-------
395
estimated to be delivered to the water courses from the livestock
operations, 65% of which comes from Ontario counties.
Soil erosion contributes approximately 3.2 tonnes/yr of phos-
phorus from Michigan to the water courses.. Based upon the
percent of total cropland erosion occurring from wind, Macomb and
St. Clair Counties should be targeted for accelerated conserva-
tion assistance.
Pesticides
Lake St. Clair geographical area is a region of potential prob-
lems regarding the movement of pesticides into the water course.
These problems are a result of an estimated 3.5 million kg being
applied to land in both Canada and the U.S. which has a high
potential to transmit the chemicals via surface runoff, fine
particulate matter carried by wind or water, and infiltration to
groundwater. Based on soil texture and drainage, approximately
70% of the St. Clair geographical area in Canada has been iden-
tified as potential problem areas with respect to surface water
contamination, and approximately 60% of the area possesses a high
risk for pollutant transfer to groundwater systems (5).
5. Atmospheric Deposition
Loadings of contaminants .to Lake St. Clair from the atmosphere
are a nontrivial portion of the total estimated load of lead and
phosphorus (see section E, modeling and mass balance considera-
tions, for further discussion). The major sources of phosphorus
are soil dust, leaf and insect debris, and industrial activity.
A large percentage of the loading may be derived from entrainment
of phosphorus-containing particles in agricultural areas. Lead
and cadmium are introduced through combustion of fossil fuel,
including exhaust from burning leaded gasoline in automobiles.
From measurements in urban and rural locations close to Lake St.
Clair, atmospheric deposition of lead was estimated to range from
4 to 8 kg/d and for cadmium from 0.8 to 1.1 kg/d (17).
The atmospheric loadings of P, NC-3, NH3, Cd, Pb, Zn and Cl to
Lake St. Clair for the years 1982 - 1985 were estimated from data
collected at the Mt. Clemens station of the Great Lakes Atmos-
pheric Deposition network. The thirty-year mean precipitation
average was used to convert concentration values into loadings,
as displayed in Table VIII-15.
Quantitative estimates of loadings of organic contaminants to
Lake St. Clair are not available. Given the quantity of inor-
ganic materials introduced to the lake from the atmosphere, how-
ever, an atmospheric source for organic pollutants is also likely
to be important.
-------
396
TABLE VIII-15
Atmospheric loadings of selected parameters to Lake St. Clair for 1982 -
1985. Mt. Clemens GLAD station is the source of data. Lake surface area
is 1101.178 km2 (430 mi2).
1982
1983
1984
1985
AVERAGE
1982
1983
1984
1985
AVERAGE
Nitrate(N03)
301 ,723
441,810
514,250
445,242
425,756
Cadmium
226
254
299
260
kg/yr
Ammonia(NH4) Total Phosphorus ( TP )
180,593 3,402
342,466 5,952
427,257 5,102
305,124 3,928
313,860 4,596
Chloride Zinc Lead
436,067 30,909
252,170 14,773 5,179
322,645 23,393 5,509
323,492 13,769 3,825
333,594 20,711 4,838
-------
397
6. Groundwater Contamination/Waste Sites
Surface Waste Sites
Active and inactive waste sites within 19 km of the Connecting
Channels were identified as part of this investigation. The
majority of sites were landfills, hazardous waste disposal sites,
and regulated storage sites. Other waste sites included trans-
portation spills, leaking underground storage tanks and contamin-
ated water wells. Underground injection wells were also iden-
tified.
Ranking of sites was based on their potential for contributing
contaminants to the Connecting Channels via groundwater. Sites
in the U.S. were ranked using the U.S.EPA DRASTIC System with
additions and minor modifications. This system assesses the
impact by evaluating the hydrogeology, waste material and the
distance from Lake St. Clair for each site. Nine U.S. sites were
ranked as confirmed or possible contamination sites within the
Lake St. Clair groundwater discharge area (Table VIII-16). In
general, these sites are in areas of sandy unconsolidated sur-
ficial materials and are near to the Connecting Channels. The
water table is generally less than 4.6m below land surface and
priority pollutants and/or inorganic contaminants are on site or
in the groundwater.
Waste disposal sites in the Ontario study area were also iden-
tified. Emphasis was placed on identifying sites that require
monitoring or remedial investigation's. Criteria for ranking and
prioritization of the sites included geologic, hydrologic, hydro-
geologic and geochemical information, on-site monitoring, waste
characterization and containment, and health and safety. No
sites in Kent County were identified that require immediate
investigation or that posed a definite potential for impact on
human health and safety. Three waste disposal sites in the area
contain only building refuse, domestic waste and commercial gar-
bage. These sites are small and not close to the lake. There-
fore, no significant impact is expected from them.
Deep Well Injections
The Safe Drinking Water Act (SDWA) of 1974 requires U.S. EPA to
provide for the safety of United States drinking water. The act
contains a set of requirements which involves the protection of
underground sources of drinking water from contamination by in-
jection well activities. Seven U.S. injection facilities are
presently authorized in the Lake St. Clair area, five of which
are salt water disposal wells and two of which are hydrocarbon
storage wells. Of the salt water disposal wells, two are cur-
rently in operation: Consumers Power injects to the Dundee Forma-
tion at 957 m and Lakeville Gas Association injects to the
-------
398
TABLE VIII-16
Confirmed or possible contamination sites in the U.S. within the Lake St.
Clair groundwater discharge areas.
1. Hwy M-29 and Michigan St. This site is a gas station with a leaking
underground tank on sandy materials near the St. Clair River and a shallow
water table.
2. Clay Township Sanitary Landfill This landfill has accepted household
and commercial wastes, and is near to the north Channel of the St. Clair
River distributary system, sandy surficial deposits, and a shallow water
table.
3. Selfridge Air National Guard Base (CERCLIS/RCRA/ACT 307) The Base site
consists of 7 individual groundwater contamination sites: 3 landfills, 2
fire training areas and 2 ramps. The landfills contain residential and
industrial wastes, solvents, and waste oils. The fire training areas
contain flammable waste (JP-4), solvents, strippers and thinners. There
have been fuel spills at the two ramps.
4. Metro Beach Incinerator This closed incinerator handled general refuse
(most likely from the Metropolitan Beach Park), and is located on the
Clinton River Delta within one-half mile from Lake St. Clair over a shallow
water table and on silty-sandy surficial material.
5. Q and L Industries (Act 307) Phthalate and lead are listed as
pollutants for this fiberboard manufacturer in Mount Clemens, Mi., and
groundwater contamination is indicated. The site is located on sandy soil
near to a shallow water table and aquifer.
6. County Line Landfill This landfill accepted household, commercial and
industrial wastes.
7. Henning Road Landfill (Act 307) The Landfill accepts domestic waste.
Groundwater contamination is not indicated in the Act 307 listing.
8. Sugarbush Road Dumpsite (CERCLIS/Act 307) This site is a solid waste
landfill with pollutants of concern being Pb, Ni, Cr, Cu and Zn. Surface
water, air and soil contamination are indicated in the Act 307 listing.
Groundwater contamination is not indicated, but there are no monitoring
wells.
9. Rosso Highway SAFE - Avis Ford This landfill accepted foundry sand.
CERCLIS: Site is listed within the information system for Superfund and is
considered for clean-up under the comprehensive Environmental
Compensation and Recovery Act of 1980 (CERCLA).
RCRA: Site has current activity under the Resource Conservation and
Recovery Act.
Act 307: Site is listed on Michigan's compilation of sites of known and
possible environmental degradation.
-------
399
Sylvania Sandstone at 733 m in Oakland county. One additional
well is presently under construction in Oakland County. Two
wells are temporarily abandoned: one to the Detroit River Group
of formations at 276 m and one to the Sylvania Formation at 588 m
Consumers Power Co. operates the two gas storage caverns in the
Salina Formation Group.
Estimates of Groundwater Discharge to Lake St. Clair
Groundwater discharges to Lake St. Clair from three hydrogeologic
units termed the shallow glacial (or shallow plus intermediate
units), glacial-bedrock interface (or regional, freshwater
aquifer), and bedrock units. The shallow glacial unit consists
entirely of Pleistocene Age glacial deposits. In southeastern
Michigan these are mostly silty-clay till and glaciolacustrine
deposits that contain discontinuous stringers of sand and gravel.
Base flow of perennial streams largely represents groundwater
discharge from this unit.
The glacial-bedrock interface unit occurs between the shallow
glacial unit and the bedrock. In general, the glacial-bedrock
interface unit discharges less water to the Connecting Channels
than does the shallow glacial unit. Environmental concerns,
however, are that high head pressures from deep waste injection
practices could cause waste fluids to migrate through fractures
or more permeable horizons in the rock. The glacial-bedrock
interface unit could thus be one pathway by which waste fluids
could reach the channels or contaminate adjacent groundwater. No
evidence exists at present that this has occurred in Michigan.
The bedrock unit is defined as the first bedrock aquifer lying
directly beneath the Connecting Channels. In the Lake St. Clair
study area, the bedrock unit includes all carbonate rocks of the
Traverse Formation which lie at depths of. 30 to 91 m beneath the
Antrim shale.
Total discharge from the three units to the Lake St. Clair study
area was estimated to be 1,315 L/s.
More direct measurement of groundwater flow to Lake St Clair was
also undertaken. Recognizing that all flow entering the lake
from groundwater must pass through its bed, the flow was calcu-
lated using the lakebed area, hydraulic gradients, and hydraulic
conductivities established by an electrical survey of the lake
sediments. The advantage of the electrical survey approach to
calculating groundwater flux is that it produces continuous meas-
urements of the hydraulic conductivity, as long as sediment is
present over the bedrock, allowing both detailed resolution of
the locations of groundwater inflow and an alternative method to
calculate the quantity. Summations of groundwater fluxes for the
entire lakeshore show a total groundwater discharge of 886 L/s.
-------
400
This estimate agrees well with that from above, estimated from
fluxes within geologic units.
Groundwater Contamination
In order to determine the concentration of contaminants in
groundwater in the Lake St. Clair area, and subsequently to cal-
culate loads from groundwater to the lake, eight monitoring wells
were installed in four groundwater discharge areas on the
Michigan shore of Lake St. Clair. Analyses of water from the
wells were made for volatile, base neutral, acid extractable and
chlorinated neutral extractable hydrocarbons, trace metals, and
other chemical substances.
Volatile hydrocarbons, if present, were consistently less than
the detection limit of 3.0 ug/L. Benzene was detected in water
from one well near Mt. Clemens at a concentration of 3.1 ug/L.
Concentrations of base neutral and acid extractable compounds,
and 13 chlorinated pesticides, were also generally below the
analytical detection limits of 0.1 to 30 ug/L and 0.01 ug/L res-
pectively. Phthalates were found in the water from all but one
well, with concentrations up to 170 ug/L (for bis (2-ethyl hexyl)
phthalate).
Some pesticides were found in four wells at levels exceeding
U.S.EPA Ambient Water Quality Criteria for Chronic Effects and
the GLWQA Specific Objectives. Lindane and total DDT were found
down-gradient from the Clay Township Landfill near the St. Clair
River delta. DDT was found also in wells near New Baltimore and
St. Clair Shores. Heptachlor was found in a well near the
Selfridge Air National Guard Base (ANGB).
Most wells exceeded the GLWQA Specific Objectives, the Ontario
(Provincial) Water Quality Objectives or the U.S.EPA Drinking
Water Primary or Secondary Maximum Contaminant Levels for total
phenols, phosphorus, pH and some heavy metals. The elevated
metals concentrations may have been due to the inclusion of fine
particulate matter in the samples, and if so, the concentrations
of metals dissolved in the groundwater may be much lower than
those reported. The well near the Selfridge ANGB contained the
highest levels of phosphorus, phenols, dissolved solids and spec-
ific conductance.
A computation of the loading of chemical substances transported
by groundwater to Lake St. Clair does not seem feasible based
upon the data currently available. Concentrations of organic
compounds were generally less than their respective limits of
analytical detection, and concentrations of trace metals were
reported higher than they would have been had the finely divided
particulate matter been excluded from the analyses.
-------
401
7. Spills
Spills reports from Michigan and Ontario information systems were
reviewed and indicate that a limited number of spills to surface
water occurred in 1986. However, in many cases the volume of the
amount spilled was not known and it is no.t possible to compare
point source effluent loadings with the loadings due to spills.
8. Contaminated Sediments
Identification
i) Organics
Depth-integrated samples (interval composites) were prepared from
sediment cores collected in 1985 (Figure VIII-5) and analyzed for
organics in order to estimate the mass of contaminants stored in
the sediments. Horizontal distributions in total storage have
patterns which are essentially congruent with the thickness of
recent sediments and form the basis for estimating total storage
in the lake by contour integration. For the sandy nonaccumula-
ting area, where cores were not collected, a value of 5 ng/cm2
was used for PCBs and HCB, and a value of 0.5 ng/cm2 was used for
OCS. These approximations were not critical since the sandy
areas contributed less than 5% of the contaminant mass for these
chemicals. Lake St. Clair sediments presently contain about 960
kg of HCB, 870 kg of PCBs and 210 kg of OCS.
These values are much higher than the contaminant masses found by
Oliver and Pugsley (40) for the St. Clair River sediments (3 kg
HCB, 20 kg OCS) indicating that Lake St. Clair is a more signif-
icant repository for chemicals than the river itself, in part due
to the much greater mass of sediments in the lake. Recent load-
ing estimates for HCB and OCS in the combined dissolved and par-
ticulate fraction at Port Lambton in the St. Clair River were 180
kg/yr for HCB and 11 kg/yr of OCS. At these rates, Lake St.
Clair sediments contain the equivalent of 5 years loading of HCB
and 20 years loading of OCS. Thus, the sediments retain signif-
icant fractions of these chemicals and, given the uncertainties
in the calculation, accumulation is consistent with sediment
reservoir residence times derived from historical studies of
metal and organic chemicals in the system and from the response
of sediments to particle-associated radionuclides.
ii) Metals
In order to estimate the total mass and anthropogenic mass of
each metal stored in Lake St. Clair sediments, the sediment cores
collected at each station in 1985 were designated to be repre-
sentative of a region of the lake. The anthropogenic mass of
each metal stored in each sediment type was calculated by sub-
tracting background metal concentrations from all concentrations
-------
402
in post-settlement sediments. In general, metal concentrations
increased above the glacial deposits.
Within the lake and its marshes, 30 to 64% of the mass of metals
stored in post-settlement sediments is anthropogenic. Storage of
anthropogenic metals is highest in the silts and clays (48-70%),
second highest in the sands (32-35%), and lowest in the marshes
(5-29%). An exception to these general statements is the high
fraction of anthropogenic lead stored in the marshes (29%), based
on the one core used to represent the marshes.
Lake St. Clair appears to be a temporary trap for some metals
(Table VIII-9). Thus, sediments and their associated contamin-
ants, appear to be transient and will eventually be transported
down the Detroit River to Lake Erie.
Classification
Using the OMOE and U.S.EPA pollution guidelines, the sediments
underlying the open water of Lake St. Clair can be classified as
only lightly polluted. Sediments at the mouths of some tribu-
taries are more contaminated.
9. Navigation
As a result of the Rivers and Harbors Flood Control Act of 1970,
which authorized the U.S. Army Corps of Engineers to construct
facilities for containment of polluted dredge spoil from the
Great lakes harbors and waterways, two diked facilities were
constructed on Dickinson Island adjacent to North Channel in the
St. Clair delta. Both sites were located on the high pre-modern
delta deposit and did not infringe on the wetlands. These
disposal sites were designed to accommodate dredgings produced
during a 10 year period, and they presently receive the materials
dredged from the St. Clair system. Navigation-related dredging,
which removes contaminated sediments and deposits them in con-
fined disposal facilities could be considered beneficial in that
the total contaminant load within the system is reduced. Impacts
of the dredging due to resuspension of contaminated sediments
during the dredging operations, and the subsequent temporary
increase in bioavailability of the contaminants, have not been
documented.
Commercial vessel operations through the shipping channel are
also believed to cause some local sediment resuspension. The
extent of influence and effects of the contaminants associated
with the resuspended particles have not been documented.
-------
403
D. DATA LIMITATIONS
A detailed discussion of data quality management for the UGLCC
Study can be found in Chapter IV. The information presented
below reflects concern for some data quality pertaining specif-
ically to Lake St. Clair.
1. Sediment Surveys
References in the text to a "1983 sediment survey" in Lake St.
Clair refer to a study conducted by the OMOE. The data have not
yet been published, nor have the methods, results or any inter-
pretation of the data been peer reviewed. Discussions with the
principal investigators, however, indicate that the samples were
obtained by bottom grab sampler, and the top 3 cm of each of 3
grabs were composited. The samples were then sent to a labora-
tory for analysis by "standard techniques". This study has the
appearance of being a valuable contribution to the knowledge of
the distribution of contaminants in Lake St. Clair sediments.
However, the data must be considered "preliminary" at this time,
and used only to support the findings of other documented
surveys, particularly the 1985 surveys conducted by Environment
Canada and by U.S. Fish and Wildlife Service.
2. Tributary Loadings
Accuracy of estimates of tributary loadings of chemical param-
eters is dependent on the responsiveness of the stream to storm
events and on the frequency of sampling. Data from a program
employing infrequent sampling will generally be biased low for
substances which increase in concentration with increasing stream
flow, such as nutrients from agricultural runoff (41). Of
various sampling strategies, flow-stratified sampling, i.e.,
emphasizing storm events, and calculations provide the most ac-
curate results. Loading data for phosphorus, nitrogen,
chlorides, lead and cadmium from the Clinton, Thames and Sydenham
Rivers were based on a combination of monthly and storm-event
sampling and included from 15 to 72 samples per year. Data for
the Ruscom, Puce and Belle Rivers were based on only 14 or fewer
samples per year, and may therefore be subject to considerable
error.
A recent analysis of the flow responsiveness of Great Lakes tri-
butaries, i.e., their potential for change in rate of flow in
relation to storm events, indicated that the Clinton River was
"stable", the Sydenham River was "event responsive", and the
Thames was intermediate between the other two (42) Estimates of
loads of phosphorus, Cd and Pb for the Thames River, with the
greatest number of annual samples, and the Clinton River, with
the most stable flow, may be expected to have about the same
-------
404
accuracy, although confidence intervals were not reported. The
estimate for the Sydenham, with the about the same or fewer samp-
les and more variable flow response, may under represent the true
load by some unknown amount.
The difficulty in calculating loads from small data sets created
the need to make loading calculations using several methods. For
Canadian tributaries, the Beale Ratio Estimator was used to ar-
rive at loads for P, Cd, Pb, and Cl. Loading calculations for
these parameters plus NC>3 were also made from the same data set
by plotting P concentration vs flow. A "best line fit" was then
drawn, and concentrations were then read off the graph for days
on which no samples were taken. Phosphorus loads on the remain-
ing Canadian tributaries were calculated using a two-strata
method. A "cut-off" line was determined by doubling the annual
mean flow. An average concentration was found for days when flow
exceeded the cutoff, and another was found for days with flow
below that value. Loads for unsampled days were calculated by
multiplying the average concentration by the flow for that day.
The values presented in this report represent an arithmetic
average of results obtained by the two methods.
Concentrations of lead, cadmium, chloride and nitrogen in
Canadian streams did not exhibit a variation with respect to
flow. Therefore, loads were calculated by averaging all samples
and multiplying by the flow.
For the Clinton River, loads were calculated using the
Stratified Ratio Estimator (43). This method is essentially a
modification of the Beale Ratio Estimator.
The average annual loads for Canadian tributaries as displayed in
Table VIII-1 represents a mixture of included data. For P, the
average unit area load is based only on data from the Sydenham
and Thames Rivers, which comprise 57% of the Lake St. Clair
watershed. Were an arithmetic average of all estimates of load-
ings from all Canadian tributaries to be used, the unit area
loading would have been reported as 3.18 kg/ha instead of 2.26
kg/ha.
For NC>3 and Cl, the loadings include an average unit area loading
from the Ruscom River, which was approximately 10 times that of
the other rivers in 1985. The average unit area loading for the
Sydenham and Thames Rivers combined for N03 and Cl was 20.5 kg/ha
and 160 kg/ha, respectively, instead of the reported 40 kg/ha and
287 kg/ha. The cause for the Ruscom River concentrations and
loads may need investigation, but the data should not be con-
sidered typical of the unit area loads for the Lake St. Clair
watershed.
-------
405
3. Point Sources
The point source monitoring data in general were developed with a
rigorously defined quality assurance program. Due to constraints
on the sampling frequency and quantity, however, a number of
shortcomings in the point source survey data limit the inferences
that can be drawn from the results of the study. Most facilities
in the Lake St. Clair basin were not sampled. The major facili-
ties closest to the lake itself, as opposed to those furthest
upstream, were surveyed, however.
One deficiency, that of a small data base consisting of one day
sampling by the U.S. and 3 to 6 day surveys by Canada, prevents
precise determination of annual loadings. The timing of the
surveys reduced the comparability of the data. The U.S. surveys
were carried out during May and August of 1986, while the
Canadian data was collected on October 1985 and March and
November of 1986. The sampling methods were also different. The
U.S. composited four grabs (one per six hours) for each facility.
Canadian samples were collected by automatic composite samplers
(one portion per 15 min.). Differences in the analytical methods
and the method detection limits used by the U.S. and Canada for
several parameters also reduced data compatibility. This defi-
ciency was particularly pronounced for PCS analyses.
Despite these limitations, the data were considered adequate for
identifying major sources of contaminants, and were used to make
conclusions and recommendations concerning specific point
sources.
4. Fish Consumption Advisories
The data upon which the fish consumption advisories for Lake St.
Clair are based were derived primarily from Canadian analyses of
samples of the edible portions of fish. This method generally
returns concentrations of contaminants less than those found in
larger skin-on fillets that the U.S. uses for its analyses of
contaminants in fish. One implication, therefore, is that if the
U.S. method for assessing contaminants in fish were used, the
fish consumption advisories may become more restrictive. Al-
though the impacts to humans of contaminants other than mercury
in fish flesh for commercially marketed fish are not quantified,
the advisories remain useful as a general guide for use by the
public who consume fish from Lake St. Clair.
-------
406
E. MODELING AND MASS BALANCE CONSIDERATIONS
1. Mass Balance Models
Four days prior to the onset of the System Mass Balance measure-
ments in the Detroit River, measurements of contaminants entering
Lake St. Clair from the St. Clair River were initiated. The
intent of starting four days before making measurements on the
Detroit River was to allow for passage of most of the St. Clair
River water through the lake. By doing so, upstream and down-
stream contaminant fluxes could be compared and conclusions could
possibly be drawn concerning whether Lake St. Clair is a source
or a sink of contaminants. It must be emphasized that the valid-
ity of comparing upstream and downstream measurements in this
mass balance calculation depends on how well the same parcel of
water was sampled at the head and mouth of Lake St. Clair. Given
winds that existed.during the sampling time, and output from a
particle transport model (developed at the National Oceanic and
Atmospheric Administration - NOAA) discussed below, we estimate
that 60-80% of the water that entered the lake, exited it on day
four. Therefore, downstream contaminant fluxes that are 20-40%
different from upstream fluxes cannot be argued to be signif-
icant. On the mass balance diagrams that follow (Figures VIII-6
through VIII-13), best estimates of point and nonpoint source
inputs have also been noted. If estimates were not available,
they are indicated with a "?" on a diagram. Loading information
was compiled with data provided by the Point and Nonpoint Source
Workgroups. Groundwater loading estimates are extremely prelimi-
nary and should be treated as such. These diagrams should there-
fore be used only to suggest possible issues that may require
further investigation. This is because of uncertainty about time
lags between the head and mouth of the Lake, and the "long term
average" character of some of loading information.
In most cases, the downstream contaminant fluxes do not differ
widely from the contaminant flux entering the lake via the St.
Clair River. In the cases of cadmium and particularly lead, it
appears that a significant portion of the lake's total load could
be coming from its tributaries. If the Thames River lead loads
are reasonably accurate, then a regulatory problem may exist.
Sediment records that indicate a net storage of lead over the
years would corroborate these observations.
A total phosphorus budget was developed for Lake St. Clair for
1975-1980 (Figure VIII-13). Phosphorus load estimates were made
for point sources and hydrological areas (Figure VIII-14).
During this period Lake Huron accounted for 52% of the total
annual load, while hydrologic area loads accounted for 43% (13).
The remaining load came from the atmosphere, shoreline erosion
and direct point sources. The Thames hydrologic area contributed
58% of the total hydrological area load, followed by the Sydenham
(17%), the Clinton (9%), the Ruscom (7%), and the Black (6%).
-------
407
Atmosphere
.71
Upstream
Input
12.3
SIS.
Point
Sources
.075
.57-.77
Clinton R.
Ground
H20
Lake
St.. Clair
.02
CANADA
Point
Sources
.-4'-6 Syden. R.
.2-2.2
Thames R.
Ground
H2O
14.3
Downstream output
ln=15.4-16.8
out=14.3
slnk=1.1 • 2.5?
FIGURE VIII-6. Lake St. Clair
total cadmium (kg/d).
Atmosphere
?
US.
Point
Sources 1-S1
9.4-15.1
Clinton R.——•
Ground
H20
1.7
Upstream
Input
450.2
CANADA
Lake
St.. Clair
472.9
Downstream output
10=465.5-471.2
OUt=472.9
sources?.4-1.7?
Point
Sources
Syden. R.
Thames R.
Ground
H2O
FIGURE VIII-7. Lake St. Clair
total copper (kg/d).
Atmosphere
.218
Upstream
Input
US.
Point
Sources -°
Clinton R. ?
Ground
H2O
CANADA
Lake
St.. Clair
-o
.252
Downstream output
ln=.218
out=.252
sources.034?
Point
Sources
Syden. R.
Thames R.
Ground
H2O
FIGURE VIII-8. Lake St. Clair
HCB (kg/d).
Atmosphere
IIS.
Point
Sources
.5
7.4-15.2,
Clinton R. aVal
13.3 46.7
Lake
St.. Clair
Upstream
Input
CANADA
Ground
H2O
18.0
Point
Sources
Syden. R.
Thames R.
123.1
Ground
~ H2O
58.8
Downstream output
1(1 = 220.7-228.5
out=55.8
store=161.9-169.7?
FIGURE VIII-9. Lake St. Clair
total lead (kg/d).
-------
408
Atmosphere
?
SIS.
Point
Sources
Clinton R.
Ground
H2O
-0
Upst ream
4.8 Input
CANADA
Lake
St.. Clair
Downstream output
in=4.8
out=7.1
source=2.3?
Point
Sources
-0
_? _ Syden. R.
_? _ Thames R.
Ground
H2O
FIGURE VIII-10. Lake St. Clair
total mercury (kg/d).
Atmosphere Upstream
S71 Input
SIS.
Point
Sources 3.2
24.9-16.4.
Clinton R."~H
Ground
H2O
1.7
Lake
St.. Clair
2.9
CANADA
Point
Sources
Syden. R.
Thames R.
Ground
H2O
499
in=595.2-603.7
out=499
sink=96-105?
FIGURE VIII-11. Lake St. Clair
total nickel (kg/d).
Atmosphere
IIS.
Point
Sources
Clinton R. 7
Ground
H2O
Upstream
Input
CANADA
Lake
St.. Clair
.0
_?
?
Sources
Syden. R.
Thames R.
Ground
H2O
.85
Downstream output
in=.89
out=.85
stores.04?
FIGURE VIII-12. Lake St. Clair
total PCB (kg/d).
U.S.
Hydrologic
Areas
Atmospheric,
Erosion, Lake Huron
Direct Point 1,621
155 r
Black
78
St. Clair
Complex 38
t
Rouge
Complex 5-
LAKE ST. CLAIR
Net Loss = 15
Canadian
Hydrologic
Areas
232 Sydenham
788 Thames
Detroit River Outflow 3,148
FIGURE VIII-13. Lake St. Clair average
phosphorus loads and losses
during 1975-80 (mt/yr).
-------
409
83W
82°00'
43*30'
43*OCT
42°00*
LAKE
ST.CLA/R
42W
Rouge ,
Complex'
'Lake\EHe
FIGURE VIII-14. Hydrological areas used in determining mass balances.
-------
410
Over the six year period examined, the lake's total input and
output of phosphorus were nearly equal. Therefore, there was no
significant net source or sink of phosphorus in the lake during
that period.
2. Process-Oriented Models
Changes of water level caused by wind are most pronounced in
shallow lakes such as Lake St. Clair. The ability to predict
wind-induced water level changes would therefore be useful, since
these changes can affect shorelines and contingent properties. A
hydrodynamic model was developed to investigate the effects of
bottom drag and wind stress on computed lake setup, and to deter-
mine the efficacy of hydro-dynamic or purely empirical approaches
to predicting water measurements. Empirical approaches by-pass
many of the calculations that are used in the hydrodynamic ap-
proach. No essential difference between the two approaches was
found, but for an empirical model to be developed, an adequate
historical data base for the site of interest must exist. The
strength of the hydrodynamic approach is that it is transferable
among lake systems.
To predict the fate and transport of contaminants in any body of
water, the movement of that water, as affected by winds or tribu-
taries, must be known or predictable. Because of this need,
several models were developed by Canadian and U.S. scientists to
predict and understand currents in Lake St. Clair. In addition,
models were developed for predicting and understanding wave dyna-
mics in Lake St. Clair since waves can resuspend sediments and
associated contaminants.
Simons and Schertzer, (Environment Canada - EC) developed a model
that predicts mean daily currents in Lake St. Clair. They found
that an important consideration in developing the model was ac-
counting for the effects of a shallow bottom on currents. Lack
of information regarding these effects has been a major impedi-
ment to the application of hydrodynamic models to shallow lakes.
They were able to develop a tentative relationship between eddy
viscosity and wind stress that aided in shallow water model
development.
Schwab and elites (NOAA) developed a particle transport model for
Lake St. Clair to answer the following questions: 1) What path
does water entering Lake St. Clair from one of the tributaries
follow through the lake before leaving the Detroit River? 2)
How long does it take? 3) How is the particle path changed by
wind-induced circulation in the lake? 4) For the meteorological
conditions during the summer and fall of 1985, what are the typi-
cal statistical distributions of these pathways? The model they
developed calculates currents on a 1.2 km grid and yields results
that are similar to those of Simons and Schertzer above. Their
-------
411
model can be used to make preliminary estimates of the spatial
distribution, transport and residence times of conservative,
hazardous spills in Lake St. Clair. This model, however, only
tracks conservative, nondispersive tracers from the mouths of the
tributaries through the lake under various wind conditions.
Even though the average hydraulic residence time for Lake St.
Clair is about nine days, the residence time for conservative
particles entering the lake from the individual tributaries
ranges from 4.1 days for the Middle Channel to over 30 days for
water from the Thames River, depending on the wind conditions.
If significant contaminant loads were to enter the lake from
tributaries that have long residence times, the impact of these
contaminants might be greater than if they entered the lake from
other tributaries.
Most of the water from the St. Clair River enters the lake
through the North Channel (35%). According to the calculations,
this water tends to flow down the western shore of the lake and
never gets into the central or eastern parts of the basin. Water
from the Middle Channel tends to remain in the western third of
the lake, almost never entering the eastern half. Water from St.
Clair Flats and the St. Clair Cutoff can be dispersed almost
anywhere in the lake to the south of the shipping channel which
connects the St. Clair Cutoff with the Detroit River. A small
amount of the St. Clair inflow. (5%) enters through Bassett
Channel. This water can pass through any part of the eastern
half of the lake depending on the wind conditions. The Thames
inflow tends to be confined to the eastern and southern shores
before reaching the Detroit River and it can take a very long
time to get there. Water from the Clinton River and the Clinton
Cutoff is most likely to follow the western shore of the lake
southward with the most probable paths within 3 km of the western
shore.
Water quality measurements made in Lake St. Clair by Leach (44,
45) showed two distinctly different areas in the lake. In the
southeastern part of the lake, the water quality was dominated by
the Thames inflow, which is a major source of phosphate and other
dissolved and suspended material. The central and western parts
of the lake possessed water quality similar to Lake Huron than to
the southeastern part of the lake. The pattern of water mass
distribution (45) is very close to the combined patterns of the
four main St. Clair River inflows and the Thames inflow. Bricker
e_t al. (46) examined the distribution of zooplankton in the
western half of the lake. They distinguished an area of biologi-
cal and physiochemical similarity along the western shore of the
lake that appeared to be influenced more by the Clinton River
than the St. Clair River. The shape of this area matches quite
well with the modelled distribution pattern for water from the
Clinton River.
-------
412
To verify the circulation model and lend credulity to currents
calculated by Schwab and elites, their model was tested by com-
paring model output to actual current data measured in Lake. St.
Clair in 1985. Two separate current data bases were gathered.
One involved the use of 5 drifting buoys which were repeatedly
launched and tracked in the lake. The other was the result of
several synoptic current surveys utilizing electromagnetic cur-
rent meters. Currents predicted by the circulation model were
used to simulate 16 drifter tracks. Most of the tracks were
about 2 days in length from various portions of the lake. In
most cases, the model simulated the tracks extremely well as did
a similar study by Hamblin et al., (26). For the entire data
set, the mean root mean square (rms) of the drifter was 25% grea-
ter than that of the calculated current track. The directions
compared favorably except for a few tracks near the mouth of the
Bassett Channel, where the model prediction was over 90 degrees
different in direction when compared with the observed track.
The comparisons between current meter measurements and model-
predicted currents were even better. In nearly 100 comparisons,
60% of the variance is explained by the model prediction. The
model again seems to under-predict the current speeds, here by
about 30%.
Contaminant transport depends in large part on the movement of
suspended particles. Therefore, accurate computation of hori-
zontal sediment transport should rely upon the accurate simula-
tion of the vertical structure of the horizontal flow field.
Hamblin et al., (26) developed such a three dimensional finite
element model for Lake St, Clair. Model agreement with observa-
tions was good near the lake bottom but poorer near the surface
and suggested that a more elaborate model would be needed to
accurately model vertical velocity profiles. The more elaborate
model would include the effect of surface waves.
An empirical model was developed to describe and understand the
relationship between waves and sediment settling and resuspension
(25). The importance of these relationships to our ability to
predict and understand the transport of contaminants is evident.
Statistical relationship between suspended matter and concentra-
tion and wave orbital velocity was computed. Integration of
computed resuspension rates provided an estimate of sedimentation
in sediment traps. The model-generated sedimentation rates com-
pared rather well with the sediment trap data.
Present Status of Physical-Chemical-Biological Models
To predict the fate and behaviour of contaminants, models that
integrate physical, chemical, and biological processes are often
needed. Two such synthesis models were developed for predicting
contaminant fate in Lake St. Clair. Halfon (EC) utilized TOXFATE
and Lang, Fontaine and Hull (NOAA) utilized the U.S.EPA TOXIWASP
-------
413
model. TOXFATE was used to predict the spatial distribution of
seven halocarbons in Lake St. Clair, and the fate of perchloro-
ethylene in the St. Clair - Detroit River system. The TOXIWASP
model was used to predict and understand the fate of the contam-
inant surrogate Cs-137, as well as PCBs and OCS. Neither of
these models could be fully tested for Lake St. Clair applica-
tions due to a limited test data set. However, these models are
based on well documented cause and effect relationships, and as
such, could be used to forecast the fate and behaviour of con-
taminants introduced to the lake in the future. Representative
results of Halfon's Lake St. Clair TOXFATE model is demonstrated
in Figure VIII-15.
Lang and Fontaine (NOAA) developed a multi-segment, generic con-
taminant fate and transport model for Lake St. Clair. The
TOXIWASP code upon which it was based was. streamlined to make it
more specific to Lake St. Clair. Because evidence of biological
mixing in Lake St. Clair was extensive, this capability was added
to Lake St. Clair version of TOXIWASP. An extremely fast version
was created that calculates steady state contaminant concentra-
tions in seconds rather than hours. Numerous programming errors
in the original code were found, corrected and passed on to the
U.S.EPA-Athens modeling group.
Lang and Fontaine (NOAA) calibrated the transport mechanisms of
TOXIWASP using chloride and meteorological data that were col-
lected during a series of cruises in Lake St. Clair during 1974.
After- obtaining reasonable agreement with the conservative
chloride ion, calibrations of contaminant dynamics was carried
out using Cesium-137. Cesium-137 was used to calibrate the
model's contaminant dynamics since Cesium-137 adsorbs to par-
ticles in a manner similar to that of many hydrophobic, organic
contaminants. Most importantly, the source function of Cesium-
137 to the lake is well known (Figure VIII-16). This informa-
tion, coupled with knowledge of the spatial and depth distribu-
tions of Cesium-137 in the sediments of the lake, provided an
excellent calibration and verification data set. Verification
results are acceptable (Figure VIII-17).
Having calibrated the TOXIWASP model for Lake St. Clair, it was
used to hindcast possible loadings of octachlorostyrene and PCBs
to Lake St. Clair. The model predicted that about 3.9 MT of OCS
had to have been loaded to the lake over a period of 12 years to
produce measured sediment concentrations (Figure VIII-18). This
finding implies that OCS was first loaded in the latter part of
1970 and is consistent with speculation to that fact. The model
also estimated that 3,400 kg of PCBs had to have been loaded to
produce measured PCB sediment concentrations (Figure VIII-19).
The model tended to under-predict the PCB values along the east-
ern and western segments of the main lake, which may indicate
additional or increased PCB sources in these areas.
-------
414
TRICHLOROETHYLENE
ng/L
OBSERVED JUNE 18-21
PREDICTED
FIGURE VIII-15. Modelled and observed distributions of trichloroethylene
1984.
-------
415
a
U
4^
on
w
D
O
-------
416
en
•4-*
c
•o
(U
v>
?
O
w
c*
D
O
E
-------
417
c
CL>
S
o
o
2
o
c
o
u
c
o
o
00
w
ft!
D
O
E
-------
418
c
W
1*
O
E
-------
419
TOXIWASP assumes a local equilibrium between the dissolved, par-
ticle-adsorbed and bio-adsorbed chemical. Hull, Lang and
Fontaine (NOAA) modified the TOXIWASP model so that kinetic,
instead of equilibrium, reactions were simulated. This was done
to determine whether the equilibrium approach was valid in all
circumstances. Equilibrium models assume implicitly that incom-
ing contaminant loads are at local equilibrium between dissolved,
adsorbed, and bioaccumulated phases. When the same load
conditions were assumed for the kinetic model, greatest devia-
tions between the two models occurred when predicting the fate of
highly hydrophobic contaminants (Kow>106). The kinetic model not
only required a longer time to reach steady state contaminant
concentrations, but also required a longer time to flush out the
resident contaminant mass after the input load was shut off.
Generally, one would expect problems with an equilibrium approach
when the time to equilibrium is longer than the residence time of
the water body in question.
Halfon (EC) used TOXFATE to predict the fate of perchloroethylene
(PERC) in the St. Clair - Detroit River system. The model sug-
gested that about 82% of the PERC would be volatilized, and the
remainder, less 1% that would remain in sediments, would enter
Lake Erie. Comparison of simulated and measured PERC concentra-
tions show reasonable agreement. Since so much of the PERC is
volatilized before it reaches the open lake, Halfon's model does
not realistically demonstrate what may happen to a nonvolatile
spill entering the lake.
In the case of a nonvolatile spill travelling the lake from the
St. Clair River to the Detroit River outflow, the dilution of the
concentration would be determined mainly by the strength of
horizontal turbulent mixing. There were no direct measurements
of horizontal diffusion in Lake St. Clair reported by any of the
UGLCCS activities. However, two investigations (17,53) have
employed a vertically integrated model of transport and diffusion
of a conservative substance, chloride, to infer an effective
horizontal diffusion coefficient of 10+5 cm^/s. Because this
quantity has been deduced from vertically averaged concentration
in the possible presence of current shear over the water column,
these authors have termed the diffusion coefficient as a disper-
sion coefficient.
The particle trajectory measurements and models reported for
August 12, 1985 by Hamblin (47) and by the Modeling Workgroup
Report (53) for September 1985 demonstrated that particles would
take about four days to cross the lake. If a slug of contamina-
ted river water had dispersed longitudinally to a length of 5 km
in the St. Clair River, then in the four day transit to the out-
flow region it would have grown by about 7 km to a characteristic
patch size of 12 km under the assumptions of average mete-
orological conditions and horizontal Gaussian diffusion. In
turn, this patch would take about two days to pass across the
-------
420
water intakes near the outflow. Finally, the average concentra-
tion would be about 20% of the original concentration entering
the lake.
3. Summary
The modeling work on Lake St. Clair has made much progress during
the study period from the water level fluctuation models (storm
surge) to the coupled contaminant-circulation models. However,
more work is required before the models could be used as
effective water management tools. Testing of the models with
parameters additional to PCB and OGS, more realistic treatment of
sediment water interaction, and linkage of the models to lake
biota are seen as necessary steps before the models can reliably
assess the ecological responses to reductions in loadings to the
lake. Although not developed for operational purposes, the mod-
els TOXFATE and TOXIWASP, with modest additional effort, could be
used to predict the trajectories and dilutions of spills of
either volatile or nonvolatile substances occurring on the lake
or entering from the rivers.
-------
421
F. OBJECTIVES AND GOALS FOR REMEDIAL PROGRAMS
The following objectives and goals are grouped according to
media. However, remedial actions are likely to have multimedia
effects. For example, elimination of point and nonpoint sources
of contaminants can be expected to reduce concentrations in
water, sediments and biota, even though direct remediation of
contaminated sediments or biota may be infeasible. Some objec-
tives may be reached, therefore, upon attainment of one or more
others.
1. Water Quality
Since the water quality of Lake St. Clair is dominated by that of
the St. Clair River, remedial programs directed towards the St.
Clair River will also improve water quality in Lake St. Clair.
Objective 1. Full implementation of recommendations for the St.
Clair River presented in Chapter VII of this
report for the elimination of industrial, munici-
pal and nonpoint sources of contaminants to the
St. Clair River, particularly HCB, HCBD, OCS, Hg,
and Pb.
Excluding input from the St. Clair River, phosphorus loadings to
Lake St. Clair are dominated by nonpoint sources. For example,
in the Thames River 93% of the loading was of the nonpoint source
type. In water samples from Lake St. Clair tributaries, nearly
all contained phosphorus in excess of the PWQO of 30 ug/L.
Improved agricultural practices such as conservation tillage,
elimination of over-fertilization and control of feedlot
effluents are identified as actions relevant to reduction of
nonpoint source loadings.
Objective 2. Reduction of phosphorus loadings from point and
nonpoint sources in Michigan and Ontario to assist
in meeting target load reductions for Lake Erie.
The Mt. Clemens WWTP was identified as having average phosphorus
concentrations in its effluent exceeding the GLWQA objective of
1.0 mg/L for municipal water treatment facilities. Municipal
treatment plants discharging to the Thames River in excess of
this guideline in 1986 were Chatham, Ingersoll (new), City of
London (Adelaide, Greenway, Oxford, Pottersburg and Vauxhall) and
the Strathroy Town Plant.
Objective 3. Necessary and sufficient technology and operation
procedures at all wastewater treatment facilities
to meet the target concentration of phosphorus in
the effluent of no more than 1.0 mg/L.
-------
422
Excessive unit area loading of pesticides from agricultural lands
into tributaries of Lake St. Clair was identified. Some areas
were identified to be of particular concern.
Objective 4. Reduction in the loadings of pesticides from all
tributaries.
Objective 5. Identification and elimination of the source of
DDT and metabolites to the Milk River.
Water quality in several tributaries was reduced by the presence
of heavy metals. Cadmium concentrations generally exceeded the
GLWQA specific objective and PWQO of 0.2- ug/L, and some were
greater than the chronic AWQC of 1.1 ug/L in the Belle, Sydenham,
Thames and Clinton Rivers. Also, some lead concentrations were
in excess of the chronic AWQC of 3.2 ug/L in the Belle and
Sydenham Rivers, and in the Thames River some exceeded the acute
AWQC of 82 ug/L.
Objective 6. Identification and elimination of all point
sources of Hg, Pb and Cd in the watersheds of the
Clinton, Thames and Sydenham Rivers.
Objective 7_ Elimination of combined storm sewer overflows
which will reduce contributions of P, Pb, Cd, Hg
and PCBs to Lake St. Clair tributaries.
2. Sediment Quality
Reductions in industrial loadings of mercury in the St. Clair
River have resulted in dramatic improvements since 1970 in the
bottom sediments. However, surface concentrations in bottom
sediments still exceed the IJC and OMOE guidelines of 0.3 ppm and
contain values classified as "polluted" by the U.S.EPA Classifi-
cation Guidelines. Since recent mercury concentrations of bottom
sediment samples do not appear to be reducing as quickly as in
the earlier studies there is some concern that unknown tributary
sources exist. The mass balance studies of Section E indicate a
net outflow of mercury from Lake St. Clair. Since the tributary
loadings are not known, it is impossible to determine the source
of the mercury.
Objective 8. Identification and elimination of continuing
sources of Hg to the St. Clair River.
Objective 9. Identification and elimination of point and non-
point sources of Hg to Lake St. Clair tributaries.
-------
423
Of the other metals, only zinc and copper exceed the OMOE guide-
lines in the sediments of the open lake and would result in a
classification of sediments as moderately polluted.
Objective 10. Reduction in heavy metals concentrations in sur-
ficial sediments of Lake St. Clair to levels sup-
porting a classification of "not polluted" by
OMOE, U.S.EPA and IJC Guidelines.
The sediment surveys revealed that PCBs did not exceed the guide-
lines in the open lake. However, guideline concentrations were
exceeded in some of the tributary sediments including the Cot-
trell Drain, the mouth of the cutoff channel of the Clinton River
and the Sydenham River. Other organic contaminants with specific
guidelines such as HCD, OCS and pesticides were identified in
sediments from the open lake and tributaries. In general, the
sampling of all tributary sediments was incomplete, so there
could be cases of excesses of certain compounds not reported or
cases of compounds that were sampled which have no guidelines.
Objective 11. Elimination of DDT in sediments at the mouth of
the Milk River.
Objective 12. Identification and elimination of sources of PAHs
in sediments from the Milk River, Cottrel Drain,
Clinton River and Prog Creek.
Objective 13. Reduction in PCB concentrations at the mouths of
Lake St. Clair tributaries such that the sediments
would be classified as "not polluted" by OMOE,
U.S.EPA and IJC Guidelines.
3. Biota and Habitat
The most significant impaired use of Lake St. Clair waters is the
restriction in the consumption of sports fish. A joint fish
consumption advisory between Ontario and Michigan remains in
effect for the larger specimens of 18 species of sports fish
(33). Levels of mercury in excess of Canadian governmental
guidelines have been identified as the main contaminant respon-
sible for restricted fish consumption. Because the concentra-
tions of mercury in the tissues of sports fish have declined
dramatically since 1970, programs to control the major historical
sources of mercury appear to be satisfactory. However, since
tributaries were not monitored, smaller, uncontrolled sources
could be contributing to the loading.
Objective 14. Reduction in mercury concentration in Lake St.
Clair fish to less than 0.5 mg/kg, and subsequent
elimination of the fish consumption advisory based
on mercury contamination.
-------
424
Objective 15. Continued reduction in PCB concentrations in fish
to meet the GLWQA specific objective of 0.1 mg/kg
for protection of birds and animals which consume
fish.
In addition to being an important sports fishery, Lake St. Clair
is a major duck hunting area. The habitat necessary for wildfowl
resting, feeding and breeding is provided by the extensive wet-
lands around Lake St. Clair particularly in the Lake St. Clair
Delta. More than 9,000 km of wetlands were lost to shoreline
development in Lake St. Clair between 1873 and 1968. Losses are
most evident in the Clinton River, the St. Clair River Delta and
the eastern shore of the lake. In 1979 the state of Michigan
prohibited the modification of a wetland over 5 acres in size to
restrain encroachment into the wetland areas. In Ontario, sub-
sidies for engineering projects still encourage drainage of wet-
lands and their conversion to agricultural use. However, tax
relief that favors retention of the wetlands has recently (1987)
been granted to wetland owners. Although diked Ontario wetlands
are effectively managed for waterfowl hunting, there is a loss of
other wetland functions, particularly those related to fish prod-
uction.
Objective 16. Preservation of remaining wetlands surrounding
Lake St. Clair, and protection of them from
further diking, filling or other forms of destruc-
tion.
4. Management Issues
In the Clinton River, the concentration and impact of contamin-
ants are sufficiently severe for the area to be recognized as an
IJC "Area of Concern". A Remedial Action Plan is in the process
of being developed by the State of Michigan for restoring benefi-
cial uses of the area. This plan will contain details of the
problems, their extent and causes, and a schedule for remedial
actions to be implemented. Plans for further monitoring for
results of the actions will also be included.
Objective 17. Full implementation of the Remedial Action Plan by
Michigan and other responsible agencies for clean-
up and restoration of uses in the Clinton River.
-------
425
Although the Thames River is not presently one the IJC Areas of
Concern, many agricultural and industrial contaminants have been
identified in the water and sediments, and impaired uses were
identified that are similar to those for 'the Clinton River. The
absence of the Thames River on the AoC list should not imply that
the area is contaminant-free.
Objective 18. Preparation and implementation by Ontario of a
Plan for the restoration of impaired uses in the
Thames River. The Plan should address issues of
agricultural runoff of nutrients and pesticides,
CSOs in the watershed, and sources of heavy metals
in the tributary.
-------
426
G. ADEQUACY OF EXISTING PROGRAMS AND REMEDIAL OPTIONS
1. Projection of Ecosystem Quality Based on Present
Control Programs
In general, the ecosystem quality in Lake St. Clair is adequate
for the maintenance of a desirable biological community that
includes the production of sport fish. Impairment of the bio-
logical communities due to contaminants appears to exist only in
localized areas around the mouth of some tributaries (although
some contaminant levels in fish are sufficient to force the is-
suance of a fish consumption advisory by Michigan and Ontario),
and the loadings of agricultural nutrients have not caused severe
eutrophication problems. Loss of habitat due to wetlands de-
struction, however, has been extensively documented.
The specific concerns addressed in Section B, above, relate most-
ly to contaminants in the Lake St. Clair basin, and can be
grouped into three major categories: nonpoint source loading of
contaminants and nutrients, contaminants in tributary water and
sediments, and contaminants in fish. Of these categories, insuf-
ficient data exist to determine trends in the loading of con-
taminants from nonpoint sources, including tributaries. However,
the concentration of mercury in the edible portions of northern
pike, white bass and yellow perch from Lake St. Clair, and of
PCBs in walleye from 1970 through 1984 have been declining at a
geometric rate (7), indicating that control programs for these
two contaminants have been at least partially effective. Evi-
dence for continuing loadings of nutrients, pesticides, PCBs, and
heavy metals implies that the rate of decline in contaminant
burdens in fish could be greater were no additional contaminants
entering the system.
Although the impact of the loading of the UGLCCS parameters to
Lake St. Clair directly may appear to be minimal, consideration
must be given to the ultimate impact on Lake Erie populations.
Lake St. Clair may be storing HCB and HCBD, but it is a source
for PCBs and total phosphorus. These contaminants are then
transported through the Detroit River and should be accounted as
loadings to Lake Erie.
2. Assessment of Technical Adequacy of Control Programs
Present Technology
In 1985, inputs of nine of the UGLCCS parameters were determined
to be significant, resulting in impacts to either water, sediment
or biota quality. These were cadmium, copper, cyanide, lead,
mercury, nickel, PCBs, phosphorus and zinc. In general, dis-
charge of these parameters from point sources was not controlled
by limitations or objectives. All of the surveyed point sources
-------
427
were municipal facilities, and all were subject to discharge
limitations mainly for conventional parameters. However, for
many of the parameters, point sources were not the most signifi-
cant contributors. Rather, the largest loading was obtained from
unidentified sources discharging through tributaries.
The control of phosphorus has been the main approach of the U.S.
and Canada to remediating the eutrophication of the Great Lakes.
All municipal plants surveyed in the Lake St. Clair basin had
average concentrations less than 1 mg/L, except the Mt. Clemens
WWTP. The GLWQA Objective, the Canadian Municipal effluent Ob-
jective, and the standard Michigan permit limit for phosphorus is
1.0 mg/L monthly average in sewage plant effluent. The Mt.
Clemens WWTP exceeded the 1 mg/L average frequently in 1986 ac-
cording to self-monitoring data. An expansion and improvement of
the facility is underway (1987) which will enable the plant to
meet the limitation.
Excluding input from the St. Clair River, the Thames River
provided the largest loading of phosphorus to Lake St. Clair,
exceeding the contributions made by the point sources by a factor
of about 16. Similarly, the Sydenham and Clinton Rivers exceeded
the point source loadings by factors of 7 and 3 respectively.
Atmospheric loading to the lake was less than 5% that from the
Clinton River. This indicates that these rivers were receiving
substantial inputs of phosphorus from other sources, and that
controls were not adequate or effective. The most probable route
is drainage of phosphorus from agricultural uses and livestock
operations. The application rates in Michigan and Ontario were
found to be 2 and 3 times the recommended rates, respectively,
and the use of conservation tillage techniques were not
widespread.
Likewise, excluding input from the St. Clair River, the Thames
River provided the largest loading of cadmium to Lake St. Clair,
almost twenty times greater than all point sources combined. Of
the three point sources that were found to discharge cadmium,
none did so to the Thames River. The loading from the Sydenham
River was 34 times greater than accounted for by the Wallaceburg
WWTP, and the loading from the Clinton River was 11 times that of
the two WWTPs that discharged Cd. None of the facilities had
site-specific permit limits or objectives for Cd. However, the
evidence indicates that all three rivers were receiving signifi-
cant inputs of cadmium from other sources, perhaps from air depo-
sition or use of cadmium-contaminated phosphate fertilizer (48).
Estimated loading of Cd to Lake St. Clair from the atmosphere was
approximately the same at that from each of the Sydenham and
Clinton Rivers.
The Thames River also provided over 100 times the loading of Pb
than all the surveyed point sources combined, and three times the
loading from the St. Clair River. The Clinton and Sydenham
-------
428
Rivers each contributed more than 10 times the quantity of lead
than did the point sources, and the atmospheric loading was es-
timated to be similar to that of the Clinton and Sydenham Rivers.
Clearly the loading of lead to Lake St. Clair from unidentified
sources in the tributary basins was more significant than from
the point sources, which did not have effluent limitations or
objectives for lead.
Mercury contamination in Lake St. Clair has resulted largely from
historical inputs through the St. Clair River. However, inputs
may still be occurring, as evidenced by sediment surveys and by
the mass balance calculations presented in Section E, above.
Although none of the point sources surveyed had effluent limita-
tions or objectives for the discharge of mercury, point source
loadings accounted for only 0.0157 kg/d of an estimated 2.3 kg/d
source in the Lake St. Clair basin. The source could include the
contaminated sediments themselves. Loading estimates from the
tributaries and atmosphere were not available for this study.
The Clinton River also contributed significant loads of PCBs to
Lake St. Clair. Both the Warren WWTP and Mt. Clemens WWTP serve
large communities with substantial industrial bases, and both had
industrial pretreatment programs in place. Neither reported
specific sources of PCB in their service areas, and neither had
permit limits for PCB at the time of the survey. PCBs were not
found in three Ontario WWTPs. Although the Canadian MDL was
1,000 times greater than that in the U.S.., the PCB concentrations
in the U.S. sources were much higher than the Canadian MDL.
Michigan and Ontario both recommend zero discharge of PCB.
Michigan is now using a water quality based effluent limit of 1.2
X 10~5 ug/L in some NPDES permits, the allowable effluent guide-
line calculated using the State's Rule 57(2). The level is below
any current MDL, so the permits also contain an interim limit of
detection at 0.2 ug/L, the MDL commonly achieved with routine
monitoring methods. The permittee is further required to develop
a plan to meet the water quality based limit.
The Warren WWTP, Mt. Clemens WWTP, Rochester WWTP and Pontiac
WWTP all operate an industrial pretreatment program, receiving
waste water from industries in their area. Due to the quantities
of contaminants coming from these facilities, however, the pre-
treatment requirements of these facilities and/or the compliance
by the contributing industries with the requirements may be
suspect.
Similarly, the Chatham WWTP receives industrial waste water, and
it provided the largest loading of oil and grease and the third
largest loading of nickel to the Lake St. Clair Basin. The
quality of the waste water it receives may also be suspect and
not in compliance with the Ontario By-Law to control the receipt
of contaminants from industrial sources.
-------
429
Best Available Technology
Discussions concerning the adequacy of "best available
technology" (BAT) for reducing or eliminating loadings of con-
taminants to Lake St. clair are premature until specific sources
of the loadings are defined. No direct industrial discharges
occur to Lake St. Clair, but elevated levels of contaminants were
found in the water and sediments of many tributaries, implying
that sources may exist upstream. Should specific sources of
contaminants be identified, then an assessment of the impact of
BAT may be made for that industry on the receiving stream and on
Lake St. Clair,
Because phosphorus is found to be coming from agricultural prac-
tices, the implementation of conservation tillage and reduced
fertilizer application rates should greatly reduce the magnitude
of the loadings of P to the system. Likewise, reductions in
phosphorus loadings from municipal and industrial effluent, if
needed, can be achieved with improved facility design and opera-
tions. Urban nonpoint source runoff, however, may be more dif-
ficult to control.
Additional efforts are needed to identify the sources of mercury
loadings to Lake St. Clair. If internal loadings from the con-
taminated sediments are found to be significant, active control
technology might be infeasible. Techniques for dealing with in-
place polluted sediments is a topic for current research, and
demonstration projects are expected to be established within the
next several years by U.S.EPA. However, technology for treating
contaminated sediments is expected to be applicable to localized
areas, including harbors and restricted tributary mouths, but not
appropriate for a whole lake basin. Given the rather short resi-
dence time of sediments in Lake St. Clair, in the order of 10
years, the problem of contaminated sediments could be resolved
for Lake St. Clair through natural proces-ses. However, continued
problems would be expected in the western basin of Lake Erie.
3. Regulatory Control Programs Applicable to Lake St. Clair
A detailed discussion of regulatory programs in the UGLCCS
regions may be found in chapter III. The following programs have
particular impact on Lake St. Clair. The Clinton River is one of
the Areas of Concern as designated by the International Joint
Commission. As part of the effort to develop and implement a
Remedial Action Plan (RAP) for the river basin, the State of
Michigan has begun intensive remedial activities in the area
(49). All major NPDES permits in the Clinton River basin were
reviewed and new water quality based or technology based effluent
limits (whichever was more restrictive) were developed in 1985.
Metals, organics and conventional pollutants were included. A
pretreatment program for process industrial wastewater was im-
-------
430
plemented throughout the Clinton River basin as of 1987, and
upgrades to four WWTPs were completed in 1986 and 1987. Full
details of the remedial programs and schedule for implementation
will be included in the RAP, which is expected to be submitted to
the IJC in 1988.
Where stormwater is determined to impact water quality in
Michigan, the stormwater provisions (section 405) of the U.S.
Water Quality Act of 1987 will be implemented to correct the
problem. The State 305 (b) report will be reviewed in 1988 to
determine if any of the Upper Great Lakes Connecting Channels
areas are impacted by stormwater runoff.
Some technical and educational programs for farmers are in exis-
tence. For example, a Canadian Federal and Provincial effort
called the Soil and Water Environmental Enhancement Program
(SWEEP) encompasses all aspects of soil and water conservation.
Within the SWEEP program, a provincial program called the Ontario
Soil Conservation Environmental Protection Assistance Program
exists which will financially assist the farmer in implementing
soil and water conservation practices with up to 67% funding. A
Land Stewardship Program has also recently been announced to
assist farmers in the implementation of conservation techniques.
All of these programs should assist in achieving reduced phos-
phorus and pesticide contamination in streams.
The preservation of wetlands in Lake St. Clair has been assisted
by three relatively recent laws enacted by the State of Michigan:
1) The Great Lakes Submerged Lands Act (1955) which prohibits
constructing or dredging any artificial body of water that would
ultimately connect with a Great Lake, and which requires a permit
from MDNR to fill any submerged lands, including Lake St. Clair;
2) Shorelands Protection and Management Act (1970) which desig-
nates wetlands adjacent to a Great Lake as environmental areas
necessary to preserve fish and wildlife; and 3) The Goemaere-
Anderson Wetland Protection Act (1979) which regulates wetlands
through several laws relating to shorelands and submerged lands
(36).
-------
431
H. RECOMMENDATIONS
A. Industrial and Municipal Point Source Remedial Recommendations
1. Ontario and Michigan should incorporate the Great Lakes
Water Quality Agreement's goal of the virtual elimination
of all persistent toxic substances into their respective
regulatory programs.
2. The City of Mt. Clemens should determine the source of
PCBs, total phenols and mercury in the WWTP effluent and,
through pretreatment or in-plant controls, reduce the con-
centrations of these pollutants to acceptable levels.
Effluent limitations for these parameters should be con-
sidered. Phosphorus concentrations in the effluent should
be lowered to meet the 1 mg/L Great Lakes Water Quality
Agreement objective.
3. Site specific effluent limitations for total cadmium, total
copper, total chromium and total nickel to protect the
water quality for the Sydenham River and Lake St. Clair
should be developed for the Wallaceburg WWTP. The opera-
tion of the plant should be optimised to meet the Ontario
industrial effluent objective of 10 mg/L for ammonia.
4. The Warren WWTP should determine the source of PCBs in its
effluent and take the necessary steps to reduce the con-
centration to acceptable levels.
B. Nonpoint Source Remedial Recommendations
5. Agricultural areas with high rates of wind erosion need to
be targeted for assistance due to the characteristics of
wind transported soil (fine textured, high enrichment
ratio, and high organic matter content) and its ability to
transport nutrients and agrichemicals. The relatively low
erosion rates and high percentage of wind erosion in com-
bination make conservation tillage the most practical con-
servation practice to be recommended. The primary reasons
for this are the effectiveness of residue cover in reducing
wind erosion and the low cost of implementing the practice.
Conservation tillage is recognized as being highly cost-
effective and physically effective in areas of sandy soils
where wind erosion is a problem. If conservation tillage
were applied to all cropland eroding over the soil toler-
ance level, with a resulting compliance with the tolerance
level, a 32% reduction in phosphorus loading from cropland
could be achieved.
6. Rural landowners need to implement,' with the assistance of
Federal, State and Provincial governments, a comprehensive
-------
432
soil and water management system in order to control, at
source, the contribution of conventional and organic pol-
lutants including manure and pesticides to surface and
groundwater. Specifically:
a. Agricultural and conservation agencies need to accele-
rate the implementation of control technologies through
technical, financial and information/education
programs. There is a need for extension, education and
incentives to persuade farmers to implement conserva-
tion management systems including cropping, tillage and
structural practices, nutrient and pesticide management
technology,-thereby reducing the movement of soil,
conventional pollutants and contaminants off their land
into the waterways.
b. Environmental and agricultural agencies should assess
the adequacy of existing controls, regulations and
permits for the use of fertilizer and pesticide
products.
c. Specific programs, especially in Macomb County, MI,
should be directed at reducing the excessive levels of
phosphorus fertilization, improving the management of
animal waste disposal and storage, and educating pest-
icide users with respect to handling, application and
storage of pesticide products.
7. Future assessment and control of agricultural nonpoint
sources of pollution would be facilitated by compatible
Federal, State and Provincial monitoring data and more
frequent flow-weighted tributary monitoring data. The
small water quality monitoring data set available for tri-
butaries indicated the need for increased sampling for all
parameters, especially flow weighted data. The lack of
samples in high flows created difficulty in calculating
representative loads as well as understanding seasonal
patterns of pollutant transport. More samples on high flow
days would improve the basis for pollution control strat-
egies .
8. Macomb and St. Clair Counties, Michigan, should be targeted
for fertilizer management. U.S.EPA Region V has requested
the USDA-SCS Michigan State Office to develop standards and
specifications for a nutrient, best management practice
that would protect ground and surface waters as well as
sustain crop production. The Michigan Departments of Agri-
culture and Natural Resources are developing a joint action
plan to manage livestock waste problems that includes best
management practices for proper animal disposal that gives
attention to air and water pollution from concentrated
animal operations. This program may require a system of
-------
433
permits for concentrated feeding operations.
9. The CSOs from municipal wastewater treatment plants should
be intensively surveyed to determine their contribution of
pollutant loadings to the surface waters. In the long term
(due to enormous cost) combined sewers in all munici-
palities should be eliminated. In the interim, the munici-
palities should institute in-system controls to minimize
the frequency and volume of overflows.
10. The Michigan Pollution Emergency Alerting System and the
Ontario Spills Action Centre spills reports should be im-
proved so that all information on recovery, volume (if
known) and final resolution are fed back to the central
reporting system to complete each report for inventory
purposes.
11. The Superfund Site Investigations to be undertaken at
Selfridge ANGB should focus on groundwater and surface
water runoff impacts upon Lake St. Clair and the Clinton
River. In the event that this site is not included on the
U.S. National Priorities List, the 'State of Michigan should
place high priority upon cleanup on this site.
12. Michigan should require groundwater monitoring as a permit
condition for the Sugarbush solid waste landfill.
13. Michigan should include groundwater monitoring as part of
the RCRA Generators permit for G and L Industries.
C. Surveys, Research and Development
14. Data interpretation would be facilitated by the development
of more complete water quality objectives for the organic
pollutants and pesticides that are used extensively by the
agricultural industry. Currently, water quality objectives
do not exist for many parameters that are measured. Al-
though meeting water quality objectives does not guarantee
"no impact" of a contaminant, the objectives do provide a
point of reference for assessing the relative potential for
negative impacts of various contaminants in the aquatic
system.
15. The presence of organic contaminants (PCBs, HCBs and OCS)
in the Canadian tributaries illustrates the need to locate
the contaminant sources.
16. The cadmium content of the phosphate fertilizer that is
being used on agricultural lands should be determined.
-------
434
17. A study of atmospheric deposition of organic contaminants,
particularly PCBs, to Lake St. Clair and to the tributary
watersheds would provide quantitative information on load-
ing of these contaminants to the lake. The loading esti-
mates are important for mass balance calculations and the
identification of unknown sources of the contaminants.
18. Urban runoff was identified as being a potentially major
nonpoint source of many parameters,including PCBs, oil and
grease, zinc, mercury, copper and nickel. The loadings
from urban runoff, however, were based on contaminant con-
centrations from Canadian urban areas outside of the Lake
St. Clair basin. Therefore, the loading information
provide only a general potential for urban runoff to con-
tribute contaminants to Lake St. Clair. A study should be
performed to determine the contribution actually made by
urban runoff on the Michigan shore where the shoreline is
more urbanized than is that of Ontario.
19. The sediments near the mouth of the Clinton, Sydenham and
Thames Rivers contain contaminants that may be impairing
benthic communities. Studies are needed to document
possible impairment of benthic communities of these sites.
Appropriate actions to remedy any observed problems will
need to be defined. Techniques and technologies for remedi-
ating in-place polluted sediments should be developed.
20. Recognizing that the biological effects of a substance are
dependent in part on the chemical species of that sub-
stance, studies should be conducted' to identify the
chemical species and valances of the heavy metals in Lake
St. Clair and its tributaries. For those forms which are
present but for which toxicity information is lacking in
the literature, toxicity and bioaccumulation experiments
should be conducted on appropriate target organisms.
21. The evaluation of the point source data has been conducted
on a parameter by parameter basis. In order to assess the
quality of whole effluents, it is recommended that biomon-
itoring studies, both acute and chronic, be conducted at
the major facilities (Wallaceburg WWTP, Chatham WWTP,
Warren WWTP, and Mt. Clemens WWTP).
22. An inventory of all point sources, hazardous waste sites,
urban and rural runoff, and spills discharging or poten-
tially discharging to the Clinton River should be col-
lected. These facilities, sites or incidents should then
be examined for their potential to contribute chemicals to
the Clinton River.
23. A more complete analysis of sediment, water and biota
quality along the entire stretch of the Clinton River is
-------
435
needed. Such information would establish the locations of
sources of contaminants.
24. The Thames and the Sydenham Rivers were found to be major
contributors of phosphorus, ammonia, lead and cadmium. An
inventory of all point sources, hazardous waste sites,
urban and rural runoff and spills discharging to these
rivers should be collected. These facilities, sites or
incidences should then be examined for their potential to
contribute chemicals to the rivers.
-------
436
I. LONG TERM MONITORING
1. Purposes for Monitoring and Relationships Between UGLCCS and
Other Monitoring Programs
A presentation of the purposes for monitoring and surveillance
activities is included under Annex 11 of the 1978 GLWQA, and a
discussion of considerations for the design of a long term moni-
toring program can be found in Chapter 7 of the Report of the
Niagara River Toxics Committee (50); Because the focus of the
UGLCC Study was toward remedial actions to alleviate impaired
uses of the Connecting Channels System, long term monitoring
recommendations will likewise focus on the evaluation of trends
in environmental quality in order to assess the effectiveness of
remedial actions. In general, post-UGLCCS monitoring should be
sufficient to 1) detect trends in system-wide conditions noted by
the UGLCCS, and 2) detect changes in ambient conditions which
have resulted from specific remedial actions. Monitoring pro-
grams should be designed to specifically detect the changes in-
tended by the remedial actions so as to ensure relevance in both
temporal and spatial scales.
Two major programs sponsored by the IJC also contain plans for
long term monitoring: the Great Lakes International Surveillance
Plan (GLISP) and the Areas of Concern Remedial Action Plans (AoC-
RAPs). The GLISP for the Upper Great Lakes Connecting Channels
is presently incomplete, pending results of the UGLCC Study, but
it is expected to provide monitoring and surveillance guidance to
U.S. and Canadian agencies responsible for implementing the pro-
visions of the GLWQA that include general surveillance and
research needs as well as monitoring for results of remedial
actions.
Lake St. Clair is not one of the AoCs, although the Clinton River
in Michigan is, and a RAP is being developed by Michigan for the
Clinton River. The RAP will present details of uses impaired,
sources of contaminants, specific remedial actions, schedules for
implementation, resources committed by Michigan to the project,
target clean-up levels, and monitoring requirements. Results and
recommendations coming from the UGLCC Study will be incorporated
extensively into the RAP, which will then be the document that
influences Michigan programs in the Clinton River. The recommen-
dations for long term monitoring that are presented below are
intended for consideration and incorporation into either or both
the GLISP for the Upper Great Lakes Connecting Channels, and the
RAP for the Clinton River.
-------
437
2. System Monitoring for Contaminants
Water
Knowledge of the concentrations of the principal contaminants in
the water of Lake St. Clair should be used to indicate general
exposure levels for the biota, to identify changes and trends
over time in the concentration levels, and to be used for general
assessment of contaminant impacts. The parameters to be moni-
tored include phosphorus, PCBs, mercury, lead, and cadmium. Near
tributary mouths, concentrations of ammonia, total phenols, pest-
icides, Cu, Ni and PAHs should also be determined. Monitoring
stations should be located to coincide with identified water use
areas, such as biota habitat, and with contaminant entry points
to the lake. Suggested locations include the mouth of the St.
Clair River at Port Lambton, around the St. Clair Delta, at the
mouth of the Clinton, Sydenham, and Thames Rivers, and at the
head of the Detroit River. Sampling frequency should be
influenced by the variability in contaminant sources. Spring
high flow conditions and late summer low flow conditions would be
expected to bracket the normal seasonal variability in flow that
could influence measured contaminant concentrations.
A mass balance approach to contaminant monitoring will help to
identify any changes in the contaminant mass over time, and it
will provide the basis for targeting future remedial actions by
providing a comparison of the magnitude of the sources. A mass
balance analysis should be conducted approximately once every
five years, assuming that some effective remedial action has been
implemented against one or more sources such that the total load-
ings of contaminants, or the relative contribution of the sources
to the loading, has changed. The sources to be measured should
include:
1) Head and mouth transects. The number and location of
stations should relate to measured and predicted plume
distributions. Suggested locations include the mouth of
the St. Clair River at Port Lambton and the head of the
Detroit River. Dispersion modeling and past sampling
results should be used to predict contaminant concentra-
tions and therefore to establish appropriate collection and
analytical methodology. Both dissolved and particulate
fractions should be analyzed. The quantity of suspended
sediment flux should also be measured.
2) Municipal and industrial point sources. No direct in-
dustrial sources are considered to be major contributors of
contaminants to Lake St. Clair. The principal municipal
sources all discharge to tributaries. Thus, special moni-
toring consideration should be given to the Sydenham,
Thames and Clinton Rivers to fully address municipal load-
ings of the contaminants.
-------
438
3) Tributaries. Efforts should be focused on seasonal and
storm event loadings of contaminants to Lake St. Clair from
the Clinton, Sydenham and Thames Rivers. Tributary mouth
stations should be sampled and analyzed for both dissolved
and sediment-associated contaminant loadings.
4) CSOs and Urban Runoff. To provide an estimate of con-
taminant mass loadings expected during storm events, oc-
casional studies on selected urban drainage areas should be
conducted, particularly for the Michigan shoreline.
5) Groundwater inflow. The quantity and quality of potential
contaminant releases from waste dispos.al sites adjacent to
Lake St. Clair or its tributaries should be determined.
6) Sediment transport. Efforts to mea.sure and model sediment
transport to, within and from Lake St. Clair should be
continued. The quantity of contaminants being desorbed from
the sediments should be determined in order to assess load-
ings from these in-place polluted sediments.
7) Atmospheric deposition. Monitoring of wet and dry atmos-
pheric deposition to Lake St. Clair should continue, and
should be expanded to include organic contaminants. Vola-
tilization losses of organics should also be quantified.
Sediments
Monitoring of sediments for concentrations of contaminants should
be conducted periodically throughout Lake St. Clair in order to
assess both the trends in surficial contaminant concentrations
and the movement of sediment-associated contaminants within the
Lake. The grid used by the U.S. Fish and Wildlife Service during
the 1985 survey would be appropriate for consistency in sampling
sites and sediment composition. An analysis of sediment chem-
istry including bulk chemistry, organic and inorganic contamin-
ants, and particle size distribution should be conducted every
five years, in conjunction with a biota survey (see "habitat
monitoring" below).
In Lake St. Clair, particular attention should be given to sedi-
ment concentrations of PCBs and mercury. Additional stations
should also be established at the mouth of the Clinton, Sydenham
and Thames Rivers and at Chenal Ecarte to track effects of
remedial actions in the tributary watersheds to reduce loadings
of these materials.
Because the grid stations are distributed throughout the river
reach and are associated with appropriate habitat for a sensitive
benthic invertebrate (Hexagenia), the periodic survey will allow
assessment of 1) contaminant concentrations in the river sedi-
-------
439
merits throughout the river reach, 2) relative movement of the
contaminants within the river sediments between surveys, and 3)
correlation of contaminant concentrations with benthic biotic
communities.
The sediments at any stations established at the mouths of tribu-
taries to Lake St Clair should be monitored for organic and
inorganic contaminants on an annual or biannual basis when sig-
nificant remedial actions are implemented within the watershed of
the tributary. In order to trigger the more frequent sediment
monitoring program, the remedial actions should be expected to
measurably reduce loadings of one or more particular contaminants
via the tributary.
Biota
Long term monitoring of concentrations of contaminants in biota
will provide a time series useful to track the bioavailability of
contaminants to selected representative organisms. Three long
term monitoring programs are already in place and should be con-
tinued:
i) Annual or Bi-Annual Monitoring of Sport Fish.
This program should focus especially on PCBs, mercury and/or
other contaminants (e.g. dioxins and dibenzofurans) that are
considered to be known or suspected health hazards. -The monitor-
ing should be continued regardless of the differences that may be
observed between acceptable concentrations or action levels that
may be established by governmental agencies and the measured
contaminant concentrations in the fish flesh. As a link between
human health concerns and integrated results of remedial programs
to reduce contaminants in the UGLCCS system, this program is
critically important.
ii) Spottail Shiner Monitoring Program.
This program is designed to identify source areas for bioavail-
able contaminants. In locations where spottail shiners contain
elevated levels of contaminants, additional studies should be
conducted to identify the sources of the contaminants. Some
upstream studies in tributaries may be required. Spottails sho-
uld also be employed to confirm that remedial actions upstream to
a previous survey have been effective in removing or reducing the
loading of one or more contaminants.
iii) Caged Clams Contaminants Monitoring.
Caged clams should continue to be used at regular time intervals,
perhaps in conjunction with spottail shiners, to monitor inte-
grated results of remedial actions to reduce contaminant loadings
-------
440
to the water. Clams may be located at tributary mouths and down-
stream of suspected source areas. Repeated assays from the same
locations should confirm results of remedial actions.
3. Sources Monitoring for Results of Specific Remedial Actions
Remedial actions intended to reduce concentrations and/or load-
ings of contaminants from specific point sources generally re-
quire monitoring for compliance with the imposed criteria or
standards for permitted contaminants. The monitoring may be
conducted by the facility or by the regulating agency, whichever
is applicable, but attention must be given to the sampling
schedule and analytical methodology such that mass loadings of
the contaminants can be estimated, as well as concentrations in
the sampled medium. Monitoring of the "nearfield" environment,
i.e., close downstream in the effluent mixing zone, should be
conducted regularly to document reductions in contaminant levels
in the appropriate media and to document the recovery of impaired
ecosystem processes and biotic communities. Such monitoring may
be required for a "long time", but over a restricted areal
extent, depending on the severity of the impact and the degree of
reduction of contaminant loading that is achieved.
For Lake St. Clair, seven actions were recommended that would
affect specific sources of contaminants, and that would require
site-specific monitoring for compliance or other effects of the
action at the following locations: Macomb and St. Clair Counties,
Michigan (fertilizer management); Mt. Clemens WWTP (PCBs,
phenols, mercury, phosphorus); Wallaceburg WWTP (Cd, Cu, Cr, Ni,
ammonia); Warren WWTP (PCBs); Selfridge Air National Guard Base
(several contaminants) ; Sugarbush landfil-1, Michigan (groundwater
monitoring); and G and L Industries, Michigan (groundwater moni-
toring) .
Other recommendations for specific contaminant sources involve an
assessment of the present conditions or a study to quantify con-
centrations or loadings: quantify CSOs from municipal waste
water treatment plants, identify sources of organic contaminants
in tributaries; determine Cd content of phosphate fertilizer,
measure atmospheric deposition of organic contaminants; measure
loadings of contaminants from urban runoff; conduct biomonitoring
studies at WWTP's; inventory point sources and waste sites dis-
charging to the Clinton River; analyze sediment, water and biota
quality along the Clinton River; and inventory point sources and
waste sites discharging to the Sydenham and Thames Rivers. Each
of these items requires a specific program of data collection and
analysis. Additional needs for longer term monitoring may be
identified as a result of these studies.
-------
441
4. Habitat Monitoring
Habitat monitoring should be conducted to detect and describe
changes in the ecological characteristics of Lake St. Clair
through periodic analysis of key ecosystem elements. The follow-
ing items are recommended:
a) The abundance and distribution of the mayfly Hexagenia
should be determined every five years. The grid used by
the U.S. Fish and Wildlife Service during the 1985 survey
would be appropriate for consistency in sampling sites each
survey. An analysis of sediment chemistry, including bulk
chemistry, organic and inorganic contaminants, and par-
ticle-size distribution, should be conducted for samples
taken concurrently with the Hexagenia survey. These data
will provide information on the quality of the benthic
habitat for a common pollution sensitive organism that
would serve as an indicator species of environmental
quality.
b) Quantification of the extent of wetlands along Lake St.
Clair should be conducted every five years, in conjunction
with the Hexagenia survey. Aerial photography or other
remote sensing means would be appropriate to discern both
emergent and submergent macrophyte beds that are important
as nursery areas for larval fish and other wildlife. Veri-
fication of areal data should be conducted by inspection of
selected transects for plant species identification and-
abundances. Changes in wetland areas should be correlated
with fluctuating water levels and other natural documented
influences so that long term alterations in wetlands can be
tracked and causes identified.
-------
442
J. REFERENCES
1. OMOE (Ontario Ministry of the Environment) 1975. Great Lakes
Shore Damage Survey, Toronto, Ontario 97 pp.
2. Edwards, C.J., P.L. Hudson, W.G. Duffy, S.J. Nepszy, C.D.
McNabb, R.C. Hass, C.R. Liston, B. Manny and W.D. Busch.
1986. Hydrological, morphometrical, and biological
characteristics of the connecting rivers of the
International Great Lakes: a review. Contribution XXX.
National Fisheries Centre-Great Lakes. U.S. Fish and
Wildlife Service. Ann Arbor, Michigan
3. Quinn, F.H. 1976. Detroit River flow characteristics and
their application to loading estimates. J. Great Lakes Res.
2(1):71-77.
4. Poe, T.P., C.O. Hatcher, C.L. Brown and D.W. Schloesser.
1986. Comparison of species composition and richness of
fish assemblages in altered and unaltered littoral habitats.
J. Freshwater Ecol. 3(4): 525-536
5. Wall, G.J., E. A. Pringle and w.T. Dickinson. Agricultural
Pollution sources Lake St. Clair - Canada. UGLCC Study Non-
point Source Workgroup Level 2 report.
6. Chan, C.H., Y.L. Lau and E.G. Oliver. 1986. Measured and
modelled chlorinated contaminant distributions in St. Clair
River water. Water Poll. Res. J. Can. 21(3):332-343.
7. EC/MOE (Environment Canada/Ontario Ministry of the
Environment). 1986. St. Clair River Pollution Investigation
(Sarnia area). Canada/Ontario Agreement Report, January 28,
1986. Toronto, Ontario. 135 pp.
8. Johnson, G.D. and P.B. Kauss. 1987. Estimated Contaminant
Loadings in the St. Clair and Detroit Rivers - 1984. OMOE,
Great Lakes Section, Water Resources Branch, November 1987.
Toronto, Ontario.
9. Munawar, M. and I.F. Munawar. 1987. Phytoplankton of Lake
St. Clair, 1984. Great Lakes Laboratory for Fisheries and
Aquatic Science Report. Fisheries & Oceans Canada. Canada
Centre for Inland Waters. Burlington, Ontario.
10. Sprules, W.G. and M. Munawar. 1987. Plankton spectrum and
zooplankton of Lake St. Clair, 1984. Great Lakes Laboratory
for Fisheries and Aquatic Sciences Report. Fisheries and
Oceans Canada. Canada Centre for Inland Waters. Burlington,
Ontario.
-------
443
11. Schloesser, D.W. and B.A. Manny. 1982. Distribution and
relative abundance of submersed aquatic macrophytes in the
St. Glair-Detroit River ecosystem. U.S. Fish Wildl. Serv.,
Great Lakes Fish. Lab., USFWS-GLFL/AR-82-7. Ann Arbor, Mich.
49 pp.
12. Hudson, P.L., B.M. Davis, S.J. Nichols and C.M. Tomcko.
1986. Environmental studies of macrozoobenthos, aquatic
macrophytes, and juvenile fish in the St. Glair-Detroit
River system. U.S. Fish Wildl. Serv., Great Lakes Fish. Lab.
Admin. Rep. 86-7. 303pp.
13. Edwards, C.J., P.L. Hudson, W.G. Duffy, S.J. Nepszy, C.D.
McNabb, R.C. Hass, C.R. Listen, B. Manny and W-D Busch.
1988. Hydrological, morphometrical, and biological charac-
teristics of the connecting rivers of the International
Great Lakes: a review. Can J. Fish. Aquat. Sci. 44. (In
press).
14. Lyon, J.G. 1979. Remote sensing analyses of coastal wetland
characteristics: The St. Clair Flats, Michigan. Proc. 13th
Symp. Remote Sensing of Environment. Mich. Sea Grant Rep.
MICHU-56-80-313.
15. Manny, B.A., D.W. Schloesser, S.J. Nichols and T.A. Edsall.
1988. Drifting submersed macrophytes in the upper Great
Lakes Channels. U.S. Fish and Wildlife Service, National
Fisheries Centre-Great Lakes.
16. Griffiths, R.W. 1987. Environmental quality assessment of
Lake St. Clair in 1983 as reflected by the distribution of
benthic invertebrate communities. Aquatic Ecostudies, Ltd.
Kitchener, Ontario 35 pp.
17. GLI (Great Lakes Institute). 1986. A case study of selected
toxic contaminants in the Essex Region. GLI, Univ. of Winds-
or. Vol. 1: Physical Sciences. Parts One and Two, July,
1986. Windsor, Ontario.
18. Goodyear, C.D., T.A. Edsall, D.M.O. Demsey, G.D. Moss and
P.E. Polanski. 1982. Atlas of spawning and nursery areas of
Great Lakes fishes. U.S. Fish Wildl. Serv. Ann Arbor, MI
FWS/OBS-82/52, 164 pp.
19. McCullough G.B. 1985. Wetland threats and losses in Lake
St. Clair. pages 201-208 in H.P Prince and P.M. D'ltri,
eds. Coastal Wetlands, Lewis Publishing Co., Chalsea,
Michigan.
20. McCullough, G.B. 1982. Wetland losses in Lake St. Clair and
Lake Ontario, pages 81-89 in A. Champagen, ed., Proc.
Ontario Wetlands Conf., Ryerson Polytech. Institute.
-------
444
Toronto, Ontario September 1981.
21. Rukavina, N.A., 1987. Status report of UGLCCS, Lake St.
Clair Bottom Sediment data. Level I, Report to the IJC.
22. International Joint Commission. 1982. Guidelines and Regis-
ter for Evaluation of Great Lakes Dredging Projects. Report
of the Dredging Subcommittee to the Water Quality Programs
Committee of the Great Lakes Water Quality Board. 365pp.
23. Oliver, E.G. and R.A. Bourbonniere. 1985. Chlorinated con-
taminants in surficial sediments of Lakes Huron, St. Clair
and Erie: implications regarding sources along the St. Clair
and Detroit Rivers. J. Great lakes Res. 11:366-372.
24. OMOE, Unpublished.
25. Sediment Workgroup Report, 1987 Geographical area report,
Lake St. Clair. UGLCCS Level II Report.
26. Hamblin, P.P., P.M. Boyce, P. Chiocchio and D. S. Robertson,
1987. Physical measurements in Lake St. Clair: Overview and
preliminary analysis. National Water Research Institute
Contribution 87-76
27. Robins, J.A. and E.G., Oliver, 1987. Accumulation of fall-
out cesium-136 and chlorinated organic contaminants in
recent sediments of Lake St. Clair. In Modeling Workgroup
Report (53).
28. MDNR (Michigan Department of Natural Resources). 1985. Non-
point Assessment for Small Watersheds. Staff report, Surface
Water Quality Division, Lansing, Michigan.
29. Leuck, D. and B. Leuck. 1987. survey of Great Lakes Bathing
Beaches 1986. OMB No. 2090-003. U.S.EPA, Great Lakes
National Program Office, Chicago.
30. Baker, David B. 1987. Pesticide Loading into the St. Clair
River and Lake St. Clair in 1985. Final Report. U.S.E.P.A.
Grant R005817-01. Great Lakes National Program Office,
Chicago.
31. Wall, G.J., E.A. Pringle and T. Dickinson. 1987.
Agricultural Sources of Pollution, Lake St. Clair. Executive
Summary of the Nonpoint Source Workgroup, Level 2 reports.
32. Lundgren, R.N., editor. 1986. Fish contaminant monitoring
in Michigan. Report of EPA 205j Grant. Michigan Dept. of
Natural Resources. Lansing, Michigan.
-------
445
33. OMOE/OMNR (Ontario Ministry of the Environment/Ontario Mini-
stry of Natural Resources). 1987. Guide to eating Ontario
sport fish. Ministry of the Environment, Ministry of Natu-
ral Resources, Toronto.
34. GLI (Great Lakes Institute). 1987. Organochlorinated com-
pounds in duck and muskrat populations of Walpole Island.
University of Windsor, Ontario.
35. Amundson, T.E. (UNDATED). Environmental Contaminant Monitor-
ing of Wisconsin Wild Game 1985-86. Bureau of Wildlife
Management, Wisconsin Department of Natural Resources,
Madison, Wisconsin.
36. Herdendorf, C.E., C.N. Raphael and E. Jaworski. 1986. The
Ecology of Lake St. Clair Wetlands: A Community Profile.
U.S. Fish Wildlife Service. Biol. Report. 1985 (7.7). 187
pp.
37. Point Source Workgroup. 1988. Geographic Area Report - Lake
St. Clair. UGLCCS Level 2 report.
38. Pugsley, C.W., P.D.N. Herbert, G.W. Wood, G. Brotea and T.W.
Obal. 1985. Distribution of contaminants in clams and sedi-
ments from the Huron-Erie corridor. I. PCBs and octachloro-
styrene. J. Great Lakes Res. 11(3):275-289.
39. MDNR (Michigan Department of Natural Resources). Undated.
Progress Summary-Activity E.8. Draft UGLCC Study report,
Nonpoint Source Workgroup Level 2 Report for Lake St. Clair.
40. Oliver, E.G. and C.W. Pugsley. 1986. Chlorinated Contamin-
ants in St. Clair River sediments. Water Poll. Res. J. Can.
21:368-379.
41. Richards, R.P. and J. Holloway. 1987. Monte Carlo studies of
sampling strategies for estimating tributary loads. Water
Resources Res. 23 (10):1939-1948.
42. Richards, R.P. 1988. Measures of flow variability and a new
classification of Great Lakes tributaries. Report, U.S.EPA
Great Lakes National Program Office, Chicago 40 pp.
43. Dolan, D., A. Yui and R. Geist. 1981. Evaluation of river
load estimation methods for total phosphorus. J. Great Lakes
Res. 7(3):207-214.
44. Leach, J.H. 1972. Distribution of chlorophyll a and related
variables in Ontario waters of Lake St. Clair. pp 80-86.
In Proc. 15th Conf. Great Lakes Res., Inst. Assoc. Great
Lakes Res.
-------
446
45. Leach, J.H. 1980. Limnological sampling intensity in Lake
St. Clair in relation to distribution of water masses. J.
Great Lakes Res. Vol 6 141-145.
46. Bricker, K.S., Bricker F.J., and J.E. Gannan, 1976. Dis-
tribution and abundance of zooplankton in U.S. waters of
Lake St. Clair, 1973. J. Great Lakes Res 2:256-271.
47. Hamblin, P.P., P.M., Boyce, J. Bull, F. Chiocchio and D.S.,
Robertson, 1987. Reports to UGLCCS Workgroups. National
water Research Institute Contribution 87-87.
48. Hammons, A.S., J.E. Huff, H.M. Braunstein, J.S. Drury, C.R.
Shriner, E.B. Lewis, B.L. Whitfield and L.E. Towill. 1978.
Reviews of the Environmental Effects of Pollutants: IV.
Cadmium. EPA Publication No. EPA-600/1-78-026. Office of
Research and Development, Cincinnati, Ohio.
49. GLWQB (Great Lakes Water Quality Board). 1987. 1987 Report
to Great Lakes Water Quality Board, Appendix A, Progress in
Developing Remedial Action Plans for Areas of Concern in the
Great Lakes Basin. Report to the International Joint Com-
mission, Windsor, Ontario.
50. Niagara River Toxics Committee, 1984. Report on the
Niagara River Toxics Committee to U.S. U.S.EPA, Environment
Canada, OMOE and N.Y. DEC.
51. Marsalek, J. and H.Y.F. Ng. 1987. Contaminants in Urban
Runoff in the Upper Great Lakes Connecting Channels Area.
NWRI contribution No. 87-112. National Water Research In-
stitute, Burlington, Ontario.
52. Marsalek, J. and H.Q. Schroeter. 1984. Loadings of selected
toxic substances in urban runoff in the Canadian Great lakes
Basin. NWRI Unpublished Report. National Water Research
Institute, Burlington, Ontario.
53. Modeling Workgroup, UGLCCS. 1988. Geographical area
synthesis report. Draft May 1988, T.D. Fontaine (Chairman),
NOAA-Great Lades Env. Res. Lab. Ann Arbor, MI. 96p.
-------
CHAPTER IX
THE DETROIT RIVER
A. STATUS OF THE ECOSYSTEM
1. Ecological Profile
Watershed Characteristics
The Detroit River makes up the lower 51 km of the connecting
channels between Lakes Huron and Erie. An international boundary
divides the Detroit River about equally into United States
(Michigan) and Canadian (Ontario) waters (Figures II-5 and IX-1).
The Detroit River is a hydrologically and ecologically distinct
ecosystem compared to Lake St. Clair and the St. Clair River (-1) .
It is limnologically mesotrophic and supports cold water fish
from September to June. The Detroit River provides important
habitat for fish, birds and the bottom dwelling life on which
they feed. It is also an important source of potable water, with
drinking water intakes near Belle Isle, Windsor, Amherstburg and
Wyandotte (2). Water is also used to supply a major industrial
complex consisting of automobile, steel and chemical companies.
The St. Lawrence Seaway utilizes the Detroit River for commercial
shipping. This portion of the Seaway is presently the busiest in
the upper Great Lakes, involving shipments of iron ore, coal,
limestone, gypsum, oil, and wheat.
The topography of the Detroit River basin is flat, broken only by
the valleys of the Rouge River and a few lesser tributaries. Low
moraine deposits and beach ridges of ancestral Lake Erie provide
slight relief. Land elevations range from 214 m above sea level
near the tributary head waters to approximately 174 m along the
Detroit River. The relative relief of the lake plain is 1 to 5
m/km3, and most slopes are less than 3%.
The Detroit River courses through Pleistocene glacial drift
-------
448
. - LAKE
/ HURON
Port
Huron
Marysville
Marine
.-' City
MICHIGAN
UPPER
DETROIT
RIVERA*
.Windsor .':.'••'.•"• • ".
ONTARIO
LOWER
DETROI
FIGURE IX-1. The Huron-Erie corridor.
-------
449
underlain by Paleozoic sedimentary rock. The sedimentary rock
beneath the river is the Detroit River Formation (primarily dolo-
mite) which outcrops intermittently in the navigation channels
east of Grosse lie. On top of the bedrock is a mantle of glacial
drift 0 to 30 m thick.
Lake plain soils are poorly drained loam and clay loams, which
developed on former lake bottoms or lacustrine clay sediments.
Sandy ridges mark former shorelines, and on the Michigan side, an
isolated sand sheet marks remnants of the glaciofluvial delta of
the post-glacial Huron River. When drained and tiled, the loamy
lake plain soils are agriculturally productive. Many surface and
subsurface soils are moderately permeable (0.25 and 1.27 cm/hour)
with high surface runoff coefficients causing the local streams
to be storm event responsive.
The Ontario shoreline, except for the City of Windsor and its
docks, is less disturbed than the Michigan shoreline. North of
the Canard River there are scattered marinas, canals, and private
boat slips. In places, Ontario farmers have encroached upon the
wetland margins of the Detroit River and its tributaries. Thus,
a green buffer zone exists only intermittently between the farm
fields and the riverine ecosystem. Access to the water for com-
mercial navigation, business, pleasure boating, fishing and hunt-
ing is important locally on both sides of the river.
Hydrology
Nearly 98% of the Detroit River flow enters from Lake Huron via
the St. Clair River and Lake St. Clair. The river discharge
averages 5,300 m-Vsec and ranges from a low of 3,200 m-Vsec to a
maximum discharge of 7,100 m-Vsec. The Fleming Channel in the
upper Detroit River, north of Peach Island, accounts for 77% of
total river flow. Flow distribution in the lower river is rela-
tively complex downstream of Fighting Island, as several channels
separate or combine the flow (2,3,4).
Flow velocities average 0.49-0.88 m/sec, but mid-surface veloci-
ties can be nearly twice that rate. Surface currents near the
Ambassador Bridge and in the Amherstburg Channel reach 1.2 m/sec,
while the Trenton Channel flow averages 0.6 m/sec.
Detroit River water depth and velocity are directly affected by
water levels in Lakes St. Clair and Erie, which vary seasonally
and annually. Lake Erie seiches and Lake. St. Clair ice jams may
also produce changes in Detroit River water levels and currents.
The river slope is relatively uniform, and falls 0.9m over its
51 km length. The average time of passage for water through the
Detroit River is about 19 to 21 hours.
-------
450
The Rouge River, the main tributary to the Detroit River, drains
about 121,000 ha in Michigan, and consists of upper, main, middle
and lower branches. The stream is very event-responsive and
frequent flooding occurs along the middle Rouge. Its mean annual
discharge is 26 m3/sec, with over 75 percent of it draining
through urban areas, collecting considerable stormwater runoff,
overflow from combined sewers during wet weather, and over 500
million gallons per day (mgd) of waste water from municipal and
industrial facilities. The lower Rouge is partially lined with
concrete, so runoff rapidly reaches the Detroit River during
storms.
Other tributaries include the Ecorse, Canard and Little rivers
and Turkey Creek. The Ecorse River tributary drains 11,556 ha in
Michigan, occupied by 2 communities with a total population of
198,000 in 1980. The Ecorse River has two open channel tributa-
ries, the North Branch and the South Branch (or Sexton-Kilfoil
Drain). These branches join approximately 1 km upstream from the
confluence of the Ecorse and Detroit rivers near Mud Island.
Ontario's Little River empties into the Detroit River at its
mouth, by Peach Island. It drains approximately 5,750 ha of
agricultural and industrial land. Turkey Creek enters the
Detroit River just north of Fighting Island, draining 2,960 ha of
primarily agricultural land in Ontario. The Canard River enters
the Detroit River in Ontario, south of Windsor and east of Grosse
lie. It is a turbid, slow moving stream which discharges into
diked wetlands just north of its mouth, and drains approximately
20,000 ha of primarily agricultural land (5). Other minor tribu-
taries also exist, such as Monguagon Creek (in Michigan, by the
northern end of Grosse lie) and Conners Creek (in Michigan, by
the eastern end of Belle Isle).
Effluent from the Detroit area wastewater treatment plants
(WWTPs) discharge over 32 m3/sec (1985), a volume equal to the
combined tributaries flowing into the Detroit River. The Metro-
politan Detroit WWTP alone discharges 30 m^/sec near the mouth of
the Rouge River (6).
Habitats and Biological Communities
The Detroit River ecosystem can be divided into an upper stretch
(upstream of the Rouge River) and a lower river stretch. The
Detroit River's biologic zones include deep channels, shallow
water/nearshore zones, and terrestrial zones. Deep channel
environments generally have water depths exceeding 7 m, relative-
ly high flow velocities, and coarse sediments. Since the river
channels are also used for shipping, the high sediment load and
lack of anchorage prevent macrophyte growth. Macrophytes and
associated periphyton and invertebrates are most abundant in the
shallow water-nearshore zone, seldom occurring at depths greater
than 4 m. The terrestrial biological zone includes undeveloped
-------
451
island habitat, coastal wetland and riparian environments along
such less developed tributaries as the Canard River. The
Wyandotte National Wildlife Refuge is located in the Detroit
River, off the northern tip of Grosse lie. This Refuge encour-
ages shorebirds and waterfowl feeding, nursery and nesting activ-
ities. Stony, Celeron, Grassy and Mud Islands provide shorebird
habitat.
The coastal wetlands and large, emergent and submersed macrophyte
beds along the Detroit River were nearly continuous in colonial
times. They now exist only in 31 small isolated remnants cover-
ing 1,382 ha (7). Most of the remaining vegetation along the
river consists of submersed macrophytes because the land formerly
occupied by the swamp-scrub-meadow communities along the ter-
restrial river margin has largely been converted to other uses.
Fifty-four percent (748 ha) of the remaining wetlands are in
Ontario. The single largest wetland, immediately north of the
Canard River, is functional only along its outer, undiked mar-
gins. Functional wetlands also exist along the open water mar-
gins of a few islands.
A number of biological surveys have documented the biotic com-
munities in the river (7,8,9,10,11,12,13,14,15,16). Although it
is not well understood how the various trophic levels relate to
one another, enough information exists to describe species com-
position, standing crop and biomass for a variety of primary and
secondary producers.
i) Macrophytes
At least 21 submersed macrophyte taxa occur in the river, domin-
ated by Vallisnera, Chara, Potamogeton, Myriophyllum and Heteran-
thia. Stands are typically composed of 2 or 3 species but as
many as eleven have been recorded in a single stand. Chara is
the only taxon consistently occurring in monotypic stands. The
lower depth limit for plant colonization is not established, but
most stands occur in water less than 3.7m deep. In the Detroit
River, the area of the river bed between shoreline and the 3.7m
depth contour is about 99 km2, 72% of which is occupied by sub-
mersed plants. The wetlands and submersed macrophyte beds con-
stitute the most critical areas for primary and secondary produc-
tion for plants, fish and birds, and are the most stable habitat
in the ecosystem (17). Their invertebrate populations include
clams, snails, midges, caddisflies, mayflies, amphipods, spring-
tails, and worms. Juvenile yellow perch and adult northern pike
have been- observed feeding along the wetland shoreline among the
submersed macrophytes. These areas are also heavily used for
spawning by numerous fish species. No detailed studies of spec-
ies composition, distribution, and relative abundance of emergent
macrophytes have been completed, although wetland communities
have been mapped by remote sensing. Over 95% of the emergent
beds occur in the lower river.
-------
452
The St. Glair-Detroit River system produces about 264,000 tons of
plant biomass each year, of which 19% originates in the Detroit
River. Most of the plant biomass in the Detroit River is pro-
duced by submersed macrophytes.
ii) Phytoplankton
Phytoplankton standing crop and production values is assumed to
have phytoplankton biomass and daily production similar to Lake
St. Clair. Eighty two phytoplankton species are present in the
river at low density (about 500 cells/ml), and are dominated by
diatoms that are common in Lake Huron in July and August. Blue-
green algae that are common in Lake St. Clair at that time domin-
ate the Detroit River phytoplankton. No periphyton studies have
been conducted to date, but a recent study in a wave exposed
breakwater in western Lake Erie indicates that diatoms, green
algae and red algae may be common over-wintering taxa in the
Detroit River. Filamentous green algae can be expected to domin-
ate during summer months.
Current information is inadequate to determine how much of the
planktonic production of the river is used by river biota. If
only moderate amounts of this biomass is retained, then the
littoral plant complex of emergent and submersed macrophytes and
macrozoobenthos are the main standing stock in the river. From
calculations of drifting macrophytic plants, it appears that the
Detroit River is a large source of detrital organic matter that
•supports productivity in western Lake Erie.
iii) Zooplankton
Detroit River zooplankton studies are not yet completed, but
zooplankton composition and abundance seem to resemble those
found in Lake St. Clair. Cladocera and several species of Cyc-
lops and Diaptomus dominate the zooplankton in Lake St. Clair.
Difflugia is the most common protozoan, and Conochilus, Keratel-
la, Polvarthra. Synchaeta, and Brachionus are the most common
rotifers. Maximum numbers of zooplankton may be expected between
June and September. A study of foods eaten by larval yellow
perch during passage through the Detroit River revealed that
zooplankton, including copepod nauplii, older cyclopoids and
copepods, cladocera and rotifers were eaten. Hence, zooplankton
are likely the critical food resource for larval fish.
iv) Macroinvertebrates
The Detroit River benthic macroinvertebrate community includes
over 300 species. Oligochaetes, chironomidae, gastropoda, ephem-
eroptera, trichoptera and amphipoda dominate the biomass. Chir-
onomidae are common throughout the system while oligochaetes are
dominant in the lower river. Hydropsychid caddisflies are the
dominant trichoptera and Hyalella is the most common amphipoda.
-------
453
Hexagenia is the most common mayfly, but density is lower in the
Detroit River (88/m2) than the St. Clair or the St. Marys Rivers
(95/m2 and 199/m2), respectively. Detroit River benthic produc-
tion (5.4 g ash-free dry weight/m^/yr) is lower than the St.
Clair River and Lake St. Clair (7.0 and 6.8 g ash-free dry
weight/m3/yr) with the annual production (440 metric tons ash-
free dry weight/yr) equal to about 2% of the combined annual
Detroit River phytoplankton, periphyton, macrophyte and zoo-
plankton production (7,14,16).
v) Fish
The present Detroit River fish populations are a mixture of
natural and introduced (exotic) species. Among the exotic fish
is the common carp, which was introduced in 1883 in western Lake
Erie. From there, it spread through the Detroit River to the
upper Great Lakes, destroying beds of wild celery and wild rice,
the preferred food of native waterfowl. Large carp populations
continue to inhabit the Detroit River. Rainbow smelt and ale-
wife, introduced in 1932, spread through the Detroit River and
upper lakes. Alewives now comprise the bulk of forage fish in
all the Great Lakes. The sea lamprey spread through the Detroit
River to the upper Great Lakes in the 1940s, greatly reducing
populations of desirable fish, such as the lake trout. The most
recent exotic Detroit River fish, the white perch, was introduced
into Lake Erie in 1953 and now hybridizes with native white bass.
The Detroit River fish community presently has approximately 60
resident or migrant species, 32 of which use mainly the lower
river along the islands and the mainland shoreline for spawning
(18,19,20,21,22).
The Detroit River and its tributaries are. important spawning,
feeding and nursery areas for many species that support major
fisheries in the river and Lakes Huron and Erie. There are 60
recorded resident or migrant fish species in the Detroit River,
32 of which spawn in the river. Townet catches of larval fish in
the Detroit River in 1977-1978, 1983-1984 and 1986 show that the
river is a nursery ground for at least 25 species of fish. Most
abundant were alewife, rainbow smelt, and gizzard shad. Other
species were much less abundant.
The river is part of a complex migration route for walleye and
yellow perch, important recreational fish species, which move
between Lake St. Clair and Lake Erie. Large walleye spawning
runs once occurred in the lower river, the reduction of which is
attributed to pollution and sedimentation. In the 1970s, spawn-
ing was again documented, and walleye larvae were collected in
several locations in the lower 16 km of the Trenton Channel and
the main river. Recently, yellow perch spawning has been ob-
served in the Trenton Channel and near the mouth of the Detroit
River in some areas previously used by walleye.
-------
454
The Detroit river once supported a large commercial fishery for
lake whitefish, lake herring, walleye, lake sturgeon, black bass,
northern pike, muskellunge and carp. Overfishing, pollution and
dredging contributed to the Detroit River commercial fishery
decline (23,24,25).
Sport fishing is still an important activity in the Detroit
River. In 1985, an estimated 1.4 million hours were spent har-
vesting approximately 1.4 million fish (22). The lower river
harvest was 980,200 while the upper river was 440,600 annually.
Dominant species were white bass (63%), walleye (12%), yellow
perch (10%), and freshwater drum (7%).
A larval fish passage study from Lake St. Clair to Lake Erie was
conducted along the Detroit River at 17 transects, 2.5 km apart
(Figure IX-2)(22). Thirteen larval fish taxa were observed.
Larval fish densities of walleye, yellow perch and white bass/
white perch greatly increased in the mid-Trenton Channel (tran-
sect 12-13), suggesting spawning and rearing activities in the
vicinity. Yellow perch showed a strong lateral distribution with
greatest densities along the western near-shore, decreasing
toward the main channel with lowest densities along the eastern
shore. Surprisingly, the area containing the highest density of
larval yellow perch coincides with the highest concentration of
environmental contaminants in water or sediments. White
bass/white perch and rainbow smelt did not exhibit significant
east-west density gradations. Longitudinal distribution patterns
were evident for larval bloaters, burbot and deep water sculpin.
Deep water densities of these species were greatest in the upper
Detroit River, but were present throughout, probably being trans-
ported from Lake Huron and Lake St. Clair. Walleye and white
bass/perch were not found, and yellow perch and rainbow smelt
exhibited relatively low abundances in the upper river. Yellow
perch, white bass/white perch, rainbow smelt and walleye larval
densities were greatest in the lower river.
vi) Waterfowl
At least 3 million waterfowl migrate annually through the Great
Lakes region, which is situated at the intersection of the Atlan-
tic and Mississippi flyways. An estimated 700,000 diving ducks,
500,000 dabbling ducks, and 250,000 Canadian geese migrate across
Michigan each fall (1) .
Important species of nesting ducks in the Detroit River wetlands
include mallards, blue-winged teal, black ducks and, if nesting
boxes are provided, wood ducks. In the past, 24 species of ducks
regularly fed in the river. Each year, thousands of waterfowl,
including scaup, goldeneyes, canvasbacks, black ducks, redheads,
and mergansers congregate on the river to forage sediments.
Major concentrations of feeding ducks are often found in littoral
waters around Belle Isle, Grosse lie and Mud, Fighting, Sugar and
-------
455
MICHIGAN
FIGURE IX-2. Detroit River water sampling transects and 24-hour
sampling locations.
-------
456
Celeron islands. Preferred foods vary among species. Mergansers
feed primarily on fish, whereas American goldeneyes prefer crayf-
ish, clams, and other invertebrates. Many diving ducks feed on
submersed aquatic plants and their associated communities.
A recent survey of eelgrass tubers, a preferred food of many
waterfowl, indicated that over the past 35 years, tuber densities
have decreased substantially, resulting in a net loss of 4.6 x
109 tubers in the lower river. This large loss of eelgrass
tubers in the Detroit River explains in part why fewer waterfowl
now use the Michigan migration corridor.
Climate
The Detroit River area enjoys a mid-continental climate, with
cold winters and relatively short hot summers, moderated somewhat
by the Great Lakes. The average first frost is on October 21 and
the average last freezing temperature is on April 23, with an
annual growing season of 180 days. Precipitation averages about
76 cm per year, including 40 cm of snow. Prevailing winds are
from the southwest, and average 16 km/hour.
During late autumn and early winter, water from Lake Huron cools
rapidly as it flows through shallow Lake St. Clair. As a result,
ice often enters the Detroit River from Lake St. Clair before it
begins to form in the Detroit River itself. Before the 1930s,
most of the Detroit River was ice covered in winter, but now
large volumes of heated effluents entering the river usually
prevent the upper river from freezing over, except between Belle
Isle and the Michigan mainland. Extensive slush ice still devel-
ops in the lower river, especially in the broad shallow expanses
adjacent to the islands. In general, ice may now be found in the
river from early December to mid-March, but main navigation chan-
nels remain ice-free. Minor ice jams occur in the Detroit River
with the breakup ice moving south from Lakes Huron and St. Clair
from late March to early May. Easterly winds can also cause Lake
Erie ice to reverse into the lower Detroit River. Monthly water
temperature data show that the highest water temperatures occur
in August, with an average of 22.2°C. In the shallow nearshore
areas of the lower river, water temperatures may attain 25.2°C.
Lowest temperatures occur in January-February, sometimes reaching
0°C.
2. Environmental Conditions
Water Quality
The Detroit River area is heavily industrialized and densely
populated. Industrial and municipal raw water is taken from the
river then returned after use. Due to its varying channel width
-------
457
and depth, berms and islands, the Detroit River is hydrologically
complex, a fact which influences water quality and modifies the
human impact on the Detroit River system.
Information on water quality was obtained as part of this study
(26). To obtain a reliable data set which could provide a mean-
ingful interpretation while minimizing the need for analyses,
water sampling transects across the river were used. Figure IX-3
shows the location of the upper (DT 30.8W and DT 30.7E) and lower
(DT 8.7W and DT 9.3E) transects and the major tributaries. The
upper transects are at Peach Island near Lake St. Clair, upstream
of Detroit and Windsor. The lower transects are near Grosse lie,
upstream of the Livingston Channel and Stoney Island in the east,
and near the lower end of the Trenton Channel on the west. The
lower transect was designed to avoid the influence of Lake Erie,
and in the process was located upstream of two industrial facili-
ties, General Chemical at Amherstburg and McLouth Steel, Gibral-
tar. Therefore, water quality data for the lower transect does
not reflect these facilities. In addition, loadings from Frank
and Poet Drain, which serves several permitted Michigan industri-
al discharges, were also excluded (26). Figure IX-4 describes
the flow distribution in the channels of the Detroit River, and
shows that approximately 21% of the total Detroit River flow
passes through the Trenton Channel and approximately 26% and 47%
through the Livingston and Amherstburg channels, respectively
(27).
Three additional, partial river width water quality monitoring
transects were established in the Trenton Channel between Grosse
lie and the Michigan shore at Point Hennepin (A), just south and
parallel to the Grosse lie toll bridge (C), just south and paral-
lel to the Grosse lie Parkway Bridge off the Monsanto Breakwall
(D). Michigan's monthly Detroit River water sampling transect at
the mouth of Detroit River between Bar Point and Maple Beach (DT
3.9) is also shown (Figure IX-3).
i) Cross-Channel Variations in Water Quality
Cross-channel variation of water quality occurs where large
volumes of low concentrations or smaller volumes of higher con-
centrations of substances are discharged to the river. Cross-
channel variations were demonstrated by dye studies below the
Detroit WWTP outfall (Figure IX-5) (28). The upper Detroit River
between Belle Isle and Fighting Island has a relatively constant
channel width and depth where little or no cross-channel mixing
occurs. In contrast, the lower river section is broken up into
three major channels and several shallow embayments. There, and
downstream of these islands and structures, increased cross-chan-
nel mixing may occur due to the generally lower current veloci-
ties, eddies below these structures, and wind driven currents
cross and counter to the normal current direction.
-------
458
DT30.8W
Michigan
USA
Rouge
River
Ecorse
River
DT12.0W,
Monguagon
Creek
DT8.7W
Little
River
Ontario
CANADA
M
•DT3.9
FIGURE IX-3. Detroit River mass balance sampling transects.
-------
459
77%
DETROIT
Ambassador Bridge
51%
26%
Navigation
—' Channel
36%
FIGURE IX-4. Flow distribution in the Detroit River (27).
-------
460
00
cs
C3
&
<—
o
>1
0
o>
-C
E
o
I
X
w
Qi
D
O
-------
461
Cross-channel variation in concentrations of some organochlorine
contaminants (for example PCBs and chlorobenzenes) between water
in the upper Detroit River and the Detroit River mouth has been
shown (Figure IX-6). Organochlorine concentrations are similar
at the head of the river along both the Michigan and Ontario
shores (about 0.5 ng/L at stations 399 and 379, respectively)
(29). Proceeding downstream, higher levels are found along the
Michigan shore, with levels up to 209 ng/L (station 346), com-
pared with 0.5 ng/L across the river. Station 269 (17 ng/L), on
the Canadian side, may be influenced by U.S. sources as this
station is well within the 50% flow panel of the Detroit River.
ii) Longitudinal Variations in Water Quality
The flow of the Detroit River ranges from 3,200 m3/sec to 7,100
m3/sec, constituting a large water mass. To detect statistically
significant changes in water quality between the river head and
mouth, inputs or sinks of such substances must be substantial.
Due to natural fluctuations between seasons, shipping and dredg-
ing activities and both natural and man-induced fluctuations of
in-coming water quality, any quantitative and even qualitative
interpretation of data is difficult. Only a statistical evalua-
tion of many samples will allow definite conclusions. That sam-
pling intensity was not achieved in this study for most data, and
comparisons made are primarily relative comparisons. Evaluation
of relative changes in water quality parameters does not require
absolute values, but compares the relative abundance or absence
of materials, and may indicate temporal or spatial differences.
Polychlorinated Biphenyls (PCBs):
Qualitatively, the composition of PCBs in Detroit River water
changes from the upper to the lower Detroit River transects (Fig-
ure IX-7). For nine commonly observed PCB homolog series (com-
prising approximately 100 of the theoretically possible 210 PCB
isomers), a decrease of the lower chlorinated homologs (with one
to four chlorines per biphenyl molecule) and an increase of the
higher chlorinated homologs (6 to 10 chlorines per molecule) is
observed as one moves downstream. Considering the stability of
PCBs, it can be concluded that the observed change in PCB homolog
distribution is due to inputs of higher chlorinated PCBs along
the river stretch (26).
The observed qualitative changes in PCB composition are also
supported by quantitative observations. PCB concentrations in
water averaged approximately 0.6 ng/L at four stations above and
below Belle Isle on both sides of the river from a 1985 survey
(26). Downstream, at several locations along the Ontario side,
PCB concentrations increased to approximately 1.0 ng/L, while PCB
concentrations on the Michigan side in and downstream of the
Trenton Channel increased to levels as high as 3.4 ng/L. In the
-------
462
STATION NUMBER
21} 214 231 223 224 255 311 314 330 346 3*2 384 3M
o-
5-
KJ-
IS-
20-
2S-
1*| I
H ~!
(2 :
[
^
\
:
:
75|
11
•
:
; :
j :
i ;
i \
; i
j E
r*"
ta
1\
:
:
i
0
11
H
K
ff
•
S WATER
53 SUSPENDED SOLOS
= • SURFCIAL SEDMENT
1 • HO SAMPLE
_ i-i ND NOT DETECTED
1 @30 109| |38
™ i^ I CUB i e •* B*
210 212
269 28O
STATION NUMBER
353 370 379
OrfuiocMorine contaminants (OCst in tnair, aupmdtd solidt, aid nir/Idof Jtdimaai of tht Detroit Rivtr.
CoHOiarmiom in 10* nf kg ' (itdimmtt. aaptndtd solids) and Iff' ng L1 (water), mptctitely.
STATION NUMBER
213 214 231 223 224 25S 311 314 330 346 ' 352 384 399
I'.Jo
WATER
CS SUSPENDED SOLIDS
& SURFICIAL SEDIMENT
• NO SAMPLE
ND NOT DETECTED
"^
"^J^'~'N-3£5
AMHERSTBUDG ^V
zs-
20-
^
tj
K 10-
Z
O 5-
I
"V, =
»• -a^- ^^^
\SSvWWO«
1
[i
•
:••
3-1
3.
\c*x^!
n
i]
fy
IL
Jr.
^
N
210 212 289 280 353 370 379
STATION NUMBER
Polynuclnr aromatic hydrocarboni in w«r«r, suspended solids, and surflcial sediments of the Detroit
River. Concentrations in It' ng kg ' Isedimentst. IV ng kg ' (suspended solids), and IVngL ' (water), respectively.
FIGURE IX-6. PCBs, CBs, PAHs and OCS in Detroit River water, suspended solids and
surficial sediments (29).
-------
463
STATION NUMBER
213 214 231 223 224 255 311 314 330 34< 352 384 3M
5-
10-
20-
2ft-
y*
Fj
y
y*
LJ
g
3
-
_
*£
ML
: "
; |
t E
* E
-
•
E
:
i B
IEV1.LE
I |
' n
I i
\
34
r,«
;
j
-
:
:
E
0
i
•
Fi1
H
1
i
•
!
E
:
;
i
i
r |
28|
& WATER
£3 SUSPENDED SOt.CS
a SUBFC1AL SEDIMENT
• NO SAMPLE
NO NOT DETECTED
1
|
: B
!
a
i
I
i
!
\ Hi ll a
^^
\1
N
_
i.
20 212 2«B 2SO 363 370 378 3"
STATION NUMBER
Polychlorinaitd biphinyls (PCBi) in wour. susptndtd solids, and surficial udimtna of llu Detroit Him.
Conctninuioni in If ag kf ' (uditiuntt, susp*nd*d solulsl and Iff' ng L ' (watir), rtsptalvtly.
STATION NUMBER
213 2M 231 223 224 255 311 314 330 34< 352
384 3W
8 WATER
Q SUSPENDED SOLOS
• SURFCIAL SEOMENT
• NO SAMPLE
NO NOT DETECTED
Chlarobtnuna In wattr. aaptndtd solids, and surficial stdlmmls of llu Oaroit Klvtr. Conctnlra-
lioiu In If at *l' (udimtna. luiptndtd totids) and If' nf L ' twatr), rtsptctivety.
FIGURE IX-6. (Cont'd.) PCBs, CBs, PAHs and OCS in Detroit River water,
suspended solids and surficial sediments (29).
-------
464
U
UJ
CO
z
oc
UJ
O
3D
O
U
UJ
CO
Z
cc
i-
cc
111
a.
a.
(U
5
2
«^
CD
O
a
o
c,
r-'
i
X
D
O
ivioi jo % sv amvA DOIOWOH
-------
465
Detroit River System Mass Balance Study (30) , total PCB concen-
trations averaged 1.4 ng/L (plus or minus 0.6 ng/L) at the head
of the river and 3.3 ng/L (plus or minus 1.3 ng/L) at the mouth,
based on composite samples across the entire river at each re-
spective transect. Total PCB concentrations in whole water sam-
ples from tributaries averaged 45.4 ng/L in the Rouge River, 47.9
ng/L in Turkey Creek, 33.3 ng/L in the Ecorse River and 7.6 ng/L
in the Little River (Table IX-1). In the Trenton Channel Mass
Balance Study (31), total PCBs in whole river water ranged from 1
ng/L to 385 ng/L. The highest concentrations were found along
the western shore of the Trenton Channel, with daily variations
ranging from 6.8 ng/L to 15.7 ng/L.
PCB concentrations throughout the Detroit River exceeded
Michigan's Rule 57(2) allowable level of 0.02 ng/L, the Ontario
Provincial Water Quality Objective (PWQO) of 1 ng/L and the
U.S.EPA Ambient Water Quality Criteria (AWQC) for Human Health
(based on fish and water consumption) of 0.079 ng/L, and some
locations (e.g., Trenton Channel) exceeded the U.S.EPA chronic
AWQC of 14 ng/L.
In suspended solids, PCB levels were at or below 50 ng/g at most
locations on both sides of the river, except at two stations on
the Michigan side, below Belle Isle and at the lower end of the
Trenton Channel, where they reached 280 ng/g. Concentrations
measured on suspended solids at the head of the Detroit River
averaged 428 ng/g, largely due to one elevated measurement. A
single suspended sediment sample collected in 1985 from the
Canard River had a very high PCB concentration of 11,760 ng/g,
but other data suggest that the Canard River is only an intermit-
tent PCB source (32).
Chlorobenzenes:
Several of the 5 possible chlorobenzene homologs are commonly
found in aquatic systems, of which hexach'lorobenzene (HCB) is
probably the most widely distributed congener. In Detroit River
water, Chlorobenzenes ranged from 0.3 to 1.0 ng/L at stations
above Belle Isle and at all but two Ontario stations (maximum
approximately 2 ng/L, Figure IX-6). On the Michigan side, chlor-
obenzene levels were somewhat higher, particularly at the mouth
of the Rouge River, where chlorobenzene levels reached 25.9 ng/L
(Figure IX-6). However, HCB concentrations were only 0.28 ng/L,
indicating other Chlorobenzenes are present. In a later study,
concentrations of HCB remained virtually the same from the head
(0.31 ng/L) to the mouth (0.33 ng/L) of the Detroit River (Table
IX-1). In another survey, HCB in water and/or suspended particu-
lates showed essentially the same HCB concentrations on both
shores and at upstream and downstream transects. These results
indicate small or intermittent sources of HCB along the Michigan
side of the Detroit River, perhaps from the Rouge River, with
important background concentrations of HCB entering the Detroit
-------
466
41
+j
j
0
41
•o i
•H
J^
J2
U
1
•o
0 i
oi
2
o
JS
ft
*
o !
JS <
a '•
I
09 '
p^
flj !
*J 1
.
41
O
0
b |
b
O '
•••s '
0 •
fl -4
— »
>. co
41 00
c at
« —
N 1 i
C •* 1
41 00
X CB
O ~*
b —
O
—1 OB
X X
U -P
0 3
X 0
41 B
•-i 0 !
X *>
C 3
41 X
X -H
0. b
X
b
•o o
41 •*»
•P 0
o a
c
f4 TJ
b e
O 0
£ ••
U X
rl 3
o* a
44 TJ
O fl
0
«
e -a
o 0
-rt 41
•P X
b b
P «
C >
4) -rt
0 OS
C
O -P
O -H
O
C b
4) 4)
X O
—
CO M
U B
X —
i
i
1
^ !
J J
< CO \
H CO M
O U B
H flu —
„
J J
< Q-v
H «t no
g- r_^ . — .
•
CO
S
at
O ^
ij 0* J
< CO >»
HOW
O B 3
HO.—
U
Q
M — -
OS —
O ^
x e
o —
!
•^
•J i-3 '
i < O ^-
! H Z »
O M 3
H N — '
-3 —
! -J M J
•t *-^
HOW
1 O w 3
j HZ-
| ^
i 0* — *
J D J
•< U ^
H OS W I
O W 3
£* Z ^^
OS -1
•J U V.
< a. w
H CU 3
O O —
HU
3-
J HH tJ
< Z x,
HOW
O ^1 3
i
z
o
^f
t**
^J
«-< en oi
• • •
O O O
^
* V CO
^ CO *
CO
^ CO CO
1 1 1
^
CO •> O
" • •
CO >O 00
— oo
c- * r-
to to c-
m
CM O CM
CM n
• • ^*
r* CO i^
t- in CM
01 « to
0 i-l CM
00 CO CM
0 O O
o o o
o o o
en ^* m
CM to o
*H ••* C4
m 10 «
CM n co
o o o
o o o
X
•O *> b
0 3 41
4) 0 >
X I •-«
1 1 OS
•rf -rt 41
O O 09
b b b
•P «i O
4} 4) O
Q Q U
X
CM
•
O
^
^
•
IO
1
o
•
CM
O
"
CM
o>
to
CO
•
t-
to
PH
o
"
l>
^H
o
o
en
o
t-
fn
to
o
CM
b
41
^l
OS
41
t»
3
£
CM
0
^
a>
c*
N
n
m
i
M
^^
0
o
o
O c-
* 1
S5
n —
M
— »
o
CD
to oo
10 —
O 10
••1 -^
CM
•
PH
CM
to
C*
oo
to
1-1
o
o
oo
CO
•
^«
c*
to i
CD CM
PH •
O —
it
4)
41
b
O
^
41
b
3
i-t
O
P4
IO
•
*•
rt
1
CO
^
p^
0
o
^*
O CM
• 1
CO CM
t- IO
•* —
M
tO
1-H
00 CM
CO tO
en «
*^
c-
•
m
c-
«
O
Z
to
1-1
o
o
^f
fmt
•
IO
(^
-^
^J*
o
00 1
1O CM
o •
" O
o —
b
41
.^4
OS
41
•P
•P
•H
|
i
1
M
e
i
CO
f
M i
o '
IO
IO
1
(•
a
1
1
N
10
CM
1
t^
CM
I
!
| |
i
I
1 i
1
|
^
• !
O
CM
O
b
41
•H
as
T>
b
0
e
o
b
1
41
41
CO
^
•o
o
B
O
b
£ t
e
gg 0
41 -H
3 P
-4 0
« fl
> b
O
44
• B
•o -<
+J ^H
0 0 •
c o -a
•H 41
00 +J *>
0 00 b
P P 0.
A 0 41
41 P b
U 00
X 4)
41 T3 3
B -H
— * *
to >>
CM W 41
•«> O 00
rt 0
1-P O b
b -0 «
O O >
ax 0
4) -P
as 41 o
s z
a
3 W>
O B
b •« -
M-H —
b e ""
00 —
2 09
.-. t)
0 -^ ttj
3 0
b -0 0
41 41 *
« b >>
» 0 J3
a
41 4i •a
X b 41
b b
COO —
•H 4« a to
41 N
•0 -P b —
41 b
•P O 09 09
b a 41 >>
O 41 3 41
at.-i >
41 0 • b
3 41 00
00 O b *)
0 b 4) O O
M 3 OJ 3
< .£ O *> -P
0. b -1 41
U O T3 *i
• » -0 0
m B P
• >. 0 O 4)
3 -P CM
•H b 0
a -t 41 it b
000, 41
b 3 ao >
b »3 Z <
-. M n •*
-------
467
River from upstream. Data from a 1984 study, however, indicated
increased HCB concentrations on suspended sediments, from ap-
proximately 3.5 ng/g at the river head to approximately 15 ng/g
at the Detroit River mouth.
Other Organochlorine Compounds:
A variety of additional organochlorine contaminants (OCs) are
frequently observed in Detroit River water and seston samples.
Among these are DDT and its environmental metabolites, commonly
referred to as total DDT, hexachlorocyclohexane (three isomers),
chlordane (two isomers), heptachlor epoxide, endosulfan (two
isomers), dieldrin, endrin, methoxychlor, and octachlorostyrene
(OCS). These compounds, collectively referred to as OCs, were
found at concentrations of 0.3 to 0.5 ng/L in upper Detroit River
water on both shores (Figure IX-6). Significantly higher OC
concentrations were observed at many downstream stations on the
Michigan side, with values as high as 20 ng/L at the mouth of the
Rouge River. OCS levels, however, were virtually constant
throughout the river at 0.005 to 0.008 ng/L in water and at 2.0
to 4.3 ng/g on particulate matter as found in another survey.
These data indicate sources of OCS are primarily upstream of the
Detroit River but important loadings of other OC compounds occur
along the Michigan side of the Detroit River (26,32,33).
Polynuclear Aromatic Hydrocarbons:
Polynuclear aromatic hydrocarbons (PAHs) are byproducts of incom-
plete combustion of fossil energy resources. PAHs are also as-
sociated with petroleum refining and steel-making operations
(coking, in particular). Consequently, their presence in air and
water in urban and industrial areas is not surprising. At the
head of the Detroit River, PAH concentrations of 100-200 ng/L
were found in water. Higher concentrations were observed at
several downstream stations along the Ontario, and particularly,
the Michigan side of the river, with values as high as 6,100 ng/L
(Figure IX-6). Based on the high concentrations of PAH that were
found at the mouth of the Rouge River and sampling locations
immediately downstream, large sources for PAHs appear to exist in
the Rouge River area (26,30,31). Water samples from the Ontario
tributaries (Turkey Creek, Little River and the Canard River)
obtained during 1984 revealed no PAHs were present at the limit
of detection used (34). There is no appropriate ambient water
quality guideline with which to compare PAH concentrations in
Detroit River water.
Total Trace Metals, Total Phosphorus and Filtered Chlorides:
A 1987 survey of selected trace metals (copper, cadmium, mercury,
nickel, and zinc), phosphorus and chloride concentrations result-
ed in the following general conclusions (.Table IX-1) (26,30).
-------
468
Total cadmium concentrations increased from the head to the mouth
of the Detroit River from a mean of 0.023 ug/L to a mean of 0.035
ug/L. In general, Detroit River water concentrations were below
relevant ambient water quality guidelines. The Trenton Channel
Mass Balance Study found total cadmium concentrations ranging
from 0.7 ug/L to 0.77 ug/L (data not shown in Table IX-1) in the
vicinity of the Grosse lie free bridge along the western shore of
the Trenton Channel, three of the four times it was sampled.
These concentrations exceeded Michigan's Rule 57(2) allowable
level of 0.4 ug/L (assuming a water hardness of 100 mg/L calcium
carbonate). High cadmium concentrations were found in the Rouge
River (2.06 ug/L), the Canard River (0.2-0.4 ug/L), Turkey Creek
(0.196 ug/L in one study and up to 3 ug/L in another), the Ecorse
River (0.084 ug/L) and the Little River (.0.058 ug/L in one study,
and up to 0.4 ug/L in another). Concentrations in the Rouge
River, Turkey Creek and the Canard River exceeded the Great Lakes
Water Quality Agreement (GLWQA) specific objective and the PWQO
of 0.2 ug/L, and concentrations in the Rouge River and Turkey
Creek exceeded Michigan's Rule 57(2) allowable level.
Total copper concentrations were slightly higher at the Detroit
River mouth than at the river head (1.64 ug/L vs. 1.29 ug/L).
Total copper concentrations in the tributaries were between two
and six times higher than in the Detroit River, with the Rouge
River levels highest at 7.1 ug/L. In general, both Detroit River
and tributary copper concentrations were below relevant guide-
lines, with the exception of the Rouge and Little rivers, which
slightly exceeded the GLWQA specific objective and the PWQO of
5 ug/L.
Total mercury concentrations in Detroit River water did not show
any change between river head and mouth (both 0.008 ug/L). Total
mercury concentrations in the Detroit River and in the Trenton
Channel ranged from 0.024 ug/L to 0.449 ug/L. Tributary mercury
concentrations were approximately double those in the Detroit
River, except in the Ecorse River, where they were lower. These
concentrations generally exceeded the U.S.EPA chronic AWQC of
0.012 ug/L.
Total nickel concentrations in the Detroit River showed little
change between upper (0.97 ug/L) and lower (1.1 ug/L) Detroit
River transects. Nickel concentrations in the Ecorse and Rouge
rivers, and Turkey Creek were from two to eight times the Detroit
River level, with the highest concentration in Turkey Creek (8.8
ug/L). Especially high concentrations of nickel were noted in
the Little River (676.2 ug/L)(26). With the exception of the
Little River, all Detroit River and tributary concentrations of
nickel were below ambient water quality guidelines. Little River
exceeded U.S.EPA chronic, Ontario and Michigan ambient water
quality guidelines.
-------
469
Total lead concentrations were all below the method detection
limit (MDL) of <0.1 ug/L in the Detroit River head and mouth
transects. Several locations in the Trenton Channel contained
total lead concentrations ranging from 3.24 ug/L to 10.61 ug/L,
which exceeded Michigan Rule 57(2) allowable levels (3.0 ug/L)
and the U.S.EPA chronic AWQC (3.2 ug/L). The highest concentra-
tion was upstream of the Grosse lie toll bridge along the western
shore of the Trenton Channel (transect A, Figure IX-3). Tran-
sects C and D also have total lead concentrations exceeding
guidelines along the western shore of the channel. Total lead
concentrations in Ontario tributaries were determined for the
Little River (3-13 ug/L), the Canard River (3-30 ug/L) and Turkey
Creek (3-33 ug/L). These tributaries all contain total lead
concentrations above guidelines (26,35). Concentrations of total
lead in Michigan tributaries were not available for this report.
Total zinc concentrations increased between upper (1.2 ug/L) and
lower (3.3 ug/L) Detroit River transects. Each of the tributar-
ies also had high mean zinc concentrations, with the Ecorse River
having the least (14 ug/L) and the Rouge River the highest (167
ug/L) total zinc concentrations. With the exception of the Rouge
River and the Little River (74 ug/L), water concentrations were
below ambient water quality guidelines. Little River concentra-
tions of total zinc exceeded GLWQA specific objectives (30 ug/L).
Rouge River total zinc concentrations exceeded this guideline and
also the U.S.EPA chronic and acute AWQC.
Total phosphorus concentrations were nearly twice as high at the
Detroit River mouth (15.7 ug/L) compared to the river head (8.6
ug/L). Total phosphorus concentrations in the major Detroit
River tributaries were much higher than concentrations in the
Detroit River.
Filtered chloride concentrations increased from 6.7 mg/L to 8.4
mg/L between upper and lower Detroit River transects. The lower
Detroit River transect was located above General Chemical, a
major chloride loading source discussed later, and therefore this
loading was not reflected in the Detroit River mouth transect
value shown in Table IX-1. The filtered chloride concentrations
in the Detroit River tributaries were one to two orders of mag-
nitude greater than the Detroit River head. Total chloride con-
centrations (not shown) did not increase between the head and the
mouth. The drinking water guideline for chlorides (250 mg/L) was
exceeded in Turkey Creek and North Drain.
Nutrients, Dissolved Gases and Microorganisms:
The basic plant nutrients in the Detroit River include
phosphates, nitrates, and silicates. Dissolved oxygen and the
metals iron, sodium, calcium, magnesium, manganese and aluminum
are also present in sufficient quantities. The oversupply of
phosphate, chloride and ammonia has decreased substantially over
-------
470
the past 20 years.
Dissolved organic carbon (DOC) and particulate organic carbon
(POC) are often many times greater than the organic carbon found
in living plankton, macrophytes, and fauna produced in streams.
DOC measurements available from Lake Huron, the St. Clair and
Detroit Rivers are in the range of 2-3 g/m-3. The POC entering
the St. Clair-Detroit River system from Lake Huron is about 0.7
g/m3. An average of 1.4 g/m3 was measured at the mouth of the
St. Clair River, and up to 2.0 g/m3 were found in Lake St. Clair.
A single POC sample from the mouth of the Detroit River was 3.8
g/m^. Suspended solids increased by a factor of six between Lake
Huron and Lake Erie, and bed load POC has not been studied, so
3.8 g/m3 may underestimate POC in the Detroit River.
Although not measured during these studies, fecal coliform bac-
teria are of concern in the Detroit River because fecal coliform
bacteria standards and criteria have been violated on both sides
of the river. The Ontario objective is 100 counts/100 ml and the
Michigan standard is 200/100 ml fecal coliform bacteria. Beaches
have been closed or not developed because of this continuing
problem.
Water Bioassays:
Seven day chronic bioassays measured the impacts of Detroit River
near-bottom water on Ceriodaphnia. Reproductive success was
significantly reduced (mean young produced/female) relative to
Lake Michigan controls at all four test sites. Station 83 near-
bottom water collected along the southwestern shore of Fighting
Island produced the greatest reduction in the number of young
produced/female (70 to 100% reduction) followed by stations 34
(along the west shore of the Trenton Channel), 53 (at the south-
ern tip of Grosse lie and 30CR (in Monguagon Creek). These re-
ductions were most severe from July to September (36).
Considering both exceedences of water quality and impacts on
biota, the pollutants of concern in water of the Detroit River,
or that of its tributaries, include PCBs, chlorobenzenes, PAHs,
total cadmium, total mercury, total lead, total zinc, and total
phosphorus, in addition to fecal coliform bacteria.
Biota
i) Phytoplankton, Macrophytes and Zooplankton
Detroit River phytoplankton communities consist of low densities
(500 cells per ml) of 82 species dominated by diatoms (8,10,37).
Summer blue-greens contribute to phytoplankton community, but
Detroit River picoplankton, a large component of the phytoplank-
ton biomass, were not surveyed.
-------
471
Activity causing habitat loss, such as filling or dredging, water
or sediment contaminants or simply continuous elevated suspended
solids that reduce macrophyte production, reduces desirable fish
and wildlife production in the Detroit River and western Lake
Erie. Macrophyte production was estimated at 16,410 metric tons
of ash-free dry weight/yr (12). Only 25% is from emergents re-
flecting the limited habitat presently available.
Detroit River zooplankton populations (potential larval fish
food) were 85% copepods with other zooplankton populations at
very low relative abundances. Zooplankton densities were greater
during the night than the day with typically patchy distribution
with peak numbers between June and September (36,38). Zooplank-
ton are a critical component in the diet of many larval and some
juvenile fish. Poor diversity or depressed zooplankton produc-
tion is likely to result in poor fish year classes during natur-
ally occurring or contaminant related stressful conditions.
ii) Benthic Macroinvertebrates
Diversity and abundance of benthic macroinvertebrates are lower
in the deep, fast flowing areas of the river because the sub-
strate is either difficult to adhere to or burrow into. Shal-
lower, uncontaminated zones containing macrophytes are likely to
yield the greatest diversity. The greatest densities are reached
in strongly enriched,, unconsolidated sediments where oligochaetes
are often monotypic.
The Detroit River benthic community upstream of Zug Island is
diverse and dominated by pollution intolerant organisms with the
exception of the Windsor shoreline. Adjacent,to Zug Island, the
community is severely impacted, and downstream, especially in the
Trenton Channel, the community is dominated by pollution tolerant
oligochaetes (13,15,39). The Ontario shoreline is considerably
better as evidenced by the presence of pollution intolerant may-
flies (11,15).
Schloesser, et al. (40) demonstrated an inverse relationship
between Hexagenia abundance and visible oil in sediments of the
Connecting Channels. Edsall et al. (41) found Hexagenia averag-
ing 2,086 mg dry wt/m^/yr at three locations where sediment con-
taminants did not exceed sediment guidelines, but only 364 mg dry
wt/m^/yr where as many as seven contaminants exceeded these
guidelines.
Native Detroit River Lampsilis radiada siliguoidea, at 4 stations
along the Ontario shore, contained lead and cadmium ranging from
3 to 9 and 3.5 to 6.2 mg/kg respectively (42). PCBs ranged from
73 to 196 ug/kg at these same locations. Octachlorostyrene (OCS)
in clams ranged from 31 to 57 ug/kg, 70 to 285 times higher than
sediment concentrations.
-------
472
Caged Elliptic compalanta placed in the Detroit River for 18
months accumulated HCB and OCS and a variety of organochlorine
pesticides (43). Highest levels were found along the western
Detroit River shore near Conners Creek, the lower Trenton Channel
and the Rouge River. PCBs were the major organochlorine clam
contaminant, ranging from 20 to 293 ug/kg along the Michigan
shore; clams from the Ontario shore had much lower concentrations
(Figure IX-8).
Polynuclear aromatic hydrocarbons (PAHs) were also reported in
caged clams at elevated levels along the Michigan shoreline and
downstream in the Trenton Channel ranging from 136 to 772 ug/kg.
Along the Ontario shoreline PAHs ranged from 52 to 274 ug/kg.
iii) Fish
Five fish species were collected from six sites in the lower
Detroit River and examined for external lesions, necropsied for
internal abnormalities and tissues removed for histological
examination (Figure IX-9) (44). Several neoplasms and pre-neo-
plastic lesions were found in Detroit River brown bullhead,
walleye, redhorse sucker, white sucker and bowfin. Bullhead and
walleye were the only two species exhibiting dermal/oral neo-
plasms at 14.4 and 4.8 %, respectively. Other species exhibited
liver neoplasms with highest incidence observed for bowfin at
15.4%. In bullhead, no relationships between dermal/oral and
liver tumors were found. Tumor incidence was age/size related
since tumors were present in bullheads over 25 centimeters and in
walleye over 50 centimeters. Of the six sites examined, bull-
heads at Point Hennepin and Gibraltar Bay, exhibited the greatest
tumor incidence at 36.4% and 33.3%, respectively. Bullheads near
Mud Island north of the Trenton Channel and in the lower end of
the Trenton Channel did not exhibit tumors.
In this study (44), bile was analyzed for benzo(a)pyrene (BaP)
and its metabolites. All species had BaP or its metabolites in
their bile. Walleye and redhorse sucker contained the greatest
BaP concentrations, with concentrations in bullhead substantially
lower. The greatest BaP concentrations were in bowfin and red-
horse sucker from Point Hennepin and in brown bullhead, walleye,
and white sucker from Mud Island.
Contaminants exceeding relevant guidelines were found in the
flesh of fish in the Detroit River. PCBs were found in carp,
with concentrations exceeding the Ontario1 Ministry of Environment
(OMOE) and Ontario Ministry of Natural Resources fish consumption
guidelines and the U.S. Food and Drug Administration action level
of 2 ppm, as well as the GLWQA specific objective of 0.1 ppm
(Figure IX-10). PCBs in young-of-the-year spottail shiners were
found at significantly (p<0.01) higher concentrations along the
Michigan shoreline than along the Ontario, suggesting Michigan
inputs of PCBs (45). High concentrations of mercury were found
-------
473
629
294
552
Michigan
58
695
Total PCB
Concentrations
ng/g, wet wt.
FIGURE IX-8. Total PCB concentrations in Detroit River caged clams.
-------
474
Michigan
Fish Sampling
Locations
n
FIGURE IX-9. Fish sampling locations for tumor analysis.
-------
475
PCS pom
Carp 6.7
HCB ppm
Carp 0.1
m __ . .^. —— Spottail .006
.____.__-. — .<—• Spottail .007
Spottail 03
Carp 10.7
Spottail .007
Spottail .009
Spottail 2.6
t
FIGURE IX-10. PCBs and HCB concentrations in Carp and Spottails shiners.
-------
476
in the edible portion of several species of fish (rock bass,
freshwater drum and walleye). Concentrations were above both the
GLWQA specific objective and the Ontario fish consumption ad-
visory of 0.5 ppm (46,47). Other chemicals, such as HCB, OCS,
chlordane and DDT metabolites, were uniformly distributed in
Detroit River spottail shiners, suggesting a diffuse source (45).
iv) Birds
Thirteen wintering lower Detroit River diving ducks (7 lesser and
3 greater scaups and three goldeneyes) were analyzed for organic
chemical contaminants (48). Total PCBs ranged from 2 to 20
mg/kg, indicating significant bioaccumulation. Highest mean
concentrations of other residues in ducks were 1.7 mg/kg hexa-
chlorobenzene in goldeneyes, and trans-nonachlor (0.33 mg/kg) and
4,4' DDE (1.3 mg/kg) in greater scaups. Similar chemical resi-
dues were also found in some tern species. Concentrations of
total PCBs in Detroit River seston (5.2 mg/kg) and oligochaete
worms (0.44 mg/kg) mg/kg) were also noted.
Herring gull eggs from Fighting Island contained high PCB and HCB
concentrations in 1985 and 1986 studies. Detroit River herring
gull eggs contained the lowest concentrations of dieldrin, hepta-
chlor epoxide, photomirex, oxychlordane and alpha hexachloro-
cyclohexane in the Great Lakes (49).
Detroit River waterfowl surveys completed in 1982 showed dramatic
declines in merganser and black ducks, and dramatic increase in
canvasbacks and redheads since 1974 (50). It was postulated that
loss of emergent macrophytes caused by high Great Lakes water
levels caused this reduction in dabbling ducks.
In summary, the pollutants of concern in Detroit River biota
include PCBs, PAHs, HCB, OCS, mercury, lead, cadmium and oil and
grease. Other biota concerns include habitat alteration and fish
tumors.
Sediment Quality
i) Sediment Characteristics
Sediments in the Detroit River are generally sandy, consolidated
clay or bedrock because of the relatively high flow velocities.
Sediment particle size analysis conducted in 1980 revealed that
surficial sediments were generally sand, but gravel dominated
areas of high velocity along the Detroit waterfront, the entrance
of the Trenton Channel and the upper Amherstburg Channel. Fine-
grained samples were collected in slow waters near tributary
mouths. Silts and clays were found downstream of Zug Island, in
the Rouge River, the Trenton Channel near Trenton and the Detroit
River mouth (51,52).
-------
477
Detroit River sediment thickness over bedrock revealed a maximum
sediment thickness of 33 m near Belle Isle, which declined stead-
ily southward to nearly zero in the Trenton Channel and zero in
the main channel (53) . The outer and Amherstburg Channel silt
layer averaged 0.45 to 0.50 m near Lake Erie and zero in the
Amherstburg Channel at Bois Blanc Island and in the Ballard Reef
Channel.
The Michigan Detroit River tributaries which were not sampled in
1982 were sampled in 1985, revealing fine-grained, anthropogenic
sediments frequently of sludge-like consistency (54,55). Samples
in Monguagon Creek and downstream of the Rouge River contained
very fine sands, silt, and coarse sand and gravel. The upper
Rouge River sediments were coarser than elsewhere, consisting of
medium to fine sands with little very fine sand sediments.
Conners Creek sediments also had only minor amounts of fine to
very fine sands. Studies conducted in 1986 at 47 sites (56,57),
generally confirmed the earlier findings.
ii) Sediment Transport
Detroit River average main channel velocities are 0.49 to 0.88
m/sec, but surface velocities may be nearly twice that rate in
the main channels (0.9 to 1.2 m/sec)(58). Sand is transported in
the main channels when the velocity exceeds 0.42 m/sec, while
along the shore and in shallow water areas, where velocities may
drop to 0.25 m/sec or less, sand deposition occurs. Navigation
channel bottoms are scoured by currents leaving few sediments to
resuspend, and no significant relationships between ship passage
and turbidity has been found (59).
A field portable shaker device was used to measure sediment
resuspendability at eight Trenton Channel locations from Mon-
guagon Creek to Celeron Island. Lick et al. predicted that
resuspension could occur regularly in the Trenton Channel (60).
Direct instantaneous measurements of flow velocity, turbidity and
sediment concentration at four locations in the Trenton Channel
using instrumented towers assisted the above researchers (61).
iii) Navigation and Dredging
Until recently, the entire Detroit River commercial navigation
system was dredged by the U.S. Army Corps of Engineers (USCOE) to
a depth of 8.2 m below low water datum. At present, the Ontario
portion of these channels are dredged by Public Works Canada
under contract to Transport Canada. Before enactment of the
Rivers and Harbors Act of 1970, nearly 3 million m3 of dredged
materials were disposed of in the open lake at two sites in Lake
Erie south of the Detroit River mouth (62). In addition, an
unknown amount of Detroit River dredged materials were placed in
Lake St. Clair, near the head of the Detroit River. Since 1970,
about 30,100 m3 of polluted dredged materials were placed on
-------
478
Grassy Island. From 1979 to 1984, 3.1 million m^ of dredged
material were deposited in the Pointe Mouillee confined disposal
facility (CDF) near the Huron River mouth (58). In 1985, 814,000
m^ of polluted Detroit River material was scheduled for disposal
in the Point Mouillee CDF. Rouge River sediments, since 1950,
have been placed on Grassy Island (62). Some polluted dredged
materials were also disposed of along the lower Raisin River
prior to 1979. Mud Island, a small containment site near Grassy
Island, was also used for dredged material disposal.
iv) Sediment Contamination
Results of the six major surveys conducted since 1982 include
contaminant chemistry at approximately 135 sites (51,54,55,63,64,
65,66,67,68,69,70,71,72,73). For ease of presentation, the
Detroit River was divided up into seven subareas (Figure IX-11).
Because the purposes for the survey, sampling gear, analytical
methods, depth of sample collection, compositing techniques and
sampling locations varied considerably between the studies, com-
parison of these data from year to year may not be entirely
valid. However, an attempt was made to make some comparison.
Organics - Polychlorinated Biphenyls:
High total PCS concentrations were found by six surveys in all
subareas except subarea 7 (Table IX-2, Figure IX-11). The high-
est mean sediment PCB concentrations were found in subarea 2,
just below Belle Isle, where 5 of 10 samples exceeded 10,000
ug/kg in 1986. These were associated with sewer system outfalls,
and indicate that combined sewer overflows have historically
been, and may still be, an important source of PCBs (64).
The 1984 analyses of Oliver and Pugsley (74) noted localized
areas of high concentrations of PCBs downstream of the Detroit
WWTP and the Rouge River (in subarea 3), at concentrations higher
than reported in 1980 (75), assuming the methodologies of the
1980 and 1984 studies were comparable. Comparison of 1982 and
1985 collections are supportive of the conclusion that subarea 2
sources were more significant than the Rouge River. Rouge River
sediments collected at the mouth in 1986 revealed total PCBs up
to 3,500 ug/kg (76). Samples collected downstream of the Detroit
WWTP outfall and off the Rouge River mouth in 1985 and 1986
revealed PCBs up to 2,840 ug/kg near Zug Island (28). Concentra-
tions up to 3,800 ug/kg were found in the Ecorse River (subarea
4). The highest concentrations in the navigation channel (sub-
area 4) was 140 ug/kg, between Grosse lie and Fighting Island
(77). Sediments analyzed from along the Windsor waterfront
showed PCB concentrations ranging from less than 1 ug/kg to 370
ug/kg.
Sediment collections made in 1982 and 1985 also indicate PCB
sources in subarea 6, the Trenton Channel. Highest levels were
-------
479
FIGURE IX-11. Detroit River sub-areas for sediment sampling.
-------
480
1 en
i! S
M fc
CO
•
3
II
•
1
H i i
"^ Cdl
so ,i
>
•H
Q! ''
•P
•t-t
0 ,(
CD X
CM ^J
i i cu
C CzJ
CM .H "
1
X 01
M C
0
Ed •"
-1 «•> II
CC «
< t-
^* *^ 1
C ''
Cj Ed
O ' C
c z
0 O
o
—
«
O ..
a. i'
— ^
&"•»
C
flj p*
X il Z
0, 0
•M 2
•° '1
u
4J
03
C
•H
«
0
s
u
—4
o
a.
1 1
^
' «5
i ! [j
[i
1
'1 CC
3
oooo^Joco^Joinco^
tn t* -« 1 1-1 w as co ! o co co co
CO i— t
it fH •* CM
in n CM
1
1
i
1
!
O fH fH CO
^1 fH fH
i
O O ^H 9) C*«
^> *-«
O N *™t W O5
N CD 31 M
O 3S ^ »0
C*4 1—1
« XJ ^ t- Tf
CO ^ «5 O
r— «-« CO C»
as (N c^
: o t- x o fH
OOOfHOJO O.HCM
en en 9>
fH fH fH
o o to to in
- 0 to 0
O CM to t-
M to OS CM
ii
en m CB
P- fH CM
CO 00 00 t^ OS
i-i m co en
CM en 10 co
CM fH
c^ en *•* oo en
•* N O O
to CM as to
en fH to CM
°
i i
i
l
I
J
O 10 O C* ^
?J CO i^ W
*f 1-* in
O O *** CJJ X
O O 0 O
^J
z
X Z <
•« H" W Q
O O t- O n
O 1-1 Tf fH 1C
o en ^H in
** rt O
z
in m fH to ^
»H CM m CM
oo m m
fH
O O c- (• c-
o c- o m
oo as o to
V tO fH
jj
< w Ui O ' < M U Q
ZZXCCOiXZXCCO
i
1
]
i
i
1 * CM
x z x c co
O3 O •* 00 0
eo p- in
O CM CO
O i™* CM tO C—
O •» CO CM
O T OS
i
e- O fH o t»
in to n fH
os as co
o to in oo c—
O to fH to to
co to t- en
— i to * O
en CM to
o en CM to c-
N 00 O i O CM •» CO
co ii i ^ to *
00 fH CM
oo to t- co oo
oo o fH in
fH
£
in in to co
to fH CM
a o oo to oo
fH fH m oo
in fH M
i oasocMO ! TT x en x co
i as •* os t
m CM •* ii i
O O iA UD IA
r* OS C»J ^
00 O 00
OJ Tf CO
fH
'
o o m CM in
o as ^* m
O CM CO CO
^H
3;
X Z <
< M W Q
x x x c ca
to
!
! i
a o o CM o
to •* o to
c- os as
i-i in in
fH
Z
xz <
< M (d Q
z z z e co
m -H o ^
fH fH
o o to to oo
c- m CM oo
oo to in en
co CM t- en
fH
0 !0 0 fH
to to in CM
0
f— 1
t- M m oo to
CM CM 00 O
so en to
CM fH
i
i
j
!
I
z
X Z <
•« n [d Q
2
XZ <
< M u a
XZXCCO i XXXCcy)
!
j
in to
|
c-
i
1 1
^J
OH
O
X
t.
01
C
X
0
b
CO
4J
c
o
OJ
z
c
X
CO
11
u
••
0
01
Oi
ce
--
cfl
(.
3
CO
z
q_
0
.
^j
0.
D
C
a
an
•fH
f.
U
•f4
X
4}
.j
• >»
CO
CO "O
^ 1)
g
01 t-
3 O
U t.
•rt Cl)
> a.
X H)
0 b
C dj
CS 3
0) t.
11 (X 01
o. -QU
e c >
a) co M
M 3
01 01
O -*
01 M
t. 3 -H
0) eg o!
B NO)
3 -f4 i
e x E-
II
C H <*
-------
481
in Monguagon Creek (13,870 ug/kg), although very high PCB con-
centrations were found near BASF/Federal Marine Terminal Prop-
erties and below McLouth Steel, near Trenton. Tributary data
collected in 1985 also targets Monguagon Creek as a PCB source,
with concentrations in the creek of up to 1,530 ug/kg (55).
Since PCB concentrations of up to 9,130 ug/kg were reported in
the Trenton Channel proper, other sources are contributing PCBs
to the Trenton Channel in addition to Monguagon Creek (56).
Bottom sediments from Ontario tributaries obtained during the
1984-1985 survey revealed PCB concentrations of 1,305 ug/kg, 248
ug/kg and 20 ug/kg at the mouths of Turkey Creek, the Little
River and the Canard River, respectively (5).
Many of these PCB sediment concentrations in the Detroit River
and its tributaries, in Michigan and Ontario (particularly ad-
jacent to and downstream of Detroit, Windsor and Amherstburg and
in the Trenton Channel), exceed dredging guidelines. Guidelines
exceeded include the OMOE dredging guidelines (50 ug/kg), the
U.S.EPA dredging guidelines (10,000 ug/kg) and are higher than
the guidelines recommended for Lake Erie by the Dredging Sub-
committee of the Great Lakes Water Quality Board (up to 252
ug/kg).
Hexachlorobenzene:
Sediments collected in 1982 and 1985 in subareas 3, 6 and 7 con-
tained hexachlorobenzene (HCB) exceeding 100 ug/kg. Concentra-
tions' of HCB in 1985 downstream of Monguagon Creek ranged from 26
to 140 ug/kg. Inputs from the St. Clair River are probably minor
since loadings between the St. Clair River mouth and the head of
the Detroit River were reduced at least 95%. Increases noted
within the Detroit River may arise through diffuse or unknown
minor inputs. The highest concentrations of HCB were found, in
Michigan at the mouth and downstream of the Rouge River and in
the Trenton Channel; and in Ontario adjacent to Amherstburg and
east of Fighting Island. There are no dredging guidelines for
HCB.
Polynuclear Aromatic Hydrocarbons:
Polynuclear aromatic hydrocarbon (PAH) analyses were performed on
Detroit River sediments in 1982 and 1985. Total PAH values
ranged from 620 to 265,000 ug/kg along the Michigan shore down-
stream of Belle Isle. High total PAH levels (up to 125,000
ug/kg) were also reported in the lower Rouge River. In 1985,
PAHs were reported in the Detroit Dearborn Channel and all
Michigan Detroit River tributaries, ranging from a low concentra-
tion of 600 ug/kg to a high concentration of 600,100 ug/kg in
Monguagon Creek. Most tributary PAH samples were dominated by
3-, 4-, and 5-ring PAH compounds. Two-ring naphthalenes were
found in appreciable quantities only in the Monguagon Creek and
-------
482
the Rouge River. There are no dredging guidelines for total
PAHs.
Phenols:
Phenols ranged from nondetectable to 44,000 ug/kg in localized
areas within subarea 6 along the Michigan shore. High levels
were generally found in subareas 1, 2, and 3 near the Edward C.
Levy Company. There are no dredging guidelines for total phen-
ols.
DDT and Metabolites:
DDT analyses were performed on Detroit River sediments collected
in 1982 and 1985. In 1982, the highest total DDT concentrations
were found near Belle Isle (2,265 ug/kg). In 1985, total DDT was
highest in subarea 1. DDT and metabolites were found in all 1985
samples ranging from 7 to 482 ug/kg (Conners Creek). High levels
of total DDT were also found in the Rouge River mouth and Trenton
Channel, suggesting recent additions that have not been degraded.
Sediments from the mouths of Ontario tributaries generally con-
tained less than 5 ug/kg p'p'-DDT, while breakdown products p'p1-
DDE and p'p'-DDD approached maximum levels of 36 ug/kg and 20
ug/kg, respectively. There are no dredging guidelines for DDT or
its metabolites.
Other Pesticides:
Approximately 34 other pesticides were analyzed in sediments in
1985, 14 of which were found in bottom sediments. Alpha-chlor-
dane, gamma-chlordane, dieldrin and methoxychlor were most com-
monly found. Highest dieldrin levels were found in subarea 5, at
the Canard River mouth (30 to 55 ug/kg). Methoxychlor and gamma-
chlordane were highest in sub-area 3. Maximum levels in bottom
sediments for methoxychlor were 86 ug/kg while gamma-chlordane
levels were 10 ug/kg.
Several chlorinated pesticides were found in the Detroit River
sediments collected in 1985 with highest levels in Monguagon and
Conners Creek sediments. Highest levels of trifluralin (19
ug/kg) were present in the Frank and Poet Drain and the only
occurrences of DCPA (Daethai) were in the Ecorse River and the
Detroit River Dearborn Channel, a tributary to the Rouge River.
Dieldrin (14 ug/kg) was highest in the Detroit-Dearborn Channel,
while aldrin was found primarily in the Rouge River and Conners
Creek sediments.
Beta-BHC concentrations were elevated at Belle Isle (170 ug/kg)
and near the Ecorse River (195 ug/kg) in 1982 collections.
Gamma-chlordane was found throughout the study area with peaks at
Conners Creek and the Ecorse River. Concentrations of other
-------
483
pesticides in sediments showed no distinct relation to potential
sources.
Phthalate Esters:
Phthalate esters were found in 14 of the 20 Detroit River tribu-
tary samples in 1985. Highest levels were found on the Michigan
side in Conners Creek, the Rouge River and near the Federal
Marine Terminals and BASF properties (17,600 ug/kg). There are
no dredging guidelines for phthalate esters.
Volatile Organic Compounds:
Volatile organic compounds were found in 15 of 20 sediment samp-
les analyzed from the Detroit River tributaries in 1985. Di-
chloromethane appeared in 9 of the 20 samples ranging from 0.8 to
6.9 ug/kg in Monguagon Creek where the great variety of volatile
organic compounds were found. Highest concentrations were found
in subarea 7, in the Frank and Poet Drain. There are no dredging
guidelines for specific volatile compounds.
Metals - Mercury:
Mercury analyses were performed on sediments collected in 1982,
1985 and 1986. The highest levels in subarea 6 (Trenton Channel)
were located below the mouth of Monguagon Creek near the Edward
C. Levy Company (55.8 mg/kg). However, a 1985 sample in Mon-
guagon Creek (1.5 mg/kg) indicated that Monguagon Creek was not a
prominent mercury source. Mercury analyses of sediments in sub-
area 6 exceeded 3.0 mg/kg, while bottom sediments in subarea 1
exceed 2.5 mg/kg. U.S.EPA and Ontario dredging guidelines for
mercury were exceeded at many sampled locations along the
Michigan and Ontario shores throughout the length of the river.
Lead:
Lead concentrations exceeded 200 mg/kg in subareas 1, 2 and 6 in
1982 and 1985. Tributary sediment levels were highest in Conners
Creek and the Detroit-Dearborn Channel of the Rouge River, rang-
ing from 500 to 750 mg/kg, but declined downstream to less than
100 mg/kg in subarea 1. Sediment lead concentrations for samples
collected in 1982 and 1985 were similar at subarea 6 above Eliza-
beth Park Canal (1,750 mg/kg). Dredging guidelines were exceeded
along most of the Michigan shore and downstream of Windsor and
Amherstburg in Ontario.
Arsenic:
Sediment data for 1982 and 1985 indicate that Detroit River sedi-
ments contain approximately 10 mg/kg arsenic throughout, with
elevated levels of 36 and 54 mg/kg found at Elizabeth Park Canal
and the Rouge River, respectively. The uniformity of the data
-------
484
suggests no major point or nonpoint sources of arsenic to the
Detroit River; however, dredging guidelines for arsenic were
exceeded.
Cadmium:
Peak cadmium concentrations were in subareas 1, 3 and 6, ranging
between 25 and 96 mg/kg. Cadmium concentrations in suspended and
bottom sediments were approximately equal, perhaps indicating a
persistent local source. Dredging guidelines for cadmium were
exceeded along the full length of the Michigan shore (especially
adjacent to Detroit and in the Trenton Channel) and adjacent and
downstream of Windsor and Amherstburg.
Copper:
Sediment data from 1986 show copper peaks exceeding 100 mg/kg in
subareas 2,3,4, and 6. Sediment data for 1985 showed generally
higher copper levels in subarea 1 and 3, than in 5 or 7 (approxi-
mately 100 mg/kg versus approximately 50 mg/kg). In 1982 and
1985, copper values exceeded 700 mg/kg in subarea 3, Turkey Creek
and the Rouge River. Dredging guidelines for copper were ex-
ceeded along the Michigan and Ontario shores, specifically ad-
jacent to the cities of Detroit, Windsor and Amherstburg and in
the Trenton Channel.
Zinc:
Sediment data for 1986 indicate levels of zinc exceeding 500
mg/kg in subareas 2 and 6. The 1982 and 1985 sediment data show
zinc exceeding 1,000 mg/kg in subareas 1, 2, 3 and 6. The Rouge
River, Conners, Turkey and Monguagon Creeks all appear to be
contributing zinc to the Detroit River. Dredging guidelines for
zinc were generally exceeded at the same locations as for copper.
Chromium:
Sediment data for 1986 indicate chromium levels exceeding 100
mg/kg in subareas 2 and 6. The 1985 sediment data show tributary
sediments as chromium sources in subareas 1 and 3, where suspen-
ded and bottom sediments contained greater than 300 mg/kg total
chromium, indicating a continuing source. Chromium levels were
nearly twice as high in the Detroit Dearborn Channel of the Rouge
River as the lower Rouge River sediments. The 1982 chromium
peaks were not apparent in the 1985 subarea 6 sediments samples,
perhaps indicating some source control. Dredging guidelines were
exceeded at several locations in the Detroit River (as per cop-
per) .
-------
485
Nickel:
High nickel levels (500 mg/kg) were found in bottom sediments
from the Ontario tributary in subarea 1, the Little River. Sedi-
ment nickel levels exceeded 50 mg/kg in subareas 2, 3 and 6 in
1986, while 1985 data indicate subareas 1 and 3 as having high
nickel contamination. The high nickel levels found during the
1982 survey in subareas 4 and 6 were not evident in 1985 data.
Dredging Guidelines were exceeded at several locations (as per
copper).
Manganese:
Manganese levels exceeding 1,000 mg/kg were found in subareas 3,4
and 6 (the Rouge and Ecorse Rivers and Monguagon Creek) in 1985,
which was about the same as in 1982. High manganese in subarea 7
in 1982 was not reported in 1985, but 5,000 mg/kg manganese was
reported in the Ecorse River in 1985 that was not noted in 1982.
Dredging guidelines for manganese were exceeded along the
Michigan shore. Manganese concentrations in Ontario sediments
were not determined.
Iron:
Sediment concentrations of iron from the 1982 survey reached
180,000 mg/kg above Elizabeth Park (subarea 6). Iron levels
along the Michigan shore were very high in 1982, with some sta-
tions in all subareas exceeding 25,000 mg/kg. The highest iron
concentration found during the 1985 survey was 120,000 mg/kg from
the Ecorse River. Dredging guidelines were exceeded along the
Michigan shore. Iron concentrations were not determined for
sediments along the Ontario shore.
Cobalt:
Cobalt was analyzed in 1982, 1985 and 1986. The 1986 cobalt
concentrations were relatively uniform with a slight increase
downstream. Highest levels (over 10 mg/kg) were found in subarea
6. The 1982 samples were also relatively uniform, although
slightly higher than 1986 samples. The highest cobalt levels
were found in the 1985 tributary samples in subarea 3 in the
Detroit Dearborn Channel (17 mg/kg). No exceedences of dredging
guidelines were noted.
Nutrients and Conventional Pollutants - Cyanide:
In 1982, cyanide levels exceeding 10 mg/kg were present in sub-
areas 1,3 and 6. In 1985, high cyanide concentrations were pres-
ent in subareas 1 and 3 (Conners Creek and Detroit Dearborn
Channel). Lower levels were found in the Lower Rouge and
Monguagon Creek, indicating that sources other than Monguagon
Creek were responsible for high levels found in subarea 6 in
-------
486
1986. Exceedence of dredging guidelines for cyanide occurred in
Michigan and Ontario adjacent to Detroit, Windsor and Amherstburg
and in the Trenton Channel.
Oil and Grease:
The highest oil and grease levels found during the 1986 survey
were reported in subarea 6 with concentrations over 24,000 mg/kg.
In 1985, oil and grease levels were highest in subareas 1 (44,800
mg/kg) and 3 (28,600 mg/kg), and generally decreased downstream
from the Detroit River head to its mouth. In 1982, peak oil and
grease levels exceeding 30,000 mg/kg were present in subareas 1,
2,3 and 6. Dredging guidelines for oil and grease were exceeded
in many areas, primarily along the Michigan shoreline adjacent to
and downstream of Detroit and in the Trenton Channel, as well as
adjacent to the cities of Windsor and Amherstburg.
Total Phosphorus:
Most total phosphorus concentrations in sediments were lower than
5,000 mg/kg. Along the Michigan side, phosphorus levels up to
6,200 mg/kg in 1982 were found in subarea 6, whereas the highest
level in 1985 (6,200 mg/kg) was found in the Detroit Dearborn
Channel. Exceedences of phosphorus dredging guidelines were
noted in the majority of samples analyzed in both Michigan and
Ontario.
Ammonia:
The 1982 concentrations of ammonia exceeded 500 mg/kg in subareas
1,3,4, and 6 with highest levels (1,400 mg/kg) in the Rouge
River. In 1985, ammonia levels were below 500 mg/kg in all sub-
areas except subarea 1, where 900 mg/kg was found in Conners
Creek. Dredging guidelines for ammonia were exceeded along the
Michigan shore. Ammonia concentrations were not determined for
sediments from the Ontario shore.
v) Sediment Bioassays
Certain Detroit River depositional zone sediments have demonstra-
ted a range of toxicity to various forms of aquatic life, and
some Detroit River sediments have been tentatively classified as
hazardous waste. Figure IX-12 shows the status of macrobenthic
communities along the Detroit River. Bacterial bioluminescence
(Phosphobacterium phosphoreum) assays (MicrotoxR) conducted on
Detroit River sediment porewater provided dose-response relation-
ships with degree of toxicity inferred by a decrease in light
emission. Figure IX-13 indicates that localized western near-
shore Trenton Channel stations caused a 50% reduction in bio-
luminescence with less than 100% porewater while other stations
elicited lesser responses and 30 percent of the stations were
nonresponsive (78).
-------
487
Status of Macrobenthos
Communities
Normal
Intermediate
Severely
Impacted
kilometers
i./4/ff £/?/£
FIGURE IX-12. Macrobenthos distribution in the Detroit River.
-------
488
n
CO
DETROIT
RIVER
1986
— 42°15'
— 42°10'
— 42°05'
n
CO
Microtox
Toxicity
Statement
Great
Moderate
Slight
None
42°15'
42°051 —
ORtAT LAKES INFORMATION SYSTEM
DEPARTMENT Of NATURAL RESOURCES
LAND ATJO WATER MANAGEMENT DIVISION
, ID
O
n
CO
O
O
FIGURE IX-13. Detroit River sediments porewater Microtox toxicity.
-------
489
Mutagenic potential of sediment extracts were measured by the
bacterial Salmonella/microsome assay (Ames test). Some mutagen-
icity was noted at 28 of 30 Detroit River stations, with the most
strongly mutagenic sediments from the Trenton Channel (Figure IX-
14). Moderately mutagenic sediments were primarily concentrated
in the lower river near Lake Erie (44).
Bacterial and phytoplankton bioassays were conducted on control
sediments and water along the west end of Fighting Island and the
southern end of Grosse lie, measuring changes in the rate of food
uptake in bacteria and phytoplankton photosynthesis. Bacterial
uptake rates were suppressed by control and contaminated sedi-
ments when sediments exceeded 12 to 1,200 ppm of suspended sol-
ids. At 120 ppm suspended solids, control sediments inhibited
uptake by 50% whereas contaminated Trenton Channel sediments
inhibited uptake by 75%. The impact of sediments on phytoplank^
ton was similar to bacteria, but less accentuated (36).
Daphnia pulicaria feeding was generally inhibited 50 to 75% by
Detroit River elutriate with an approximately three fold decrease
in ingestion rate at station 34, downstream of McLouth Steel near
Trenton. Slight feeding suppression of the control at stations
83 (along the west shore of Fighting Island) and 53 (at the
southern tip of Grosse lie) were reported at high elutriate con-
centrations (36) .
The acute toxicity of Detroit River sediment porewater to Daphnia
magna was demonstrated in a study where ten of the thirty sta-
tions in the Trenton Channel caused 50% mortality in a 96-hour
exposure to 50% or less concentration of porewater (78).
Ten day Chironomus tentans growth tests using whole sediments
found the greatest growth inhibition (up to 95%) along the west-
ern near-shore Trenton Channel. Growth rates for these stations
ranged from 0.02 to 0.08 mg/day, whereas reference stations and
three other stations ranged from 0.48 to 0.53 mg/day (36).
Stylodrilus was used to determine avoidance response to Detroit
River sediments. In control sediments, all worms burrowed and
remained buried with no mortality. At other stations, 70% of the
worms remained buried, but a slight incre.ase in mortality rate
was evident. At station 34, downstream of McLouth Steel near
Trenton, only 10% remained buried, with a 53% mortality (36).
Chironomus tentans respiration, undulation, turning and crawling
movements and rest responses to Detroit River sediments showed
significant differences in escape, respiration and rest respon-
ses, relative to Lake Michigan control sediments. Escape time
was higher and respiration and rest time were lower at these
stations compared to the Lake Michigan sediments (36).
-------
490
Michigan
Mutaoenic Potential
Strongly mutagenic
Weakly mutagenic
Non mutagenic
ri
FIGURE IX-14. Mutagenic potential of Detroit River sediments (Ames test).
-------
491
Feeding rates of larval channel catfish exposed to Detroit River
contaminated and control sediments and sediment porewater indi-
cate the greatest inhibition of feeding rates occurred from ex-
posure to Trenton Channel sediments. There were no differences
in feeding rates when porewater and water column assays were
completed on Trenton Channel stations (36).
Late-eyed stage rainbow trout eggs were injected with serial
dilutions of Detroit River sediment extracts; all sediment ex-
tracts increased embryo mortality two to three fold relative to
the solvent carrier control. Incubated eggs and fry were moni-
tored but increased mortality was not evident in the early sac
fry stages. One year after injection, 3% of the survivors'
livers exposed to Monguagon Creek sediment extract at 100 ug/egg
had liver neoplasms (44).
Schloesser et al. (40) demonstrated an inverse relationship bet-
ween Hexagenia abundance and visible oil in Detroit River sedi-
ments. Edsall et al. (41) found Hexagenia averaging 2,086 mg dry
wt/m-Vyear at three locations where sediment contaminants did not
exceed dredging guidelines, but only 364 mg dry wt/m^/year where
as many as seven contaminants exceeded these guidelines. Both
studies indicate that sediment contaminants had notable negative
impacts on the benthic community.
In summary, sediments of the Detroit River were found to be
severely impacted by a variety of compounds, including PCBs, HCB,
PAHs, total phenols, total cyanide, oil and grease, total phos-
phorus, ammonia and metals (total mercury, total lead, total ar-
senic, total cadmium, total copper, total zinc, total chromium,
total nickel, total manganese, total iron). In addition, some
non-UGLCCS parameters were also found in sediments (pesticides,
phthalate esters and volatile organic compounds). Several tribu-
taries appear to be sources of many of these contaminants. Toxic
effects of the sediments and sediment porewater on benthic biota
were also noted by a variety of toxicity tests.
-------
492
B. SPECIFIC CONCERNS
The specific chemicals which are impacting the Detroit River
ecosystem, as determined in this study, and other concerns, are
identified in this section. They are summarized in Table IX-3.
1. Conventional Pollutants
In the past, severe oxygen depletion in the Lake Erie hypolimnion
was associated with excessive inputs of phosphorus, and correct-
ive action was undertaken by most jurisdictions to reduce phos-
phorus loadings. Since the Detroit River is the major tributary
to Lake Erie, all phosphorus loadings from the Detroit River are
considered important. Concentrations of total phosphorus in the
Detroit River have steadily decreased since the late 1960s and
are presently below 20 ug/L. Tributary concentrations, however,
still currently exceed ambient water quality guidelines.
Chloride concentrations in the Detroit River water were relative-
ly constant, and not excessive; however, one industry which was
found to be discharging high levels of chlorides (i.e., General
Chemical) was not represented by the water quality survey. High
chloride levels may encourage the growth of halophilic phyto-
plankton in the Great Lakes which could cause a shift in the
phytoplankton community and upper trophic levels.
Fecal coliform bacteria are of concern because fecal coliform
bacteria standards and criteria are routinely violated on both
sides of the river. Beaches along both shores have been closed
or not developed because of this continuing problem. Although
not demonstrated in this study, ammonia is also problematic,
since calculated levels of nonionized ammonia have periodically
exceeded the chronic criteria for coldwater fisheries (0.02 mg/L)
along the western Detroit River shoreline.
Phosphorus and ammonia concentrations in sediments exceeded
dredging guidelines at a number of locations in the Detroit River
and in some tributaries.
2. Organic Contaminants
Polychlorinated biphenyl (PCB) concentrations in the Detroit
River were found at concentrations exceeding guideline levels.
Although the levels are below acutely toxic concentrations, high
persistence and bioaccumulative properties of PCBs may (and in
fact has) resulted in bioaccumulation of PCBs in aquatic organ-
isms. Similar findings are made for several organochlorine
compounds, including hexachlorobenzene, dieldrin, heptachlor,
heptachlor epoxide, chlordane and endosulfan. The effects of
these contaminants may not be found in the Detroit River itself
-------
493
TABLE TX-3
Specific concerns and use impelrnents in the Detroit River, 1 988
Impairment or Concern Causes of Impairment Location where Impairment Probable Sources of Contaminants
or Concern or Concern is Found Causing Impairment or Concern
Consumption advisory for Body burdens of PCB or Carp-whole river; other Upstream of Detroit River Watershed;
carp, rock bass, walleye Mercury I Other organo- species-1imited river sec- point anrt nonpoint sources; food chain
and freshwater drum chlorine compounds and tions for some larger sizes
aome pesticides mav be
present, but have no
criteria
Changes in fi sh species Toxic concentrations of Prinarilv U.S. shoreline Point and non-point sources; food
compos1tion and fish organic compounds, heaw downstream of Rouge River, chain; habitat changes
toxicity metals and possibly am- Detroit River Tributaries
Monia in water and
sediments
Tumors and deformities PNA's, PCR and other Primarily lower river and Point and nonpoint sources; food
in fish organochlorine contami- downstream of the Rouge River chain
nants, perhaps heavy
metals
Elevated body burdens of PCB, HCB and other or- Primarily lower Detroit River Upstream Detroit River watershed;
organic contaminants in ganochlorine compounds point and nonpoint souces; food
waterfowl and forage fi sh chain
Elevated concentrations PCB, HCB and other or- U,S, Detroit River shoreline Point and nonpoint sources; food
of organic contaminants ganochlorine compounds and Fighting Island chain
in bird livers and egg*
Loss of f i sh and wi |dIife Ru I khenrii ng , t"i I I i ng , Pn mnri I v «1 ong r he U.S. Poi nt and nonpoi nt sources; dredg i ng/
habi tat dr*»d«i ng nnvi eat ion Detroi t River Shore 11 ne and f 11 1 ing
channe I s ; organ1 rs find i n navi gat ion rhanne t s
heavy met A 1«;; contmerr i a I
and t ndtistn al rteve 1 opment
Loss of aquatic animals Contaminants ' urbani- All U.S. Shore1ine and most Point and nonpoint sources; urbaniza-
zation' habitat loss of the Canadian shoreline tion? dredging/filling
Phvtoplankton popui ation Chi orides and heavy Chlondes-J ower Detroit River Industrial and municipal discharges
changes metals esppci«1 Iv Canadian shore1ine
Henvv m«ta M-pn man I y U.S.
shoreline and Trenton Channel
Zooplankton toxicity Organic compounds and Where sediments are heavily Industrial and municipal discharges
heavv metals contaminated) in Trenton
Channe1 espec i a11y
Benthic macroinverta- Heavv metals and organic From Zug Island downstream Point and nonpoint sources
brate commumtv changes compound contamination in along the U.S. shoreline,
sediments and water, also Trenton Channel and Windsor
nutrient enrichment and shoreline
oil and grease
Aesthetic degradation- Nutrients, BODS, oil and Near shore, U.S. Michigan Primarily municipal discharges and
eutrophication grease and organic and side, downstream of CSO's CSO's
CUl. rupn luai 1UI1 5 IWBC aiiu urgaiui; aiiu »*uc , UWWIIB v i c-«n w L \*au a
tributaries
Sediment contamination Organic and heavv metals Primarilv depositional zones
and potential loading and phenols near the U.S. shore, Trenton
to the water column Channel, lower Detroit River
and local i zed spots
Contaminated Groundwater Organics, heavy metals At waste disposal sites
les Point and nonpoint sources; CSO's
.on
Contaminated Groundwater Organics, heavy metals At waste disposal sites Primarily local industrial waste or
1 oad ings pheno I s , ot he r 7 sp 1 1 1 s
Loss of total bodv con- Fecal coliform bacteria Tributaries and both shores CSO's, stormwater, municipal WWTP's
tact recreation of the entire length of the and septic tank leachate reaching
Detroit River into Lake Erie tributaries
Added cost of treatment Excessive concentrations Primari lv along the U.S. Industrial and municipal discharges
to industry and agricul- of contaminants from shoreline downstream of the
ture other dischargers leaves Rouge River
little asm mi lative
capacity for
other dischargers
Potential con tarn mat! on Primari ly organic At publ ic drinking water Upstream industrial discharges
of public potable water chemicals and spills of supply intakes throughout waste disposal sites, spills from
supplv materials the Detroit River ships, and WWTP bypasses and upsets
-------
494
but in Lake Erie, particularly its Western Basin. Significant
concentrations of polynuclear aromatic hydrocarbons (PAHs) enter
the Detroit River at and near the Rouge River mouth. There is no
water quality guideline for PAHs for aquatic life; however, many
of these compounds are known or suspected animal or human car-
cinogens.
Fine-grained sediments in the river are excessively contaminated
by a variety of organic contaminants. Several areas along the
Michigan shore contain excessive PCB concentrations. Organo-
chlorine contaminants other than PCBs are also found in most
Detroit River and tributary sediments. DDT and its metabolites,
dieldrin, methoxychlor, chlordane, trifluralin, hexachlorocyclo-
hexane and hexachlorobenzene are present. Polynuclear aromatic
hydrocarbons (PAHs) have been found at high concentrations in
Detroit River sediments. Excessive phenols were present in sedi-
ments of the Trenton Channel. High concentrations of phthalates
were present in many sediment samples from Detroit River tribu-
taries, particularly Conners Creek and the Rouge River. Exces-
sive concentrations of oil and grease are present in many Detroit
River depositional zone sediments, and have degraded benthic
macroinvertebrate communities (24).
Fish from several stations in the lower Detroit River had ele-
vated levels of certain, organic chemicals. PCB concentrations
exceed consumption guideline levels in the edible portion of
Detroit River carp. Consequently, the Michigan Department of
Public Health has issued a consumption advisory for these fish.
Several Detroit River fish species exceed the GLWQA objective of
0.1 mg/kg (wet weight) total PCBs in whole fish tissue. OMOE has
also issued a fish consumption advisory for Detroit River carp
because of elevated body burdens of PCBs.
Waterfowl contain elevated PCB levels and other persistent or-
ganic chemicals. There are no existing criteria for a consumption
advisory to protect children and women of child-bearing age from
the potential effects resulting from consumption of these birds.
Herring gull eggs collected from Fighting Island in 1985 and 1986
contained high concentrations of PCBs and PAHs, and contained
several other organochlorine pesticides.
Native and caged Detroit River clams showed increased levels of
PCBs, PAHs and several organochlorine pesticides. Some PAHs
found in Detroit River sediments are probable human carcinogens,
and are thought to be responsible for some liver, lip and dermal
tumors in fish.
3. Metals
Concentrations of metals measured in water during the study were
generally all below the ambient water quality guideline, with the
-------
495
exception of mercury, which exceeded Michigan's Rule 57(2) allow-
able levels throughout the river. Generally, water in the
Trenton Channel was of a poorer quality than other portions of
the river. During the 1986 Detroit River System Balance Study,
some localized areas exceeded water quality guidelines for iron
(GLWQA specific objective) cadmium, lead and mercury (Michigan's
Rule 57(2) allowable level). Water quality in the Little River,
Rouge River, Turkey Creek, the Canard River and Ecorse River is
impaired with respect to certain metals.
Heavy metal contamination of Detroit River sediments is found in
most depositional areas, with concentrations of many metals ex-
ceeding guidelines. Lead, cadmium, copper and zinc levels are
significantly elevated in the Rouge River and Turkey Creek and in
Detroit River sediments downstream of their confluences. High
levels of chromium and nickel are present in the Little River.
Manganese and especially iron are strongly elevated in Trenton
Channel sediments and other Michigan nearshore and sedimentary
zones.
Overall, certain Detroit River sediments are severely degraded by
heavy metals, especially in the Trenton Channel. This contamina-
tion may reduce or eliminate the viability of Detroit River and
Lake Erie sediments as substrate for benthic organisms. Desorp-
tion of contaminants and re-solubilization through chemical and
biological processes make an unknown portion of these chemicals
available to higher aquatic organisms.
OMOE has issued a fish consumption advisory on several fish
species because mercury concentrations exceed 0.5 mg/kg in the
edible portion of the larger sizes of these fish. Native and
caged Detroit River clams showed increased levels of several
metals, particularly lead and cadmium.
4. Habitat Alterations
Eighty-five percent of the wetlands and littoral zones along the
Michigan Detroit River shoreline have been eliminated by filling,
dredging and bulkheading. Aquatic plants which live only in the
littoral zone provide food, substrate, cover and nursery produc-
tion for aquatic organisms, and drive the production and energy
flow through the aquatic ecosystem. Loss* of the littoral zone
results in the loss of large segments of the upper trophic lev-
els, including fish. Habitat loss was the major factor, along
with pollution and overfishing, in the demise of the Detroit
River commercial fishery around the turn of the century. Large
areas of shallow water and marshes associated with tributaries
are still found on the Ontario shore, below Fighting Island.
Seventy percent of the remaining littoral zone is occupied by
submerged plants, macrophytes and other wetland plants.
-------
496
In the Detroit River, upstream of Zug Island, the benthic com-
munity is diverse and dominated by pollution intolerant organ-
isms, except along the Windsor shoreline. Adjacent to Zug Island
the community is severely impacted, and downstream, especially in
the Trenton Channel, the benthos is dominated by pollution toler-
ant oligochaetes.
Overall, aquatic biota, especially benthos, show detrimental
responses to contamination of Detroit River sediments with or-
ganic and inorganic substances, particularly in the lower river
and in the Trenton Channel. Laboratory tests with sediments and
sediment extracts indicate higher toxicity and increased mutagen-
icity on a variety of native species. Fish species diversity and
fecundity may also be negatively affected in some areas.
-------
497
C. SOURCES
This section discusses contaminant inputs from point and non-
point sources in the Detroit River which were analyzed between
1984 and 1987.
1. Point Sources
Introduction, Qualifications and Criteria
During 1985, 1986 and 1987 the Michigan Department of Natural
Resources (MDNR), OMOE, U.S.EPA and Environment Canada collec-
tively monitored flow and effluent quality of major direct and
indirect point source dischargers to the Detroit River (direct
sources are those which discharge directly to the river and in-
direct sources discharge to the river via tributaries or drains).
Nine municipal treatment plants and 20 industrial facilities were
sampled over a 24 hour period (Michigan sources) or 3 to 6 days
(Ontario Sources) during 1985 and 1986. Composite samples were
analyzed for conventional pollutants, metals and trace organics,
including the list of contaminants chosen for the UGLCC Study
(Chapter I, Table 1-1). Table IX-4 presents the industries sur-
veyed and the parameters which are regulated in their effluent.
Table IX-5 presents the municipal facilities and their regulated
parameters. Figures IX-15 and 16 show the locations of these,
and other, Industrial and municipal facilities along the Detroit
River.
Shortcomings limit the inferences that can be drawn from the
survey, including the small data base, differences in survey
timing, and differences in sampling and analytical methods. The
U.S. surveys were performed in May, and July through September,
1986, while the Ontario data were collected between October and
December, 1985. The U.S. composited four grab samples (1 every 6
hrs), while Ontario samples were collected by automatic composite
samplers (1 portion every 15 min).
Differences in detection limits further hinder comparisons. The
U.S. generally used lower detection limits than did Canada, al-
lowing calculated loadings from Michigan facilities with no cor-
responding loadings from Ontario facilities for some parameters
(e.g., OCS and HCB). Consequently, the percent of the total
point source loadings to the Detroit River for some parameters
(depending on corresponding flow volumes) may be skewed towards
Michigan dischargers.
Flows
There were a total of 75 known point sources discharging 9,233 x
to the Detroit River basin in 1986. Three Detroit