HomeMy WebLinkAboutAPA1951-
-
.-
'I
.....4
INCREASES IN FISH MERCURY LEVELS IN LAKES FLOOQED BY THE
CHURCHILL RIVER DIVERSION.NORTHERN MANITOBA
by
R.A.Bodaly,R.E.Hecky and R.J.P.Fudge
Department of Fisheries and Oceans
Freshwater Institute
501 University Crescent
Winnipeg,Manitoba R3T 2N6
UNIVERSITY OF ALASKA ....' .
ARCTIC ENVIRONMENTAL INFOFtMAilt:>N
AND DATA CENTER
707 A STREET
ANCHORAGE,AK 99501
TK
,/4;)5
0tlS c>sE
*~~A-J3
no.Iq-;-l
Running head:Fish mercury increases in flooded ·lakes
~l".
-----_._-
I'••
ABSTRACT
Bodaly,R.A.,R.E.Hecky,and R.J.P.Fudge.1983.Increases in fish mercury
levels in lakes flooded by the Churchill River diversion,northern
Manitoba~Can.J.Fish.Aquat.Sci.-
Reservoir creation has often been inferred as a cause of elevated fish
mercury concentrations.Increases in fish muscle mercury levels,occurring
.....coincidentally with flooding,are documented for three lakes affected by the
Churchill River diversion for which pre-and post-impoundment data were
available.For example.pike ([sox lucius)and walleye (Stizostedion vitreum)
muscle mercury levels from Southern Indian Lake,which was increased in-
-
surface area by 21%by flooding in 1976,increased from baseline values of 0.2
-0.3 ~g 9_1 prior to flooding to 0.5 -1.0 ~g_l in 1978-1982.Muscle
it"
mercury levels from predatory species (pike and walleye)from all ten lakes
tested in the Churchill,Rat and Burntwood valleys flooded by the Churchill
-diversion are near to or exceed the current export marketing limit of 1.0 ~g
9_1 •Because mercury levels in fish from nearby unf100ded lakes have not
shown recent increases,atmospheric fallout of the metal does not appear to be
the cause of the problem.Also,there are no known industrial sources of
mercury in the area and no agricultural activity.Post-impoundment mercury
-..-..-
"ld'"
~
ooo
l!)
l!)
'"("I)
("I)
levels in predatory fish appeared to be related to the flooded terrestrial
area compared to pre-impoundment lake area.They were highest (l.lS -2.90 ~g
g_1)in Rat and Notigi lakes which were increased in surface area by 282%were
lower (0.60 -1.53 ~g 9_1 )in lakes immediately below Notigi Reservoir
increased in surface area by 31-37"',and were lowest (0.45 -;1.03 !-lg g_l)in
W\Southern Indian and Wuskwatin Lakes,increased in surface area by 13-21%.
-ARLIS
Alaska Resources
Library &Information SerVices
Anchorage.Alaska
.....Fish mercury levels responded quickly to impoundment,increasing
noticeably within 2 - 3 yrs.Declines in mercury concentrations had not,in
"""general.taken place within 5 - 8 yrs of impoundment,with the exception of
\'Ihitefish (Coregonus clupeafonnis)from Southern Indian lake.
It is hypothesized that observed fish mercury level increases were due to
the bacterial methylation of naturally occurring mercury found in flooded
soils.
Key words:lake whitefish;walleye;northern pike;mercury,impoundment,
river diversion;mercury methylation •
...,.
....
.....
....
1.
Table
LIST OF FIGURES
Area of Churchill River diversion in northern Manitoba including lakes
and sample sites referred to in text •
LIST OF TABLES
1.Changes in water level and surface area of several lakes affected by
the Churchill River diversion project.,
....
....
-.
2.
3.
4.
Fish mercury levels (~g 9_1 )from lake whitefish,walleye and northern
pike from Southern Indian Lake.
'..Fish mercury levels (~g 9_1 )from lake whitefish,walleye and northern
pi ke taken from Rat and Burntwood Ri ver bas i n 1akes f1 aoded by the
Churchill River diversion.
Mean and range of total mercury concentrations in various possible
source materials at Southern Indian Lake and Notigi Reservoir •
....
- 1 -
INTRODUCTION
Fish mercury levels in excess of 0.5 IJ.g g_l have generally been accepted
as evidence of industrial pollution,however fish with high mercury levels
occur in many pristine lakes unaffected by industrial sources of mercury
(e.g.Koirtyohann et ale 1974;Wobeser et ale 1970;Johnels et al.1967).
This natural variability in fish mercury concentrations from unpolluted
environments makes it difficult to interpret high mercury levels in fish from
infrequently sampled lakes.Although several studies have implicated
reservoir formation as the cause of high fish mercury concentrations observed
after flooding (Potter et al.1975;Abernathy and Cumbie 1977;Cox et al.
1979;Bruce and Spencer 1979)we do not know of any published studies which
present both pre-and post-impoundment data on fish mercury concentrations to
verify that increases have actually occurred.We present such information
here and suggest possible mechanisms based on our knowledge of the
environmental changes which accompanied impoundment.
MATERIALS AND METHODS
STUDY AREA
The majority of the flow of the Churchill River in northern Manitoba
(Fig.I),was diverted into the Nelson River basin for hydroelectric purposes
by a series of lake and river manipulations over the period 1974-1978.The
point of diversion was Southern Indian Lake,flooded 3 m above the mean lake
level in 1976.Water was diverted out of Southern Indian Lake via a diversion
channel,down the Rat River valley,through a control structure at Notigi_
Lake~into the Burntwood River and then to the lower Nelson River (Fig.1).
The Notigi control structure flooded lakes in the Rat River valley,including
Issett,Pemichigamau,Rat and Notigi lakes,from 8 -15 m,over the period
~
I
""'"I
....
-
- 2 -
1974-76.Lakes downstream of Notigi Lake on,or connected to the Burntwood
River,such as Wapisu,Footprint and Wuskwatim,have experienced a rise 1n
water 1eve 1s of 3-5 m due to increased Burntwood Ri ver flows.Table 1 gives
the changes in water levels and surface areas of lakes affected by the
diversion.Bodaly et al.(in press)give a more detailed description of the
Churchill River diversion project.
COLLECTION OF DATA
Sampling of fish for muscle mercury determination was carried out in two
different ways:survey sampling and commercial sampling (Bodaly and Hecky
1979).Survey samples were captured by graded mesh experimental gill nets,
and,in the case of Southern Indian Lake,samples were ~eparated according to
the region of the lake fished •.For individual fish,fork length was measured
(to the nearest 5 nm)and mercury concentrati on was detenni ned from a porti on
of muscle taken from the caudal peduncle area.In the case of commercial
samples,fish were removed from time to time from commercial catches and were
classified only as to lake of origin.A sample of at least five fish weighing
no less than 6.8 kg was taken for each detennination.Fillets,one from each
fish,were combined and homogenized prior to mercury determinations (in
triplicate).Mercury concentrations were determined according to Hendzel and
Jamieson (1976)who reported an analytical.precision of ±O.025 ~g g_1 at 0.5
~g g-l ~sh muscle tissue.Survey samples from Southern Indian Lake and
Issett Lake were collected by the authors.Other data is from McGregor (1980)
~~~
and ~sh;n9 ~~j S~rtiees,Department of Fisheries and Oceans,
Winnipeg,Manitoba (unpublished data).
Water samples were taken from various locations on Southern Indian Lake,
the Churchill River,Pemichigamau Lake and Notigi Lake in September 1978 and
".
""'"- 3 -
detection limit of the total mercury analysis in water is 5 x 10-6 lJ9 g_1 and
of the methyl mercury ana 1ys is is 0.2 x 10-6 ~g_1.
Samples of unflooded bank materials were collected from various locations
adjacent to the Churchill River diversion (Fig.1)in 1981.Subsamples were
dried to constant weight·at 105°C,ground with a mortar and pestle.and passed
through a 1.0 rrm mesh screen •.A weighed portion was digested with aqua regia,
brought toa boil,simmered for one minute,cooled and made to 50 mL volume.
The analysis for mercury was completed with the semi-automated procedure of
Armstrong and Uthe (1971).Surface sediment samples were collected in the
summer of 1980 with an Ekman dredge and treated similarly to the bank
July 1981 (Fig.1).Samples for total mercury were collected from 1 m depth
~using a van Dorn sampler and returned to the laboratory in 300 ml glass
reagent bottles.Sample preservation,extraction and analysis followed
closely that of Kopp et ale (1972).Ten L surface water samples for
methylmercury determination were collected in polyethylene carboys.The
methylmercury was extracted from acidified water into benzene.Subsequent
analysis followed the method of Uthe et al.(1972).The methodological....
[
material.Suspended sediment was collected by continuous flow centrifugation
in August 1980 from known volumes of lake water and subjected to the same
treatment except that screening was not necessary.A Sorvall RC2-B centrifuge
was used at 14,500 RCF with a flow rate of approximately 50 mL min-1•
RESULTS
Increases in fish mercury levels coincident with flooding
Both pre-and post-impoundment fish mercury data are available for
Southern Indian Lake on the Churchill River at the point of diversion,for
Issett Lake at the upper end of the Notigi reservoir,and for Wuskwatim Lake,
-
-
-
.....
- 4 -
on the Burntwood River below Notigi reservoir.These data demonstrate that
mercury levels in fish increased significantly soon after flooding in all
three lakes.
Muscle mercury levels in lake whitefish (Coregonus clupeaformis),
northern pike (Esox lucius)and walleye (Stizostedion vitreum)from Southern
Indian lake have increased substantially from before the impoundment of the
lake to after lake impoundment (Table 2).For example',mean mercury
concentrations in lake whitefish survey samples from Region 4 were higher
after impoundment,with means of 0.22, 0.10,0.14,0.08 and 0.11 ~g_l in
1978 to 1982 respectively,as compared to a pre-impoundment mean of 0.05 jJ.g
9_l in 1975 (Table 2).Similar increases have occurred in lake whitefish from
The Channel,Camp 9,and Region 6 (Table 2).Although levels from 1975 were
determined from samples stored,frozen for approximately three years before
analysis,there has apparently been little effect due to storage because 1975
levels reported here are comparable to the lake whitefish mean mercury
concentration of 0.05 ~g g_l determined prior to flooding (1969-1973)from 6
samples removed from commercial shipments (Table 2).Unfortunately,mercury
levels in commercial shipments have not been monitored since 1973.Hendzel
(personal communication)reports no detectable changes in mercury
concentrations in fish tissue stored frozen for many months.Frozen storage,
if accompanied by severe dehydration,might increase mercury concentrations on
a wet weight basis,and therefore the mercury concentrations observed for the
1975 whitefish samples represent maximum estimates.
Samples of the two predatory fish species landed by the Southern Indian
Lake commercial fishery,walleye and northern pike,also show post-impoundment
increases in muscle mercury levels as compared to pre-impoundment levels
(Table 2).Mercury levels in walleye commercial samples were relatively
-
-
- 5 -
stable at 0.2-0.~~g g_l over the period 1971 to 1977 but were much higher
(0.57-0.75 ~g g_l)in 1978-1982.Northern pike mercury levels were somewhat
higher than walleye before lake impoundment,in the range 0.25-0.35 ~g g_l
over the period 1971-1973.Levels in pike may have been elevated in 1976-1978
to 0.4-0.5 ~g g_l and means in 1979-1982 (0.67-0.95 ~g 9_1 )were much above
pre-impoundment levels.
Mean mercury levels in walleye from Wuskwatim Lake were relatively stable
over the pre-impoundment period 1970-1977 at 0.25-0.44 ~g g_l but increased to
0.76,1.00,and 0.89 ~g g_1 in 1979,1980 and 1981 respectively (Table 3).
Whitefish mercury concentrations also increased,rising from 0.08 i.l9 9_1 in
1970 to 0.33 ppm in 1981 (Table 3).Mercury concentrations in lake whitefish
muscle frcm ·Issett Lake doubled from·a mean of 0.15 ~g_1 in 1975~prior to
Churchi 11 Ri ver di versi on,to a mean of 0.32 ~g g_1 in 1978,after Churchi 11
River diversion (Table 3).
It is well known that mercury concentrations in fish tend to increase
with fish size (Scott and Armstrong 1972;Scott 1974;Huckabee et a1.1979),
however,increases in mercury levels in fish in new impoundments noted here
were not due to changes in the average size of fish sampled.Significant
changes in the average size of survey samples have,in general,not occurred
and where changes in fish size occurred,average mercury levels did not
usually follow average fish sizes (Tables 2 and 3).In the case of commercial
samples,mean fish size tends to be held rather constant by the use of one
size of commercial gill net mesh.Furthermore,significant correlations
between fish size and mercury concentrations were not the rule for pre-or
post-impoundment survey samples from the Churchill River diversion area;
significant correlations were observed in a majority of survey samples only
for pike (14 of 22 samples)whereas significant correlations were observed in
-.
....
....
,"""
....
- 6 -
a minority of samples of whitefish (7 of 20 samples)and walleye (9 of 24
samples).
Regional differences in post-impoundment mercury levels
Only post-impoundment data were available for other flooded lakes located
on the route of the diverted Churehill River.All of these lakes contained
fish with mercury levels much higher than expected background concentrations
(Table 3).In general,mercury levels in predatory fish after impoundment
were highest in lakes now covered by the Notigi reservoir,were moderately
high in lakes flooded by diversion flow downstream of Notigi control
structure,and were lowest in Southern Indian Lake (Tables 2 and 3).Mercury
levels in predatory fish from Notigi reservoir lakes ranged from ~.6-2.9 ~g
g_1,while the comparable range for lakes below Notigi was ·0.6-1.5 (.Lg g_1 and
for Southern Indian Lake was -0.4-1.0 ~g g_1.Mercury levels in whitefish
tended to be higher in lakes below Notigi reservoir and lowest in Southern
Indian Lake •
Time course of elevated fish mercury levels
Mercury levels in predatory fish became elevated within 2-3 years of lake
impoundment and there was no indication of general declines fram peak levels
within 5-leyears of impoundment.Predatory species from Southern Indian lake
showed elevated mercury levels by 1978,two years after flooding.There were
2notrendstowardsdeclininglevelsovertheperiod1978-198t (Table 2).
Mercury concentrations in walleye were relatively stable both in survey and
commercial samples over this period.Survey samples of pike taken from
specific regions of the lake show generally increasing mercury levels over
2-1978-8&,and mercury levels in commercial shipments were highest in 1981
.-- 7 -
(Table 2).In lakes now covered by the Notigi reservoir,impounded over the
period 1974-1976,high fish mercury levels were evident by 1977-1978 when the
lakes were first sampled (Table 3).Mercury levels in predatory fish showed
.-~"fA.u.tc1EnJCA..-Hu..~d/q"rt-/qez)W·
l-sam pling was irreguiar (Table 3).In lakes downstream of the Notigi control
structure,flooded by diversion flow in 1976-1978,elevated fish mercury
levels were evident by 1977-1979 (Table 3).As with Notigi reservoir lakes,
sampling was irregular and post-impoundment trends are not obvious.
Mercury levels in whitefish from Southern Indian Lake also responded
.....quickly to flooding and were elevated by 1978,two years after lake
impoundment,when first sampled (Table 2).However,in contrast to mercury
levels in predatory fish from all impounded lakes in the area,whitefish
mercury levels from Southern Indian Lake.decreased consistently over the
period 1978-1982,although pre-impoundment levels had not been reached by
1982.
-
-
Water,Soil,and Sediment Mercury Concentrations
.Tota 1 mercury concent rati ons in water collected from 17 1ocat ions (Fi g.
1)in 1978 and 1981 were <5 x 10-6 ~g g-1,the limit of detection with the
methods and sample volumes used.Similarly,13 of the 14 analyses for
methylmercury in water were below the limit of detection,0.2 x 10-6 ~g g_1,
while a concentration of 0.4 x 10_6 ~g g_1 was observed in the forebay of the
Notigi Reservoir.
Terrestrial soils are a possible source of mercury in new impoundments.
Soils underly approximately three-quarters of the surface area of Notigi
Reservoir (Table 1)and eroding banks are a continuing source of terrestrial
.~oo
material to the Southern Indian lake reservoir (Hecky et al.,this VOlum]:.
Bank materials generally consist of three recognizable horizons:an upper
- 8 -
.-
1ayer of moss,1itter and humus,a second 1ayer of hi gh 1y organi c surface
-.soils (soil horizon A)and the inorganic subsoil (soil horizon C).The
greatest mass of material brought into suspension due to shore erosion is of
fine grained silts and clays which originate from extensive glacio-lacustrine
deposits surrounding the lake (Newbury and McCullough this volume,p.DO)•
.-Mercury concentrations in the upper,organic rich soil horizons are clearly
higher than in the inorganic C horizon (Table 4).Lake sediment mercury
concentrations are similar to the eroding,inorganic bank materials but lower
than the organic horizo~(Table 4}.Suspended sediments collected from lake
water are substantially higher in mercury concentration than surface sediments
i"""'
I
-
collected from the lake bottom (Table 4).
DISCUSSION
The present data show that increases in fish mercury concentrations have
occurred coincidentally with increases in water level and that high mercury
levels have been observed after inundation in all lakes flooded by the
Churchill River diversion.Pre-flooding data.available for Southern Indian
Lake,Issett Lake and Wuskwatim Lake show increased fi sh mercury
concentrations soon after increases in water levels.Lakes on the diversion
route between Issett and WusKwatim,for which there were no pre-diversion data
(Rat,Notigi,Wapisu and Footprint lakes)show high fish mercury
concentrat~ons in years immediately following diversion and flooding,and
these concentrations presumably represent increases over pre-flooding
concentrations.Increases in fish mercury concentrations appear to be
restricted to lakes flooded for the Churchill River diversion and there is no
suggestion that similar increases have occurred in undisturbed lakes over the
same time period throughout northern Manitoba.In fact,there are over 30
.....
..-
.....
""'"
-9 -
lakes in northern Manitoba (north of the 55th parallel)for which fish mercury
levels have been detennined from corrmercial shi rxnents during t~hiCh .
show no trends of increasing fish mercury levels (data from ~~hing i~~
DUv.~
lRd\,l~tl"j'Ser l /iee3',)Department of Fisheries and Oceans,Winnipeg,Manitoba).
This indicates that the recent increases in fish mercury levels in Southern
Indian Lake and lakes on the Rat and Burntwood Rivers are probably not
directly due to atmospheric fallout.Mercury analyses on a 20 cm sediment
core from Southern Indian Lake show a slow,constant increase in Hg
concentration from the base of the co~e to the top,resulting in a 2X top to
bottom differenti al in mercury concentrati on (G.J.Brunski 11,unpubl.data).
but there is no evidence of drama~ic increase in ~tion prior to
flooding.An approximate doubling in mercury flux to the sediments since 1900
has been identified at a number of "pristine"locations and has been
attributed to an increase in atmospheric fallout of mercury because of
industrialization (Kemp et ale 1978;Weiss et al.1975).The observed
increase in the Southern Indian Lake core is consistent with this apparent
global trend.Modern deposition rates of mercury in Southern Indian lake
(G.J.Brunskill,pers.comm.)are below the pre-modern deposition rates for
the upper Great Lakes (Kemp et al.1978)and are similar to the deposition
rates on the Greenland glacier (Weiss et al.1975).Suspended sediments -
collected in Southern Indian lake after "impoundment (Table 4)are enriched in
mercury relative to older deposited sediments;this may reflect a recent
change in mercury availability in the lake.
There are no known point sources of mercury and no agricultural activity
in the Churchill diversion area which could supply a sudden surge in mercury
deposition beginning in 1976.Although it is not possible to rule out the
possibility that all the geographic areas listed above might have unusual
-
-10 -
geological formations which provide a rich local source of mercury,it seems
unlikely that mercury-rich mineral formations are the ultimate cause of the
elevated fish mercury concentrations in all these reservoirs.Source
materials at Southern Indian Lake tend to be low or average in their mercury
content when compared to similar materials analyzed elsewhere (Andren and
Nriagu 1979;Andersson 1979).The modern mercury deposition rate in Southern
Indian lake prior to flooding in 1976 is among the lowest reported in the
literature (G.J.Brunskill,pers.comm.).We conclude that there is no reason
to be1ie~e that the Southern Indian Lake-Notigi region has unusually high
mercury concentrations in source material.Instead it seems that reservoir
formation and associated inundation of land in itself has led to higher fish
i
I •mercury concentrat10n.
The hypothesis that reservoir formation can lead to elevated fish
mercury'levels was initially made by Potter et a1.(1975)and Abernathy and
Cumbie (1977).A similar hypothesis has more recently been made by others
(Bruce and Spencer 1979;Waite et a1.1980;Meister et a1.1979;Cox et a1.
1979).These hypotheses have emphasized that reservoirs provide new sources
of mercury in inundated soils (Abernathy and Cumbie 1977;Meister et a1.1979)
or increased availability of naturally transported mercury because of
increased retention of inflowing material (Potter et al.1975).Problems
associated with elevated fish mercury levels in newly impounded reservoirs are
quite widespread in North America.They have been reported in Labrador,
Saskatchewan,Manitoba,Arizona,Illinois,South Carolina.{references above},
Utah (Smith et a1.1974).Idaho (Benson et al.1976;Kent and Johnson 1979>-
Mississippi (Knight and Herring 1972).and elsewhere.It now seems clear that
this is a widespread phenomenon which has come to the attention of fisheries
workers in the last decade due to the recent introduction of routine testing
of fish for mercury levels.
'.
....
.....
.....
.....
-11 -
Mercury in fish muscle exists predominantly in the organic or
methylmercury fonn (Westoo 1966).The methylation of inorganic mercury is
known to be primarily bacterially mediated in nature (Beijer and Jernelov
1979;Bisogni 1979).Increased bacterial production due to degradation of
flooded terrestrial vegetation,peat and humus probably serves to promote
mercury methylation;Furutani and Rudd (1980)showed that an increase in
microbial substrate resulted in increased rates of mercury methylation •
We hypothesize that,in lakes flooded by the Churchill River diversion,
elevated fish mercury levels were due to the bacterial methYlatioJof
naturally occurring mercury found in flooded soils.The apparent relationship
over the Churchill River diversion system between the increase in lake surface
\"
area and mercury levels in predatory fish,where fish mercury levels were
highest in the extensively flooded Notigi reservoir,supports this
hypothesis.The primary source of mercury was probably the upper,organic soil
horizon because mercury levels in this soil layer were much higher than in
inorganic subsoil layers.Inorganic subsoil is apparently not acting as a·
major source of mercury because the addition of large amounts of inorganic
subsoil to the water column through shore erosion in Southern Indian Lake
(Newbury et al.1978)did not result in fish mercury levels approaching those
found in the Notigi reservoir where shoreline erosion was negligible but
increase in area flooded was greater.
Water mercu~y levels throughout the Churchi 11 River di versi on system were
very low.This has been reported for other new reservoirs where fish mercury
levels were elevated (Cox et al.1979;Potter et al.1975;Abernathy and
Cumbie 1977)and was expected because the geochemistry of inorganic mercury
strongly favors association with particulate phases (Cranston and Buckley
1972;Hannan and Thompson 1977),and the biogeochemistry of methylmercury
.....
-
-
....
....
-12 -
strongly favors association with biota owing to its aqueous and lipid
solubility and affinity for sulfhydryl groups {Carty and Malone 1979}.In
Southern Indian Lake,water concentrations were <5 x 10_6 ~g_l (or 5 pg g-l)
while mercury concentrations on suspended sediments were 0.20 ~g g_l.a
concentration factor of at least 2 x 10 5 ,.and mercury concentrations in pike
were approximately 0.60 ~g g_l,a minimum concentration factor of 6 x 10 5 •
These concentration factors are simil ar to those reported in the 1 iterature
(Potter et al.1975).
Average muscle mercury levels in predatory species (walleye and pike)
exceed the Canadian marketing standard of 0.5 I1g g_l (and usually the U.S.
standard of 1.0 119 g_l)in every lake on the Churchill,Rat and Burntwood
rivers flooded by the Churchill River diversion project.The widespread
nature of the high fish mercury level -new reservoir association makes it
imperative that elevated fish mercury levels be considered in all impact
assessments of proposed reservoirs.
ACKNOWLEDGMENTS
Many persons assisted in field collections and we would especially like
to thank Helen A.Ayles (Metcalfe)for collection of the 1975 whitefish
samples and Neil E.Strange for collection of the 1981 soil samples.The
Southern Operations Directorate.Fisheries and Oceans,Winnipeg provided
unpub 1i shed data and performed the fi sh mercury ana·l yses.A.·Lutz performed
the soil mercury analyses.
REFERENCES
Abernathy,A.R.,and P.M.Cumbie.1977.Mercury accumulation by largemouth
bass (Micropterus salmoides)in recently impounded reservoirs.Bull •.
Environ.Cantam.Taxicol.17(5):595-602.
.....
-
-
~
!
I""'",
l
-13 -
Andersson.A.1979.Mercury in soils.pp.79--112.l.!l J.D.Nriagu (ed.)The
biogeochemistry of mercury in the environment.Elsevier/North Holland.
Amsterdam.
Andren.A.W.and J.O.Nriagu.1979.The global cycle of mercury.pp.1-21-
In J.O.Nriagu (ed.)The biogeochemistry of mercury in the environment.
Elsevier/North.Holland,Amsterdam.
Armstrong,F.A.J.,and J.F.Uthe.1971.Semi-automated determination of
mercury in animal tissue.At.Absorpt.Newsl.10:101-103.
Beijer,K.and A.Jernelov.1979.Methylation of mercury in aquatic
environments,pp.203-210..ill J.O.Nriagu (ed.)The biogeochemistry of
mercury in the environment.El sevier/North Holland,~lsterdam.
Benson,W.W.,W.Webb,0.101.Brock,and J.Gabica.1976.Mercury in catfish
and bass from the Snake Riyer in Idaho.Bull.Environ.Contam.Toxicol.
15(5):564-567.
Bisogni,J.J.,Jr.1979.Kinetics of methylmercury formation and
decomposition in aquatic environments,pp.211-230.In J.D.Nriagu (ed.)"
The biogeochemistry of mercury in the environment.Elsevier/North
Holland,Amsterdam.
Bodaly,R.A.,and R.E.Hecky.1979.Post-impoundment increases in fish
mercury levels in the Southern Indian Lake reservoir.Manitoba.Can.
Fish.Mar.Serv.MS Rep.1531:iv +15 p.
Bodaly,R.A.,D.M.Rosenberg,M.N.Gaboury,R.E.Hecky,R.W.Newbury,and K.
Patalas.In press.Ecological effects of hydroelectric development in
northern Manitoba,Canada:The Churchill-Nelson River diversion.In
P.J.Sheehan,D.R.Miller,G.C.Butler and Ph.Bourdeau (eds.)Effects of
pollutants at the ecosystem level.SCOPE,John Wiley and Sons,
Chichester,New York,Brisbane.Toronto.
.....
--
-14 -
Huckabee,J.W.,S.A.Janzen,B.G.Blaylock,Y.Talmi,and J.J.Beauchamp.
1979.Methylated mercury in brook trout (Salvelinus fontinalis):
Absence of an ~vivo methylating process.Trans.Am.Fish.Soc.107(6):
848-852.
-15 -
Johnels.A.G.,T.Westermark,W.Berg,P.I.Persson,and B.Sjostrand.1967.
Pike (Esox lucius L.)and some other aquatic organisms in Sweden as
indicators of mercury contamination in the environment.Gikos 18:
323-333.
Kemp,A.L.W.,J.D.H.Wi 11 iams,R.L.Thomas and M.L.Gregory.1978.Impact of
'man's activities on the chemical composition of the sediments of Lake
Superior and Huron.Water,Air,and Soil Pollution 10:381-402.
Kent,J.C.,and D.W.Johnson.1979.Mercury,arsenic and cadmium in fish,
water,and sediment of Pmerican Fa 11 s Reservoi r,Idaho,1974.
Pesticides Monitoring J.13(1):35-40.
Knight,L.A.,and J.Herring.1972.Total mercury in largemouth bass
(Micropterus salmoides)in Ross Barnett Reservoir,Mississippi -1970
and 1971.Pesticides Monitoring J.6(2):103-106.
Koirtyohann,S.R.,R.Meers,and L.K.Graham.1974.Mercury levels in fishes
from some Missouri lakes with and without known mercury pollution.
Environ.Res.8:1-11.
Kopp,J.F.,M.C.Longbottom,and L.B.Lobring.1972 •."Cold vapour"method
for determining mercury.Am~Water Works Assoc.J.64:20-25.
McCullough,G.K.1981.Water budgets for Southern Indian Lake,Manitoba,
before and after impoundment and Churchill River diversion,1972-79.
Can.MS Rep.Fi she Aquat.Sci.1620:iv +22 p.
McGregor,.G.W.G.1980.SUlIiIlary of mercury levels in lakes on the Churchill-
Rat-Burntwood and Nelson River systems from 1970 to 1979.Can.Data
Rep.Fi sh.Aquat.Sci.195:iv +16 P •
Meister,J.F.,J.DiNunzio,and J.A.Cox.1979.Source and level of mercury
in a new impoundment.Am.Water Works Assoc.J.1979:574-576.
,-
i
-
-16 -
Newbury,R.W.,K.G.Beaty,and G.K.McCullough.1978.Initial shoreline
erosion in a permafrost affected reservoir Southern Indian Lake Canada,
p.833-839.1Q Proceedings of the Third International Conference on
Permafrost,July 10-13,1978.Edmonton,Canada,Vol.l.
Potter,L.,D.Kidd,and D.Standiford.1975.Mercury levels in Lake
Powell.Bioamplification of mercury in man-made desert reservoir.
Environ.SCi.Tech.9(1):41-46.
Scott,D.P.1974.Mercury concentration of white muscle in relation to age,
growth and conditions in four species of fishes from Clay Lake,Ontario.
J.Fish.Res.Board Can.31:1723-1729.
Scott,D.P.and F.A.J.Armstrong.1972.Mercury concentraiton in relation to
.
size in several species of freshwater fishes from Manitoba and
Northwestern Ontario.J.Fish.Res.Board Can.29:1685-1690.
Smith,F.A.,R.P.Sharma,R.I.Lynn,and J.B.Low.1974.Mercury and
selected pesticide levels in fish and wildlife in Utah:1.Levels of
mercury,DDT,ODE,Dieldrin,and PCB in fish.Bull.Envir.Contam.Tox.
12(2):218-223.
Uthe,J.F.,J.Solomon,and B.Grift.1972.Rapid semi-micro method for the
determination of methyl mercury in fish tissue.Assoc.Off.Anal.Chern.
J.55:583-589.
Vitkin,N.,1979.1979.Review of suspended sediment sampling program in
Manitoba.Province of Manitoba,Winnipeg,December 1979.
Wa ite,D.T.,G.W.Dunn,and R.J.Stedwi 11.1980.Mercury in Cookson
Reservoir (East Poplar River).Sask.Envir.Wat.Poll.Con.Branch.
W.P.C.23:19 p.
Weiss,H.,K.Bertine,M.Koide,and E.D.Goldberg.1975.The chemical
composition of a Greenland glacier.Geochimic et Cosmochirn Acta 39:
1-10.
.....
....
-17 -
Westoo,G.1966.Determination of methylmercury compounds in foodstuffs.
I.Methylmercury compounds in fish,identification and determination.
Acta.Chern.SCand.20:2131.
Wobeser,G.,N.D.Nielsen,R.H.Dunlop and F.M.Atton.1970.Mercury
concentrations in tissue of fish from the Saskatchewan River.J.Fish •
Res.Board Can.27:830-835 •
.'
Table 1.Changes in water level and surface area of several lakes affected by the Churchill'River diversion project.
Areas and levels are estimated under long term mean levels prior to diversion and under projected mean levels
after full Churchill River diversion.Sources:McCullough (1981);Brown (1974);Vitkin (1979);D.Windsor,
Manitoba Hydro,pers.comma
================"",======-,----=
250.6 258.2
248.1 258
247.8 258
251.1 258
249.0 258
247.8 257.9
242.0 257.2
239.9 243.2 '
239.0 242.6
231.0 233.0
Lake
Southern Indian
Notigi Reservoir]
Issett
Karsakuwigamak
Pemichigamau
Central Mynarski
West Mynarski
Rat
Notigi
Wapisu
Threepoint &
Footpri nt
Wuskwatim .
Pre-impoundment
level (m)
255.0
Post-impoundment
1eve1 (m)
258.0
Pre-impoundment
area (km 2 )
1977
153
3.7
18.8
19.3
11.5
6.2
78.4
15.1
49
75
70
Post-impoundment
a rea (km 2 )
2391
584
67
103
79
-=
Relative
change
%
+21
+282
+37
+31
+13
1 Pre-impoundment water area for Notigi is the sum of the several lakes (listed under Notigi Reservoir)which existed
before impoundment.~~
J I I m I L «J i -(t I I L
~Table 2.fish mercury levels (~g got)from lake whitef1sh.walle.ve and northern pike from Southern Ind1an Lake.
Region of lake given for survey samples (see methods for d1fference between survey and conrnercial
sample).See fig.1 for location of region of lake.
Mean fork
Mean mercury Range of length(cm)
concentration mercury of survey
Species Region Year (~g g-1)concentration *sampl es sample
Whitefi sh The Channel 1975 1 0.06 0.03-0.12 50 33.2
1978 0.30 0.06-0.60 17 42.3
1919 0.25 0.04-0.55 26 33.9
1980 0.21 0.02-0.42 24 33.5
1981 0.20 0.04-0.34 25 34.2
1982 0.09 0.03-0.27 25 40.1
Camp 9 1975 2 0.05 0.03-0.08 25 33.5
1979 0.13 0.06-0.26 40 40.1
1980 0.13 0.04-0.59 28 37.9
1981 0.10 0.04-0.35 24 35.1
1982 0.09 0.03-0.27 25 40.1
Region 4 1915 0.05 0.02-0.10 25 31.6
1978 0.22 0.09-0.38 16 33.7
1979 0.10 0.06-0.30 68 37.0
1980 0.14 0.05-0.37 27 37.0
1981 0.08 0.03-0.32 67 35.8
1982 0.11 0.06-0.21 25 39.0
Region 6 1975 0.07 0.03-0.12 25 32.9
1979 0.31 0.05-0.55 30 38.6
1980 0.20 0.04-0.44 20 33.3
1981 0.14 0.03-0.38 26 32.6
1982 0.11 0.01-0.36 36 32.5
(Coomercial)1969 0.05 1
1970 0.05 0.02-0.08 2
1972 0.11 1
1973 0.02 0.02-0.02 2
Walleye The Channel 1979 0.47 0,25-2.19 30 33.3
1980 0.56 0.29-2.04 33 39.3
1981 0.55 0.35-1.22 32 38.7
1982 0.45 0.23-1.41 24 38.9
Camp 9 1979 0.59 0.32-1.80 11 40.6
1980 0.53 0.37-0.76 14 38.8
1981 0.45 0.35-0.51 5 35.2
1982 0.47 0.32-0.55 25 40.6
Region 6 1978 0.80 0.45-1.20 15 43.5
1979 0,47 0.06-1.14 21 35.5
1980 0.59 0.33-1.03 28 37.6
1981 0.64 0.32-1.94 26 38.2
1982 0.78 0.31-1.66 25 35.5
(COIIII1erci a 1)1971 0.19 0.15-0.22 6
1912 0.21 0.18-0.23 3
1973 0.28 0.20-0.35 3
1975 0.30 0.22-0.38 2
1976 0.24 0.20-0.32 4
1977 0.26 0.23-0.30 2
1978 0.57 0.33-1.12 7
1979 0.75 0.47-1.21 6
1980 0.54 0.35-0.92 8
1981 0.62 0.26-1.26 20
1982 0.62 0.51-0.61 5
Pike The Channel 1979 0.57 0.29-0.89 35 49.7
1980 0.57 0.05-1.11 38 53.2.....1981 0.64 0.42-1.00 25 52.0
1982 0.77 0.48-1.09 25 54.2
Camp 9 1979 0.58 0.36-1.10 35 55.4
1980 0.61 0.41-1.01 31 56.5
1981 0.66 0.43-0.95 24 54.4
1982 0.68 0.43-0.96 25 60.3
Region 4 1979 0.49 0.30-1.20 .54 52.8
1980 0.63 0.45-0.91 28 53.9
1981 0.72 0.19-1.13 25 55.2
1982 0.63 0.35-0.82 24 58.2
Re9ion 6 1978 0.77 0.28-1.72 15 66.6
1979 0.59 0.42-1.21 60 53.0
1980 0.78 0.42-2.55 34 53.7
1981 0.89 0.55-1.15 25 55.4
1982 0.96 0.38-1.54 28 55.2
(C~rcial)1971 0.26 0.24-0.29 4
1972 0.32 0.24-0.40 5
1973 0.30 0.26-0.33 3
1976 0.47 0.25-1.02 10
1977 0.43 0.42-0.45 2
1978 0.50 0.25-0.83 7
1979 0.88 0.53-1.51 9
1980 0.67 0.45-1.12 14
1981 0.95 0.53-2.04 28
1982 0.90 0.25-1.99 11
r~Th1s sample is a cllllC1ned sample fT'Olll Re9ions 2 and 6 (see F1g.1).
2 This sample ..as taken fT'Olll Region 2 (see F1g.1).
"..------"---_._--
..-
Table 3.Fi sh mercury 1eve 1s (~g g-l)from 1ake wh;tef;sh.wa 11 eye and northern p;ke taken fram Rat and Burntwood
River basin lakes flooded by the Churchill River diver<s;on.See Fig.1 for 1ocat;on of 1akes.Issett,
Rat and Notigi laKes are part of Notigi reservoir;other lakes are downstream of Notigi Reservoir.
Type of sample and ntallber Qf samples (see Methods);s also indicated.
Mean fork
Maan me rcury Range of
Type of length (em)
concen t rat i on mercury of survey
lake Spec;es Year (Ilg g_l)concentra ti on sample samples sample--Issett 101M tefish 1975 0.15 0.02-0.30 survey 24 36.9
1978 0.32 0.17-0.40 survey 5 31.0
1982 0.21 0.12-0.36 survey 25 38.4
_an eye 1978 1.52 1.24-1.95 survey 5 38.9
1982 0.79 0.20-2.52 survey 19 34.7
Pike IS7a 0.61 0.37-0.74 survey 5 57.3
1982 0.90 0.36·1.75 survey 26 59.6
Rat IIhl tef;sh 1978 0.40 0.26-0.59 survey 5 43.9
1978 0.37 commerc;al 1-1980 0.32 0.15-0.61 survey 24 42.2
1980 0.34 cOlllT\l!rc;a1 1
lia 11 eye 1978 2.54 2.39-2.67 commercial 5
1978 2.56 2.17-3.51 survey 26 44.9
1979 2.32 1.68-3.29 survey 25 45.0
1980 1.15 0.41-3.37 survey 22 40.8
1980 1.15 conrnerci a1 1
Pike 1978 2.14 2.04-2.25 commercial 5
1978 2.05 1.47-2.49 survey 24 69.8
1980 2.32 commercial 1
't Notigi Wllitef;sll 1980 0.12 0.07-0.25 survey 6 42.1
1981 0.23 0.12-0.71 survey 38 30.9
Walleye 1978 1.41 0.19·2.91 survey 19 42.2
1978 1.32 0.26-2.32-commercial 3
1980 2.90 2.11-3.47 survey 4 45.2
'i""'"1980 2.59 comnercial 1
1981 1.88 0.95-2.55 survey 29 41.6
1982-1.23 C.98-1.63 survey 6 50.3
1982 1.11 ccmnercial 1
P;ke 1977 1.59 conmercial 1
~1980 1.95 1.62-2.29 survey 5 78.8
1981 1.70 0.24-2.82 survey 50 58.1
1982 1.85 1.32-2.38 survey 10 73.8
1982 2.06 1.91-2.21 collllll!rc1a 1 2
Wap;su Walleye 1977 1.17 0.14-3.03 survey 91 41.1
1977 1.33 0.80-1.81 COlllllercia I 3
Pike 1977 1.08 0.32-2.25 survey 38 67.2
1977 1.53 1.48-1.61 cOl1lllerc;a1 3
Footprint Waneye 1978 0.82 0.29-2.61 survey 40 38.2
1980 0.92 0.31-1.72 survey 12 39.4
1981 1.10 0.71-1.76 survey 30 37.6
Pike 1978 0.60 0.28-0.90 survey 36 45.7
1980 .1.38 0.82-3.37 survey 8 52.6
1981 1.12 0.83-1.74 survey 14 52.0-Threepoint IoIl1i tefi sll 1980 0.56 0.34-0.91 survey 10 42.9
1981 0.23 0.11-0.41 survey 16 33.3
Walleye 1980 1.18 0.62-1.81 survey 10 40.3
1981 1.35 0.84-2.05 survey 42 38.1
Pike 1980 1.28 0.49-3.05 survey 10 68.8
F'"1981 1.33 0.44-2.27 survey 28 51.7
Wuskwat;m 101M tefisll 1970 0.08 cOllllll!rcial 1
1981 0.33 0.18-0.78 survey 28 39.8
Walleye 1970 0.34 cemnerci aI 1
1971 0.25 cOl1lT.ercial 1
1973 0.40 conmercial 1
1974 0.44 conmercial 1
1975 0.35 cOll'lllerci a1 1
1976 0.26 cOllllll!rcial 1
1977 0.38 conmercial 1
1979 0.76 0.25-2.18 survey 90 41.5
1980 1.00 0.78-1.41 survey 19 41.3
1981 0.89 0.62·1.36 survey 34 39.9
Pike 1979 0.91 0.21-5.31 survey 75 56.0
1980 1.03 0.79-1.21 survey 7 60.5
1981 0.80 Q.47-1.98 survey 25 51.4
Mystery Walleye 1979 1.13 0.53-1.75 survey 33 46.2
~1ke 1979 0.79 0.23-1.64 survey 45 54.1
Table 4.Mean and range of total mercury concentrations in various possible
source materials at Southern Indian Lake and Notigi Reservoir.
Total mercur1 content
(~g g-)
~Source Date sampled n Mean Range
Moss/l itter/humus July-Aug 1981 83 0.095 <0.005-0 .290
Soil A horizon July-Aug 1981 47 0.090 <0.005-0.220
Sail C horizon July-Aug 1981 60 0.041 <0.005-0.180
Lake sediment July-Aug 1980 28 0.036 0.010-0.070
~
Suspended sediment Aug 1980 4 0.255 0.120-0.360
Water Sept 1978 and 17 <5 x 10-6
July 1981
....------~-
-.
!
-
..
SOUTHERN
INDIAN
LAKE
~
~
~L
\~
NOTIGIL.
DIVERSION
CHANNEL
.Water Sample.;.Total Mercury
+Water Sample-Total Mercury and
Methyl Mercury
•Bank Material
III Sediment Core Sample
,.,
f ,-;'
.~....j
i