vvEPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth MN 55804
EPA-600/3 78-067
July 1978
Research and Development
Influence of Turbidity
on Fish Abundance
in Western Lake
Superior
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. 'Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-78-067
July 1978
INFLUENCE OF TURBIDITY ON FISH ABUNDANCE
IN WESTERN LAKE SUPERIOR
by
William A. Swenson
Department of Biology and
Center for Lake Superior Environmental Studies
University of Wisconsin-Superior
Superior, Wisconsin 54880
Grant No. R802455
Project Officer
J. Howard McCormick
Research Branch
Environmental Research Laboratory-Duluth
Duluth, Minnesota 55804
ENVIRONMENTAL RESEARCH LABORATORY-DULUTH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
DULUTH, MINNESOTA 55804
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory-
Duluth, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or recommenda-
tion for use.
ii
-------
FOREWORD
From the standpoint of the area affected and the tonnage of material in
the water, turbidity from silt and clay erosion is probably one of the most
significant water pollution problems in the United States. Because of the
multiple inputs of suspended material producing turbidity from an enormous
number of man's activities, the control of turbid water is an extremely
expensive and elusive matter. While there is general agreement among
biologists that turbid water has adverse effects on aquatic communities, there
is little information on which to prove such effects. On the South Shore of
Lake Superior in the western end of the lake, there is a band of red clay which
erodes as a result of wave action on the shore line and man's activities on
the land. This red clay gives an aesthetically displeasing appearance to the
otherwise clearwater lake and much public concern has been expressed about it.
Various committees and agencies have tried to find solutions to this problem.
The high cost of control measures has frustrated the recommendations of such
groups.
The study reported here was initiated to see what adverse effects if any
the suspended material might have on the aquatic community in order to provide
justification for the cost of control measures. While strongly contested by
some of the experts, this report proposes a fascinating relationship between
turbidity caused by red clay, smelt and the herring population in Lake Superior.
Turbudity was thought to favor herring abundance through the input of nutrients
to support plankton food supplies and to reduce predation by lake trout which
avoid turbid water. After the introduction of smelt into Lake Superior, there
seems to be an inverse relationship between the numbers of smelt and the
abundance of herring. Evidence has been gathered to suggest that smelt may be
more predacious on the young herring in turbid water than in clear water, and
that turbidity is now having an adverse effect on the herring population
through enhanced smelt predation. While not conclusively proven, such a
relationship offers insight into the subtle ways in which turbidity could
have a major effect on the species considered most beneficial to man.
Donald I. Mount, Ph.D.
Director
Environmental Research Laboratory-Duluth
ill
-------
ABSTRACT
This research project was developed to improve understanding of the in-
fluence of turbidity on fish populations and the mechanism through which its
effects are induced.
Field and laboratory studies emphasized measurement of behavioral re-
sponse of fish and resulting changes in fish species interrelationships in
western Lake Superior. Direct effects of red clay turbidity on survival and
growth of larval lake herring (Coregonus artedii) were also measured.
Field measurements demonstrated that light penetration in western Lake
Superior is reduced significantly even at very low levels of red clay tur-
bidity. Zooplankton and fish abundance and distribution were influenced by
turbidity. Zooplankton abundance and distribution was highest near the sur-
face in red clay plumes. Smelt (Osmerus mordax) move into the upper 12 m of
water in response to turbidity where their predation on larval fish increases.
Predation by smelt on larval lake herring was identified as a factor
contributing to the decline of the formerly abundant western Lake Superior
lake herring population and the commercial fishery which depended upon it.
Walleye (Stizostedion vitreum vitreum) and lake trout (Salvelinus
namaycush) demonstrated opposite responses to turbidity. Walleye concen-
trated in turbid water where food availability was apparently greater. Lake
trout showed partial avoidance to turbidity in the lake and in laboratory
turbidity gradients.
This report was submitted in fulfillment of Grant R-802455 by the Uni-
versity of Wisconsin-Superior, Center for Lake Superior Environmental Studies
under the sponsorship of the U.S. Environmental Protection Agency. This
report covers the period from 4 May 1973 to 31 October 1976, and work was
completed 12 February 1978.
IV
-------
TABLE OF CONTENTS
Foreword Hi
Abstract Iv
Acknowledgments , vii
Introduction i
Conclusions ..... 4
Recommendations , 6
Methods
Field Studies
Western Lake Superior Fish Sampling 7
Analysis of Catch Records 11
Black Bay, Ontario, Fish Sampling 12
Zooplankton Sampling 12
Depth, Temperature, Turbidity and Light Measurements 13
Laboratory Studies
Turbidity Gradients 15
Larval Herring Bioassay ..... 19
Predation Studies 19
Results
Turbidity and Temperature in Western Lake Superior 21
Light Intensity in Western Lake Superior 21
Zooplankton Abundance and Distribution 25
Fish Abundance and Turbidity
Spacial Variations in Fish Abundance 30
Temporal Variations in Fish Abundance 38
Influence of Turbidity on Smelt Populations
Smelt Growth and Turbidity 38
Smelt Distribution and Water Temperature 38
Smelt Distribution and Turbidity 42
Smelt Food and Predation 46
-------
Influence of Turbidity on Growth, Survival and Distribution
of Larval Herring 53
Influence of Turbidity on Walleye
Walleye Abundance and Distribution 53
Response of Walleye to Laboratory Turbidity Gradients .... 55
Walleye Feeding 55
Influence of Turbidity on Lake Trout
Lake Trout Distribution in Lake Superior 61
Response of Lake Trout to Laboratory Turbidity Gradients ... 61
Influence of Turbidity on Other Fish Species 61
Discussion 65
References 69
Appendix Tables 73
Appendix Figure 83
vi
-------
ACKNOWLEDGMENTS
Mr. James Jonasen assisted in all phases of the field and laboratory
work. Ms. Mary Balcer performed the zooplankton work, shared responsibil-
ities for the laboratory studies including data analysis, and developed the
graphic illustrations used in this report. Mr. Wayne Schaefer performed
the computer analysis of field data. Mr. Douglas Standen assisted in the
field sampling and feeding studies. The special performance of these stu-
dents is gratefully acknowledged.
The project officer, J. Howard McCormick, actively participated in
some phases of the field work and took primary responsibility for the lab-
ratory predation studies. I gratefully acknowledge his exceptional effort
and continued support throughout the project. Equipment and laboratory
space were provided through Mr. Bernard Jones of the EPA Environmental
Research Laboratory-Duluth. Robert Drummond, Richard Carlson, and Walter
Dawson of the laboratory staff provided valuable technical assistance to
the laboratory phases of the study.
vix
-------
INTRODUCTION
The influence of turbidity on fish in lakes has not been critically
measured. Field studies on the problem are inadequate and contradictory.
Turbidity in Lake Erie has been cited as the cause for both fishery decline
(Langlois 1941)and high fish production (Doan 1941, Van Oosten 1945).
Studies on post larval stages of many fish species show levels of turbidity
greatly exceeding those found in lakes are required to influence survival
or growth under controlled laboratory conditions (Cordone and Kelley 1961;
Herbert and Merkens 1961). However, because suspended solids generally
alter nutrient and light conditions in natural systems, behavioral responses
of fish are implied. Effects of turbidity on distribution, feeding or other
aspects of behavior may be significant to the success of responding species
populations within the community and may indirectly influence nonreactive
species by altering relationships with responsive populations.
This study was undertaken to measure the influence of turbidity on dis-
tribution, feeding and interrelationships between major fish populations in
western. Lake Superior and to identify the mechanisms through which effects
of turbidity on fish are induced. Western Lake Superior provides an excel-
lent environment for measuring the general effects of low levels of turbidity
on fish populations because turbidity varies spatially and temporally in
what is generally a stable system with respect to other physical conditions,
such as oxygen. Decline of lake herring (Coregonus artedii) and increased
abundance of rainbow smelt (Osiaerus mordajc) in western Lake Superior has
resulted in significant reductions in commercial fish production. Relation-
ships between changes in fish stocks and turbidity were studied to provide
information essential to fishery management.
Turbidity in western Lake Superior results primarily from erosion of
glacial-lacustrine red clays deposited during an earlier high water stage
of the lake. These unconsolidated sediments are most prominent in northern
Wisconsin and are thickest at the western end of the lake near Superior,
Wisconsin. The sediments occur in a continuous zone from Superior, eastward
along 75 km of shoreline to Port Wing, Wisconsin, and cover approximately
3,600 km2 (Red-Clay Inter-Agency Committee, 1972; Figure 1).
Startz et^ al. (1976) identified major sources of clay and distribution
of turbidity using Earth Resources and Technology Satellite (ERTS) images,
settling rates and measurements of turbidity in the drainage basin. They
found that erosion of shoreline bluffs by storm wave activity is the prin-
cipal source of turbidity. Approximately 2.3 x 106 metric tons are eroded
from the Douglas County shoreline annually. Douglas County includes one-
half of the shoreline characterized by exposed clay bluffs. Approximately
-------
N
Duluth
Superior
0 10
Kilometers
Figure 1. Map of western Lake Superior identifying the red clay soil formation (barred) and
field sampling stations (numbered).
-------
5.6 x 10 metric tons are resuspended from the lake bottom, and 3.2 x
metric tons are added by stream erosion annually. Average turbidity for
the 1972-1975 study period was estimated at less than 1 Formazin_Turbidity
Unit (FTU) in mid-April (Sydor 1975). Turbidity rose to approximately 5
FTU from mid-April through May in association with ice breakup, then de-
cline slowly through the summer (average of 3.5 FTU). In November and
December, turbidity averaged 8 FTU as a result of autumn storms (Sydor 1975)
Turbidity was higher in red clay plumes but rarely exceeded 50 FTU except
in the nearshore wave surge zone.
Bahnick (1977) estimated that slightly over 300 metric tons of ortho-
phosphate are released annually to Lake Superior water from suspended red
clay soils. Red clay was found to contribute 20.7 x 10^ metric tons of
dissolved solids, 19.7 x 103 metric tons of alkalinity, 14.4 x 103 metric
tons of silica, 3.5 x 1Q3 metric tons of potassium and smaller quantities
of various metals including iron (64 metric tons), aluminum (76 metric
tons), zinc (<8 metric tons) and copper (3 metric tons) (Bahnick 1975).
-------
CONCLUSIONS
Red clay turbidity does not directly influence survival of even the
most sensitive life stages or txsh in western Lake Superior.Red clay does
cause dramatic changes in the quality and intensity of light even at low
turbidity levels. Behavioral regprmsps of fish to turbidity and associated
changes in light have a major influence on important fish populations.
C T.^]ff> trout (Salvelinus namaycush) demonstrated some avoidgnJcS^to turbid
water in Lake Siiper-i^i- *nA -j^ ^ aboriifinry fiirtrlffify CTrfHIpj^"""'" contrast to
jallevp fSi-.-iy.o^t-.p.riinn vitreum vitreum) which prefer turbidwatlg?' Turbidity
results in increased walleye production injwestern Lake Superior by reducing
light intensity which directly enhances their feeding success ancL by causing
rainbow smelt to become pelagic, increasing walleye food avaifability. Wall-
eye fed almost exclusively on^smelt in Lake Superior and have been found to
require low light intensities and dense pelagic prey populations in order to
maintain high food consumption rates (Swenson 1977).
Rainbow smelt apparently became abundant in western Lake Superior dur-
ing the mid 1940's when commercial fishermen reportedly captured quantities
in large mesh gill nets set for other species.-'- Commercial fishing for
rainbow smelt was initiated in the early 1950's. During the period of in-
creasing smelt abundance, which commercial catches suggest continued into
the late 1950's, the valuable lake herring population underwent a sharp
decline. High plankton densities identified by this study and increasing
herring growth rates (Lake Superior Herring Subcommittee 1973) suggest that
the herring decline did not result from food competition with smelt. How-
ever, the results of this study show that juvenile and adult smelt move into
the upper 12 m of water under turbid conditions, the zone formerly occupied
by larval herring. Cannibalism by pelagic smelt was found to induce high
mortality. Smelt were also found to prey on larval herring in the labora-
tory and in Black Bay, Ontario, where both species are presently abundant.
Turbidity apparently contributed indirectly to the decline of lake herring
in western Lake Superior when smelt became a part of the community. Effects
of turbidity were induced through its influence on the distribution and
feeding behavior of smelt.
Personal communication from Mr. Stanley Sivertson, President, Sivert-
son Fishery, Duluth, Minnesota, and Mr. George King, Lake Superior Manage-
ment Coordinator, Wisconsin Department of Natural Resources.
-------
Prior to introduction of smelt, the findings indicate turbidity pro-
moted lake herring production by stimulating high zooplankton densities in
near surface waters where larval herring concentrate. If lake trout pre-
dation influenced herring survival, low abundance of lake trout in turbid
water zones may have been a factor resulting in the former high abundance
of herring in western Lake Superior. Addition of smelt to the community
induced negative effects which resulted in reduction in the herring stock
and significant economic loss to the commercial fishery.
Behavioral response of fish to the reduced light intensity associate^
with turbidity appears to represent the primary mechanism through which
turbidity intiuences individual species populationst interspecific rela-
tionships, tish production and economic value of fish in lakes. This study
o± Lake Superior stocks shows that the effects of ^turbidity are dependent
upon the fish species complex and that small changes in species composition
will greaCly alter the influence of turbidity.
-------
RECOMMENDATIONS
Erosion of red clay in western Lake Superior represents a natural phe-
nomenon accelerated by man's activities. Although partial control is tech-
nically feasible and is important with respect to several uses of the lake,
it would be unrealistic to reduce turbidity to a level which would affect
the Lake Superior fish community. Rather, the fish community should be
managed directly to minimize the negative impact and maximize the beneficial
aspects of the problem. Commercial exploitation of smelt should be stimu-
lated by appropriate management agencies to control smelt population levels,
increase economic returns and stimulate increased lake herring abundance.
Success of ongoing planting programs to reestablish commercial concentra-
tions of herring might be improved by stocking during periods of low tur-
bidity, by stocking areas of reduced smelt density or by growing herring in
the hatchery to a size less vulnerable to smelt predation. Walleye could
be managed to reduce survival of young-of-the-year and juvenile smelt which
concentrate in the near shore zone.
Smelt were introduced into the Great Lakes with little knowledge of
the potential negative effects which have resulted through their interaction
with lake herring. Although western Lake Superior herring apparently have
adapted to red clay turbidity, this study indicates that smelt reduced the
ability of herring to survive in the turbid water zone of western Lake
Superior. The results demonstrate the need for strict control measures to
curtail future species introductions.
Because turbidity induces significant effects on resource populations
through its influence on behavior and species interrelationships, it is
recommended that future research go beyond toxicology studies and concen-
trate on the analysis of behavior and community dynamics.
-------
METHODS
FIELD STUDIES
Western Lake Superior Fish Sampling
Information on western Lake Superior fish populations was collected by
bottom trawl, midwater trawl, seine and hydroacoustical techniques from six
stations in western Lake Superior (Figure 1). Stations 2 and 4 are charac-
terized by sandy, clay and organic substrates conducive to bottom trawling
and were sampled intensively. Sampling zones included 3-5 km parallel to
shore and up to 12 km off shore. Stations 1, 5 and 6 to the northeast and
southeast respectively (Figure 1) were characterized by rocky bottom in-
shore which prevented or restricted sampling by bottom trawl. Turbidity is
generally higher at Stations 4 and 5 due to erosion of red clay deposits
along the shoreline and inflow of turbid waters from the Nemadji River.
Stations 1 and 6 are characterized by low turbidity. High variability in
turbidity level occurred at Stations 2 through 5 which facilitated measure-
ment of behavioral response of fish.
During 1973 and May-August 1974, trawling was restricted to water less
than 15 m deep. During September through November 1974 and June through
October 1975-1976, trawling was extended to depths of 40 m. On most sam-
pling days during the May 1973 through August 1974 sampling period, a 3 mm
bar mesh bag seine was used to sample the 0.5-1.2 m depth zone (Table 1).
Bottom trawling was conducted at depths of 1.8-4.7, 4.8-7.6 and 7.7-15 m
with a 7.6 m headrope semiballoon trawl constructed of 18 mm bar mesh with
a 6 mm mesh cod liner. Daylight seining and bottom trawling were followed
by night trawling with a 6.1 m headrope level trawl constructed of 6 mm
mesh with a 691ym cod liner. The smaller net was trawled at the surface
and on the bottom in the 1.8-4.7 m depth zone; at the surface, 3 m from the
surface and at the bottom in the 4.8-7.6 m depth zone; and at the surface,
3 and 6 m from the surface and on the bottom in the 7.7-15 m depth zone
(Table 1).
After August 1974, sampling was conducted primarily during daylight
hours at depths exceeding 20 m using a 9.5 m headrope semiballoon trawl
constructed of 76 mm stretch mesh with a 6 mm cod liner (Table 1). During
most days bottom trawling with the 9.5 m trawl was followed by midwater
trawling (6.1 m net) near the surface and at 6, 12 and 18 m from the sur-
face (Table 1).
Bottom trawling with the 6.1 m net was conducted at Station 6 in Little
Sand Bay, Apostle Islands ational Lakeshore, during June, July and August
-------
TABLE 1. FIELD SAMPLING EFFORT
Month
May
June
July
Aug.
Sept.
Oct.
Year
1973
1975
1976
1973
1974
1975
1976
1973
1974
1975
1976
1973
1974
1975
1976
1973
1974
1975
1976
1973
1974
1975
1976
No.
Days
2
1
5
7
7
5
5
7
11
12
6
8
5
4
3
7
5
1
—
8
6
3
1
Station
No.
4-5
4
2-4
2-4
1-2-4-5
1-2-4-6
2-4
2-4
1-2-4-5
2-4-6
1-2-4
1-2-4
1-2-4-5
2-4-6
4
2-4-5
1-2-4-5
4
—
2-4
2-4-5
2-4
4
Fishing
Effort
Trawl Size
Seine
8
—
—
18
10
4
—
16
10
4
—
12
8
2
—
12
2
—
—
16
2
—
—
6.1m
—
—
—
68
48
31
9
46
53
39
16
57
38
27
6
44
38
7
—
53
16
—
—
7.6m
3
—
16
18
14
—
—
17
15
—
—
28
9
—
—
21
9
—
—
18
—
—
—
9.5 m
—
1
7
—
—
22
11
—
—
33
7
—
—
6
6
—
—
2
—
—
15
2
4
Plankton
Samples
—
—
49
41
48
—
—
37
53
—
—
57
38
—
—
44
38
—
—
54
16
30
—
Nov.
1974
Totals
121
125
596
172
116
505
-------
o
1975. A total of 20 standard hauls were made during 3 days at depths of
1.8-4.7, 4.8-7.6 and 7.7-15 m.
Standard 10 min tows of the 7.6 m bottom trawl averaged 0.77 km and
filtered an estimated 3,428 m . Ten minute tows of the 9.5 ra trawl aver-
aged 0.68 km and filtered an estimated 3,800 m^ volume. Standard 10 min
tows with the 6.1 m trawl averaged 0.81 km and filtered an estimated 1,800
m . Trawling distance and speed were measured for each haul by a meter
mounted in a tow (Figure 2). Estimates of water volumes were derived from
the distance measurements and SCUBA diver measurements of trawl openings.
Fish were counted in the field. Separate counts were made of adult
and immature individuals of most species. Representative samples of smelt,
walleye, burbot (Lota lota) and lake trout were measured in the field.
Scale and stomach samples of larger individuals were collected for analysis
in the laboratory. Stomach samples and whole fish were preserved in 10%
formalin.
Length frequency distribution and age determination from scales were
used in estimating abundance of Age 0, I, II and older smelt in trawl
catches. Smelt scales mounted in a drop of water were magnified SOX and
aged following the criteria for annulus identification of McKenzie (1958)
and Bailey (1964). Stomachs from 858 smelt and 269 walleye were analyzed
in describing their diets. Invertebrates and fish found in the stomach
samples were identified using keys by Brooks (1957), Eddy and Hodson (1961)
and Edmondson (1959) and by comparison to larval fish and plankton samples
collected from the lake.
Midwater trawling was conducted at night to reduce net avoidance.
Occurrence of older smelt in midwater trawl catches suggests the gear was
partially successful in sampling larger fish. Although it is probable that
some escapement occurred even at night, it was assumed that escapement at
night was independent of turbidity and catch data could be used as unbiased
estimates of changes in relative abundance at a location. Catch data were
analyzed by regression analysis. Catch per 100 m^ or the percentage of the
total catch taken at each sampling location in a vertical column or hori-
zontal stratum were considered the dependent variable in regression models
in which turbidity and temperature at the location were independent vari-
ables.
Influence of turbidity on fish distribution during daylight hours was
interpreted from bottom trawl catches and analysis of chart recordings from
a Raytheon Model DE 731 fathometer. Fathometer chart records were made in
association with bottom trawling and during selected dates by traveling
known distances through and adjacent to red clay plumes. Temperature and
turbidity observations were recorded with the chart records. The number of
o
Sampling in Little Sand Bay was supported by National Park Service
Contract CX2000-5-0034.
-------
Figure 2. Probe tow with distance meter mounted forward. The tow and
transmissometer probe are resting upon the Hydro-Products Model 410 Br
transmissometer and Yellow Springs Instrument Model 43 temperature system
used in the study.
10
-------
"fish targets" recorded by the instrument per 1.5 min interval were counted
and averaged by 3 m depth intervals. Counts were made for 3-6 m, 6.1-9.1 m,
9.2-12.2 m and the average 3 m for all depths exceeding 12.2 m but at least
2 m from the lake bottom. The percentage of the total number of "fish tar-
gets" occurring in each stratum was calculated and made the dependent vari-
able in regression models in which turbidity was the independent variable.
Analysis of Catch Records
Variation in gill net catches obtained during 1973-1974 by the Wiscon-
sin Department of Natural Resources was used with trawl catches from this
study in defining relationships between turbidity and fish abundance. Gill
nets 10.97 m long, consisting of 7 equal length sections of mesh graduated
from 38 to 178 mm, stretch measured, were fished for approximately 20 h at
15 western Lake Superior stations (King and Swanson 1974). Gill net sta-
tions were classified as shallow turbid, shallow clear, deep turbid or deep
clear from secchi disc readings and sampling depth measurements made during
setting or lifting. Classification was also assigned according to station
location in relation to the zone of red clay erosion and turbidity. Two
shallow water turbid and two deep water turbid stations located west of Port
Wing, Wisconsin, were compared with catches from four shallow clear water
stations and seven deep clear water stations located east of Port Wing.
Commercial catch records for Minnesota District M-l (Smith et al. 1961)
were analyzed to identify the influence of turbidity in spawning streams
and in Lake Superior on abundance of smelt. Percentage deviation in mean
catch per unit of fishing effort (Hile 1962) by the Minnesota pound net
fishery, during the periods 1952 through 1976 and 1957 through 1976, were
used as indices of smelt abundance. Data for 1972 were omitted due to
unusual ice conditions which interferred with smelt spawning and influenced
fishing success in the harbor.
Precipitation and wind records from the Duluth, Minnesota, weather
station were used to develop an index of relative turbidity. Percentage
deviation from mean January through May precipitation during the period
1949 through 1972 was used as an index of turbidity in spawning streams.
Based on Sydor (1975), indices of turbidity during May through July and
May through September were developed by counting the number of storm days
occurring each month. Storm days were defined as any day, occurring in a
sequence of at least three consecutive days, in which the fastest mile of
wind exceeded 15 mph and was from the north, northeast or east. Deviations
from the average number of storm days for the 1949 through 1972 period were
used as the index of relative turbidity in the lake each year.
Correlation analysis was used to measure relationships between the
smelt abundance index and lake or stream turbidity indices three or four
years earlier. The approach is based on the assumptions that year-class
strength is determined during the first year of life, that a single factor
is significant in controlling year-class success and that catch is depen-
dent on smelt starting their third (Age II) or fourth (Age III) summer of
11
-------
life.
Black Bay, Ontario, Fish Sampling
Smelt and larval herring were collected from Black Bay during May 4
and 5, 1973. Although Black Bay represents a clear water environment,
smelt and lake herring are abundant and sampling in the bay provided an
opportunity to identify whether smelt predation will occur if herring lar-
vae are available. Samples were collected approximately one-third of the
way up the bay from the main lake and 1.6 km off shore in an area identi-
fied by local commercial fishermen as an important herring spawning ground.
Herring larvae were collected in 5 tows with 1/2 and 1 m diameter larval
nets with 530ym and 750ym mesh. Depth of sampling was determined by warp
length and angle. Smelt were collected during daylight hours with 5 tows
of a 5 m headrope, 25 mm stretch mesh otter trawl. Night samples were col-
lected with two 15 x 2.7 m, 13 and 16 mm stretch mesh gill nets set at
depths of approximately 12 to 18 m. Gill nets were set overnight for 15 h,
picked and reset for 6 h during the day. Larval lake herring were identi-
fied from characteristics described by Fish (1932).
Sampling on Black Bay was not continued after 1973 because the U.S.
Bureau of Sport Fisheries and Wildlife and the Ontario Ministry of Natural
Resources initiated a research program to measure smelt and herring inter-
actions in Black Bay during 1974.
Zooplankton Sampling
Zooplankton distribution was studied to identify the influence of
turbidity on food availability. Samples were collected using a 16 liter
Kemmerer bottle constructed of PVC with transparent end caps. During 1973
and 1974, zooplankton were collected immediately after most midwater trawls
at the trawling depth. During autumn 1974, Stations 1, 2, 3, 4 and 6 were
sampled allowing comparison of abundance levels over a broad area. During
1975, plankton were collected in clear and turbid areas within and adjacent
to distinct red clay plumes at Stations 1, 2 and 3 (Figure 1). Zooplankton
were collected on October 10 and 16, 1975, at the surface, 6.1, 12.2 and 18
m from the surface in water 20-25 m deep.
To determine if zooplankton could avoid the decending sampler in clear
water, comparisons were made between 30 paired samples. One member of each
pair was collected by lowering the Kemmerer directly to the desired depth.
The other pair member was obtained by lowering the sampler below the sam-
pling depth and then raising it to the collection depth. Comparisons were
also made between 30 paired samples collected by the Kemmerer bottle and by
a 25 liter Schindler sampler (Schindler 1969) constructed from clear PVC.
Samples were reduced to 200 ml by filtering through a 50um screen and
preserved in 5-10% formalin. Further concentration was performed in the
laboratory prior to examination in a Sedgewick-Rafter cell at 50 to 100 X.
12
-------
Complete counts were made, except on samples collected during 1973 when high
concentration required splitting. Two subsamples, each equaling 25% of the
total sample, were removed by pipette after agitation and counted. Identi-
fication was made using keys by Brooks (1957), Eddy and Hodson (1961) and
Edmondson (1959). Density is estimated as number/Hr*.
Depth, Temperature, Turbidity and Light Measurements
Bottom depth was measured and recorded by fathometer (Raytheon Model
DE-731). Cable length and wire angle were measured to estimate depth of the
midwater trawl and probe tow. Accuracy of depth estimates for the upper 6
m was verified to within 0.5 m by a SCUBA diver's capillary depth gauge. A
Vexilar Model 510 fathometer, mounted in a second vessel, was used to verify
estimated sampling depth of the trawl at 12 and 18 m.
Turbidity and temperature were monitored during or after each trawl or
plankton sample by probes carried in the tow. Measurements were made at the
sampling depth for midwater trawls or within 2 m of the bottom for bottom
hauls in water under 13 m. Resistance on the wires leading from the probes
to the deck monitors prohibited lowering the tow beyond 13 m during trawling.
Temperature and turbidity were therefore measured after each trawl by lower-
ing the tow to the trawling depth. During trawls at depths exceeding 13 m
the tow was pulled at various depths (0.5-10 m) to provide measurements of
distance and the depth and density of near-surface red clay turbidity plumes.
Water temperature was measured by a Yellow Springs Model 43 telether-
mometer or Rustrak Model 2133 recording telethermometer. Turbidity was de-
fined from measurements of percentage light transmittance over a 10 cm path
using a Hydro-Products Model 410 Br Transmissometer. Light transmittance
(T) readings less than 90% were converted to Formazin Turbidity Units (FTU;
American Public Health Association 1971) using the relationship:
FTU
= 84.5413 + (-1.5894) (T) + (0.0077) (T2)
The relationship was based on 83 paired measurements of the two tur-
bidity indices (Figure 3). Correlation analysis showed 95% of variation in
percentage light transmittance and FTU is explained by the relationship.
Because estimates of FTU derived for light transmittance readings exceeding
90% were high, a separate relationship for observations between 90 and 100%
transmittance was developed and is given by the equation:
FTU = 27.7976 - 0.2733 (T)
A portable Nephelometer, Ecologic Model 104, was used to measure turbidity
(FTU).
Twenty-two water samples of known FTU were filtered through 0.45ym fil-
ters and weighed to determine their suspended solid concentration (American
Public Health Association 1971) in order to identify the relationship be-
tween FTU and suspended solid level (ppm) (Figure 3). Correlation analysis
13
-------
o
CO
o
•o
d>
3
CO
80
60
40
20
5 80
U.
* 60
"c
± 40
20
N
O
20 40 60 8O 100
Formazin Turbidity Units (FTU)
20
40
60
80
100
Percentage Transmittance (%T)
Figure 3. Relationships between percentage light transmittance (T) and
Formazin Turbidity Units (FTU), and between Formazin Turbidity Units and
suspended solid concentration (ppm).
14
-------
demonstrated 98% of the variation in suspended solid concentration (ppm)
and FTU readings are explained by the relationship:
ppm = 0.1552 -f 0.6089 (FTU)
Surface light intensity (footcandles) was measured with each trawl or
seine using a Photovolt Model 200 Photometer. During 1975 and 1976 light
intensity profiles were obtained with a Kahl Submarine Photometer (Model
268WA-320) equipped with clear, red, blue and green color filters. The in-
strument provided measurement of light energy in yw/cm /nm for clear, red,
blue and green wave bands (Appendix Figure 1).
LABORATORY STUDIES
Turbidity Gradients
The responses of walleye and lake trout to laboratory turbidity gradi-
ents were studied to define turbidity preference ranges of the two species.
Responses to the range of turbidity levels found in western Lake Superior
were measured under day and night conditions at the U.S. E.P.A. Environ-
mental Research Laboratory-Duluth.
Two 55 liter (125 x 25 x 19.5 cm) and two 135 liter (152 x 31 x 31 cm)
electrode chambers partially partitioned to restrict mixing of water enter-
ing each of 4 sections but to insure free movement of fish (Spoor and
Drummond 1972; Figure 4) were used in the study. Turbid water was mixed
in a 200 liter head tank by spraying water over a clay source (Figure 4).
Turbidity was controlled at approximately 50 ±10 FTU by a photoelectric
cell and a light located on opposite sides of the tank. Light reaching the
sensor at reduced turbidity caused the sensor to activate a pump which
sprayed water over the clay source (Figure 4) until light received by the
sensor was reduced sufficiently by increased turbidity to cause the photo-
cell to break the circuit. Turbid water flow was directed to a 185 liter
constant head mixing tank and then to a 70.4 liter manifold. The mixing
tank reduced variation in turbidity caused by cycling of the pump in the
source tank. The manifold and a similar structure connected to a clear
water source distributed water to the four sections of each gradient chamber
through calibrated standpipes and funnels providing proportional dilution
and flow control (Figure 4). Flow into each section ranged between 100-300
ml/min for the smaller chambers and 200-500 ml/min in the larger chambers.
Turbidity level varied between experiments and ranged between 5-51 FTU
within the chambers (Table 2). In order to isolate the influence of prefer-
ence for a specific area of the chamber from response to turbidity, midway
through each experiment the chambers were flushed with clear water and the
gradient was reversed. Experiments ran 6-10 days (Table 2).
Lake trout and walleye were captured during the 1974-1975 field sam-
pling program and acclimated in the laboratory until feeding commenced
(approximately two weeks) prior to testing. Walleye ranged in size from
126-264 mm TL (12-139 g) and lake trout from 128-275 mm (15-141 g) (Table
15
-------
Clay Turbidity
Source Tank
Clear Water
Source Tank
CS
I
Mixing
Tank
Diluter
LJUUU
V Y V V
Gradient
Chamber
1
1
1
1 1
1 1
1 1
Figure 4. Continuous flow turbidity system. In the clay turbidity
source tank, water from the inlet (I) is pumped (P) over a clay
source (CS) and aerated (A) when a sensor (S) is activated by light
(L). Inflow (I) exceeds outflow to the mixing tank and excess is
drained (D). Turbid water from the mixing tank and clear water from
a second source tank are directed to diluters and metered by stand-
pipes into gradient chambers.
16
-------
TABLE 2. LABORATORY GRADIENT TEST SUMMARY
Test
Ln
(ram)
Fish
Wt
(g)
Temp
Range
(°C)
Turbidity
(FTU)
Avg
Min
Salvelinus
227
155
156
228
177
130
128
151
140
195
275
146
156
257
195
95
34
32
96
46
15
15
28
24
44
141
30
36
132
55
9.6-10.5
9.6-10.4
9.6-10.3
9.6-10.5
7.9-11.1
7.8-11.0
7.9-11.0
8.8-13.8
8.6-13.8
8.0-13.9
8.0-13.9
7.9- 9.7
7.6- 9.1
7.0- 8.5
7.0- 8.5
16.4
8.0
5.9
9.4
8.5
5.5
7.5
12.3
17.3
8.0
13.5
10.4
14.0
14.0
18.9
Avg
Max
namaycush
35.2
37.0
37.6
37.4
31.5
34.8
43.5
42.5
44.4
40.0
41.2
44.7
46.8
51.4
46.0
No.
Days
6
6
6
6
9
9
9
10
10
10
10
7
7
7
7
Duration
Hours
Day
33
26
26
48
21
21
88
76
87
89
74
74
73
73
Observ.
Night
22
40
38
37
—
29
29
69
66
75
75
54
54
54
54
Stizostedion vitreum vitreum
213
242
264
230
249
250
167
197
239
206
142
126
196
194
73
90
139
82
109
121
31
55
97
58
22
12
50
59
9.8-11.8
9.5-10.2
9.5-10.2
13.0-15.1
12.8-15.0
12.8-15.0
15.5-20.9
15.2-20.7
15.2-20.7
15.2-20.8
15.0-20.1
15.0-20.1
15.0-20.2
15.0-20.2
13.8
13.4
15.0
13.0
12.0
16.9
16.3
12.0
12.2
13.4
12.6
15.6
14.7
15.6
41.2
43.6
46.7
37.7
46.0
42.9
45.5
44.7
43.0
42.1
44.4
44.4
46.1
49.7
8
8
8
8
8
8
9
9
9
9
10
10
10
10
49
52
54
55
76
84
60
52
71
34
55
59
59
63
52
61
59
60
45
49
56
50
60
37
44
49
44
44
17
-------
2). Turbidity and temperature of Lake Superior water entering the gradient
chambers were measured twice daily. Lake trout and walleye experiments were
conducted at 7.0-13.9 C and 9.5-20.9 C respectively (Table 2). Temperature
variation within and among experiments corresponded with changes in Lake
Superior. Temperatures usually did not vary within experimental chambers,
at any point in time, by more than 1 C. The laboratory lighting system
(Drummond and Dawson 1970) simulated a natural diurnal pattern. Light in-
tensity ranged from 37.6-64.6 lux over the chambers throughout most of 13 h
photoperiod.
Location of test fish in the gradients was monitored 15 min/h, 24 h a
day by physiograph (Narco Model PMP-4 1460 or Gilson Model ICT-5). A pair
of electrodes in each of the four chamber sections was connected to a
physiograph channel through a rotary switch which switched the physiograph
to another chamber every 15 min. The system was sensitive to most activity
including movement associated with maintaining station or breathing (finning
and opercular movements). Location of fish in the chambers was determined
by the channel and associated chart record identifying activity in a chamber
section (Spoor and Drummond 1972).
The length of time experimental fish were located in a chamber section
was counted for 600 sec periods from each 15 min monitoring period. Peri-
ods less than ten seconds were considered to represent movement through a
section rather than selection of a location and were not counted. Differ-
ences in the time a fish resided in a given chamber section during high and
low turbidity periods, caused by reversing the gradient midway through the
experiment, were measured. The difference in the time fish resided in a
specific chamber section was related to the change in turbidity within the
section resulting from gradient reversal to determine effects of turbidity
on distribution. Clear water flowed through the chambers at the start, end
or midway through the experiment when activity was also monitored. Three
fish did not move from the initial chamber section under turbid or clear
water conditions. It was concluded that the unusual performance of these
fish resulted from failure to acclimate to experimental conditions; there-
fore, the data were not used in the analysis. Counts were analyzed separ-
ately for light and dark phases of the 24 h cycle by regression analysis.
Counts of the number of times fish changed chamber sections during the
600 sec monitoring periods were used to determine day-night activity pat-
terns of walleye and lake trout during clear and turbid water periods.
Activity of lake trout was also monitored in 3 liter electrode chambers
(Spoor et al^. 1971) at four levels of turbidity ranging from 54 to 6 (FTU) .
Tests using the small chambers were conducted with three fish held at a
constant turbidity (FTU) and one control. Activity associated with fin and
body movements was monitored at 54, 28, 9 and 6 FTU. The percentages of
the monitoring time lake trout were active under turbid water and in the
control chamber (0 FTU) were compared. Lake trout were acclimated to the
test turbidity for 1.5 to 76 h prior to monitoring. Laboratory conditions
were similar to those described for the gradient experiments. Plow rates
through the experimental chambers ranged from 100-130 ml/min.
18
-------
Larval Herring Bioassay
Direct influence of red clay turbidity on survival, growth and distri-
bution of larval herring was measured in the laboratory by holding larvae
for 62 days at 9 turbidity levels. The bioassay study was supported pri-
marily through a grant from the University of Wisconsin, Sea Grant College
Program, and is described in detail by Swenson and Matson (1976). Partial
support was derived from this EPA program. Because the results represent
an integral part of the analysis of red clay turbidity on western Lake
Superior fish populations, methods and results are outlined in this report.
Turbid water source tanks, clear water source tanks and manifolds
(Figure 4) were used to deliver a constant flow of Lake Superior water at
400 ml/min to each of 20, 70.5 liter (61 x 20.3 x 61 cm) test chambers.
Eight concentrations ranging from 6 to 46 FTU were run in replicate. Four
control chambers were maintained between 0 to 2 FTU. Turbidity levels in-
cluded in the experiment covered the normal range for western Lake Superior.
A feeding system (Anderson and Smith 1971a) distributed approximately
400 brine shrimp (Artemia salina)/h to each chamber 12 h each day. Light
from two 15 watt incandescent bulbs located 10 cm above each chamber was
passed through translucent fiberglass covers to reduce glare and increase
dispersion over the water surface. A constant 13 h photoperiod was main-
tained throughout the tests. Light intensities averaged 15.3 lux in the
upper 10 cm of the water column. Intensity 51 cm below the surface was
reduced by 34% in the lowest concentration (1-6 FTU), 48% in intermediate
concentrations (12 to 28 FTU) and 53% in the higher concentrations (34 to
46 FTU).
Turbidity and temperature were monitored twice daily. Temperature
changes of 3 to 8 C during the study corresponded to similar changes in
Lake Superior. Mean daily temperatures between chambers within replicates
did not vary by more than 1 C. Weekly measurements of oxygen, pH and con-
ductivity showed pH increased slightly with turbidity. Oxygen was main-
tained near saturation.
Counts and measurements of herring larvae were made from photographs.
Behavioral response was defined from direct counts of larvae in upper,
intermediate and lower sections of the chambers (Swenson and Matson 1976).
Predation Studies
Several attempts at maintaining adequate numbers of smelt in the labo-
ratory failed. However, limited information on smelt predation was obtained
from two Age I-i- smelt, 62 and 80 mm T.L., captured during November 1973.
The two young smelt were maintained in a 20 liter aquarium receiving a con-
tinuous flow of Lake Superior water and fed brine shrimp at the E.P.A.
Enviromental Research Laboratory-Duluth until 29 March 1974 when predation
studies were initiated. The studies measured relative preference of the
two smelt for brine shrimp, lake herring eggs (embryos) and lake herring
19
-------
larvae. The rates at which lake herring larvae were eaten by smelt under
light and dark conditions were also estimated. Information on the rate of
gastric digestion at 12.6 C was obtained by sacrificing one smelt 1-3/4 and
the other 4 h after feeding.
Estimated rates of gastric digestion from this study and by Foltz
(1974) were used with information on the occurrence of young smelt in older
smelt stomachs and estimates of smelt density in Lake Superior to determine
the importance of smelt cannibalism on survival. Estimates of the number
of young smelt consumed daily and monthly were calculated by correcting the
percentage occurrence of young smelt in the stomachs of older smelt for the
rate of digestion and the average number of young smelt found in smelt sto-
machs. The corrected percentage occurrence value is an estimate of the per-
centage of the smelt population (Age II and older) which ate the equivalent
of one Age 0 smelt daily. Multiplying the value by the density of Age II
and older smelt and by the days in a month gave a gross estimate of the
number of young smelt consumed monthly by older members of the population.
The number of young smelt consumed monthly by older smelt/100 m-' and the
average density of young smelt (number/100 m^) were used to estimate the
percentage mortality resulting from cannibalism during various sampling
months. Food consumption of walleye and the number of smelt consumed daily
by walleye/100 m^ were estimated using the method described by Swenson and
Smith (1973).
20
-------
RESULTS
TURBIDITY AND TEMPERATURE IN WESTERN LAKE SUPERIOR
Turbidity and temperature of near-surface and near-bottom waters in
western Lake Superior were monitored during 1973 through 1976. The highest
turbidity occurred at Stations 3 and 4 located closest to the red clay
source area (Table 3; Figure 1). Monthly averages for Station 4 often
exceeded 30 FTU in contrast to Station 1 along the north shore where tur-
bidity did not exceed 7 FTU and Station 2 where monthly average turbidity
seldom exceeded 10 FTU except in the wave surge zone (1 m depth).
Measurements at the surface, 3, 6 and 9.1-15 m, made during 16 days of
1973, showed turbidity was highest during June, September and October when
temperature averaged less than 11 C (Table 3; Figure 5). Surface and bot-
tom temperature in the 1 to 15 m depth zones averaged above 14 C during
July and August when turbidity was usually low (Table 3; Figure 5).
Both turbidity and temperature decreased with distance from shore.
Near-surface temperature dropped slowly, but bottom water temperature de-
clined rapidly as depth increased offshore (Table 3). The rate of decline
in turbidity was slower in near-bottom than in near-surface waters. At
depths exceeding 3 m, turbidity of near-bottom water often exceeded that
of water near the surface (Table 3; Figure 5). During 1975 and 1976 a
total of 22 temperatures and turbidity profiles were recorded in water
exceeding 15 m (Appendix Table 1). Although turbidity was comparatively
low due to distance from shore and the mild climatic conditions encoun-
tered during 1975 and 1976, turbidity in bottom waters averaged approxi-
mately 6 FTU and often exceeded surface water turbidity. Increased tur-
bidity below the surface appears to be caused by settling of suspended
solids, resuspension of solids from the lake bottom and sinking of higher
density turbid water to levels where increased density associated with the
clay load reaches equilibrium with cold high density bottom water. Occur-
rence of higher turbidity in offshore bottom water suggests red clay tur-
bidity is more extensive spatially and temporally than surface water plumes
would indicate.
LIGHT INTENSITY IN WESTERN LAKE SUPERIOR
Turbidity in western Lake Superior significantly reduced light pene-
tration even at every low concentrations. Light profile measurements made
during 1975 and 1976 on 21 days (Appendix Table 2) showed the depth of 1%
surface incidence was reduced from approximately 16.5 m in clear water
21
-------
TABLE 3. NEAR SURFACE (S) AND BOTTOM (B) WATER TEMPERATURE AND TURBIDITY
Temperature is °C. Turbidity is FTU (in parenthesis).
KJ
Station No.
Bottom
Depth (M)
June
1973-S
1973-B
1974-S
1974-B
1975-S
1975-B
1976-S
1976-B
July
1973-S
1973-B
1974-S
1974-B
1975-S
1975-B
1976-S
1976-B
August
1973-S
1973-B
1
9-12
1 3
13(11) 13( 8)
13(10)
5(5)
12(3)
—
—
12(7)
17(1)
16(1)
16(2)
15(38) 12(11)
11( 9)
__
— —
18 ( 4) 14 (
15 (
18 ( 3) 14 (
14 (
11 (
—
20 ( 6) 18(
19 (
5)
5)
3)
3)
8)
5)
4)
2
6
12 (
11 (
6)
9)
12(10)
8( 9)
__
IK
14 (
12 (
9(
8(
—
18 (
17 (
1)
7)
5)
4)
4)
5)
4)
4)
9
13( 8)
IK 7)
9(10)
6( 7)
11(11)
—
14( 6)
11 ( 5)
ll( 4)
8( 5)
IK 4)
5( 3)
—
19 ( 4)
16( 4)
12+
—
_ _
12(7)
8(9)
14(2)
—
—
14(5)
5(6)
19(1)
16(1)
—
1 3
13(36) 12(37)
12(35)
19(15) 13(23)
12(24)
— —
13( 9)
18(29) 17( 8)
16(16)
17(14) 15(16)
13(16)
19(18)
19(21)
20(49) 19(42)
19(29)
4
6
13(27)
12(29)
12(20)
19(14)
11(19)
12 ( 6)
18(13)
14(29)
15(14)
13(12)
12( 9)
16( 6)
18 ( 7)
17(16)
9
12(17)
11(11)
13(27)
9(15)
11(30)
—
17( 5)
13( 8)
15( 9)
12(11)
15(28)
9( 8)
19(17)
14(20)
19 ( 9)
16(10)
12+
—
—
9( 7)
14( 3)
5( 6)
—
—
18(10)
9( 6)
17(14)
4( 5)
—
(continued)
-------
TABLE 3. (continued)
to
Station No.
Bottom
Depth (M)
1974-S
1974-B
1975-S
1975-B
1976-S
1976-B
September
1973-S
1973-B
1974-S
1974-B
1975-S
1975-B
1976-S
1976-B
October
1973-S
1973-B
1974-S
1974-B
1975-S
1975-B
1976-S
1976-B
1
9-12 1 3
12(3) 18( 2) 14 ( 5)
14 ( 3)
— — —
16( 5)
— —
— —
17(14) 18( 9)
16 ( 7)
9( 4)
7(2) — 9( 3)
— — —
—
— — —
—
IK 5) 6( 5)
10 ( 4)
— — —
—
— — —
—
— — —
—
2
6
16( 4)
IK 4)
18 ( 2)
—
—
—
18( 8)
15 ( 8)
9( 4)
10( 3)
—
—
—
—
6( 5)
10 ( 4)
—
8( 4)
—
—
—
—
9 12+
16( 4)
10( 3)
18(2)
5(4)
— —
—
18( 9)
14 ( 7)
10 ( 4)
10( 3)
— —
—
— —
—
11( 6)
9( 4)
— —
8(4)
— —
—
— —
— —
1 3
18(50) 15(30)
15(31)
— —
—
— —
— —
16(31) 16( 7)
15( 7)
12(118) 11(77)
11(70)
— —
—
— —
—
11(47) 10(42)
10(44)
8(30)
8( 7)
— —
—
— —
9( 4)
4
6
14(19)
14(32)
18 ( 2)
14 ( 7)
22( 5)
—
16 ( 8)
15( 7)
7(40)
11(60)
—
—
—
—
10(38)
10(43)
—
8( 5)
—
—
—
9( 5)
9 12+
14(13)
10( 4)
18 ( 3)
4(10)
19 ( 3) 18( 2)
— —
16( 8)
15 ( 5)
4(18)
8(40)
13( 1)
IK 7)
— —
—
10(29)
10(43)
— —
9( 5)
IK 1)
11( 1)
— —
—
-------
40
20
Surface
40
20
3 meters
3;
la
^
H 40
•o
o
O
^ 20
9
60
40
20
6.1 meters
9.1 meters
\ ;^_
June
July
August
1973
September October
Figure 5. Station 4, turbidity (dashed lines) and temperature (solid
lines) during 1973 at the surface, 3, 6.1 and 9.1-15 m in 9.1-15 m water
columns.
24
-------
(0-2 FTU) to an average of 2.5 m in turbid water (10-12 FTU).
2
Relationships between the log _ of light intensity (yw/cm /nm) and
water depth showed the rate of lignt penetration decreased with increased
turbidity (Figure 6). However, the degree of change in the rate of pene-
tration (penetration coefficient) was found to be higher at loxj turbidi-
ties (Figure 7). The rate of penetration is estimated by the equation:
PC = a-
where
PC = rate of light penetration
ppm = suspended solid concentration in parts per million
a = 0.1968
c = 0.2261
Measurements of penetration by blue (400 to 530 ma) and red (595 to
730 nm) spectra (Appendix Figure 1) made at five red clay concentrations
showed rate of penetration was higher for blue light only in clear water.
With 0.6 FTU (0.5 ppm) the slopes of least square regression lines de-
fining rate of penetration were +0.743 for blue light and +0.627 for red
light (Table 4). At 2.9 FTU (1.8 ppm) rate of penetration of red light
exceeded that for blue (Table 4). At 10.7 FTU (7.7 ppm) rate of penetra-
tion of blue light was reduced to 10% of clear water, whereas rate of
penetration by red light was 41% of clear water. As a result of the high
rate of absorbance of shorter wave energy in turbid water, low concentra-
tions of red clay caused a dramatic change in the quality and intensity of
light in western Lake Superior.
ZOOPLANKTON ABUNDANCE AND DISTRIBUTION
Plankton densities estimated from samples obtained by lowering the 16
liter Kemmerer bottle were compared with those obtained from samples col-
lected by raising the Kemmerer and with a Schindler sampler. Analysis of
variance showed estimates of abundance were not influenced by the sampling
procedures (P > 0.1).
Species composition and abundance of western Lake Superior zooplankton
at Stations 1, 2, 4, 5 and 6 were compared. Comparisons were based on
samples collected during September and October 1975, within a 32 day period
when temperatures ranged from 6.9-10.8 C. Zooplankton composition was gen-
erally similar throughout the western end of the lake although certain
species of minor importance with respect to abundance were not found at all
sites (Table 5). Rotifers were the dominant group at all sampling sites.
Conochilus, Kellicottia, Polyarthra and Keratella were the dominant genera
(Table 5). Although composition was similar, the number of organisms dif-
fered greatly between locations. Stations 1 and 6, characterized by clear
water, had few zooplankton when compared to the other sites. Stations 2
and 4 at the extreme western end of the lake averaged 14,200 more
25
-------
n
Light (uw/cm /nm)
N3
.01
Figure 6. Light penetration in western Lake Superior at different concentrations of red clay
turbidity.
-------
.80
-60
o
"o
o
o
JO
*-
o
0)
Q_
.40
.20
8
Turbidity ( ppm)
Figure 7. Relationship between change in rate of light penetration
and turbidity in western Lake Superior.
27
-------
TABLE 4. LIGHT PENETRATION AND TURBIDITY RELATIONSHIPS
Regression constants defining relationships between light intensity (Y),
(yw/cm2/nm) and depth X, (m) for blue and red spectra at five concentra-
tions of red clay turbidity in western Lake Superior. Light intensity is
estimated by the equation: Y = abx.
Turbidity
(FTU) (ppm)
1.0 (0.5)
2.9 (1.8)
4.1 (3.1)
10.7 (7.7)
12.3 (8.5)
Color
Blue
Red
Blue
Red
Blue
Red
Blue
Red
Blue
Red
Regression
Ordinate
Intercept
a
+10.16
+ 4.64
+12.67
+11.35
+ 7.81
+ 6.75
+ 6.23
+ 5.24
+10.95
+18.59
Constants
Regr.
Coef .
b
+0.743
+0.627
+0.437
+0.454
+0.371
+0.432
+0.071
+0.255
0.077
0.213
28
-------
TABLE 5. ZOOPLANKTON ABUNDANCE AT FIVE STATIONS DURING 1974
•3
Number of zooplankton/m represent averages of four to six samples col-
lected at 3 m intervals from 10 m water columns.
Sampling Location
Rotifera
Conochilus sp.
Asplanchna spp.
Kellicottia longispina
Keratella cochlearis
Polyarthra spp.
Synchaeta spp.
Cladocera
Bosmina longirostris
Daphnia galeata mendotae
Daphnia pulex
Holopedium gibber urn
Leptodora kindtii
Others
Adult Copepoda
Cyclops spp.
Diaptomus spp.
Epischura lacustris
Limnocalanus macrurus
Others
1
9/16/74
2,922
1,625
16
453
16
78
734
640
312
328
—
—
651
31
406
203
16
2
9/13/74
12,687
6,396
188
3,521
260
1,833
489
3,042
292
2,625
62
31
—
2,375
375
1,969
21
10
4
9/11/74
10,551
1,125
238
2,825
1,962
3,087
312
1,787
825
862
62
12
12
12
1,837
525
1,262
25
25
and Date
5
9/26/74
11,437
4,385
229
2,719
1,865
1,146
802
979
312
635
10
21
—
1,562
385
1,156
10
10
6
10/13/74
6,042
1,917
219
2,125
73
396
1,312
1,458
906
438
62
52
3,146
1,906
1,104
62
31
42
Copepoda nauplii
297 2,417 2,800 1,177 2,687
Total Zooplankton
4,516 20,521 16,975 15,156 13,333
29
-------
zooplankton/m3 than Station 1 on the north shore, 3,600 more zooplankton/
m^ than Station 5 on the south shore and 5,400 more zooplankton/m-5 than
Station 6 located in the Apostle Islands area (Table 5).
Sampling in clear and turbid water zones within Stations 2 and 3
(October 10, 1975) and at Station 4 (October 16, 1975) showed the greatest
zooplankton densities occur in turbid water (Table 6; Figure 8). A signif-
icent positive relationship between zooplankton density and turbidity was
identified for all major groups except Copepoda (Table 7).
A significant increase in zooplankton in surface waters (0.1-1.0 m) at
higher turbidities was demonstrated by relationships between the percentage
of zooplankton in surface samples and turbidity. Percentage in surface
water was calculated from the total number of zooplankton found in all
samples in the water column and the number in surface samples. Rotifers
and Cladocerans showed the greatest increases in surface waters at higher
turbidities (Table 7). Surface abundance of zooplankton and light inten-
sity in surface water were negatively correlated (Table 7). The correla-
tion was the strongest for rotifers.
During both 1973 and 1974, zooplankton densities were usually higher
at Station 2 having lower average turbidity than Station 4 characterized
by higher turbidity. Differences appeared to result from higher densities
of rotifers at Station 2 (Table 8). Variation between Stations 2 and 4
fails to support a conclusion that zooplankton densities are directly in-
fluenced by turbidity. The zooplankton studies did show that the highest
food availability for plankton feeding larval fish occurs at the extreme
west end of the lake (Stations 2, 3, 4 and 5} and in near-surface waters
of red clay plumes.
FISH ABUNDANCE AND TURBIDITY
Spacial Variations in Fish Abundance
Trawl and seine catches from the six locations sampled during this
study and experimental gill net catches at 15 Wisconsin Department of
Natural Resources survey stations (King and Swanson 1974) were analyzed
to identify relationships between relative abundance of fish and red clay
turbidity in western Lake Superior. Trawl and seine catches during 1973-
1976 were summarized by geographic location, gear type and vertical loca-
tion (bottom or midwater; Table 9). Average gill net catches for the
period 1973-1974 were summarized according to sampling depth and turbidity
(Table 10). The sampling programs demonstrate the occurrence of 38 species
of fish and suggest smelt, longnose suckers (Catastomus catastomus) and
walleye are most abundant. Troutperch (Percopsis omiscomaycus), shiners
(Notropis atherinoides, N_. hudsonius) and sculpin (Cottus cognatus, C.
rici, C^. bairdi) were common in trawl catches but were not captured by gill
nets. ~Burbot, white suckers (Catastomus commersoni), round white fish
(Prosopium cylindraceum), lake herring and chubs (Coregonus hoyi) were
common in gill net catches (Table 10) and occurred or were common in trawl
30
-------
TABLE 6. ZOOPLANKTON ABUNDANCE IN TURBID AND CLEAR WATER DURING 1975
o
Number of zooplankton/m represent averages of six or eight paired samples
taken at the surface and at 6 m intervals in 16 to 20 m water columns.
Zooplankton
and
Turbidity
Turbidity (ppm)
Rotifera
Conochilus sp.
Asplanchna spp.
Kellicottia longispina
Keratella cochlearis
Polyarthra spp.
Synchaeta sp.
Others
Cladocera
Eubosmina coregoni
Bosmina longirostris
Daphnia galeata mendotae
Daphnia pulex
Copepoda (Adult)
Cyclops bicuspidatus
Cyclops vernalis
Diaptomus spp.
Limnocalanus lacustris
Mesocvcloos macrurus
Sampling Location
Station
1.75 0
7,234 4
1,856
203
1,516 1
1,656
687
1,523 1
7
1,101
62
211
805
23
6,148 3
1,812
8
4,328 2
—
2
.5
,211
359
70
,554
367
531
,289
39
250
—
62
188
—
,351
977
—
,312
47
16
Station 3
7.67
13,676
875
157
1,750
6,916
1,135
2,812
31
3,051
541
1,176
1,177
55
5,427
1,427
63
3,874
—
3.12
6,851
1,101
54
1,507
1,797
Ilk
1,586
32
1,227
78
226
890
31
5,610
1,687
8
3,906
8
—
Station 4
8.5
21,248
729
292
843
14,177
1,510
3,676
21
9,509
2,791
4,708
1,999
—
4,260
1,812
—
2,447
—
—
1.8
6,019
760
260
1,000
1,771
1,322
906
—
2,645
166
1,687
792
—
3,479
961
—
2,510
—
—
Copepoda nauplii 3,812 2,391 5,729 3,343 3,354 1,562
Total 18,295 10,203 27,884 17,024 38,364 13,654
31
-------
12
ex
Q
Stcrtion 3
(4.5)
4.0)
9.5)
8
16
24
32
40
12
o.
0>
O
(1.5)
Station 4
(2.0)
,
//
f(7.0)
^(10.5)
(8.0)
8 16 24 32
Zooplankton (Thousands/m 3)
40
Figure 8. Zooplankton density in low (solid line) and high (dashed line)
turbidity zones of Stations 2, 3 and 4. The number associated with each
plotted point is the turbidity (FTU) at each sampling point.
32
-------
TABLE 7. ZOOPLANKTON DENSITY, TURBIDITY, DEPTH AND LIGHT
PENETRATION RELATIONSHIPS
Regression constants are for relationships between zooplankton abundance
indices (Y), red clay turbidity or light penetration (X) in western Lake
Superior.
Model
and
Zooplankton
Group
Regression
Coefficient
(b)
Intercept
(a)
Coefficient
of Variation
(r2)
Students-t
for
b = 0
(t)
Zooplankton/m (Y) and Turbidity (X)
Rotifera 1802.1
Cladocera 795.9
Copepoda (adult) 90.4
Copepoda (nauplii) 242.6
2871.5
-147.6
4406.9
2345.5
,88
,54
.06
.37
8.72**
3.42**
.76
2.45-
Percentage Zooplankton in Surface Samples (Y) and Turbidity (X)
Rotifera
Cladocera
Copepoda (adult)
Copepoda (nauplii)
1.61
2.96
.38
-.02
19.6
17.7
16.9
22.0
.43
.33
.03
.00
2.72*'1
2.22*
.54
-.02
ry
Percentage of Zooplankton in Surface Waters (Y) and Light (pw/cm /nm; X)
Rotifera -1.06 29.9 .38 -2.84*=
Cladocera - .41 32.2 .03 - . 54
Copepoda (adult) - .37 20.8 .14 -1.27
Copepoda (nauplii)
- .20
23.6
.03
- .58
"Indicates statistical significance at P<0.05.
"""Indicates statistical significance at P<0.01.
33
-------
TABLE 8. ZOOPLANKTON ABUNDANCE AT TWO STATIONS DURING 1973-1974
o
Number of zooplankton/m are averages of five samples collected at the surface and 3
intervals in 6.1 to 9,1 m water columns.
Zooplankton
and
Turbidity
Turbidity (ppm)
Rotifera
Cladocera
Copapoda (adult)
Copepoda (nauplii)
4
8/16/73
15,0
38,020
15,870
6,946
9,156
-•••" 2
8/19/73
2,0
153,714
20,952
19,619
9,904
Sampling Location and Data
4
9/1/73
8,0
21,692
3,896
14,133
9,502
2
9/2/73
3.0
72,653
19,238
37,605
11,319
4
10/14/73
25,0
25,857
15,810
10,993
10,585
2
10/19/73
2.5
15,765
7,163
14,272
9,789
Total
69,992
204,189 49,223 140,815
63,245
46,989
-------
TABLE 9. AVERAGE TRAWL CATCH (Number 100/m3)
u>
Year
Station
Gaar
Type
Location
Qftmiru.*. tnordax (yy)''4
0, mordax (older)
Stizoatadion vitreum v,
Salvelinua namaycueh
Salmo trutta
S. ^ardnari
Coragonus, sp.
Prosojjium jiylindraceum
Catoatomua catostomua
C, commerfloni
NptrQpjLB ap.
Percopsifl omiacomaycua
Cottua ip.
AjLoajL j^Beudgharengija
Coueaius plumbeua
Perca flavaacena
Pereina eaprodea
EtheoBtojna ni^rum
Gasterosteidae
EBOX lucius
1 ctaluruH ap.
Pomoxls nigromaculatus
- ._ ..w...,- .-, ,-~~^ 111 »li «Tl k l — - . -. . _--- -f-
AmbloplJ.te^ rujieatria
Cyprinus *'fl*"g.i£
Lota lota
1973-74
2 and 4
6.1m
Trawl
Bottom
1.8-15 m
22,167
6,667
.267
.004
...
__
< .001
.113
,016
.040
1.122
.279
. .014
.053
—
< .001
< . 001
<.001
< . 001
< . 001
. 009
1975
6
6.1 m
Trawl
Bottom
1.8-15 m
17.185
13.347
.127
—
--
._
.089
.465
.005
1.410
.958
,066
.033
—
1,101
1.360
~_
1973-74
2 and 4
7.6 m
Trawl
Bottom
1,8-15 m
4.544
3,194
,296
<,001
<,001
<.001
<,001
.105
.044
,066
.898
,040
<.00l
.014
—
<.001
< . ooi
<.001
<,001
-- .001
.016
1975-76
2 and 4
9.5 m
Trawl
Bottom
< 15 m
,534
7,368
.104
.002
—
—
,020
.080
.042
.002
.316
.032
.012
.001
__
—
.MS
1975-76
2 and 4
9.5 m
Trawl
Bottom
> 15 m
,075
3,230
.026
.018
..
—
.018
.062
.004
.002
.001
.134
.002
< .001
--
_
.00/4
(yy) refers to Ap,t> 0.
-------
TABLE 9. (continued)
OJ
Year
Station
Gear
Type
Location
Osmerus mordax (yy)^
0. mordax (older)
Stizostedion vitreum v.
Salvelinus namaycush
Salmo trutta
S. gardneri
Coregonus sp.
Prosopium cylindraceum
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Perca flavescens
Percina caprodes
Etheostoma nigrum
Gasterosteidae
Esox lucius
Ictalurus sp.
Pomoxis nigromaculatus
Ambloplites rupestris
Cyprinus carpio
Lota lota
1973-76
2 and 4
6.1m
Trawl
Midwater
1.8-15 m
2.393
.818
.002
.002
< .001
.063
.014
.009
< .001
< .001
< .001
< .001
< .001
.006
1973-76
1
6.1m
Trawl
Midwater
1.8-15 m
1.082
.020
.001
.001
.002
.003
.001
1975-76
2 and 4
6.1 m
Trawl
Midwater
> 15 m
.242
.036
.002
< .001
.001
.019
< .001
1973-74
2 and 4
Seine
Bottom
.5-1 m
213.900
21.800
1.089
.044
5.860
.039
26.070
.233
.293
.100
.035
.007
.019
1.308
.123
.014
t
(yy) refers to Age 0.
-------
TABLE 10. AVERAGE GILL NET CATCH PER SET AND PERCENT SPECIES ABUNDANCE (IN PARENTHESIS)
u>
Zone Type
Depth Range (m)
Secchi (average) (m)
Station No.'''
Osmerus mordax
Stizostedion vitreum v.
Salvelinus namaycush n. (Ad)
S. namaycush n. (ju)
S. namaycush siskowet
S. fontinalis
Salmo gairdneri
S. trutta
Oncorhynchus kisutch
Splake = S.f x S.n
Coregonus artedii
(C. hoyi
C. clupeaformis
Prospium cylindraceum
Lota lota
Catostomus catostomus
C. commersoni
Alosa pseudoharengus
Acipenser fulvescens
Esox lucius
Perca flavescens
Cyprinus carpio
Semotilus atromaculatus
Total
Shallow Turbid
3.0 to 12.8
5.5
15 and 17
66. 0( 19.4)
74. 6( 21.9)
0.3(> 0.1)
0.2( > 0.1)
0.3( > 0.1)
27. 8( 8.2)
0.8( 0.2)
34. 6( 10.1)
127. 0( 37.2)
8.8( 2.6)
0.2( > 0.1)
0.3( > 0.1)
340.9(100.0)
Shallow Clear
3.6 to 16.4
15.5
14,19,22,23
27.6(11.1)
27.7(11.1)
0.6( 0.2)
1.7( 0.7)
0.2( 0.1)
1.7( 0.7)
0.5( 0.2)
0.2( 0.1)
2.3( 0.9)
9.5( 3.8)
6.4( 2.6)
32.5(13.0)
3.6( 1.5)
72.3(29.0)
60.6(24.3)
0.6( 0.2)
0.3( 0.1)
0.2( 0.1)
0.3( 0.1)
0.3( 0.1)
249.1(99.9)
Deep Turbid
12.8 to 19.5
9.5
19 and 25
34. 5( 38.0)
1.3( 1.4)
1.6( 1.8)
0.8( 0.9)
5.1( 5.6)
5.2( 5.7)
4.8( 5.3)
2.1( 2.3)
11. 0( 12.1)
19. 2( 21.1)
5.0( 5.5)
0.2( 0.2)
0.2( 0.2)
91.0(100.1)
Deep Clear
14.6 to 60
21.7
1,8,9,11,12,18,28
37. 0( 28.0)
5.6( 4.2)
5.8( 4.4)
0.3( 0.2)
5.5( 4.1)
9.2( 7.0)
1.3( 1.0)
25. 2( 19.0)
2.3( 1.7)
40. 0( 30.2)
0.4( 0.3)
132.6(100.1)
t
Station locations are identified by King and Swanson (1974) and are suggested in the text.
-------
catches (Table 9).
Comparisons of 6.1 m bottom trawl catches in turbid areas (Stations 2
and 4) with catches from stations characterized by clear water (Station 6)
suggested that abundance of several species was influenced by turbidity.
Although trawl catches of smelt, white sucker and troutperch were generally
similar, catches of longnose sucker, round whitefish, Gasterosteidae (pri-
marily Pungitius pungitius), sculpin and darters (Etheostoma nigrum) were
higher at Station 6 (Table 9). Trawl catches of walleye and burbot were
higher at the more turbid Stations 2 and 4. Gill net catches from turbid
and clear water locations support the trends suggested by trawl catches
for most species and indicate abundance of lake trout (particularly juve-
niles) is lower in the turbid water zones (Table 10). Gill net catches
provided some evidence that smelt and chubs are more abundant in turbid
water. Comparison of midwater trawl catches from Station 1 (1973-1976)
with catches from Stations 2 and 4 (Table 9) provides additional evidence
that smelt abundance is higher in turbid water.
Temporal Variations in Fish Abundance
Analysis of commercial catches of herring and smelt from Minnesota
district M-l confirmed previous findings by Anderson and Smith (1971b)
which showed abundance (CPE) of lake herring has declined in the Duluth-
Superior area and abundance of smelt has increased (Figure 9). Abundance
of the two species is negatively correlated (r = 0.5, P>0.05). Smelt
abundance was not correlated with spring runoff 3-4 years previous. No
relationship was found between percentage deviation in smelt abundance
(CPE) for a given year and percentage deviations in turbidity indices for
May through June or May through October 3-4 years previous.
INFLUENCE OF TURBIDITY ON SMELT POPULATIONS
Smelt Growth and Turbidity
High abundance of Age 0 smelt and intensive sampling during 1973 per-
mitted estimation of growth at Stations 2 and 4 (Figure 10). Comparison
of mean length measurements suggested growth was similar. Comparison of
average lengths calculated for five months of sampling at Stations 1, 2
and 4 suggested Age 0 smelt in the more turbid Stations 2 and 4 were 9%
longer than those at Station 1 but differences were not significant
(P> 0.05).
Smelt Distribution and Water Temperature
Catches in bottom trawls and seine hauls showed juvenile and adult
smelt (Age II and older) move offshore to depths exceeding 10 m after June
when near-shore waters exceed 11 C. The association with temperature is
suggested by higher June catches (Table 11).
38
-------
600
o
o
— 400
x
o>
o
O 200
Lake Herring
;\
"*• «•—-x
200
100
O
O
O
o>
JC
a>
u
o
•o
c
1950
I960
1970
O 400
O
O
S
o
o 2OO
O
Smelt
1950
I960
Year
3000
2000
1000
O
O
o
9
o
c
o
•o
c
3
JO
1970
Figure 9. Catch (solid line) and abundance (CPE, dashed line) of lake
herring and smelt in western Lake Superior, Minnesota Statistical
District M-l. (See Hile(1962) for statistical district boundaries.)
39
-------
50
40
E
30
o 20
o
10
Station 2 x
Station 4 •
• •
x
* x x
* i *
June
July
August September October
Figure 10. Average total length of Age 0 smelt at Stations 2 and 4 during 1973.
are calculated from measurements of 10-100 individuals.
Averages
-------
TABLE 11. BOTTOM TRAWL CATCH BY SAMPLING MONTH
Catches are averages for 100 m captured by 7.6 m bottom trawl at stations
2 and 4 during 1973 and 1974
Species
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jv)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
June
.016
1.150
1.420
1.150
.067
.100
—
<.001
.050
.250
.016
.883
.016
July
.133
.750
2.166
.533
.067
.083
—
<.001
__
.066
.033
—
.150
—
Aug.
4.70
1.450
.480
.580
.133
.200
<.001
—
.116
.033
—
.417
—
Sept.
12.47
3.366
.570
.433
.417
.150
<.001
—
—
.033
.133
.066
2.400
.166
Oct.
6.26
.770
.030
.300
.300
.067
—
—
.020
.016
.433
.383
—
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pornoxis nigromaculatus
Cyprinus carpio
.016
<.001
<.001 <.001
.016 .033
<.001
<.001 <.001
<.001
<.001
41
-------
Correlations between smelt catch/100 m3 at sampling locations within
Stations 2 and 3 demonstrated teirperature preference influences distribu-
tion and varies with age. Correlations between temperature and catches of
Age 0 smelt captured by seine and by trawling on the bottom, at midwater
locations and near the surface were positive (Table 12). The analysis de-
monstrated that distribution of Age I smelt was not influenced by tempera-
ture. Correlation coefficients defining variability between catches of
Age I smelt and temperature were both positive and negative and were not
significant (Table 12). Catches of Age II and older smelt in bottom
trawls (6.1 and 7.6 m nets) were negatively correlated with temperature
demonstrating preference for colder water (Table 12).
Smelt Distribution and Turbidity
Sight Distribution: Response of saelt to turbidity was analyzed by
multiple regression models in which smelt density (number/100 m3) estimated
from 6.1 m trawl catches was the dependent variable, and temperature and
turbidity were independent variables. Partial regression coefficients
estimating rate of change in catch with temperature followed the general
trends described previously from simple correlation analysis (Table 12).
Smelt density near the surface in areas 1.8-15 m deep increased with tur-
bidity. In contrast, regression coefficients describing change in smelt
density near the bottom were negative. Partial regression coefficients
relating change in catch 3 m from the surface to turbidity were generally
positive whereas those defining change in catch with turbidity 6.1 m from
the surface, in a 7.7 to 15 m column, were positive for some age groups
and negative for others. The density changes suggest that smelt respond
to change in turbidity by vertical movements. However, variation in
catches between sampling nights was high, and few coefficients were sig-
nificant (P>0.05).
To control variability in catch between sampling nights, smelt in sur-
face, midwater and bottom trawls (number/100 m3) were totaled by age group
for three of the horizontal depth zones sampled (1.8-4.7, 4.8-7.6, 7.7-15
m), and the percentage of the total captured at each vertical location was
calculated. Multiple regression analysis was used to measure the influence
of turbidity and temperature on the percentage of saelt captured in each
vertical stratum. Percentages generally increased with turbidity near the
surface and at midwater locations but declined near the bottom (Table 13;
Figure 11). Coefficients estimating rate of change in percentage abundance
with turbidity were significant for most tests with fish older than Age O.
Percentage abundance of Age O smelt was generally high at surface, midwater
or bottom locations and did not change appreciably with turbidity. Abun-
dance of Age II and older smelt increased from 1 to 32 in surface waters
and from 8 to 24Z at 3.1 m with increase in turbidity from 0 to 15 FTU
(Figure 11).
42
-------
TABLE 12. SMELT DENSITY AKD TEMPERATURE CORRELATIONS
Correlations between water temperature and trawl catches of Age O, I, and
Age II or older smelt are given for surface, nddvater and bottom sailing
points in four depth zones.
Sampling
Years
Age 0
1973
1973
1973, 1974
1973
1973
1973
Age I
1973
1974
1973
1974
1973
1973
1973
Gear
Type
Seine
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Trawl
Size
7.7 m
6.1 ra
6.1 m
6.1 m
6.1 o
6.1 m
6.1 »
6.1 m
6.1 B
6.1 B
6.1 8.
6.1 m
6.1 IB
Number
(Obs.)
57
56
116
20
21
21
82
44
83
59
28
29
29
Location (m)
Net
Bottom
Bottom
0.5 IB
3 B
3 B
6 B
Bottom
Bottom
0.5 a
0.5 B
3 IB
3 B
6 m
Bottom
0.5-1 a
1.8-15 m
1.8-15 m
4.8-7.6 HI
7.7-15 »
7.7-15 m
1.8-15 a
1.8-15 B
1.8-15 B
1.8-15 n
4.8-7.6 m
7.7-15 B
7.7-15 m
Correlation
Coefficient Sign.
(r) Level
+.24
+.28
+.18
+.43
+.33
+.55
-.10
+.11
+.07
+.08
-.03
-.31
+.14
0.1
0.05
0.01
0.1
ss
0.05
ss
ss
NS
NS
NS
NS
NS
Age II and Older
1973, 1974
1973, 1974
1973
1973
1973
Trawl
Trawl
Trawl
Trawl
Trawl
7.6 IB
6.1 m
6.1 B
6.1 B
6.1 B
141
142
28
29
29
Bottoa
0.5 B
3 B
3 B
6 B
1.8-15 *
1.8-15 a
4.8-7.6 B
7.7-15 m
7.7-15 B
-.17
-.01
-.28
-.21
-.10
0.05
ss
NS
NS
NS
43
-------
TABLE 13. PERCENT SMELT ABUNDANCE, TURBIDITY AND TEMPERATURE RELATIONSHIPS
Regression constants are given for relationships between the percentage of
smelt sampled in a water column at a depth Y, turbidity X-p and temperature
X2.
Age Class
and
No. Obs.
Age 0
38
36
37*
111
36
37*
37*
39
35
37*
111
Age I
37
35
35*
107
35
35*
35*
38
35
35*
108
Age II and
33
32
32*
102
32
37*
39*
33
32
37*
102
Location
Sampling
Depth
0.5
0.5
0.5
0.5
3.1
3.1
6.1
Bottom
Bottom
Bottom
Bottom
0.5
0.5
0.5
0.5
3.1
3.1
6.1
Bottom
Bottom
Bottom
Bottom
Older
0.5
0.5
0.5
0.5
3.1
3.1
6.1
Bottom
Bottom
Bottom
Bottom
Bottom
Depth
1.8-4.7
4.8-7.6
7.7-15
1.8-15
4.8-7.6
7.7-15
7. 7-15
1.8-4.7
4.8-7.6
7.7-15
1.8-15
1.8-4.7
4.8-7.6
7.7-15
1.8-15
4.8-7.6
7.7-15
7.7-15
1.8-4.7
4.8-7.6
7. 7-15
1.8-15
1.8-4.7
4.8-7.6
7. 7-15
1.8-15
4.8-7.6
7.7-15
7.7-15
1.8-4.7
4.8-7.6
7.7-15
1.8-15
Regression Constants
Intercept
a
+49.526
+28.685
+27.153
+34.694
+34.383
+ 1.684
+ 4.573
+30.769
+88.551
+41.476
+34.694
- 1.233
- 0.682
- 3.019
- 3.606
+12.844
+ 5.893
+ 4.796
+97.567
+88.551
+86.467
+79.586
+16.624
- 0.744
-12.224
+ 0.281
+34.116
+11.463
+20.045
+98.715
+52.190
+69.936
+61.680
Turbidity
bl
-0.083
+0.232
+0.031
+0.108
+0.061
+0.096
+0.210
-0.344
-0.341
-0.132
+0.108
+0.486
+0.102
+0.295
+0.370
+0.418
+0.751
+0.352
-0.485
-0.341
-0.536
-0.337
+0.395
+0.086
+0.472
+0.327
+0.542
+0.628
+0.543
-0.649
-0.339
-0.712
-0.370
Temp
b2
-1.059
-0.833
-0.989
-0.972
+0.018
+1.535
+1.498
+2.420
-0.159
-0.356
-0.972
+0.523
+0.058
+0.384
+0.400
-0.129
-0.410
+1.488
-0.235
-0.159
-1.868
+0.072
-1.112
+0.040
+0.802
-0.172
-1.602
-0.308
+0.253
+0.320
+2.572
-0.818
+1.330
Statistical
Significance
bl
NS
NS
NS
NS
NS
NS
NS
NS
0.05
NS
NS
0.05
0.1
NS
0.01
0.05
0.01
NS
0.05
0.05
NS
0.05
0.01
0.01
0.05
0.01
0.05
0.05
NS
0.01
NS
0.1
0.05
b2
NS
NS
NS
NS
NS
0.1
NS
NS
NS
NS
NS
0.1
NS
NS
0.01
NS
0.05
NS
0.1
NS
NS
NS
0.05
0.1
0.1
0.01
0.05
0.1
NS
0.01
NS
NS
0.05
'"Regression models for 7. 7-15 m water column were used to develop Figure 11.
44
-------
e 3
*—^
J3
Q.
Q 6
E 3
ex
£ 6
Percent
o 4p
i•ir~i
Age 0
. Age I
. Age 3T +
0 FTU
15 FTU
30 FTU
Figure 11. Vertical distribution of smelt in water 9.1 ra deep at
night. Percentage of Age 0, Age I and older smelt located near the sur-
face, midwater and on the bottom is described for three turbidity levels
(FTU).
45
-------
Distribution during the day: Analysis of 7.6 m bottom trawl catches
showed similar trends occurred during daylight hours. Catches at 4.8 to
7.6 m showed abundance declined by 12 and 23% with increase in turbidity
from 0 to 15 and from 15 to 30 FTU respectively. At depths of 7.7 to 15 m,
bottom catches of Age II and older smelt declined by 53 and 95% with in-
creases in turbidity from 0 to 15 and from 15 to 30 FTU respectively
(P<0.05). Estimates of change in catch were calculated by multiple re-
gression models in which temperature, the second independent variable, was
held at 0 C.
Fathometer records made during daylight hours support a conclusion
that reduction in bottom catches resulted from vertical movement by smelt
toward the surface. Percentage of fish targets increased in 6.1 to 9.1 m
and 9.2 to 12.2 m strata with increased turbidity and declined in deeper
water (Figure 12).
Smelt Food and Predation
Food Habits; Copepoda and Cladocera were the major foods of smelt
during all years sampled. Copepoda occurred in over 60% of the smelt
captured and were more abundant in smelt under 110 mm T.L. (Appendix
Table 3). Fish, primarily Age 0 smelt, occurred in approximately 50% of
the older smelt sampled during September 1973 and 1974 (Tables 14 and 15).
Fish were negligible in 1975-1976 smelt diets when sampling occurred off
shore, where Age 0 smelt density was low (Appendix Table 3). During both
1973 and 1974 occurrence of larval fish in the diet of smelt was high in
fish captured off the bottom by midwater trawl and in those captured on
the bottom by bottom trawl. Occurrence of cannibalism in pelagic smelt
increased from an average of 0% during June to 53% during September (Table
15) and for demersal smelt from 2% during June to 46% during September
(Table 14).
Correlations between the number of larval fish or number of inverte-
brates in smelt stomachs and turbidity (FTU) at the collection location
failed to show that turbidity directly influenced the quantity or quality
of food eaten by smelt. The number of food items, the number of larval
fish and the number of invertebrates found in smelt stomachs/g of smelt
were similar for individuals taken from turbid and clear water zones.
Smelt Cannibalism: Estimates of the number of young smelt consumed
daily and monthly by Age II and older smelt/100 m^ were calculated from
estimates of frequency of occurrence of Age 0 smelt in Age II and older
smelt stomachs, multiplied by 2, and the average density of Age II and
older smelt (number/m3) in the 1.8-15 m depth zone. Although limited data
from this study indicates digestion of larval fish would be completed in
approximately 4 h, at 12.6 C, Foltz (1975) found digestion of invertebrates
by smelt at 8 C required 24 h to reach 57%. It was assumed that young fish
could be completely digested in from 12 to 24 h and multiplying by 2 would
roughly correct for the average occurrence of 1. 6 Age 0 smelt found in
stomachs and for the digestion of young smelt in something less than 24 h.
46
-------
100
3.0- 6.0m
o 50
CO
100
6.1-9.1m
50
10
Y= 1.4 + 2.3X
10
lOOr
| 50
CO
9.2 - 12.2 m
7.7+ 4.9 X
100
50
5 10 0
Turbidity (FTU)
> 12.2m
. Y=87.3-(6.9X)
5 10
Turbidity (FTU)
Figure 12. Relationships between the percentage of the total number of
fish targets recorded by fathometer in four depth zones and turbidity.
Simple regression formulas are provided for three significant relation-
ships (P <0.05).
-------
TABLE 14. FOOD OF DEMERSAL SMELT
Food habits are given for smelt exceeding 100 mm captured by 6.1 and 7.6 m bottom
trawls on the bottom during 1973-74. Values represent percentage frequency of
occurrence and percentage by number (in parenthesis) for fish containing food.
Item
Number of Stomachs
Percentage Empty
June
57
24.6
July
78
20.5
August
68
20.6
September
100
32.0
October
120
29.2
Copepoda 88.4(81.8) 75.8(58.2) 25.9( 5.4) 26.5(19.7) 38.8(15.7)
Cladocera (total) 51.2(17.5) 51.6(41.3) 81.5(94.0) 54.4(78.6) 81.2(84.1)
Daphnia sp. 48.8(7.7) 48.4(40.6) 79,6(88.5) 35.3(54.4) 77.6(80.9)
Bosmina sp. 27.9( 9.8) 8.1( 0.2) — 5.9( 1.2) 8.2( 0.8)
Leptodora sp. 2.3(< .1) 8.1( 0.5) 3.1( 0.1) 2.9(< .1)
Amphipoda — 4.8(< .1) — 2.9(< .1) 3.5(< .1)
Pontoporeia affinis — 4.8(< .1) — 2,9(< .1) 3.5(< .1)
Isopoda 2.3( 0.1) 1.6(< .1) — 2.9(< .1) 2.4(< .1)
Mysis relicta 2.3( 0.1) 1.6(< .1) — 2.9(< .1) 2.4(< .1)
Plankton Eggs 7.0( 0.3) 1.6( 0.1) 3,7( 0.1) 7.4( 0.9) 2.4(< .1)
Insecta 11.6( 0.3) 19.4( 0.2) 13.0( 0.4) 1.5(< .1) 1.2(< .1)
Diptera 2,3(< .1) 9.7( 0.1)
Chironomidae 9.3( 0.2) 8.1( 0.1) 13.0( 0.3) — 1.2(< .1)
Ephemeroptera — 1.6(< .1) 1.9(< .1)
Unidentified — — — 1.5(< .1)
Fish 7.0( 0.1) 4.8( 0.1) 13.0( 0.1) 47.1( 0.7) 15.3( 0.1)
Osmerus mordax 2.3(< .1) 3.2( 0.1) 9.2( 0.1) 45.6( 0.7) 11.8( 0.1)
Unidentified 4.7(< .1) 3.2(< .1) 3.7(< .1) 1.5(< .1) 3.5(< .1)
48
-------
TABLE 15, FOOD OF PELAGIC SMELT
Food habits are given for smelt exceeding 100 mm captured by 6,1 m midwater
trawl during 1973-74. Values represent percentage frequency of occurrence and
percentage by number of food items (in parenthesis) for fish containing food.
Item June
Number of Stomachs 45
Percentage Empty 48.9
Copepoda 82.6(70.5)
Cladocera (total) 78.3(28.6)
Daphnia sp. 60.9(11.2)
Bosmina sp. 56.5(16.9)
Leptodora sp . 4.3( 0.5)
Amphipoda (total)
Pontoporeia affinis —
Isopoda
Mysis relicta —
Insecta 4.3( 0.1)
Chironomidae 4.3( 0.1)
Plecoptera —
Hemiptera —
Diptera
Unidentified
Fish 4.3( 0.1)
Osmerus mordax —
Notropis sp.
Cottus sp. —
July
34
25.0
77.8(91.3)
40. 7( 8.3)
37. 0( 7.6)
7.4( 0.3)
7.4( 0.4)
—
—
—
—
7.4( 0.2)
3.7( 0.1)
—
—
3.7( 0.1)
11. 1( 0.2)
11. 1( 0.2)
—
—
August
52
21.2
65.9(23
78.0(75
65.9(72
12. 2( 1
14. 6( 0
—
—
—
—
43. 9( 0
39. 0( 0
2.4(<
—
2.4(<
2.4(<
2.4(<
2,4(<
—
—
September
October
26
34.6
• 3)
.9)
.0)
• 2)
.4)
.8)
.7)
.1)
.1)
.1)
.1)
a)
29.4(47
58.8(47
41.2(46
—
11. 8(
17. 6(
17. 6(
—
—
5.9(
5.9(
—
—
—
—
64. 7(
52. 9(
5,9(
5.9(
0
0
0
0
0
4
4
0
0
.8)
.5)
.6)
.3)
.3)
.3)
.1)
.1)
,3)
.1)
.1)
.1)
42.
84.
84.
1.
3.
3.
3.
3.
3.
7.
1,
4.
8.
5.
65
12.3
K 9.3)
2(90.6)
2(90,6)
8(< .1)
5(< .1)
5(< ,1)
5(< .1)
5(< .1)
5(< .1)
0(< .1)
—
—
8(< .1)
—
3(< .1)
8(< .1)
3(< .1)
—
—
Unidentified 4.3( 0.1) — — — 3.5(< .1)
49
-------
The corrected value represents an estimate of the percentage of Age II and
older smelt which ate the equivalent of one Age 0 smelt daily. Multiplying
this corrected estimate of occurrence by the density of Age II and older
smelt gives an estimate of the number of young smelt consumed/100 m^/day.
Estimates of the number of Age 0 smelt eaten by older smelt suggested
that during June 4.3 young smelt were eaten/100 m^ whereas during September
Age II and older smelt consumed 33.5 Age 0 smelt (Table 16). Differences
were related to varying density of young smelt during the two periods.
Density of Age 0 (Age I were also included during June) in the 1.8-15 m
zone averaged 3.7 during June and 73.7 in September (Table 16; Appendix
Table 4). Monthly mortality resulting from cannibalism during August and
September was estimated by dividing the number of Age 0 smelt eaten by Age
II and older smelt by the average density of Age 0 smelt in the 1.8-15 m
depth zone (Table 16). The analysis suggested cannibalism by both pelagic
and demersal smelt is an important source of mortality during September
when larger smelt were not segregated from Age 0 smelt because of differ-
ence in temperature preference and related distributions. Segregation
during August appeared to be responsible for reduced cannibalism and mor-
tality. Mortality was not estimated for June and July because trawl
catches did not provide reliable estimates of relative abundance. During
the June-July period, small size of Age 0 smelt permitted escapement
through the trawl mesh resulting in low estimates of abundance.
Smelt Predation on Lake Herring; Although western Lake Superior field
samples showed larval fish are an important constituent of smelt diets, no
lake herring were identified in smelt stomachs. However, the probability
that smelt predation on larval herring would be observed, even if it oc-
curred, is low because larval herring are presently rare in western Lake
Superior (Table 9).
An average of 12 herring larvae were captured/min with 1/2 and 1 m
diameter nets on Black Bay, Ontario, establishing the presence of high con-
centrations of larval herring and potential for predation by smelt. The
presence of a yolk-sac and the small size (10 mm average) of larval lake
herring in the catches indicated a major hatch occurred prior to the May
4-5, 1973, sampling period. Analysis of stomachs from 63 Black Bay smelt
captured during the day by trawl and 33 smelt captured at night by gill net
demonstrated smelt will prey on larval lake herring if they are available.
The remains of one larval fish was removed from the stomach of a 158 mm
T.L. smelt captured by gill net. Comparison with herring captured in the
bay showed similarity in size, general shape, and form of the caudal fin
(Figure 13). Presence of 19 myomeres posterior to the anus, stellate chro-
matophores over the yolk-sac and ventral stellate chromatophores in the
anal region are characteristic of larval herring (Fish 1932) and were ob-
served in identifying the item as a larval lake herring.
Laboratory studies with two Age 1+ smelt suggested larval herring were
preferred as food over brine shrimp or herring eggs (embryos). Herring
eggs were not consumed and smelt stopped feeding on brine shrimp when lar-
val herring were available. Predation on larval herring was associated
50
-------
TABLE 16. ESTIMATED CANNIBALISM AND RELATED MORTALITY IN SMELT
Occurrence of Age 0 Smelt (%)
Month
June
Pelagic
Demersal
July
Pelagic
Demersal
August
Pelagic
Demersal
September
Pelagic
Demersal
No.
With
Food
—
2.3
11.1
3.2
2.4
9.2
52.9
45.6
Totalt
No.
—
1.7
8.3
2.5
1.9
7.3
32.5
31.0
Correct. tt
—
3.4
16.6
5.0
3.8
14.6
65.0
62.0
Density
Age II and Older
Smelt „
(No./ 100 m )
1.5
4.0
0.4
6.6
0.3
0.6
0.2
1.8
Predation
No./ 100
Rate
m
(Daily) (Month)
—
0.14
0.07
0.33
0.01
0.09
0.13
1.12
—
4.3
2.1
10.2
0.4
2.7
3.9
33.5
Prey Monthly
Density „ Mort.
(No./lOO m ) (%)
0.9
3.7
1.7
8.5
7.4 5
20.5 13
9.4 41
73.7 45
tpercentage of smelt stomachs with young smelt estimated from the total number of stomachs in a
sample.
ttpercentage occurrence of young smelt in older smelt stomachs. This value represents the estimated
percentage of the total number of Age II and older smelt which consumed an equivalent of one Age 0
smelt daily.
-------
Figure 13. Lake herring larvae removed from 158 mm smelt (above) and
from larval net catches (below).
52
-------
with swimming activity. The two smelt did not feed on inactive herring
larvae. Selection of active prey would result in preference for larval
fish over fish eggs or zooplankton.
During one test in which observations occurred following a 2-1/2 day
period without food, the two Age 1+ smelt consumed 70 herring larvae, 1-3
days post-hatching, in 6 h, a consumption rate of approximately 5 herring/
h/smelt. When previously fed an unrestricted diet of larval herring, the
rate was reduced to 1.5 herring/h/smelt.
Influence of light on smelt predation was tested by measuring the time
required for the two Age 1+ smelt to consume 10 herring larvae. Consump-
tion was faster under ambient laboratory light conditions where the 10
larvae were consumed in 8 min. In comparison, only 8 of 10 larvae were
consumed in 1-1/2 h when the test chamber was completely darkened.
Rate of digestion was defined by examining stomach contents of the two
Age 1+ smelt 1-3/4 and 4 h after feeding at 12.6 C. Seven larvae consumed
by the 62 mm smelt were 30% digested after 1-3/4 h. Six larvae consumed by
a 80 mm smelt were digested to a semi-liquid mass after 4 h.
INFLUENCE OF TURBIDITY ON GROWTH, SURVIVAL
AND DISTRIBUTION OF LARVAL HERRING
The 62-day larval herring bioassay showed red clay turbidity has no
influence on survival or growth at the concentrations normally occurring in
western Lake Superior. Size increased between 1.0 and 1.5 mm per week in-
dependent of turbidity. Mortality was generally low and unrelated to tur-
bidity level in test chambers (Swenson and Matson 1976) .
Herring larvae were active and located near the surface during the
feeding period but moved down from the surface after feeding. Larvae main-
tained locations closer to the surface at higher turbidities (P<0.05;
Swenson and Matson 1976).
INFLUENCE OF TURBIDITY ON WALLEYE
Walleye Abundance and Distribution
Walleye were rare in trawl or gill net catches from depths exceeding
15 m (Tables 9 and 10). Density of walleye past the first year of life
increased with turbidity in the 1.8-15 m zone (Table 17). Density of Age 0
walleye increased with water temperature independent of turbidity.
Multiple regression failed to demonstrate that temperature influenced
distribution of older walleye; however, changes in temperature with sam-
pling depth and time introduced variability which weakened the analysis.
Influence of depth and seasonal temperature variation was controlled by
ranking estimates of turbidity and temperature on a scale of 1 to 3 for
53
-------
TABLE 17. REGRESSION CONSTANTS AND SIMPLE CORRELATIONS FOR WALLEYE
DENSITY, TURBIDITY AND TEMPERATURE RELATIONSHIPS
Regression equations predict number of walleye/100 m . Estimates are based
upon 135, 7.6 m trawl samples collected at depths from 1.8-15 m during 1973
and 1974.
Regression Constants
Species
Intercept
(a)
Turb.
Coef.
bl
Temp.
Coef.
b2
Simple
Correlation
(r)
Turb.
Temp.
Walleye
Age 0
Age 0
+0.2735
-0.2237
-0.00181
-0.00142
-0.03762
-.06
+ .19*
Walleye
Older
Older
+0.0681
-0.0243
+0.00216
+0.00223
-0.00699
+. 23**
+.10
*(P<0.05)
**(P<0.01)
54
-------
catches occurring within two week periods in each of three depth zones
(1.8-4.7, 4.8-7.6 and 7.7-15 m) sampled during 1973 and 1974. Correlation
between ranks suggested that walleye density increased with water tempera-
ture (P <0.1). The rank correlation analysis and low density of walleye in
deep water zones suggest distribution of walleye exceeding Age I (>160 mm
T.L.) is dependent on temperature in addition to turbidity.
Response of Walleye to Laboratory Turbidity Gradients
Analysis of turbidity preference showed a linear increase in time
spent in a gradient chamber section with increased turbidity for both day
and night observations (Table 18; Figure 14) demonstrating that walleye
preferred the highest available turbidity, which exceeded levels charac-
teristic of western Lake Superior. Movement between chamber section showed
walleye were more active during the dark phase of the light cycle under
clear water conditions (t-test; P<0.01) but became day active in turbidity
gradients (t-test; P <0.05). Differences in activity patterns explain why
the association with turbidity was stronger for night observations (Table
18). Increased activity during the light period tended to decrease the
amount of time walleye located in turbid water because movement between
chamber sections required leaving areas of higher turbidity.
Walleye Feeding
Walleye fed almost exclusively on smelt in western Lake Superior.
Larger individuals (>J200 mm T.L.) utilized adult and juvenile smelt as
their primary food source from June through August (Table 19). Age 0
smelt represented the primary food of larger walleye during September and
October and of smaller walleye (£200 mm T.L.) from June through October
(Table 19). Estimates of food consumption rate during five days, July
through September 1973, showed walleye ate an average of approximately 2%
of their weight per day (Table 20). Comparison with estimates of daily
food consumption for Lake of the Woods and Shagawa Lake, Minnesota, wall-
eye (Swenson 1977) showed consumption by Lake Superior walleye was lower.
Relatively low density of Lake Superior prey populations represents the
probable cause for the reduced consumption rates (Swenson 1977). Estimates
of feeding by walleye/100 m^ indicate they consume an average of 0.5 indi-
vidual smelt/day during July, 0.6 smelt/day during August and 1.0 smelt/
day during September (Table 20). Changes in the number of smelt eaten per
day were related to a decrease in the average size of smelt consumed by
walleye during September. Summation of daily consumption rates for all
days in a month and comparison with average smelt densities (number/100 m )
in the 1.8 to 15 m depth zone suggests predation by walleye may be an impor-
tant source of smelt mortality (Table 20). However, because walleye dis-
tribution is restricted in comparison with that of smelt. accurate inter-
pretation of effects on smelt populations is not possible.
55
-------
TABLE 18. RESPONSE OF WALLEYE AND LAKE TROUT
IN TURBIDITY GRADIENTS
Change in the amount of time (Y = seconds) juvenile walleye
or lake trout selected a chamber section in relation to in-
creased turbidity is estimated by the formula: Y = a + bx.
Estimates are based on 56 observations on 14 fish of each
species.
Species
and
Time
Walleye
Day
Night
Lake Trout
Day
Night
Regression
Intercept
(a)
- 3.1418
-10.1108
15.2978
21.4220
Constants
Slope
(b)
+1.2163
+2.7336
-3.3578
-4.6381
Value of
F
3.1
10. 6 *
19.8*
31.1*
*(P < 0.01)
56
-------
400
200
CO
o
Q>
IO
0>
»-
0)
0)
e
-200
-400 -
-40
X
X X
X
-20
20
40
Turbidity Difference (AFTU)
Figure 14. Response of juvenile walleye to turbidity gradients. The
relationship is predicted using regression constants from Table 18.
57
-------
TABLE 19. FOOD OF WALLEYE
Values represent percentage frequency of occurrence and percentage by weight of food items (in parenthe-
sis) for fish containing food captured during 1973-1976.
en
00
Item
Older Walleye (>200 mm
Number of Stomachs
Percentage Empty
Insecta
Hexagenia sp.
Chironomidae
Fish
Osmerus mordax
Age 0
Age I
Older
Unidentified
June
TL)
43
25.5
6.2( 4.3)
3.1( 0.2)
3.1( 4.1)
100.0(95.5)
46. 8( 1.0)
75.0(78.2)
15.6(16.3)
18. 0( 0.1)
July
53
50.9
3.8( 0.1)
3.8( 0.1)
—
100.0(100.0)
11. 5( 0.4)
50. 0( 7.0)
53. 8( 92.6)
3.8( <0.1)
August September
88 78
40.9 28.2
--
—
—
100.0(100.0) 100.0(99.9)
3.8( 0.3) 48. 4( 8.7)
65. 3( 31.5) 43.9(26.5)
38. 4( 68.2) 30.3(64.7)
10. 6( 0.1)
October
67
25.3
—
—
—
100.0(99.
48. 0( 8.
24.0(20.
28.0(71.
16.0(<0.
9)
9)
0)
0)
1)
(continued)
-------
TABLE 19. (CONTINUED)
Item
Young Walleye (<200 mm
Number of Stomachs
Percentage Empty
Cladocera
Fish
Osmerus mordax
Age 0
Age I
Older
Unidentified
June
TL)
11
54.5
—
100.0(100.0)
40. 0( 7.9)
60. 0( 40.6)
20. 0( 51.5)
trace
July
38
52.6
5.8( <0
92.2(100
38. 8( 23
27. 7( 9
27. 7( 67
11. 7 ( <0
.1)
.0)
.1)
.6)
.3)
.1)
August September
56 37
40.3 48.6
10.0(<0.1)
89.9(98.4) 100.0(93.9)
36. 6( 9.7) 73.6(43.8)
50.0(83.2) 31.5(50.1)
3.3( 5.5)
20. 0( 0.6). 36. 8( 6.0)
October
29
44.8
—
75.0(94.6)
75.0(94.6)
—
—
25. 0( 5.3)
-------
TABLE 20. WALLEYE FOOD CONSUMPTION AND PREY DENSITY DURING 1973
Smelt Eaten by Walleye
Month
July
August
September
Sample
Size
(No.)
35
31
22
Mean
Wt.
(g)
129
240
148
Total
Consumption
(mg/g/day)
22.8
19.5
21.0
Smelt
(mg/m )
157
169
317
Density
(No./ 100 m3)
8.3
52.2
46.5
Daily
Consumption
(No./ 100 m3)
0.5
0.6
1.0
Monthly
Consumption
(No./ 100 m3)
15.5
18.6
30.0
-------
INFLUENCE OF TURBIDITY ON LAKE TROUT
Lake Trout Distribution in Lake Superior
Lake trout sampled during 1975 and 1976 were generally limited to
water exceeding 15 m deep. Distribution of lake trout, as described by
bottom trawl catches, (9.5 m net) appeared to be influenced by turbidity
although turbidity was low during 1975 and 1976. Catches indicated that
the number of lake trout/100 m^ in water from 15 to 40 m deep increased
exponentially with water clarity (Figure 15; P<0.05).
Response of Lake Trout to Laboratory Turbidity Gradients
The time juvenile lake trout resided in a gradient chamber section
descreased with turbidity (Figure 16). The inverse relationship between
lake trout residence time in a chamber section and turbidity was stronger
for observations made under dark conditions (Table 18). Differences be-
tween day and night appear to be explained by increased activity during
the light phase of the 24 h cycle (t-test; P<0.001). Activity during the
daylight period resulted in more time being spent in turbid section of the
gradient chamber as a result of increased movement. Activity of lake trout
in gradient chambers was higher than that of walleye.
Experiments with 15 lake trout (145 to 240 mm T.L.) held at 0, 6, 9,
28 and 54 FTU in 3 liter electrode chambers showed that even low turbidity
resulted in increased activity. Activity was higher in turbid water and
ranged from 1.5 to 7.3 times the activity of control fish held at 0 FTU
(Table 21). The results suggest that lake trout are sensitive to turbidity
as low as 6 FTU.
INFLUENCE OF TURBIDITY ON OTHER FISH SPECIES
Bottom trawl catches of longnose sucker, white sucker, and troutperch
indicated these species were abundant in western Lake Superior. Regres-
sion analysis showed that red clay turbidity did not influence their dis-
tribution. The data suggested that distribution of white suckers was
limited to warmer water in contrast to longnose suckers which appeared to
concentrate in cooler water; however, differences were not significant
(P> 0.05).
61
-------
0.03
m
E
O
O 0.02
o
o
0.01
0.00
Y-5xlO"IOx 1.20 X
20 40 60
Turbidity (% Transmittance)
80
100
Figure 15. Relationship between percentage light transmittance (turbidity index)
and the number of lake trout captured near the bottom at depths between 15 and 40 m.
-------
400
200
o
0)
in
o>
o
c
0)
k.
0>
-200
-400
-40
-20 0 20
Turbidity Difference (AFTU)
40
Figure 16. Response of juvenile lake trout in laboratory turbidity
gradients. The relationship is predicted using the regression constants
from Table 18.
63
-------
TABLE 21. LAKE TROUT ACTIVITY AND TURBIDITY
Variation
Acclimation Test Number Time in Activity
Time Duration Turbidity of Active Between Fish
(h) (h) (FTU) Fish % (SD)
1.5 1.1 54 2 46.1 6.8
4.0 0.5 28 3 30.6 19.9
76.0 1.5 9 3 28.2 11.5
49.0 4.5 6 3 18.6 3.9
1.5-76 0.5-4.5 0 4 13.5 12.5
(control)
Activity''"
vs.
Control
1.5
7.3
1.8
5.2
—
^Each test was performed with one fish in clear water (control). Division
of percentage activity for lake trout subject to turbidity by activity in
controls gave an estimate of the increase in activity associated with
turbidity levels in these tests.
64
-------
DISCUSSION
Several studies have shown that survival of advanced life stages of
many fish species is not influenced by naturally occurring turbidity levels
(see reviews by Cordone and Kelley 1961, Herbert and Merkens 1961). The
larval herring bioassay conducted as part of this project indicates that
natural levels of red clay turbidity in western Lake Superior have no di-
rect influence on survival or growth during the sensitive larval stage.
Although direct toxic effects were not identified, other phases of this
project provide evidence that red clay turbidity is important to fish pro-
duction in western Lake Superior.
The influence of red clay turbidity on fish production results from
behavioral responses apparently brought about by reduced light penetration
in turbid water. Light intensity was reduced significantly even at low
levels of turbidity due to selective adsorption of shorter wavelengths.
The highest zooplankton densities occurred in surface waters (0.5-
6.0 m) at stations characterized by red clay turbidity. High zooplankton
densities appear to result from vertical migrations in response to reduced
light penetration. Increased primary production may be stimulated by re-
lease of nutrients from red clays (Bahnick 1975). Higher primary produc-
tion near the surface in turbid waters could also be responsible for high
zooplankton densities.
Larval herring concentrated closer to the surface of laboratory test
chambers at higher turbidity. Field observations demonstrate that smelt
concentrate near the surface in turbid water (Figures 11 and 12) where
plankton density is highest (Figure 8). Concentration of zooplankton and
the fish which prey upon them in the same depth zone may be responsible
for the former high abundance of lake herring in the Duluth-Superior area
and present high abundance of smelt. Both species rely primarily on zoo-
plankton as a food source. Because smelt feeding increases at dawn and
dusk (Ferguson 1965), red clay turbidity may also influence smelt produc-
tion by extending the period of low light intensity which stimulates feed-
ing. Comparatively good growth of Age 0 smelt in turbid water stations is
evidence that turbidity has a positive influence on smelt feeding.
Although red clay turbidity may have promoted production of the his-
torically significant Duluth-Superior area lake herring stock by increasing
zooplankton availability, the results of this investigation suggest that
turbidity contributed to increased predation on larval herring by intro-
duced smelt and was indirectly responsible for the extensive and rapid lake
herring decline. Absence of larval lake herring from smelt stomachs is not
65
-------
unexpected with the present depressed herring stock in western Lake
Superior. Commercial fishermen maintain that occurrence of young herring
in smelt stomachs was common in the Duluth-Superior area prior to the her-
ring decline.3 Samples from Black Bay, Ontario, collected during 1973
verified that smelt prey on larval herring when they are available. Sub-
sequent studies by the U.S. Bureau of Sports Fisheries and Wildlife and the
Ontario Ministry of Natural Resources on Black Bay, Ontario, where herring
are still abundant, showed that during a short period, apparently associ-
ated with the hatching of larval herring, 28% of the smelt population fed
upon larval herring. Predation on larval herring was size dependent and
85% of the adult smelt contained larval herring in their stomachs. The
average number of herring found in adult smelt stomachs was 35.8 (Selgeby
et_ al., MS). Larval herring were not found in the stomachs of smelt cap-
tured by bottom trawls after the period, which lasted approximately ten
days (Selgeby £t_ al., MS). Failure of smelt to prey on larval herring
after the period of hatching may result from changes in distribution.
Herring larvae migrated toward the surface after hatching, whereas smelt
concentrated on the bottom in clear water.
In the Duluth-Superior area, observations on smelt distributions made
during this investigation and information on distribution of larval Core-
gonids (Anderson 1969) indicate that turbidity probably caused increased
contact between smelt and larval lake herring. Anderson (1969) captured
7,109 Coregonid larvae in 673 larval net tows during 1966-1968 at stations
near Duluth-Superior and in the Apostle Islands. The catches show larval
herring densities were generally higher in water over 18 m deep and that
larvae distributed throughout the water colum during April, May and early
June but concentrate from 12 m to the surface during late June and July.
Herring were not captured after July (Anderson 1969). This project demon-
strated that during periods of red clay turbidity smelt move into the upper
12 m depth stratum. Percentage occurrence of young fish in smelt diets
was high when they became pelagic although density of larval smelt was
lower at midwater locations than near the bottom. Selection of larval
fish by smelt was demonstrated in the laboratory where young smelt elected
to feed on mobile larval herring over brine shrimp or herring embryos.
Estimates of the number of young smelt consumed by older age groups and
comparison with young smelt densities indicated cannibalism represents a
significant source of mortality. Location of herring larvae further off-
shore (Anderson 1969) than young smelt increases the probability of con-
tact between pelagic Age II and older smelt during turbid water periods
and for predation on larval herring.
Analysis of commercial catch data shows inverse relationships exist
between lake herring and smelt abundance in several of the Great Lakes, as
summarized by Christie (1974) who concludes that smelt resulted in wide-
spread decline of Great Lakes herring stocks. Anderson and Smith (1971b)
Personal communication, Mr. Stanley Sivertson, President, Sivertson
Fishery Company, Duluth, MN.
66
-------
show that important Duluth-Superior herring stock declined faster than
herring inhabiting the clear waters of Lake Superior. They identified a
significant negative relationship between herring and smelt abundance in
western Lake Superior which was confirmed by this study. Although Anderson
and Smith's (1971b) general conclusion is supported by the results of this
study, they suggest that food competition was the primary mechanism result-
ing in change. High zooplankton densities, overlapping larval herring and
smelt distribution, and significant predation by smelt on larval fish sug-
gest that predation by smelt on larval herring rather than food competition
contributed to the lake herring decline. In addition, ongoing surveys of the
remnant herring stock show increasing dominance of older age groups and
increased growth rates (Lake Superior Herring Subcommittee 1973). These
observations would suggest that recruitment is failing in the presence of
an adequate food resource and that food competition is not significant.
This study did not determine whether lake herring are attracted to tur-
bid water areas. Lawrence and Scherer (1974) found white fish (Coregonus
clupeaformis) in laboratory turbidity gradients preferred water with sus-
pended drilling fluid concentrations of up to 1,000 ppm over clear water.
In western Lake Superior, herring larvae would be subject to the same cur-
rents which distribute red clay. Sydor (1975) and Startz £t_ al. (1977)
found currents vary with wind condition and form back eddies which distri-
bute turbid water along the Wisconsin south shore under northwesterly,
northeasterly or northerly winds. Turbid water spreads out along the cen-
tral axis (Minnesota District M-l) under easterly or westerly winds. High
zooplankton densities and avoidance of red clay zones by lake trout repre-
sent environmental pressures which should result in increased survival of
lake herring adapted to inhabit zones of red clay turbidity—a mechanism
which lost its survival advantage with the addition of smelt to the system.
Abundance of lake trout was lower in western Lake Superior as a result
of red clay turbidity. Net catches and laboratory results show lake trout
are sensitive to low concentrations of red clay and partially avoid turbid
water zones. Gill net catches showed differences in abundance between
clear and turbid water stations were greater for juvenile lake trout.
This variation may be attributable to sample size, location of lake trout
plantings or to the persistence of red clay turbidity in near bottom water
in the Duluth-Superior area. Turbidity of near bottom waters could isolate
young lake trout from their principal food supply which Anderson and Smith
(1971c) found was benthic crustaceans. Older lake trout fed primarily on
smelt which became pelagic during turbid water periods. Pelagic smelt
probably represent a highly available food resource for larger lake trout.
Red clay turbidity promotes production of walleye by reducing light to
levels acceptable for feeding throughout the day and by causing smelt to
become pelagic. Preference of walleye for turbid water was demonstrated by
higher densities in turbid water areas and higher residence times in turbid
water sections of laboratory gradients. Preference for turbidity is par-
tially explained by adaptation of the eye of Stizostedion sp. for reduced
light conditions (Ali and Anctil 1968). Other studies have demonstrated
that walleye are usually crepuscular and night active predators which
67
-------
require high densities (>400 mg/m ) of pelagic prey to maintain food con-
sumption between 3 to 4% body weight, the level for maximum food consump-
tion, food conversion efficiency and growth (Swenson and Smith 1973;
Swenson 1977). Walleye were more active during high light periods in
laboratory turbidity gradients and at night under clear water conditions,
suggesting that the western Lake Superior population may have developed an
activity pattern correlated with the low light conditions and pelagic dis-
tribution of prey resulting from periodic turbid water conditions.
Conditions encountered in Lake Superior made accurate estimation of
walleye feeding rates impractical during most sampling days; however,
estimates of consumption and prey density developed for five days during
1973 showed prey densities and food consumption were below optimum even in
the turbid water zone which apparently represents an area of the lake where
food availability is relatively high. Low abundance of walleye at clear
water stations could result because food availability is near the minimum
required for good production.
Walleye predation has been identified as the primary factor controlling
survival of yellow perch in Oneida Lake (Forney 1971), Lake of the Woods,
Minnesota, (Swenson and Smith 1976) and Shagawa Lake (Swenson 1977). Avail-
able information on walleye feeding rates and smelt density in Lake Superior
provide some indication that walleye predation influences smelt survival.
However, the major differences in distribution of walleye and smelt make
interpretation of predation effects impractical.
68
-------
REFERENCES
All, M. A. , and M. Anctil. 1968. Correlation entre la structure retinienne
et 1'habitat chez Stizostedion vitreum vitreum et _S_. canadense. J.
Fish. Res. Board Can. 25:2001-2003.
American Public Health Association. 1971. Standard Methods for the Examin-
ation of Water and Wastewater. 13th Ed. APHA, New York. 874 pp.
Anderson, E. D. 1969. Factors affecting abundance of lake herring
(Coregonus artedii LeSueur) in western Lake Superior. Ph.D. Thesis,
University of Minnesota. 316 pp.
Anderson, E. D. , and L. L. Smith, Jr. 1971a. An automatic brine shrimp
feeder. Prog. Fish. Cult. 33:118-119.
Anderson, E. D., and L. L. Smith, Jr. 1971b. Factors affecting abundance
of lake herring (Coregonus artedii LeSueur) in western Lake Superior.
Trans. Am. Fish. Soc. 100:691-707.
Anderson, E. D., and L. L. Smith, Jr. 1971c. A synoptic study of food
habits of 30 fish species from western Lake Superior. Univ. of Minn.
Agric. Exp. Stat. Tech. Bull. 279. 199 pp.
Bahnick, D. A. 1975. Chemical effects of red clays on western Lake
Superior. Univ. of Wis.-Superior Final Rep. EPA Project #4-005169.
(Mimeo)
Bahnick, D. A. 1977. The contribution of red clay erosion to ortho-
phosphate loading into southwestern Lake Superior. J. Environ. Qual.
6:217-221.
Bailey, M. M. 1964. Age, growth, maturity and sex composition of American
smelt, Osmerus mordax (Mitchill), of western Lake Superior. Trans.
Am. Fish. Soc. 93:382-395.
Brooks, J. L. 1957. The systematics of North American Daphnia. Memoirs,
Conn. Acad. Arts Sci. 13. 180 pp.
Christie, W. J. 1974. Changes in the fish species composition of the
Great Lakes. J. Fish. Res. Board Can. 31:827-854.
Cordone, A. J., and D. W. Kelley. 1961. The influence of inorganic sedi-
ments on the aquatic life of streams. Calif. Fish and Game 47:189-228.
69
-------
Doan, K. H. 1941. Relation of sauger catch to turbidity in western Lake
Erie. Ohio J. Sci. 41:449-452.
Drummond, R. A., and W. F. Dawson. 1970. An inexpensive method of simu-
lating diel patterns of lighting in the laboratory. Trans. Am. Fish.
Soc. 99:434-435.
Eddy, S., and A. C. Hodson. 1962. Taxonomic Keys to the Common Animals of
the North Central States. Burgess, Minneapolis, 162 pp.
Edmondson, W. T. 1959. Freshwater Biology Sec. Ed. Wiley and Sons, N.Y.
1248 pp.
Ferguson, R. G, 1965. Bathymetric distribution of american smelt,
Osmerus mordax, in Lake Erie. Univ. of Mich. Great Lakes Res. Div.
Pub. 13:47-60.
Fish, M. P. 1932. Contributions to the early life histories of sixty-two
species of fishes from Lake Erie and its tributary waters. Bull. U.S.
Bureau of Fish. 10:293-398.
Foltz, W. J. 1974. Food consumption and energetics of the rainbow smelt,
Osmerus mordax (Mitchill), in Lake Michigan. M.S. Thesis, UW-
Milwaukee. 102 pp.
Forney, J. L. 1971. Development of dominant year-classes in a yellow
perch population. Trans. Am. Fish Soc. 100:739-749.
Herbert, D. W. M., and J. C. Merkens. 1961. The effects of suspended
solid materials on survival of trout. Int. J. Air Wat. Poll. 5:46-55.
Hile, R. 1962. Collection and analysis of commercial fishery statistics
in the Great Lakes. Great Lakes Fish. Comm. Tech. Rept. 5. 31 pp.
King, G. R., and B. L. Swanson. 1974. Progress report on fish management
on Lake Superior. WDNR, Madison. 41 pp. (Mimeo)
Lake Superior Herring Subcommittee of the Lake Superior Committee of the
Great Lakes Fish Commission. 1973. A summary finding report on Lake
Superior Herring (Coregonus artedii) populations. 18 pp. (Mimeo)
Langlois, T. H. 1941. Two processes operating for the reduction in abun-
dance or elimination of fish species from certain types of water
areas. Trans. Sixth No. Am. Wildl. Conf. 189-201.
Lawrence, M., and E. Scherer. 1974. Behavioral responses of whitefish
and rainbow trout to drilling fluids. Can. Fish, and Mar. Serv. Tech.
Rep. No, 502. 47 pp.
70
-------
McKenzie, R. A. 1958. Age and growth of smelt, Osmerus mordax (Mitchill),
of the Miramichi River, New Brunswick, J. Fish. Res. Board Can. 15:
1313-1327.
Red-Clay Inter-Agency Committee. 1972. Erosion and sedimentation in the
Lake Superior Basin. 79 pp.
Schindler, D. W. 1969. Two useful devices for vertical plankton and water
sampling. J. Fish. Res. Board Can. 26:1948-1966.
Selgeby, J. H., W. R. MacCallum, and D. W. Swedberg. (MS) Predation by
rainbow smelt (Osmerus mordax) on lake herring (Coregonus artedii) in
Lake Superior. (Unpublished manuscript) 22 pp.
Smith, S. H., H. J. Buettner, and R. Hile. 1961. Fishery Statistical dis-
tricts of the Great Lakes. Great Lakes Fish. Comm. Tech. Rept. 2.
24 pp.
Spoor, W. A. , and R. A. Drummond. 1972. An electrode for detecting move-
ment in gradient tanks. Trans. Am. Fish. Soc. 101:714-715.
Spoor, W. A., T. W. Neiheisel, and R. A. Drummond. 1971. An electrode
chamber for recording respiratory and other movements of free-swimming
animals. Trans. Am. Fish. Soc. 100:22-28.
Startz, K., R. Clapper, and M. Sydor. 1976. Turbidity sources in Lake
Superior. J. Great Lakes Res. 2:393-401.
Swenson, W. A. 1977. Food consumption of walleye and sauger in relation
to food availability and physical conditions in Lake of the Woods,
Minnesota, Shagawa Lake and Western Lake Superior. J. Fish. Res.
Board Can. 34:1643-1654.
Swenson, W. A., and M. L. Matson. 1976. Influence of turbidity on sur-
vival, growth and distribution of larval lake herring (Coregonus
artedii). Trans. Am. Fish. Soc. 105(2):542-546.
Swenson, W. A., and L. L. Smith. 1973. Gastric digestion, food consump-
tion, feeding periodicity and food conversion efficiency in walleye
(Stizostedion vitreum vitreum). J. Fish. Res. Board Can. 30:1327-1336.
Swenson, W. A., and L. L. Smith. 1976. Influence of food competition,
predation and cannibalism on walleye (Stizostedion vitreum vitreum)
and sauger (S_. canadense) populations in Lake of the Woods, Minnesota.
J. Fish. Res. Board Can. 33:1946-1954.
Sydor, M. 1975. Red clay turbidity and its transport in western Lake
Superior. Final Report. U.S. E.P.A. Grant R-005175-01. (Unpublished)
71
-------
Van Oosten, J. 1945. Turbidity as a factor in the decline of Great Lakes
fish with special reference to Lake Erie. Trans. Am. Fish. Soc. 75:
281-322.
72
-------
Temperature is
and 1976.
APPENDIX TABLE 1. TEMPERATURE AND TURBIDITY PROFILES
3C, turbidity is FTU (in parenthesis) in waters exceeding 15 m sampled during 1975
Date
1975
June 21
June 23
July 1
July 4
July 10
July 15
July 16
July 16
July 22
July 30
Aug. 13
Aug. 18
Sep. 5
Oct. 13
1976
May 18
May 27
June 8
June 10
July 2
July 29
July 30
Aug. 11
Stat.
No.
2
2
4
4
2
4
2
2
2
2
2
4
4
2
4
2
4
2
4
2
4
4
Bottom
Depth
(m)
24
24
21
25
24
25
23
24
23
24
24
27
24
18
24
19
27
28
27
24
17
26
Sampling
<1
10(9)
12(5)
16(7)
20(8)
14(6)
18(6)
14(7)
14(7)
8(2)
21(4)
18(2)
18(3)
13(1)
11(1)
7(1)
.13(5)
14(1)
14(1)
16(3)
19(1)
20(3)
21(1)
3
10(9)
10(7)
14(7)
15(3)
14(6)
18(6)
10(7)
10(7)
8(2)
17(4)
18(2)
18(3)
13(1)
1KD
7(1)
11(2)
11(2)
13(1)
16(3)
18(1)
19(3)
19(1)
6.1
10(9)
10(7)
13(8)
15(3)
13(5)
15(4)
7(5)
7(5)
7(2)
11(3)
17(2)
18(3)
13(1)
11(1)
5(2)
10(2)
7(1)
13(1)
11(1)
18(1)
19(2)
19(1)
9.1
10(9)
9(6)
13(7)
13(3)
10(3)
8(4)
6(4)
6(4)
6(2)
8(3)
7(2)
7(1)
13(1)
11(1)
5(3)
9(1)
6(1)
10(1)
10(1)
18(1)
18(1)
16(1)
Depth
12.2
10(9)
9(6)
10(4)
11(3)
8(3)
6(3)
5(7)
5(7)
6(2)
7(5)
5(4)
5(1)
13(1)
11(1)
5(3)
8(2)
6(1)
6(1)
9(1)
18(1)
18(1)
13(1)
15.2
10(9)
8(5)
8(7)
8(1)
6(3)
5(3)
5(8)
5(8)
5(6)
7(6)
4(4)
5(2)
13(3)
1KD
5(3)
5(1)
6(1)
6(1)
8(1)
18(1)
17(1)
9(1)
18.3
9(8)
6(5)
7(7)
6(3)
5(5)
4(3)
5(8)
5(8)
5(6)
7(6)
4(4)
4(3)
12(5)
11(1)
5(3)
—
6(1)
5(1)
6(1)
17(1)
—
7(2)
21.3
8(8)
—
—
6(5)
4(7)
4(3)
5(8)
5(8)
5(6)
7(7)
4(5)
4(8)
12(6)
—
—
—
6(1)
5(4)
5(3)
4(1)
—
6(2)
24.4
—
—
—
5(6)
4(7)
—
5(8)
5(8)
—
7(8)
4(5)
4(8)
11(8)
—
—
—
3(1)
4(5)
4(6)
—
—
6(3)
-------
APPENDIX TABLE 2. LIGHT INTENSITY AND EXTINCTION IN RELATION TO TURBIDITY AND DEPTH
Average Bottom
Turbidity Station Depth
(FTU) Date Number (m)
.465
.465
.465
.739
1.01
1.01
1.28
1.50
1.56
1.83
2.10
2.92
2.92
3.20
3.89
4.14
4.14
4.98
5.29
6.31
10.70
12.33
7-29-76
7-29-76
5-27-76
6-8-76
10-10-75
7-30-76
5-20-76
7-29-76
8-11-76
8-11-76
5-18-76
10-10-75
10-16-75
5-18-76
6-10-76
5-18-76
10-10-75
5-27-76
7-2-76
6-16-76
10-10-75
10-16-75
2
1
2
4
2
4
4
2
4
4
4
2
4
4
4
4
3
4
4
4
3
4
22.6
18.0
23.2
26.5
23.2
16.7
10.1
22.6
8.2
24.7
23.8
22.6
18.3
13.4
5.2
6.7
19.5
9.8
4.6
7.6
15.2
16.8
Light at Sampling Depths (m)
Deck 246
74.1 35.6 20.1 10.7
78.4 35.6 22.7 12.2
64.1 14.5 5.56 2.48
41.3 9.40 4.84 2.78
28.9 9.58 5.02 3.02
69.8 2.95 — 0.48
49.9 6.13 1.24 0.25
83.4 4.48 2.74 1.48
78.4 13.2 3.33
78.4 35.9 23.1 15.8
92.6 38.5 22.2 12.4
31.2 5.51 1.52 0.56
70.1 12.4 3.25 0.98
88.4 2.99 0.35 70-3*
82.6 6.69 1.50 0.23
89.8 13.2 2.48 0.90
21.4 0.64 81-3* 20-3*
34.2 1.50 0.14 15-3*
85.5 7.41 1.03 0.14
45.6 1.75 0.19 27-4*
18.7 0.18 46-4" 25-5*
32.8 0.40 93-4* 56-5*
8
4.99
10.7
1.23
1.62
1.70
0.28
0.06
0.76
—
11.1
6.70
0.18
0.29
20-3*
—
—
40-4*
20-4*
—
—
—
—
10
3.25
—
0.64
1.00
0.98
0.24
0.01
0.41
—
7.84
3.46
59-3*
87-3*
70-4*
—
—
11-4*
—
—
—
—
—
12
2.01
—
0.31
0.61
0.57
0.14
—
0.25
—
5.27
1.97
15-3*
29-3*
20-4*
—
—
51-5*
—
—
—
—
—
14
1.71
—
0.15
0.36
0.32
0.04
—
0.14
—
3.42
1.13
45-4*
77-4*
—
—
—
21-5*
—
—
—
—
—
16
1.49
—
60-3*
0.21
0.18
36-3*
—
84-3*
—
2.05
0.67
15-4*
42-5*
—
—
—
77-6*
—
—
—
—
— —
18
1.04
—
30-3*
0.13
0.11
—
—
45-3*
—
1.17
0.38
56-5*
13-5*
—
—
—
34-6*
—
—
—
—
—
Ext.
Coef.
K
.2157
.2653
.3624
.2627
.2797
.3079
.7765
.3038
.6812
.2231
.2771
.5293
.6292
.6971
.8194
.6106
.6292
1.103
.9835
1.088
1.676
1.966
''Values given with negative exponents indicating the number of places the decimal point is to be
adjusted to the left: for example, 60-3 is .060.
-------
APPENDIX TABLE 3. FOOD OF SMELT
Food habits are for smelt captured during 1974-1976 by 6.1 or 7.6 m
trawl. Smelt diets are described as percentage frequency of occur-
rence and percentage by number of items (in parentheses). Frequency
of occurrence values are based on all stomachs analyzed.
Year
Length
1974
<110 mm
1974
> 110 mm
1975-1976
Number of Stomachs
132
254
211
Copepoda
Cladocera
Daphnia spp.
Bosmina spp.
Leptodora sp.
Amphipoda
Pontoporeia affinis
Isopoda
Insecta
Fish
Osmerus mordax
Other Fish
Unidentified
67(64.6)
39(29.6)
14( 5.4)
6( 0.2)
2( <.l)
9( 0.2)
>K <-D
43(43.7)
41(39.0)
17(14.7)
4( 0.3)
-------
APPENDIX TABLE 4. CATCH BY SAMPLING DEPTH AND MONTH
o
Catch is number/100 m estimated from seine (0.5 to 1 m) and 6.1 m trawl samples collected during
1973 and 1974 at stations 2 and 4.
June
Species
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jv)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
Ictalurus sp.
Ambloplites rupestris
0.5-1
5.35
33.10
12.00
1.90
0.30
0.20
1.65
0.03
5.72
1.8-4.7
0.40
5.40
4.95
1.00
0.20
0.15
0.40
0.35
< .01
< .01
4.8-7.6
0.30
3.75
3.35
1.40
0.01
0.15
0.05
1.15
0.05
Sampling
7.7-15
0.35
0.95
0.85
0.90
1.40
0.05
Depth (m)
0.5-1
62.05
22.95
7.50
0.05
0.25
0.05
0.05
5.55
0.75
0.15
0.17
July
1.8-4.7
7.40
11.50
9.00
0.25
0.30
0.15
0.10
1.65
0.15
< .01
4.8-7.6
0.75
4.75
8.60
0.30
< .01
0.05
0.10
0.60
0.10
7.7-15
0.25
0.90
1.30
0.10
0.05
0.80
0.05
< .01
(continued)
-------
APPENDIX TABLE 4. (continued)
Sampling
Depth (m)
August
Species
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jy)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Castostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
Ictalurus sp.
Ambloplites rupestris
0.5-1
279.2
15.75
1.05
14.00
26.10
0.15
19.35
0.25
0.45
0.75
0.12
0.55
5.719
.061
1.8-4.7
24.35
3.40
0.40
0.20
1.15
0.15
0.45
0.05
0.10
0.90
0.05
< .01
< .01
4.8-7.6
30.05
3.00
0.45
0.15
0.05
0.15
< .01
0.50
3.15
0.25
< .01
< .01
7.7-15
6.80
2.30
0.55
0.15
0.05
0.05
< .01
0.20
0.20
0.05
< .01
0.5-1
597.1
2.05
0.40
0.45
80.70
0.80
0.08
0.17
September
1.8-4.7
129.1
4.35
1.75
0.50
0.75
0.05
0.15
2.90
0.30
< .01
< .01
< .01
4.8-7.6
45.95
7.35
2.20
0.10
0.25
0.15
0.55
0.30
7.7-15
46.05
2.20
0.65
0.25
0.05
0.05
0.05
0.70
1.80
< .01
(continued)
-------
APPENDIX TABLE 4. (continued)
oo
Sampling
Depth (m)
October
Osmerus mordax (yy)
0. raordax (yl)
0. taordax (jv)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
Ictalurus sp.
Ambloplites rupestris
0.5-1
38.05
1.30
.20
1.20
.05
20.15
.10
.13
1.8-4.7
5.25
.55
.05
.15
.50
.05
.55
.40
.50
4.8-7.6
4.65
.40
.05
.20
.05
.05
.15
.55
.05
7.7-15
5.35
.25
.05
.05
.10
.85
.40
-------
VD
APPENDIX TABLE 5. CATCH BY SAMPLING DEPTH AND MONTH
O
Catch is number/100 m estimated from 7.6 m trawl samples collected during 1973 and 1974 at
stations 2 and 4.
Sampling Depth (m)
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jv)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notiropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
1.8-4.7
1.70
1.95
0.85
0.10
0.25
0.05
0.20
0.05
0.05
0.05
June
4.8-7.6
1.25
1.50
2.10
0.10
0.05
0.10
0.45
1.95
< .01
7.7-15
0.05
0.50
0.80
0.50
0.10
0.65
0.50
1.8-4.7
0.30
1.70
5.50
0.25
0.15
0.15
0.15
0.10
0.10
July
4.8-7.6
0.10
0.35
0.50
0.80
0.05
0.10
0.05
0.20
7.7-15
0.20
0.50
0.50
< .01
0.05
0.25
0.05
(continued)
-------
APPENDIX TABLE 5. (continued
00
o
Species
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jv)
0. mordax (ad)
Stizostedion vitreutn v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
1.8-4.7
8.95
2.45
0.70
0.30
0.20
0.05
0.05
0.60
< .01
August
4.8-7.6
3.90
1.10
0.40
0.45
0.05
0.35
< .01
0.10
0.05
0.45
0.05
< .01
Sampling
7.7-15
1.20
0.80
0.45
1.30
0.05
0.05
< .01
< .01
0.25
0.10
0.25
Depth (m)
1.8-4.7
18.45
4.40
0.55
0.20
0.80
0.10
0.05
0.04
0.20
0.90
< .01
< .01
< .01
September
4.8-7.6
14.35
5.30
0.70
0.60
0.35
0.15
< .01
0.05
0.05
4.55
0.20
0.05
< .01
7.7-15
4.60
0.40
0.45
0.50
0.10
0.20
0.05
0.05
1.75
0.25
0.05
(continued)
-------
APPENDIX TABLE 5. (continued)
Sampling Depth
Species
Osmerus mordax (yy)
0. mordax (yl)
0. mordax (jv)
0. mordax (ad)
Stizostedion vitreum v. (yy)
S. vitreum v. (ad)
Salmonidae
Coregonus sp.
Lota lota
Catostomus catostomus
C. commersoni
Notropis sp.
Percopsis omiscomaycus
Cottus sp.
Alosa pseudoharengus
Couesius plumbeus
Percina caprodes
Gasterosteidae
Esox lucius
Pomoxis nigromaculatus
Cyprinus carpio
1.8-4.7
8.50
1.20
1.10
0.50
0.20
0.10
1.10
0.70
October
4.8-7.6
9.30
1.00
0.30
0.60
0.10
0.10
1.30
(m)
7.7-15
1.00
0.10
0.10
0.10
0.10
0.20
0.30
-------
APPENDIX TABLE 6. TURBIDITY CONVERSIONS
Units given in this table were converted using relationships provided
in the text.
Percent
Light
Trans.
100
99
98
97
96
95
94
93
92
91
90
89
88
87
86
85
84
83
82
81
80
79
78
77
76
FTU
.465
.739
1.01
1.28
1.56
1.83
2.10
2.38
2.65
2.92
3.20
3.66
3.89
4.14
4.41
4.69
4.98
5.29
5.62
5.96
6.31
6.68
7.07
7.47
7.89
ppm
.44
.61
.77
.94
1.11
1.27
1.43
1.60
1.77
1.93
2.10
2.38
2.52
2.68
2.84
3.01
3.19
3.38
3.58
3.78
4.00
4.22
4.46
4.70
4.96
Percent
Light
Trans .
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50
FTU
8.32
8.77
9.23
9.71
10.2
10.7
11.44
11.98
12.33
13.10
13.68
14.28
14.89
15.52
16.16
16.82
17.50
18.19
18.89
19.62
20.35
21.10
21.87
22.65
23.45
24.26
ppm
5.22
5.50
5.78
6.07
6.37
6.67
7.12
7.45
7.66
8.13
8.49
8.85
9.22
9.61
10.00
10.40
10.81
11.23
11.66
12.10
12.55
13.00
13.48
13.94
14.43
14.98
82
-------
10
00
LO
8
0)
a:
400
500 600
Wavelength (nm)
700
APPENDIX FIGURE 1. Color response curves for Kahl Submarine Photometer (Model 268WA-320)
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/3-78-067
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Influence of Turbidity on Fish Abundance in Western
Lake Superior
5. REPORT DATE
July 1978 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
William A. Swenson
8. PERFORMING ORGANIZATION REPORT NO.
». PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Biology and
Center for Lake Superior Environmental Studies
University of Wisconsin-Superior
Superior, Wisconsin 54880
10. PROGRAM ELEMENT NO.
1BA608
11. CONTRACT/GRANT NO.
Grant R802455
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Environmental Research Laboratory-Duluth
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
14. SPONSORING AGENCY CODE
EPA/600/03
15. SUPPLEMENTARY NOTES
Project Officer:
J. Howard McCormick, ERL-Duluth
16. ABSTRACT This research project was developed to improve understanding of the influence
of turbidity on fish populations and the mechanism through which its effects are
induced.
Field and laboratory studies emphasized measurement of behavioral response of
fish and resulting changes in fish species interrelationships in western Lake Superior
Direct effects of red clay turbidity on survival and growth of larval lake herring
(Coregonus artedii) were also measured.
Field measurements demonstrated that light penetration in western Lake Superior
is reduced significantly even at very low levels of red clay turbidity. Zooplankton
and fish abundance and distribution were influenced by turbidity. Zooplankton
abundance and distribution was highest near the surface in red clay plumes. Smelt
(Osmerus mordax) move into the upper 12 m of water in response to turbidity where
their predation on larval fish increases. Predation by smelt on larval lake herring
was identified as a potentially important factor contributing to the decline of the
formerly abundant western Lake Superior lake herring population and the commercial
fishery which depended upon it.
Walleye (Stizostedion vitreum vitreum) and lake trout (Salvelinus namaycush)
demonstrated opposite responses to turbidity. Walleye concentrated in turbid water
where food availability was apparently greater. Lake trout showed partial avoidance
to turbidity in the lake and in laboratory turbidity gradients.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Turbidity
Zooplankton
Red clay Lake trout
Predation Lake
Rainbow smelt Superior
Lake herring Avoidance
Cisco Light
Walleye penetration
Species
interactions
06/F
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report}
UNCLASSIFIED
21. NO. OF PAGES
92
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220—1 (Rev. 4—77) PREVIOUS EDITION is OBSOLETE
84
ft U.S. GOVERNMENTPRINTINGOFFICE-1978—7 S7 -140/1410
------- |