysophyta
Chlorocloster minimus Pascher
Chrornulina pseudone'culosa Pascher
OphJocytlum deceruusn var. minor Prescott
Navlcula notha. /allace
Cyanophyta
Anabeana minutissima Lemmermann
Calothrix braunii Bornet & Plahault
MerismoaedJa nunctata Meyen
OscillatorJa geminata Meneghini
-------
6HI
5-
1 „. H6 T
H3
T6
4-
T3
O)
X
Q.
Q.
3-
2-
1-
I
H2
I
T2
H5
I
T5
I
El
I
E2
Figure 1. Results of analyses (X ± 2SE) of unfiltered water samples collected
from upstream (H) and downstream (T) positions in each channel and from two
positions in the effluent stream (E) between August, 1973* and February,
-------
o
0)
Q_
CO
o
V)
E
0
CO
O
• Channel 2
* Channel 5
0.3-
Figure 2. Mercury uptake January 22-23, 1972*, In channels receiving
an input of 1 ppb Hg as determined by upstream-downstream differences
in water concentrations.
-------
CO
O
»_
u
'E
2.0-
£ 1.5-
0
a
CO
E
o
0.5-
• Channel 3
<*> Channel 6
I
6
8
10
12
i
4
PM
I
6
AM
10
12
Figure 3. Mercury uptake January 22-23, 1974, in channels receiving
an input of 5 ppb Hg-as determined by upstream-downstream differences
in water concentrations.
-------
C
o
501
40-
30-
13
T>) locations in channels receiving a
mercury input.
-------
a.
<
lil
iioo-
1000-
9ocr
800-
70CT
600-
500-
400-
300-
200-
100-
(J
14 36
Upstream
25
14 36
Downstream
CHANNEL NUMBER
25
Figure 5. Total bacteria per ml (X ± 2SE) at two locations in the stream
channels, based on results of daily determinations made June 1, 1972,
to October 10, 1972.
-------
5.0-
4.0-
• Upstream
fe Downstream
O
CL
3.0-
to
E
E
Cn
O
u
"E
2.0-
1.0-
AM
January 29 1/31 2/5 2/14 2/28
Figure 6. Decrease In water concentrations of mercury at upstream and
downstream locations in channels having received an input of 5 ppb Hg until
9 AM, January 29,
-------
I.OH
0
Q.
O)
IE
en
E
D
i-
CD
O
i_
u
oUpstream
® Downstream
0.5-
AM PM
January 29
1/31 2/5 2/14 2/28
Figure 7- Decrease in water concentrations of mercury at upstream and
downstream locations in cnannels having received an input of 1 ppb Hg
until 9 AM, January 29, 1972*.
-------
to
vacuum
pump
f lowmeter
H2S04
<— KMn04
Figure 8. Ai- sampling system.
-------
Dec.
1973
I Channel 2
® Channel 5
* Channel 3
• Channel 6
January
1974
February
°f
-------
D
O
a
IE
*o
in
E
a
o
*_
w
0.2—
0.1-
o Channel 2
^Channel 5
I
3
I
I I
7 9
PM
11
I
3
1 \
AM
9
11
January 22-23,1974
Figure 10. Mercury releases in a 24 hour period over a two square foot
area of stream surface of channels receiving a input of 1 ppb Hg.
-------
1.0-1
D
O
Q.
CT)
en
£
CO
o
0.5-
o Channel 3
©Channel 6
I
PM
IT
7
I
9
11
11
AM
January 22-23, 1974
Figure 11. Mercury releases in a 24 hour period over a two square foot
area of stream surface of cnannels receiving an input of 5 ppb Hg.
-------
B5
*»
o
E
Q.
Q.
10-
9-
8-
7-
6-
5-
4-
3-
2-
1-
* Channel 2
© Channel 5
M ' J
A ' S ' O
1972
' D ' J ' F ' M ' A ' M
1973
Figure 12. Levels of mercury in mosquitof ish removed during the first
year from the channels receiving a mercury input of 1 ppb.
-------
D)
'55
10-
9-
8—
7-
6-
Q. •
EL
1-
Channel 2
Channel 5
June July Aug. Sept. Oct. Nov. Dec. Jan. Feb.
1973 1974
Figure 13- Levels of mercury in mosquitofish removed during the second
year from the channels receiving a mercury input of 1 ppb.
-------
20-
18-
16-
14-
12-
6
4-
2-
S Channel 3
& Channel 6
June July Aug. Sept. Oct.
1973
Nov.
Dec.
Jan. Feb.
1974
Figure 14. Levels of mercury in mosquitoflsh removed during the second
year from the channels receiving a mercury input of 0.01 ppb until
August, 1973, when the levels we increased to 5-0 ppb.
-------
cl
o
"S
£
-------
20-
15-
CD
E
Q.
CL
10-
5-
A
ON
A
1972
1973
Figure 16. Average mercury concentrations, on an ash-free dry weight basis,
in communities scraped from PVC strip suspended in the control channels.
-------
4000-J
3000—
2000—
£L
Q.
1000-
A'S'O'N'D'J'F'"'
1972
A'M'J'J'A'S'O
1973
Figure I?. Average mercury concentrations, on an ash-free dry weight basis,
in communities scraped from PVC strips suspended in channels receiving a
mercury input of 1.0 ppb.
-------
E
Q.
Q,
100-
80-
60-
40-
20-
-4000
-2000
A'S'O'N^'J'F'M'A'M'J'J'A'S'O'
1972
1974
Figure 18. Average mercury concentrations, on an ash-free dry weight basis,
in communities scraped from PVC strips suspended in the channels receiving
a mercury input of 0.01 ppb until August 1, 1973, and an input of 5.0 ppb
after this date.
-------
1500-1
1000-
500-
E
CL
CL
o .01 ppb & 5.0 ppb
10CH
20-
SONDJ FMAMJJ ASONDJ F
1972
1973
1974
Figure 19. Mercury levels, on an f>«?h-free dry weight basis, In the organic
portion of the channel sediments.
-------
20-
Q.
Q.
IOH
S ' O ' N 1
1972
F ' M1 AT
J r J ' A ' 5 ' O r N
1973
D ' J l
1974
Figure 20. Average mercury content, on an ash-free dry weight basis,
of material collected from the end screens of the control channels.
-------
2000-
E
CL
Q.
lOOO-i
.Olppb & 5.0 ppb
• 1.0 ppb
Hg
input
increased
1972
1974
Figure 21. Average mercury content, on an ash-free dry weight basis, of
material collected from the end screens of channels receiving a mercury input.
-------
E
Q.
Q.
100-
80-
60-
40-
20-
o 1.0 ppb
* 5.0 ppb
June July Aug. Sept. Oct. Nov. Dec. Jan.
1973 1974
Figure 22. Mercury concentration, on an asn-free dry weight basis, in roots
of Juncus diffusisslmus collected from the channels receiving mercury inputs.
Feb.
-------
1000-
June July Aug. Sept. Oct.
1973
Nov. Dec. Jan. Feb.
1974
Figure 23. Mercury concentration, on an ash-free dry weight basis, in leaves
of Juncus dlffusissimus collected ~-om the channels receiving mercury inpu<-
-------
400-
300-
Ift
c
o
200-
100-
Q Control
a 1.0 ppb
•it 5.0 ppb
I
5
I
10
I
15
1
20
I
25
October
Figure 24. Changes In thickness of periphyton on glass slides exposed to
different mercury levels, started September 29* 1973*
-------
o Control
D 1.0 ppb
* 5.0 ppb
200-
C
o
.5 looH
I
5
10
15
20
1
25
I
30
November
Figure 25. Changes in thickness of periphyton on glass slides exposed to
different mercury levels, started October 31, 1973.
-------
* Control
o 1.0 ppb
B 5.0 ppb
2000-
1500-
E
<0
500-
November
December
January
Figure 26. Changes in periphyton density on glass slides exposed to different
mercury levels, started October 31, 1973.
-------
Figure 27. Changes in periphyton density on glass slides exposed to different
mercury levels, started December 7, 1973.
2000-
CS
E
£
U)
"53
1500-
1000-
500-
Decernber
January
Hg
input
off
@ Control
B 1.0 ppb
•* 5.0 ppb
February
-------
2000-
1500-
E
E
X
in
1000-
500-
10 15 20 25
o Control
ta 1.0 ppb
* 5.0 ppb
Jan.
February
March
Figure 28. Changes in periphyton density on glass slides after mercury input
was stopped, started January 30, 1974.
-------
o Controi
a 1.0 ppb
* 5.0 ppb
10000-
s
E
^ 5000-
co
1VMB
••§•• •••
u
•"
1000 —
T
»
i
I
i
n
I ! ! J
«g
input
off
- I T
•v
Tr
!
t 1
L T J-
Dec 12 Jan 3 Jan31 Feb 19
1973 1974
to
-------
• Control
=» 1.0 ppb
* 5.0 ppb
3 -
CM
E
X
*-
JZ
0)
0)
£ 1-
I
.£
«/)
0
10
E
o
D)
Nov Dec
1973
Jcsn Feb
1974
f^!Jre^32' !?™phyton biomass on plexiglass plates
exposed to different mercury levels.
-------
7000-
6000-
5000-
4000-
E
0.
CL 3000-
2000-
1000-
ra 1.0 ppb
* 5.0 ppb
Nov Dec
1973
Jan
Feb
1974
Figure 31- Concentration of mercury on an ash-free dry weight
basis of periphyton scraped from plexiglass plates exposed to
different mercury concentrations.
-------
0)
O
Q_
to
•>-
IB
3.0-
2.5 —
2.0-
1.0-
0.8.
Dec 12
Control
1.0 ppb
5.0 ppb
x + 2SE
a.
0>
O
Jan 3 Jan 31
Sampling Date
Febl9
Figure 32. H1 diversity (X±2SE) of periphyton communities inhabiting
plexiglass plates exposed to different mercury levels.
-------
E
E
fl
60000-
500.00-
40000-
30000-
20000-
10000-
• Control
as 1.0 ppb
© 5.0 ppb
Dec 14
1973
Hg
input
off
Jan S
or 4
Feb 4
1974
density (X+2SE) on channel walls exposed to different
-------
u
X
to
e Control
a 1.0 ppb
*5.0 ppb
3.0-
2.5-1
2.0-
1.5-
1.0-
Dec 14
ei
Hg
i nput -L
off
I
Jan 8
Feb 5
Mar 4
-------
2.5-
0
D
•D
*; 2.0-
-0
C
1.5 -
• ••
-Q
1.0-
• Control
• 1.0 ppb
* 5.0 ppb
i
(
>
I
t
i
•
"1
<
<
l
cs 2000-
0
0
£
i
>
i
1500-
^
J2
c D
D 1000-
>
13
r:
.- 500-
•
-tr-~
I
y
y
/
/
/
y
/
j
/
/
y
^
C] Control
VA, 1.0 ppb
i
g
X
/s
o
X
1
5.0 ppb
1 >
/,
/ ,. -t
./ /o
/ i
/ x
y ]>
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
PROitG SOUTHEAST ENVIRONMENTAL RESEARCH LABORATORY
ATHENS. GEORGIA 30601
June 13, 1974
Mr. Elton Homan
Office of Toxic Substances (WH-557)
U. S. Environmental Protection Agency
Washington, DC 20460
Dear Mr. Homan:
Enclosed is an Interim Report on "The Fate of
Mercury in Artificial Stream Systems," EPA Research
Grant &R800510, by Kania, Knight, and Beyers,
University of Georgia. This report presents data on
the transport and fate of three levels of mercuric ion
(0.01, 1.0, and 5.0 ppb) introduced continuously into
our artificial stream channels located at the Savannah
River Plant, AEC, Aiken, South Carolina. The report
presents data on the transport to and accumulation of
Hg in the following areas:
• Hg removed from the water
• Hg loss at the air-water interface
• Bioaccumulation of Hg by mosquitofish
• Bioaccumulation of Hg by attached wall growths
• Accumulation of Hg in organic portion of
sediments
• Accumulation of Hg on exported detrital
material
• Bioaccumulation of Hg by roots of aquatic plant
• Bioaccumulation of Hg by leaves of aquatic
plant
-------
-2-
• Changes in periphyton density due to Hg inputs
• Diversity H1 of periphyton communities exposed
to two levels of Hg
• Diversity of aquatic insects exposed to two
levels of Hg.
This information will be published as a part of
the final grant report next spring; however, because of
the urgent need for this type fate data we are
distributing a few copies of this interim report now.
I would appreciate it however if you would respect the
prerogative of the authors until the information is
published.
We also have a paper in progress by Dr. Holm e_t
al. of our staff, giving both total and methyl mercury
concentrations in several food chain organisms. This
work was done in cooperation with Kania et al. and the
paper is expected to be completed in the near future.
Hope you find the material useful.
Sincerely,
Walter M. Sanders III, Ph.D.
Chief
Freshwater Ecosystems Branch
Enclosure
------- |