|f§
E M:
pun‘k E!) |

. - a
‘ ' E'Jil'i‘fi'
4*" "J. I\
(‘EEE'IEI ~ ‘I -
E. E..»;..I‘v

, \‘ I.II
II | -I
“I‘l'tr' :l‘hI ‘ E
I“ 'I :‘IIIIIII.
. , "
MI ‘1 . III I
IIIIIL:‘II'III “ "I“ {I'I‘I : H
(“"1 “V?“ I ‘w' ,‘

Iv Hi“ IE.l ; uI. :ij

E'E‘w... IIgI'sIII

. :IIIIIIIIII , ‘IiIEEWE‘J‘éIE
”w -... WEE? Nils {5%
. ‘I-IIII

fl EM

W:|;IE1’ WEE: EEEEKVE
Hug; W1 I'I'IIII,III

/ I

| E - E , 316*
( I? :sz I'hl': II. . III.I
. fl“? ' "‘A':',

I M»

mé :I..;)::’
": . El \(3‘ NEE
.’ > '1 .i'tII-E

- it" ‘3 "E
5 .' n '3'; ..l'.(\' ) RI: J33) ‘EH':

1 .
I f
.' a

W

q

I
I" J'. 'E'l‘}
2"::"‘I".’:‘i ‘J {a
I" I ' . . p .I , '3

'.'1"-‘ {TI Ixt
th 74"}
,4“: '\IIN~.

.“"I’I ' ’I

III‘IIHIIIiIflI'
‘1’“.2'1' I3 VI! II‘
'i‘ .

'1. I
' " 3" I E 3
'lh':'l"m )I

I. I
‘3

 

 

   

Wt. IHIWNIIjllllllllmlflHNIIHHIIHIINIIUIHUIHIHHI

LE“ 4. I"? 310388 1474
A ~$m$mw

V . UM it? I

THESlS

This is to certify that the

dissertation entitled

COMPETITION AMONG JUVENILES OF COHO SALMON,
BROOK AND BROWN TROUT FOR RESOURCES IN STREAMS

presented by

Kurt D. Fausch

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Fisheries & Wildlife

“Pa—7W

M yprofessor

Date October 23, 1981

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

S:
op to

 

,4 V.

TERIA

w.-

 

A
3‘.

 

cJUK

O.
r'.
N

'p
'.

in

"I‘JIP‘G
\-i I‘

‘5:

TB

HE f‘
r- “; , A
thb

 

 

 

 

 

MSU

LIBRARIES

 

 

 

 

nU

r.
2

«H

\.
’

 

COMPETITION AMONG JUVENILES OF COHO SALMON, BROOK
AND BROWN TROUT FOR RESOURCES IN STREAMS

BY

Kurt Daniel Fausch

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

1981

,_’»

'7l/ll'“) ,4" ,_J..

(4~ .f

ABSTRACT

COMPETITION AMONG JUVENILES OF COHO SALMON, BROOK
AND BROWN TROUT FOR RESOURCES IN STREAMS

BY

Kurt Daniel Fausch

Coho salmon (Oncorhynchus kisutch) were introduced to the
Great Lakes in 1966. Naturally reproduced juvenile coho may have
detrimental effects on juveniles of resident brook trout (Salvelinus

fontinalis) and brown trout (Salmo trutta) in Great Lakes tributaries.

 

 

I investigated the timing of fry emergence, their size at emergence,
and the growth of the three species during their first summer of
life in eight Lake Michigan tributaries. In a laboratory stream
aquarium, the competitive relationships between pairs of the three
species were studied by measuring niche shifts of the subordinate
species to more advantageous stream positions after release from
competition with the dominant species. I also developed and tested
a model of specific growth rate as a function of "potential profit"
at stream positions. Potential profit was a measure of the energy
potentially available from the invertebrate drift at a fish's posi-
tion in the stream, minus the cost of swimming to maintain that
position.

In Lake Michigan tributaries coho salmon emerged 2-3 weeks

earlier and were 6-8 mm longer than either brook or brown trout at

Kurt Daniel Fausch

emergence. On average, coho were larger than brook or brown trout

in all streams, maintaining a 6-20-mm length advantage and a 0.5-
4.0-g weight advantage through their first summer of life. In labo-
ratory experiments, coho salmon dominated brook or brown trout of
equal size, and brook trout dominated equal-sized brown trout. Coho
salmon grew at higher specific rates than either brook or brown trout
at all levels of potential profit, and ceased growing at a lower
threshold of potential profit than the trout.

These results indicate that age-0 brook and brown trout
could not gain advantageous stream positions in the face of competi-
tion from age-0 coho salmon even if the coho were of equal size. In
streams where they occur together, juveniles of coho are larger than
those of brook and brown trout, which gives the coho an even greater
competitive advantage. Instream cover that affords visual isolation
may somewhat ameliorate the competitive disadvantage of brook and

brown trout in Great Lakes tributaries where coho salmon reproduce.

To my grandfathers:
Christian David Fausch

Ralph Clark Smythe

ii

ACKNOWLEDGMENTS

I would like to thank Dr. Ray J. White, my major professor,
for guidance and support during my graduate studies. I am especially
grateful to Dr. Darrell L. King for frequent discussions aimed at
improving this research, and for helping and sometimes forcing me
to think quantitatively about my work. Dr. Earl E. Werner and
Dr. John A. King provided useful criticism about the research and
this dissertation.

I thank the many people who aided me during work in the
field, but especially Guy Fleischer for his enthusiastic help.
Andrew Nuhfer and Richard Stanford constructed portions of the
stream aquarium. I thank members of the Michigan Department of
Natural Resources for providing coho salmon eggs during 1979-1981.

This publication is the result of work sponsored by the
Michigan Sea Grant Program with a grant NA-80-AA-D-OOO72 (Project
No. R/GLF-6) from the Office of Sea Grant, National Oceanic and
Atmospheric Administration, U.S. Department of Commerce, and funds
from the State of Michigan.

I am grateful to my parents, Homer and Guinevere Fausch,
for instilling in me the desire to further my education, and for
their support during the many years of study. Deborah Eisenhauer,

my wife, deserves special thanks for sewing the drift nets and the

seemingly endless number of curtains for the stream aquarium as well

iii

as tending the fish while I was away. Finally, I owe Deb a great
debt for her efforts to further my graduate study, and for her love

and moral support during these difficult years.

iv

TABLE OF CONTENTS

LIST OF TABLES.
LIST OF FIGURES
INTRODUCTION
METHODS .

Sampling Natural Populations
Stream Aquarium. .
The Model of Growth as a Function of Net Energy Gain
Specific Growth Rate . . .
Net Energy Gain. .
The Relationship Between Specific Growth Rate and
Potential Profit . .
Experiments on Potential Profit vs. Specific Growth
Rate . . . .
l. Initial Measurements and Acclimation

2. Measuring Characteristics of Fish Positions :

3. Measuring Food and Drift Energy.
Experiments to Measure Competition
Experiment on Brook Trout vs. Brown Trout.
l. Initial Measurements and Acclimation

2. Measuring Characteristics of Fish Positions :

3. Measuring Food and Drift Energy .
Experiment on Brook Trout vs. Coho Salmon .
1. Initial Measurements and Acclimation

2. Measuring Characteristics of Fish Positions :

3. Measuring Food and Drift Energy . .
Experiment on Brown Trout vs. Coho Salmon .
l. Initial Measurements and Acclimation

2. Measuring Characteristics of Fish Positions :

3. Measuring Food and Drift Energy .
RESULTS AND DISCUSSION .

Natural Populations in Lake Michigan Tributaries.
Sympatry Between Coho Salmon and Brown Trout .
Sympatry Between Coho Salmon and Brook Trout. .
Brook and Brown Trout in Sympatry and Allopatry.
Coho Salmon in Allopatry. . .

Page
vii

viii

Specific Growth Rate as a Function of Potential Profit .

Competition Experiments .

Brook Trout vs. Brown Trout .

Brook Trout vs. Coho Salmon .

Brown Trout vs. Coho Salmon .
General Stream Positions and Behavior.
Significance of Competition Experiments . .
Relationships Between Specific Growth Rate and. Poten-
tial Profit . . .
Laboratory and Field Specific Growth Rates . .
Interactions Among Juvenile Salmonids in Great Lake
Tributaries . . .

CONCLUSIONS .
LITERATURE CITED
APPENDIX .

vi

Table

A1.

A2.

A3.

LIST OF TABLES

Page
Study sites in Lake Michigan tributaries. . . . . 7
Design of experiments . . . . . . . . . . . 27

Relationships between slopes of drift-energy-vs.-water-
velocity regressions and distance downstream from the
food source . . . . . . . . . . . . . . 46

Parameters for Michaelis-Menten relationships of speci-
fic growth rate as a function of potential profit for
juvenile salmonids . . . . . . . . . . . . 49

Summary of agonistic behavior among trout and coho
salmon. Percents of all two-minute observations where
any agonism was observed are shown, with actual numbers
of observations where agonism was observed in paren-
theses . . . . . . . . . . . . .

Chemical characteristics of water in the stream
aquarium at the beginning and end of each experiment. 87

Mean dry weight and percent ash of frozen Da hnia fed
per 3 h and per day during each experiment. SEM are
shown in parentheses . . . . . . . . . . . 33

Mean length and weight of juvenile salmonids in eight
Lake Michigan tributaries during 1979. Sample size

and half-widths of 95% confidence intervals are shown

for each date. . . . . . . . . . . . . . 89

vii

Figure

LIST OF FIGURES

Lake Michigan tributaries where natural populations
of juvenile salmonids were sampled. Perpendicular
lines were upstream barriers to fish migration in
major rivers during 1979.

Plan view of stream aquariun

The stream aquarium. The plastic diffuser panels
and the food carboy for the downstream section appear

near the top of the photograph.

Water depth (cm) in Section I (upstream). Depths
were measured at sampling points shown . .

Water depth (cm) in Section II (downstream). Depths
were measured at sampling points shown . . . .

a. Hypothetical regressions of energy from drifting
Daphnia as a function of water velocity for five
istances downstream from the food source. . .
b. Hypothetical decrease in slopes of drift-energy-
vs.-water-velocity regressions as a function of
downstream distance . . . . . . . . . .
c. Hypothetical relationship between specific growth
rate)and a critical resource for an organism (see
text . . . . . . .

Growth of juvenile coho salmon and brown trout in

.three Lake Michigan tributaries during 1979. Top

curves are body length, bottom curves are weight.
A "*" denotes a sample of three fish or less, and
bars show 95% confidence intervals on each mean .

| Growth of juvenile coho salmon, brook and brown trout

in five Lake Michigan tributaries during 1979. Top

curves are body length, bottom curves are weight. A
"*" denotes a sample of three fish or less, and bars
show 95% confidence intervals on each mean.

viii

Page

11

12

13

19

38

41

Figure

10.

11.

12.

13.

14.

15.

16.

17.

Page

Specific growth rate of coho salmon (a) and brown

trout (b) in allopatry as a function of mean poten-

tial profit at fish positions. One outlier (*) was
excluded from the brown trout relationship (see

text). . . . . . . . . . . . . . . . 43

a. Relationships between water velocity and drift
energy at five distances from the upper end of
Section I during the allopatry phase of the
brown- trout- vs. -coho- salmon experiment .
b. Slope of the drift-vs. -velocity relationship as
a function of distance from upper ends of Sec-
tions I and II during the same experiment. Bars
show 95% confidence intervals for slopes, trans-
formed to natural logs. . . . . . . . . 45

Specific growth rate as a function of mean potential
profit for brook and brown trout in sympatry (a) and
allopatry (b). One brown trout in allopatry was

excluded as an outlier (*). . . . . . . . . 51

Distribution of positions held by brook and brown

trout during sympatry (a and b) and allopatry (c and

d). Black portions of bars are positions of dominant 53
fish. . . . . . . . . . . . . . . .

Specific growth rate as a function of mean potential
profit for brook trout and coho salmon in sympatry
(a) and allopatry (b) . . . . . . . . . 55

Distribution of positions held by brook trout and

coho salmon during sympatry (a and b) and allopatry

(c and d). Black portions of bars are positions of
dominant fish . . . . . . . . . . . . . 57

Specific growth rate as a function of mean potential
profit for brown trout and coho salmon in sympatry

(a) and allopatry (b). One coho salmon in allopatry

was excluded as an outlier (*) . . . . . . 62

Distrubution of positions held by brown trout and

coho salmon during sympatry (a and b) and allopatry

(c and d). Black portions of bars are positions

of dominant fish . . . . . . . . . . . . 54

General relationships between Specific growth rate

and mean potential profit for coho salmon, brook and
brown trout in allopatry (a) and sympatry (b). . 7‘

ix

Figure

18.

19.

Ala.

A1b.

Alc.

A2a.

A2b.

A2c.

A3.

A4.

Page

Specific growth rate of juvenile salmonids in eight

Lake Michigan tributaries as a function of mean

weight. Dashed lines are negative exponential equa-

tions fit to all data for each species. . . . . . 74

Comparison of salmonid specific growth rates as a
function of mean weight in Lake Michigan tributaries
with the highest rates for individual fish in labora-

.tory experiments .

Water velocities (cm/sec) in Section I (upstream) 2.5
cm below water surface. Water velocities were measured
at points shown

Water velocities (cm/sec) in Section I 7.5 cm below
water surface. Water velocities were measured at
points shown . . . . . . . . . . . . . . 91

Water velocities (cm/sec) in Section I 12.5 cm below
water surface. Water velocities were measured at
points shown. . . . . . . . . . . . . . . 92

Water velocities (cm/sec) in Section II (downstream)
2.5 cm below water surface. Water velocities were
measured at points shown . . . . . . . . . . 93

Water velocities (cm/sec) in Section II 7.5 cm below
water surface. Water velocities were measured at
points shown . . . . . . . . . . . . . . 94

Water velocities (cm/sec) in Section II 12.0 cm below
water surface. Water velocities were measured at
points shown . . . . . . . . . . . . . . 95

Light QJE/mzlsec) at surface of Section I. Light was
measured 1 cm above water surface at sampling points
shown. The positions of mercury vapor (open circles)
and incandescent lamps (filled circles) are shown . . 95

Light (uE/mZ/sec) at surface of Section II. Light was
measured 1 cm above water surface at sampling points
shown. The positions of mercury vapor (open circles)
and incandescent lamps (filled circles) are shown .

INTRODUCTION

Salmonids have been introduced throughout the world to habi-
tats where they were not indigenous. During the last century, after
many attempts, Pacific salmon of the genus Oncorhynchus were success-
fully established in the Great Lakes, and have recently been intro-
duced to the Atlantic coast of North America and northwest Europe.
Fishery biologists in each geographic area have expressed concern
that the introduced salmon may have detrimental effects on established
populations of Atlantic salmon (Salmo salar) and resident stream
trout (Gruenfeld 1977, Solomon 1979, Taube 1975).

Since 1956, three salmon from the Pacific coast of North
America have been added to the Great Lakes fish community. Pink
salmon (Oncorhynchus gorbuscha) originating from releases in two
Lake Superior tributaries near Thunder Bay, Ontario in 1956, have
spread to all five Great Lakes (Emery 1981, Kwain and Lawrie 1981).
Juveniles of coho salmon (Q, kisutch) and chinook salmon (Q,

tshawytscha) were introduced in two Lake Michigan streams and one

 

Lake Superior stream in 1966 and 1967 respectively (Latta 1974).
Both species are now stocked annually in Michigan streams tributary
to Lakes Michigan and Huron, and some Lake Superior tributaries.
Many returning adult salmon stray to other tributaries, probably
because the juveniles are planted as advanced smolts and do not
imprint on the stream (Peck 1970). Consequently, with continued

1

introduction in Michigan and other states, coho and chinook salmon
now use many suitable Great Lakes tributaries for spawning.

The coho, chinook and pink salmon that ascend Great Lakes
tributaries, like the brook trout (Salvelinus fontinalis) and brown
trout (Salmo trutta) residing in these streams, are fall spawners
that bury their eggs in gravel redds to incubate during the winter.
The eggs hatch in late winter and young emerge from the gravel in
early spring, but each species of salmon spends a different period
growing in the nursery streams. Pink salmon migrate downstream to
a lake or ocean soon after emergence. Chinook salmon leave tribu-
taries in summer after three to six months of growth, and coho salmon
remain 12 to 15 months, smolting in the spring of the year following
hatching in Great Lakes streams.

These large salmon may affect resident trout in several ways:
(1) adult salmon and trout may compete directly for spawning sites
or later spawners could dig up redds of earlier spawners, (2) adult
trout and salmon may prey on juveniles of either species, (3) spawn-
ing fish excavating redds decrease the invertebrate food supply
(Hildebrand 1971), and (4) deteriorating salmon may spread diseases.
However, the species interaction most likely to have long-term effects
on resident trout populations is competition among juveniles of salmon
and trout for food and space in nursery streams. Pink salmon juve-
niles do not remain in tributaries to compete with juvenile trout,
and chinook salmon occupy them only for a few months, but coho salmon
juveniles may compete strongly with juvenile trout during their 12 to

15 months of residence in Great Lakes tributaries.

The purpose of this research was to study competition among
juveniles of coho salmon, brook and brown trout for resources in
streams. Salmonids in streams compete for two major classes of
resources, food and space. These fish defend territories, maintain
relatively fixed positions called focal points within the territories,
and make short forays from the focal point to catch drifting inver-
tebrates (Kalleberg 1958). However, the food and space resources
of salmonids are related in streams because more invertebrate drift
is delivered to areas of the stream with swifter currents, so that
defending a specific area ensures a fish access to the food drifting
nearby. In view of these relationships, Chapman (1966) proposed that
competition for space had been substituted for direct competition
for food among stream salmonids. Fausch and White (1981) further
proposed that salmonids should compete for positions in streams that
maximize the potential for energy intake from the drift, while
minimizing the energy cost of swinming--in essence, positions that
maximize net energy gain.

The most direct way to measure the effects that two competing
species have on each other is to measure niche shifts--that is,
changes in resource use that affect survival, growth, physiology, or
behavior--of one or both species when their competitor is removed
(Connell 1975, Diamond 1978, Sale 1979). When fish are used in such
experiments, the effects of niche shifts are usually measured in
terms of growth in weight, which is presumed to be a sensitive indi-
cator of fitness (Werner and Hall 1976). To measure the effects of

competition for advantageous stream positions among juvenile salmonids,

I compared relationships of specific growth rate in weight for indi-
vidual fish as a function of the net energy gain from drifting food
at the fish's position in the stream. Changes in these growth-vs.-
resource relationships when sympatry is compared to allopatry should
provide information about the effects that niche shifts have on the
energy available to fish when competitors are present and absent.

Measuring competition between fish in natural populations
is complicated by variation in the size of individuals of the same
age. Because fish do not grow to a unifonn species specific adult
size, as do birds for instance, the size structure of a population
has marked effects on competitive relationships where larger fish
are dominant. Therefore, I divided the research into two parts:
'(1) determining the size structure of juvenile salmonid populations
in Lake Michigan tributaries, and (2) measuring the innate competi-
tive ability of juvenile salmonids in laboratory experiments using
fish of equal size.

In view of the aggressive nature of the coho salmon reported
by Hartman (1965) and Glova and Mason (1977), my hypothesis was that
juveniles of coho salmon are superior competitors and could exclude
equal—sized juvenile brook and brown trout from advantageous posi-
tions in a stream aquarium. If this proved to be true, coho salmon
might reduce resident brook and brown trout populations in Great
Lakes tributaries, where juvenile coho were expected to have a slight

size advantage over the age-O trout.

METHODS

Sampling Natural Populations

To determine the size distribution of juveniles of coho
salmon, brook and brown trout in natural populations, I measured
the timing of emergence, the size at emergence, and the relative
growth of juvenile salmonids during the first sunmer of life in
eight Lake Michigan tributaries (Figure l). I chose study streams
to include all combinations of the three species in sympatry and
allopatry. All streams are first-to-third-order tributaries of
larger rivers draining into Lake Michigan, and support salmonids
that are naturally reproduced except where noted in Table 1. Several
streams also contained steelhead trout (Salmo gajrdneri) and chinook
salmon as well as a variety of other fishes (Table l).

I sampled populations in the five coho nursery streams approx-
imately every three weeks from April to September 1979. Three streams
without coho were added in July. On each sampling date, I captured
fISh by electrofishing (1 ampere, 175 volts DC) a 1004400-m section '
of stream chosen to yield large numbers of fish. All fish captured
were anesthetized (M5222), weighed (£0.05 g), measured ($0.5 mm TL),
and returned to the stream.

In early spring, I captured newly emerged trout and salmon

with a hand net in areas of low water velocity over silt flats
along the stream margins. On two sampling dates in August and

5

 

 

Jordan River

v Green River

Manistee River .5
Pine Creek

c Blgelow Creek

Muskegon River

Grand River ‘- 593"” 0““

80nd
Kalamazoo River reekSpringbrook
‘ Creek
5 $011"!
Creek
51- Joseph , Love Creek
River '

 

 

 

 

Figure l.--Lake Michigan tributaries where natural populations of juvenile
salmonids were sampled. Perpendicular lines were upstream
barriers to fish migration in major rivers during 1979.

TABLE 1.--Study sites in Lake Michigan tributaries.

 

44§pecies presence and abundance

 

 

 

Stream Salmonidsb Non-salmonTHSE
Stream Location ordera Trout Salmon

1. Love Creek 1'55 31711517 151: brook cd coho A 1.2.3.4,5.
steelhead A ,

2. Springbrook TlS,RlOW,518 2nd brown A 1,10,14

Creek

3. Smith Creek TlS,R9W,SZ lst brook A None

4. Sand Creek T3N,R14W,S34 lst brook Cd coho A 1,8,9,10,1l,
brown R 12,13,14
steelhead A

5. Egypt Creek T7N,R10W,S4 3rd brook A coho R 4,14,16,17
brown A

6. Bigelow Creek T12N,R12W,520 3rd brown C coho A 2,9,14,15
steelhead A chinook A

7. Pine Creek T21N,R14W,56 2nd brown A coho A 3,13
steelhead A chinook A'

8. Green River' T30N,R6W.SB 2nd brown A coho A 13
brook R chinook A
steelhead A

 

a Strahler's modification of Horton (1945).
b A-abundant, 11 or more fish captured on each date sampled; c-comon, 1-10 fish captured;
R-rarely found.
c Non-salmonid species list:
1. Green sunfish (Le omis cyanellus)
2. Rainbow darter Etheostoma caeruleum)
3 Central mudminnow (Umbri’limi)

4. Fathead minnow (Pime Hales promelas)
5. Bluntnose minnow 1mep a es notatus)
6 Emerald shiner.(Notropis atherihdides)
7. White sucker (Catostomus commersoni)
8. Burbot (Lota lota)

9. Bluegill sunfish (Lepomis macrochirus)
10. Longear sunfish (Lepomis megalotisl
ll. Blacknose dace (Rhinichthys atratilus)
12. Creek chub (Semotilus atromaculatgsfi
l3. Slimy sculpin (Cbttus cognatus)

14. Mottled sculpin (Cottus’bairdi)

15. Longnose dace (Rhinichthys cataractae)

16. Yellow perch (Perca flavescens)

l7. Fantail darter (Etfieostoma flabellare)
d Hatchery fish introduced, and few or no age-0 fish captured.

 

 

September I measured the bias in electrofishing in Pine Creek by

making three 30-m hauls downstream with a 3-mm seine.

Stream Aquarium

The recirculating stream aquarium (Figures 2 and 3) used for
competition experiments in the laboratory, unlike most stream aquaria
which are straight, circular, or ellipsoid, was constructed in a
sine-generated curve, the pattern of meandering taken by natural
streams (Leopold and Langbein 1966). The more natural riffle-pool
ratio and the associated bottom contour simulated the stream habitat
of juvenile salmonids more accurately than could be achieved with
other channel shapes.

The channel shell, constructed of clear Plexiglass, was 7.28
m long, 30 cm wide and 30 cm deep, and had no slope. I divided
the channel into two 3.64-m sections, the upstream section referred
to as Section I, and the downstream Section II, and constructed a
V-shaped trap at the downstream end of each section to retain fish
passing downstream. The channel bed was formed of 2-3-cm diameter
gravel, shaped to simulate the natural pattern of riffles and pools,
which had 8 and 15 cm maximum water depth respectively (Figures 4
and 5). Flow began at an upstream header box and followed the mean-
dering channel to a collector box. Return flow was through a 20-cm
diameter PVC pipe beneath the channel.

Stream flow was generated by air-lift pumping (Spotte 1979)
using a large air stone located in the return pipe below the header

box. Discharge was controlled by an air-pressure regulator. The

 

 

   

 

 

 

 
  

Section II

 

Typical 0F"ml
Distribution 0

of Coho Salmon 0°
in Allopatry

gravel

bar

   

 

  

Cross-sections .
for/Drift Samples

Enclosure

0

A

0.5. LO

PHOTO?!

 

Figure 2.--Plan view of stream aquarium.

 

10

.gnmcmouoga 0:» mo non on» Low: Leanna cowuomm

Emmepmczou as“ Low xonemo zoom ecu can mpmcma gmmzwewu omummpa one

.angmsam seesaw mghuu.m mcsmwm

.Esweoaam Emmgpm m;k--.m wczmvo

., ,. Cit i - 1‘1 s 4'
S (I... u. . l
. a or \rL-xsavAwuu‘ausnrv..tnumqumuhu,4uu~
‘ ‘ROIJD | I I: .a\.. .0 Q) 1 Dr 1P r .
1 . Luz. Pavlova. &o\

\t f.
:Ivu.‘ r

 

1?.

 

 

SECTIONI Upstream Water Depth

 

(. . . . . . ——gravei bar

 

 

FLOW

 

Figure 4.--Water depth (cm) in Section I (upstream). Depths were
measured at sampling points shown.

 

 

 

 

SECTION II Downstream Water Depth

 

 

Figure 5.--Water depth (cm) in Section II (downstream). Depths were
measured at sampling points shown.

 

14

velocity pattern at 50 psi, corresponding to a discharge of 0.00260
m3/sec (0.0919 cfs) is shown for three depths in Appendix Figures Al
and A2.

Water from the stream aquarium was circulated through a bio-
filter, built according to Spotte (1979), at a rate of about 20
liters/min. Temperature was controlled by a Frigid Unit cooler in
the biofilter, which effectively damped the total range of tempera-
ture oscillations to about 1 C. Oxygen was 100% of saturation
(9.9 mg/liter at 15 C) when measured in both the header and collector
boxes.

The stream was lit by five equally spaced 2.5 kW mercury
vapor lamps and ten equally spaced 100 W incandescent bulbs (Appendix
Figures A3 and A4), all diffused through translucent 1.5-mm thick
white plastic sheets hung 50 cm below the lights and 65 cm above the
water surface. A timer built according to Drummond and Dawson (1970)
maintained a 12-h photoperiod with a 30-min period of brightening
and dimming of the incandescent lamps prior to turning the mercury
vapor lamps on and after turning them off, to simulate sunrise and
sunset. The pattern of light intensity (uE/mZ/sec) measured 1 cm
above the water surface with a Li-Cor model LI-1888 integrating
photometer is shown in Appendix Figures A3 and A4.

A white false ceiling above the lamps reflected light down-
ward, and curtains reaching from the false ceiling to the table on
which the stream rested were lined with white cloth, all of which
produced a uniformly lit environment. A flap extended perpendicu-

larly from the curtain to the top of the stream wall, and below this

15

flap the curtain was lined with black cloth. The curtains had view-
ing slits positioned below the water level and spaced every 40-45 cm
along the channel to allow an observer to watch fish while remaining
concealed.

During the experiments, I simulated invertebrate drift by
continuous introduction of a suspension of Daphnia in water at the
head of each section for 12 h each day. Frozen Daphnia were thawed
and kept in suspension in 27 liters of stream water in two carboys
using airstones. The suspension drained through a 1.5-mm orifice in

about 3 h, so that the carboys were refilled four times each day.

The Model of Growth as a Function of Net Energy Gain
Growth in fish should be related to net energy gain. In

this section I develop a model of specific growth rate as a function

of net energy gain for stream salmonids. I first examine the reasons
why specific growth rate is a valid measure of growth, then propose

a method to measure net energy gain at salmonid positions in streams.
Last, I attempt to show how the relationship between specific growth
rate and net energy gain can be used to examine interspecific compe-

tition.

Specific Growth Rate

 

The rate of growth of an organism can be expressed simply as
the change in weight per unit time, in familiar units such as grams
per day. However, because the weights of individual organisms differ,
growth rate is better expressed per gram of weight, by dividing by

the mean weight of the individual during the growth period. This

16

quantity is called the specific growth rate (u), with units of l/day,

and may be calculated as follows:

A W
_ A t
H ' __
Mean W (1)
where: W = weight (g)
t = growth period (days)

This linear calculation of specific growth rate is adequate for growth
over short periods. For longer periods between weight measurements,
it may be more accurate to assume that the organism grows at an

exponential rate given by:

_ ut
Wt - Woe (2)
where: Wt = final weight (9)
WO = initial weight (g)

The specific growth rate during the growth period (t) is calculated
by converting equation (2) to natural logs and solving for u:

1" Wt ‘ In No (3)

t
Note that the growth rate at any time can be calculated from the
weight of the organism:
W . u = growth rate (g/day) (4)
For my purposes, the specific growth rate (l/day) is a better measure
of fish growth than the growth date (g/day) because it adjusts for

small weight differences among individual fish.

17

Net Energy Gain

Although the specific growth rate for juvenile salmonids can
be calculated from initial and final weight measurements, calculating
their exact net energy gain would require a complete energy budget
for each individual. However, because these fish maintain relatively
fixed focal points with respect to the stream bed, the potential for
net energy gain or "potential profit" (P) at a given stream position
can be estimated as the energy potentially available from the drift-
ing food (0) minus the energy required for swimming to maintain the

position (5), or:

P = D - S (5)
where: P = potential profit (cal/hr)
D = potential drift ener (cal/hr)
S = swimming cost (cal/ha)

If the water velocity at the focal point can be measured, the
cost of swimming at a fixed position is easily calculated, excluding
the energy required for short forays to catch the drifting food. I
used general metabolic equations for coho salmon and rainbow trout
presented in Stewart (MS) to calculate the swimming cost in calories
per hour given fish weight, water temperature, and swimming velocity.
The following equations are for a water temperature of 15 C:

0.784 e0.0186 V (6)

Coho salmon: S = 0.9906 W

Rainbow trout: S = 0.7007 WO'763 e0’0327 V (7)

where: S = swimming cost (cal/hr)
W = fish weight (g)
V = focal point water velocity (cm/sec)

During experiments in the stream aquarium, Daphnia were

introduced only at the upstream end of each section. The amount of

18

drifting food decreased with downstream distance, primarily because
fish ate it, but also because some Daphnia sank into the gravel. To
determine the energy available to fish from drifting Daphnia, I

needed to estimate the drift delivered to any point in the stream
aquariun as a function of water velocity when fish were present. To
do this, I first sampled drift and measured water velocity at various
places in five cross-sections located at 60-cm intervals along each
section (shown in Figure 2). For each cross-section, I then regressed
the values of drift energy against the corresponding velocity measure-

ments, forcing the equations through the origin:

E = a - V (8)
where: E = drift energy (cal/hr/cmz)
~ a = slope of the relationship
V = water velocity (cm/sec)

These regressions are shown hypothetically in Figure 6a.
Next, I calculated negative exponential functions to describe
the decrease in the slopes of these drift-energy-vs.-water-velocity

regressions as a function of distance downstream from the food input:

a = be"cx (9)

where: a = slope of drift-vs.-velocity regression
X - distance from food source (cm)
b and -c are fitted constants

An hypothetical curve is shown in Figure 6b. Using these two equa-
tions, the drift energy (E) in calories per hour per square centi-
meter at any point in the vertical plane of the cross-section can
be estimated if the distance downstream from the food source (X) and

the water velocity (V) are known:

19

 

 

 

 

 

 

 

 

 

 

 

 

 

A I eacm'
&‘ 2°” 120cm
5
t~
g 1806!!!
8
‘: ,0, 240cm
2
o 300
6 cm
5
oo 1- 20 so
5 Water Velocity (cm/sec)
r t f T 1
g a
i- w - .
C
.2
E
f as- i
8
a
a
O
Q
2
m A 1 1 L 1
.. 000 so 120 180 240 1100
.1: lmnmmohomimuuamna cm
a fi
50.02014 c SpecieeA .1
i 4
f. i
g0010+ 1
i ' Specleea
5 ’ ‘
E
130006
“a" f /1 l l
o 5 15

I
Resources or Mean Potential Profit (col/hr)

Figure 6.--a. Hypothetical regressions of energy from drifting

Da hnia
as a function of water velocity for five distances gown-

stream from the food source.

b. Hypothetical decrease in slopes of drift-energy—vs.-
water-velocity regressions as a function of downstream

distance.

c. Hypothetical relationship between specific growth rate
and a critical resource for an organism (see text).

20

‘CX - v (10)
Given an estimate of food delivered to any point in the
stream, I still needed to determine the drift energy available to
each fish by assigning it a foraging area in the vertical plane.
Coho salmon and brown trout I observed during a pilot study foraged
mainly in the area of maximum velocity above and in front of their
focal point, within a radius of about two fish lengths. For fish
positions in the pools, this area of maximum velocity was near the
outside wall of the stream bend where the water was deepest, shown
in the depth and velocity profiles in Figures 4 and 5, and Appendix
Figures Al and A2. To estimate the drift energy passing through
the fish's foraging area (D in equation (5)), I envisioned a semi-
circle centered at the focal point and extending to a radius of two
fish lengths (fork length), and assigned to each fish the drift
energy passing through one quarter of this semicircle (l/8wr2) at
a rate dictated by the maximum velocity measured within the two—

fish-length radius:
_ 2
D - l/8nr - E (11)
Substituting equation (8) for E, and using the maximum velocity

within two fish lengths then gives:

0 =1/811r2 (a - V) (12)
where: D = potential drift energy (cal/hr)
r = two-fish-length radius (cm)
a = slope of drift-vs.-velocity relationship
V = maximum water velocity within two fish lengths

of the focal point (cm/sec)

21

Further substituting equation (9) for 'a' gives the potential drift
energy as a function of downstream distance and the maximum water

velocity within the two-fish-length radius:

a = 1/8m~2 - be'cx - v (13)

where: X = distance downstream from food source (cm)
b and -c are fitted constants

During the laboratory experiments I measured the distances
traveled from the focal point to capture drifting Daphnia for two
coho salmon, and found the frequency of forays dropped sharply at
distances greater than l.5-2.0 fish lengths from the focal point.
Wankowski (1981) found that the area of capture upstream of positions
held by juvenile Atlantic salmon was fan shaped in the horizontal
plane, and that the capture distance varied seasonally from 1.9 to
9.9 fish lengths. Although the area of capture for stream salmonids
should be expected to vary with water velocity, particle size, hunger
level, and fish species, it appears that my "two fish length" crite-
rion may be a conservative estimate for the foraging area of juvenile
salmonids.

In summary, to relate specific growth rate of stream salmonids
to the net energy gain or "potential profit" at their stream positions,
the following data were necessary, where 5 indicates that the data
were used to calculate swimming cost (equations (6) and (7)), and
0 means that the data were used to calculate potential drift energy

(equation (13)):

22

Measured for each fish each day:

—l
0

Focal point velocity (S)

2. Maximum velocity within two fish lengths at the focal
point (D)

3. Distance of fish position downstream from the food

source (0)

Measured during each experiment:

Drift delivered as a function of downstream distance (0)
Individual fish weights (S)'

Mean fish length (0)

Water temperature (S)

ROOM—f
e e e e

The Relationship Between Specific Growth Rate and Potential Profit
Relating the specific growth rate of an organism to one or
more critical resources that can be measured, has inherent advantages
over density-dependent growth equations in the analysis of the growth
of individual organisms or of populations. Such relationships define
the growth rate of a population or of an individual in terms of a
resource required for growth, instead of in terms of numbers of
intraspecific or interspecific competitors as the familiar Verhulst-
Pearl and Lotka-Volterra models do. The specific rate as a function
of a critical resource often takes the form of the curves shown in
Figure 6c, which can be described by an equation known variously as

the Michaelis-Menten or Monod function:

 

R
u 3 11max KR + R (14)

where: u = specific growth rate
“max = maximum specific growth rate

R = resource

KR = resource level at I ”max or half-saturation
constant

23

When a threshold resource value at which no growth occurs is sub-
tracted to transform the equation so that the curve passes through

the origin, it takes the form:

11 = I‘lnax WI (15)

where: T - resource threshold at which no growth occurs

Although this approach has most often been used in biological
science to describe population growth of microorganisms, algae
(Young and King 1980), or diatoms (Tilman 1981), a brief search of
the literature (0. King, personal communication) revealed similar
relationships for specific growth rates of two species of zooplankton
grazing on phytoplankton (Lampert and Schober 1980), and for pelagic
sockeye salmon juveniles eating zooplankton (Warren 1971, p. 260).
In short, the relationship may be a general one for an organism
whose specific growth rate can be measured in terms of one limiting
resource at a time.

It is clear, however, that there are differences between
relationships of specific growth rates of populations as a function

of a resource, and the specific growth rates of individual organisms

 

using a resource. Rates for populations include births, deaths and
reproduction, whereas those for individuals describe only body growth.

If the curves of specific growth rate vs. a limiting resource
are determined for two species, A and 8, growing in allopatry, they
can be used to make predictions about competitive relationships

between the species for this resource. At high levels of resource R,

24

Figure 6c shows that species A grows faster than species B, and is
the superior competitor for resource R. As the individuals of
species A and B deplete resource R, species B ceases growing at a
higher resource level or threshold (T) than A, so that A is still
the superior competitor at low resource levels.

These types of predictions, based solely on single-species
relationships, may be appropriate for populations of microorganisms,
algae and probably some invertebrates, but ignore the additional
complex behavioral components of interspecific competition among
individuals of higher invertebrates and vertebrates. For instance,
if species B in Figure 6c tenaciously defends areas with high levels
of resource R from species A, individuals of A may be excluded to
areas where the resource is scarce, and grow at a lesser rate than
species 8. However, if A and B have access to equal amounts of
resource R, and if body size is important to competitive dominance,
species A would soon grow larger than B and be the superior competi-
tor.

The critical resource for stream salmonids in terms of the
model outlined above is the mean potential profit for individual
fish at their positions in the stream, which should be related to
their specific growth rates as shown in Figure 6c. Relationships of
specific growth rate vs. potential profit for salmonids in allopatry
integrate the effects of the basic physiology and intraspecific
behavior on specific growth rate. These same relationships calcu-
lated for two species in sympatry integrate basic physiology and

intraspecific behavior, but also interspecific behavior, and any

25

changes in intraspecific behavior due to the presence of another
species. Therefore, comparing the relationships of specific growth
rate vs. potential profit measured in sympatry with those in allopatry
should indicate changes in behavior resulting from interspecific

competition, and may reveal mechanisms to explain niche shifts.

Experiments on Potential Profit vs. Specific Growth Rate

 

I conducted two experiments in the stream aquarium to deter-
mine the relationship between potential profit and specific growth
rate for coho salmon and for brown trout grown separately in allo-
patry. In each experiment, I measured position characteristics for
each fish on each day to estimate the potential profit and related
it to their specific growth rate over an 18-day period.

To help ensure that inferences from laboratory experiments
would apply to natural populations in Great Lakes tributaries, all
juvenile salmon and trout used were hatched from eggs of returning
Lake Michigan coho salmon, or of wild trout from Michigan streams.
After hatching, eggs were transferred from incubator trays to a
gravel bed in a holding stream tank to promote normal fry develop-
ment and emergence. Because coho hatch and emerge at a larger size,
brook and brown trout were fed at a greater rate to grow all species

to equal size.

1. Initial Measurements and Acclimation
Twenty five coho salmon and brown trout were individually
finclipped, using combinations of no more than four of the following

five finclips on any one fish: tip of dorsal fin, tip Of ana‘ fl",

26

top caudal lobe, bottom caudal lobe, and adipose fin. Most of the
fish were given one to three finclips, but one four-clip combination
was used, and one fish had no finclips. Coho salmon averaged 54.1 mm
in fork length (range 50.0-57.5 mm) and 1.59 g in weight (range
1.21-2.03 g). Brown trout averaged 52.4 mm (range 47.5-54.5 mm) and
1.40 9 (range 0.99-1.63 g).~ I measured the fork lengths of fish
during all laboratory experiments because the caudal finclips inter-
ferred with total length measurements.

Fish were acclimated to the stream aquarium and food for
four days prior to the experiment (Table 2), during which the traps
were blocked to prevent fish from leaving the sections. Coho salmon
were placed in the upstream section and brown trout downstream in
this experiment (Table 2). I opened the mouths of the traps at the
downstream end of each section after the initial acclimation period
to allow downstream migration. Fish entering a trap were returned
to the head of the section, but if a fish entered a trap three times
it was removed from the experiment. 0f the original 25 fish of each
species, 17 brown trout and 22 coho salmon remained in the channel
at the end of the lB-day experiment.

Water velocity was adjusted to about 30 cm/sec on the upstream
riffles to prevent all fish from occupying the upstream end of each
section, a problem discovered during the pilot study. Streamflow
discharge measured 40 cm from the upstream end of the upper section
was .00179 m3/sec (0.0632 cfs). In all experiments, water tempera-

ture was maintained at 15 C i l C. Chemical characteristics of the

.aguoaome ecu zgumaeam comzuoa wows uaogu czoga moegh u

.mzmv s Lou uvcgmmo mcvuam o» mouasmpoua mew: gm_u o

.ucoewgmaxo mo ago an ugmwoz um: ;m_$ we ugougoa ma commogaxo sat sun vow «we; a: mo u;o_o: am: a
. .aneamczouuo .saosumnsua a

 

27

cm on N.~N ~.m~ om.o mm.c c.pe n.p¢ cop m m.o m.mp a anon» czogm

cameo. mm om ~.n~ o._N No.9 mm.c m.—e u.c¢ m— m m.m m.m~ = cospum ozou
cospam ozou .m> «sock exogm .e

mm mm c.pm o.pN ~m.o w~.o p.mm m.¢m on c— m.o_. um a cospmm ozou

ma—oo. mm Nu o.p~ o.p~ mN.c pm.o n.mm m.¢m cm o— m.o— um : pace» goose
cospmm osou .m> «soap xcocm .m

m m o.o~ c.c~ ~c.n p~.m m.om o.os e" o— m.op m a uzoeu goose

oomoo. m a o.o~ c.o~ hm.~ m~.m ¢.mm h.mo «F o— m.o— m = azoeu czosm
usogh czagm .m> “sock goose .N

pp 1 m.m— 1 oe.p 1 «.mm 1 up mp 1 v a anon“ czoem

mnpoo. op 1 p.m~ 1 mm.p 1 p.em 1 Nu m— 1 v = :oEPom cgou

o_eoca _a_u=oooa .m> ooag gozota o_C,ooam ._

 

Aumm\m5v xguaa aguma xguma xenon xuuog xenon xenon xenon .gm*$ ages; >uuoa copuas acumao—pn acme—gmaxm

1o—_< 15mm 1oppm 155m 1o 15mm 1o 15» mo 1o—~< 15mm. 1_p~m< :.
mmgogomPv acougoa ale a; we: dues gu.=m— .oz wade 4o amass: mcowuomm

52$ 3:... 3:5
aa.oac ;m_e\eooa goo: cam:

 

 

 

.wucoe.eoaxo mo curmmo11.~ u4m<p

28

water were measured before and after each experiment (Appendix Table
A1).

To measure growth, fish were starved prior to measurement
for 12 h (2000 to 0800 h EDT) before and after each experiment to
reach a standard level of gastric evacuation. I individually anes-
thetized (M5222), measured (i 0.5 mm), and weighed (i 0.01 g) each
fish in a tared beaker of water after light blotting on a cloth
towel. The specific growth rate for each fish was calculated on a

\

per-day basis to the nearest 0.5 d.

2. Measuring Characteristics of Fish Positions

On each day of the experiment, I measured three position
characteristics at the focal point of each fish to calculate poten-
tial profit: water velocity, the maximum water velocity within two
fish lengths, and distance from the food source at the upstream end
of the section. I used the mean initial length of all fish in each
experiment to calculate the two-fish-length radius. Water velocities
were measured to the nearest 0.5 cm/sec with midget Bentzel current
speed tubes built according to Everest (1967). I measured focal
point velocities at the fish's head, and maximum velocities within
two fish lengths of this point (100 cm). Distances of fish positions
from the upstream end of each section were judged from marks at 5 cm
intervals along the stream wall. When in doubt, the next distance
upstream of the fish position was recorded.

The fish in each section were chosen randomly to be measured

during the morning or afternoon each day. Because I wanted to

29

measure undisturbed positions of fish, during each measuring period

I located all the randomly chosen fish in the stream before measuring
their positions by looking through the viewing slits in the curtains,
and marked their positions in the vertical plane on the stream wall
with a wax pencil. To locate each fish position in the horizontal
plane I recorded notes about a prominent feature of a specific stone
directly below each fish. After all fish to be measured were located
in a section, the curtains were parted and the position characteris-

tics measured.

3. Measuring,Food and Drift Energy

A relatively constant food-to-fish ratio was maintained by
decreasing the amount of Daphnia introduced to each section as fish
migrated out and were removed. Coho were fed 10% of their wet weight
in thawed Daphnia (18.1 g/day at the end of the experiment) and brown
trout 11% or 13.3 g/day at the experiment's end (Table 2). Mean dry
weight of the ration for coho was 0.280 g/day (7.5% ash), and for
brown trout was 0.172 g/day (8.7% ash) (Appendix Table A2). Fine-
meshed screens removed all residual Daphnia at the end of each sec-
tion, but they disturbed stream flow and were abandoned in subse-
quent experiments.

I usually measured drifting Daphnia at the five cross-sections
shown in Figure 2 with 0.3-mm mesh drift nets measuring 5-by-5 cm at
the mouth (25 cm2) with an 18-cm long bag. However during this
experiment I sampled drift only at 60, 180 and 300 cm. I set blocks

of four or five nets simultaneously in each cross-section for 120-min

30

periods, but measured drift at only one cross-section at a time in
each section. After this experiment, drift was measured at the usual
five distances without fish in each section to provide baseline data.
Drift nets were difficult to place and retrieve without temporarily
frightening fish away from their usual positions, but the fish became
conditioned to this activity by the end of each experiment, and
maintained normal positions during drift sampling.

After each drift sampling period, I removed the nets and
measured the water velocity at the center and the edges of each net
mouth along its horizontal midline. Water velocity profiles measured
around net frames with and without nets showed that the netting
caused an 8.6% (SE 2.18) reduction in flow on average. The Daphnia
were washed out of the drift nets into a gridded petri dish and
counted under 15X magnification. The thawed Daphnia were frequently
broken when they reached the drift nets, so when counted, the pieces
were tallied and equated to the largest Daphnia found, which measured
about 2 mm. Smaller, unbroken Daphnia were similarly tallied.

Because some Daphnia were broken, I suspected that thawing
them, mixing them in water, and drifting them downstream may have
ruptured their bodies and reduced their caloric content to fish. To
determine the caloric content of the drifting Daphnia, I first con-
verted Daphnia counts to dry weight. I counted 10 samples of Daphnia
that had been thawed in stream water, circulated with an airstone
for 1.5 h to simulate treatment in the carboy, and strained in a
drift net. I then dried them at 105 C for 24 h for dry weight

measurement. Next, to convert dry weight to calories, I similarly

31

circulated, strained and dried samples of thawed Daphnia from which
five replicates of known dry weight were taken. These were combusted
in an adiabatic bomb calorimeter to calculate calories per gram dry

weight.

 

Experiments to Measure Competition

I conducted three experiments to measure competition between
pairs of the three species of juvenile salmonids. For each experi-
ment groups of fish were selected to be as uniform in length and
length distribution as possible, and were acclimated to the stream
aquarium and to foraging on drifting Daphnia. I measured specific
growth rates, characteristics of fish positions, and behavior when
equal numbers of each species were in sympatry in both stream sec-
tions, and later when the same fish were in allopatry in separate
sections. Most experimental procedures were the same as during the
experiment on potential profit vs. specific growth rate, except as

described below.

Experiment on Brook Trout vs. Brown Trout

1. Initial Measurements and Acclimation

 

Fourteen brook trout averaging 70.6 mm in fork length (range
68.0-72.5 mm) and 3.21 g in weight (range 2.87-3.36 g), and fourteen
brown trout averaging 69.7 mm (67.0-72.5 mm) and 3.23 9 (range 2.84-
3.72 g) were acclimated in allopatry to the stream and to feeding on
drifting Daphnia for five days (Table 2). I distributed fish of each
species to sections randomly for the sympatry phase. I measured fish

position characteristics for 10 days in sympatry, then separated the

32

fish for the 10-day allopatry phase, and made similar measurements.
Between the sympatry and allopatry portions of the experiment, I
removed fish from the stream sections and placed them in 30-cm square
enclosures in the header and collector boxes under low light and
water velocity, to reduce the effects of prior residence before
measurements in allopatry began. During this two-day rest period,
fish were fed a small amount of Daphnia to maintain their weight.

To calculate specific growth rates, I measured fish length and weight
on the morning before starting the sympatry phase of the experiment,
the morning of the first rest day, and the morning after allopatry

ended.

2. Measuring Characteristics of Fish Positions

I measured fish position characteristics as during the experi-
ment on potential profit vs. specific growth rate, except that the
two-fish-length radius was 140 mm during this experiment. Previously
unmeasurable velocities less than 6 cm/sec were measured with a hot-
bead thermistor flowmeter built according to LaBarbera and Vogel
(1976). Because the meter probe tip was small, I also used this
meter to measure all velocities around drift nets during this and
all subsequent experiments. Stream discharge was adjusted to 0.00300
m3/sec (0.1059 cfs) which created velocities excluding most fish

from the upstream riffles.

3. Measuring Food and Drift Energy
In each stream section, I fed fish 20.0 g of thawed Daphnia

per day throughout the experiment. This represented 9% of fish wet

33

weight in sympatry and allopatry (Table 2), and averaged 0.331 g dry
weight and 9.1% ash (Appendix Table A2). Drift was sampled for 30
min with blocks of four or five nets set at the five distances
located at 60-cm intervals from the upstream end of each section
when fish were present during sympatry and again in allopatry. All
Daphnia captured were counted and the pieces equated to the largest
Daphnia as during the experiment on potential profit vs. specific

growth rate.

Experiment on Brook Trout vs. Coho Salmon

1. Initial Measurements and Acclimation

I acclimated 20 brook trout averaging 34.3 mm in fork length
(range 31.0-37.5 mm) and 0.31 g in weight (range 0.20-0.42 g), and
20 coho salmon averaging 34.5 mm (range 32.0-37.0 mm) and 0.28 9
(range 0.19-0.42 g) to feeding on drifting Daphnia for seven days
and to the stream aquarium in sympatry for five of these days (Table
2). I assigned fish to sections randomly for the sympatry portion
of the experiment, and measured the position of each fish for only
eight of the ten days in sympatry and in allopatry, with the usual
two day rest period between. I placed brook trout in the upstream
section in allopatry, and coho salmon downstream. Stream discharge
was 0.00198 m3/sec (0.0699 cfs) and the two-fish-length radius was
70 mm.

2. Measuring Characteristics of Fish Positions
Fish positions were measured as during the experiment on

brook trout vs. brown trout. During this and the subsequent

34

experiment, I also measured agonistic behavior by recording all
agonsitic acts received or initiated by individual fish during two
minutes of observation. I chose fish randomly so that each was
measured every other day. Agonistic acts consisted of lateral dis-
plays(recorded only when initiated) and all nips, chases, and
frontal threats. Because repeated observations of the same fish
were not independent, I used McNemar's test (Gill 1978, p. 83) to

test differences in agonistic interactions among species in sympatry.

3. Measuring Food and Drift Energy

 

Fish in each section were fed 21.0 g of thawed Daphnia per
day throughout the experiment, representing 63-77% of their wet
weight (Table 2) and averaging 0.294 g/day dry weight and 8.9% ash
(Appendix Table A2).' I improved the sampling and counting of drift
during this and the subsequent experiment by using only two drift
nets at a time in each cross-section and positioning them at least
50 mm apart to prevent the flow disturbance of one net from affecting
the other. In total, eight 20-min samples were collected at each of
the five cross-sections during sympatry and again in allopatry. In
previous experiments, fish were observed to selectively capture the
larger drifting Daphnia, so that only those less than about 0.5 mm
in diameter reached the downstream end of each section. Therefore,
to measure the drift available to each fish, instead of the total

drift, I counted only Daphnia larger than 0.5 mm in mean diameter.

35

Experiment on Brown Trout vs. Coho Salmon

1. Initial Measurements and Acclimation

I selected 16 brown trout averaging 41.3 mm in fork length
(range 39.0-43.5 mm) and 0.56 g in weight (range 0.44-0.72 g), and
16 coho salmon averaging 40461nn (39.0-42.5 mm) and 0.55 9 (range
0.43-0.65 g) from larger groups of fish acclimated in allopatry for
13.5 days in the two stream aquarium sections. Fish were acclimated
to the stream aquarium and to feeding on drifting Daphnia for such
a long period because those used in previous experiments were sus-
pected of inadequate acclimation.

For the sympatry portion of the experiment, I sorted the fish
of each species into two groups of eight fish each to further equalize
length and weight, and placed the smaller fish in the upstream sec-
tion and the larger fish downstream. I measured the position of
each fish for eight of nine days in sympatry and again in allopatry,
with the standard two-day rest period between. Coho salmon were
upstream in allopatry and brown trout downstream. The discharge was
adjusted to 0.00250 m3/sec (0.0919 cfs) and the two-fish-length

radius was 82 mm.

2. Measuring Characteristics of Fish Positions
Positions and behavior of individual fish were measured as

during the brook-trout-Vs.-coh0-sa1m0n experiment.

3. Measuring Food and Drift Energy
I fed fish in each section 50% of their wet weight in sym-
patry, which was 21.0 g thawed Daphnia per day (0.299 g/day dry

36

weight) in the upstream section and 25.2 g per day (0.419 g/day dry
weight) downstream (Table 2 and Appendix Table A2). During the
allopatry portion of the experiment, I increased the food in both
sections to 27.2 g per day (0.418 g/day dry weight), which was 52%
of wet weight of coho and 80% of brown trout wet weight. Percent
ash ranged from 7.1 to 10.6 (Appendix Table A2). Drifting Daphnia

were measured as during the experiment on brook trout vs. coho salmon.

RESULTS AND DISCUSSION

Natural Populations in Lake Michigan Tributaries

 

Juveniles of coho salmon emerged from gravel redds earlier
in the spring and at a larger size than either brook or brown trout
in four sympatric populations. Coho salmon maintained a 6-20-mm
length advantage and a 0.5-4.0-g weight advantage over brook or brown
trout during their first summer of life in the four streams. Brook
and brown trout emerged at similar sizes and grew at about equal
rates in a sympatric population, and in two allopatric populations

in streams close to each other.

Sympatry Between Coho Salmon and Brown Trout

 

In the three streams having both coho salmon and brown trout,
Bigelow Creek, Pine Creek and the Green River, coho emerged earlier
and were larger than brown trout through the first summer of life
(Figure 7). Although no newly emerged coho were caught in these
three streams, in another stream, Sand Creek, having brook trout and
coho salmon, newly emerged coho averaged 35 mm in April, which coin-
cided with the size at emergence when coho were raised in the labora-
tory. Newly emerged brown trout caught in mid-May were 28.51nn in
Pine Creek and 28 mm in the Green River, 6-7 mm smaller than coho

salmon at emergence.

37

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

'00 fi I r r r I I r l I U
A. Bigelow Creek
8C)-
6(Dr
4&3r
20 J 1 L l L L l 1 1 l L o
A 3 Pine Creek g
E 5 o
g 80- l g
3
a 5.
5 60- 8 106'
.J g g
‘3 2‘
.. o 6‘
13 40*- 13 5 "
c O
1:
O
2 2c L L O
'00? C. Green River I ‘,.}:—---]J‘20
801- ,I’ /-15
1'” T/I
601- E ,.I’ /1 ‘_____.-IO
0 ” ’a’
2' ’I’ ’I”
g I’I’ ” /
40" “J ........ I’ ’/’ ‘5
20 e 1 A ’égé’ ; 1 4 e L 1 n 0
I5 3C1 N5 31 15 13C) 15 34 I5 31 15
April May June July August September
Figure 7.--Growth of juvenile coho salmon and brown trout in three

Lake Michigan tributaries during 1979. Top curves are body
length, bottom curves are weight. A "*" denotes a sample
of three fish or less, and bars show 95% confidence inter-
vals on each mean.

39

By extrapolating coho salmon growth in Pine Creek and the
Green River back to the 35-mm emergence length, I estimated that
coho emerged about three weeks earlier than brown trout in Pine
Creek (Figure 7b), and two weeks earlier in the Green River (Figure
7c). Earlier emergence and larger size at emergence gave coho salmon
a 6-8-mm length advantage and a 1-2-g weight advantage over brown
trout throughout the summer in Bigelow Creek (Figure 7a), a 10-16-mm
and 0.5-2.0-g size advantage in Pine Creek (Figure 7b), and a 10-20-
mm and 1.5-4.0-g size advantage in the Green River (Figure 7c). For
two separate dates, a difference of only 2.3 and 2.4 mm was found
between the mean lengths of coho salmon caught by electrofishing and
by three seine hauls in Pine Creek (p>.30, p>.50), indicating that
the electrofishing was a fairly unbiaSed method of sampling juvenile

salmonids.

Sympatry Between Coho Salmon and Brook Trout

Sympatric populations of coho salmon and brook trout are
difficult to find in Lake Michigan tributaries, probably because
brown trout have access to the same streams as coho salmon, and may
exclude brook trout (Fausch and White 1981) from these stream reaches.
I found naturally-reproduced brook trout and coho salmon only in
Sand Creek, a marginal stream for salmonids due to low flow and warm
water temperatures in late summer. Although adult brook trout were
common, three hours of electrofishing on each sampling date often

produced less than 10 juveniles.

4O

Coho salmon began emerging in mid-April in Sand Creek at
35 mm, while brook trout emerged two to three weeks later at 29.5 mm
(Figure 8a). In early May, newly emerged coho were 35 mm and brook
trout were 26.5 mm in Minnie Creek, a stream near Bigelow Creek that
I sampled only once. Thus coho salmon are 5.5-8.5 mm larger than
brook trout at emergence. In Sand Creek, coho maintained an 8-14-mm
length advantage and a 2-3-g weight advantage over brook trout

throughout the summer (Figure 8a).

Brook and Brown Trout in Sympatry and Allopatry

Brook and brown trout were of nearly equal size during their
first summer of life in Egypt Creek, a sympatric population, and in
Springbrook Creek and Smith Creek, two allopatric populations in
streams close to each other (Figure 8b). Brook trout emerged at
29.5 mm in Egypt Creek, and were 1-5 mm longer and 0.5 g heavier
than brown trout through the summer. On three separate sampling

dates, one juvenile coho salmon was captured in Egypt Creek.

Coho Salmon in Allopatry

Coho salmon grew larger in Love Creek than in other streams
(Figure 8c), probably because of enrichment from an agricultural
watershed and favorable water temperatures. Love Creek also held
age-I and older brook trout of hatchery origin, but only one age-0
wild brook trout was caught. Coho grew large enough by June to equal
the lengths of the smallest of these age-I brook trout.

In summary, coho salmon were always larger than trout, while

brook and brown trout were about of equal size when sympatric. The

41

 

100 r I r I’ T I v f u
A. Sand Creek T ””1-
’L “““ I “““ I. T
80" COMO SALMON --- I 5” ’1"
BROOK TROUT ---- 3’: I.-...-{-°""“"--~._o.,-’
I ,-
60 - 880'" TWT —’” .>" I . IO
1' I
1’: .’.
I .’

FEgypt Creek (D J1

 

g 1001- Smith Creek® g
3 Spring break a
‘ Creek Q g
3 BO ’

c i
° 9..
.J a
3 6° " z
3 ..
1- a
c 4dr , V
O

O

2

to

0

I4 Q
l

 

'20 ' C. Love Creek

 

 

 

I00-
80 -
’l
60 1- ” ‘_..- « 10
I” III..-
P”? ’1’
40- ,e’ . Lengthotonewlld .5
’x Q age-0 brook trout
2c d’ 1 1 1 L e L L 1 1 0
I5 30 15 3| I5 30 I5 5| l5 3| l5
April May June July Auguet September

Figure 8.--Growth of juvenile coho salmon, brook and brown trout in

five Lake Michigan tributaries during 1979. Top curves
are body length, bottom curves are weight. A "*" denotes

a sample of three fish or less, and bars show 95% confi-
dence intervals on each mean.

42

main determinants of the coho size advantage in each stream were
emergence two to three weeks earlier and 6-8 mm longer than either
brook or brown trout. Mean lengths and weights of fish from each
stream on each sampling date are shown with 95 percent confidence

intervals in Appendix Table A3.

Specific Growth Rate as a Function of Potential Profit

 

In the stream aquariun, the specific growth rates of coho
salmon and brown trout grown in allopatry increased with mean poten-
tial profit at fish positions (Figure 9). Growth of coho approached
a maximum rate at high potential profits, but only two brown trout
grew and one was excluded as an outlier for reasons explained below.
The relationships can be described by Michaelis-Menten or Monod
functions (see equation (15)), where specific growth rate approaches
some maximum as potential profit increases, but not enough data are
available for brown trout at high potential profits to describe the
function accurately. However, these results confirm that the speci-
fic growth rates of juvenile salmonids are related to energy con-
straints dictated by a fish's stream position. Studies of salmonid
behavior show, in turn, that the stream position is determined by
constraints within the social hierarchy (Jenkins 1969). Therefore,
it is not surprising that the dominant coho salmon in this experi-
ment held positions that had the highest potential for net energy
gain (Figure 9a). Moreover, the next three coho in the dominance
hierarchy held positions that provided successively lower potential

profit. These patterns provide evidence to support the hypothesis

 

* r T f I V r f r r I v

0.040 A. Coho Salmon

0.030

LJLIAAAALA

0.020
Rainbow Trout Metabolism

 
 
   
    

erU'VtUV'VIII'I

AAILIALJAII

 

 

(l/day)

 

 

 

(1&310
o Coho Salmon Metabollem
0.000;
' 1
I I
-QOIO - all
1 I
-00020 ' 1 1 L 1 I 1 1 l I 1 L 1 "

 

0.040 B. Brown Trout

Specific Growth Rate In Weight

 

 

 

0.010
/ 11
0.000
b O O :1
I ° 1
“0.010 1
l l l l L L l l L l L l ‘

 

 

-202468|0|2|4l6|8202224
Mean Potential Profit (cthr)

Figure 9.--Specific growth rate of coho salmon (a) and brown trout (b)
in allopatry as a function of mean potential profit at fish
positions. One outlier (*) was excluded from the brown
trout relationship (see text).

44

of Fausch and White (1981) that salmonids compete for stream positions
that maximize net energy gain.

The relationship between potential drift energy and water
velocity described in the Methods section (equations (8) through (13)
and Figures 6a and 6b) generally held true for this and all subse-
quent experiments, especially at the three sampling points in the
pools (120-240 cm downstream from the food source), but the drifting
Daphnia were not uniformly distributed at the 60-cm distance. Drift
was usually sampled during 1100-1300 h during each experiment, but
a few samples from 0800-0900 h showed that fish captured a greater
proportion of the drift early in the day. These early-morning
samples and some others biased by equipment failure were not used.

The best drift-vs.-velocity relationships were achieved
during the experiment of brown trout vs. coho salmon competition
(Figure 10a). The decline in slopes of drift-vs.-velocity relation-
ships with downstream distance for the two channel sections during
the allopatry portion of this experiment were fit to negative exponen-
tial equations, shown in Figure 10b. Table 3 shows the slope-vs.-
distance equations used to estimate potential drift energy for each
experiment.

Relationships for drift without fish are only shown in Table
3 for the first experiment, because too few drift samples were taken
without fish during subsequent experiments to allow accurate slope-
vs.-distance curves to be fit. However, all samples showed that

more drift was available without fish than when fish were present.

45

 

 

 

 

 

 

 

 

 

0.80
A
(3
Distance Downstream from Food Source (cm) 0
0.60 1- 60 o 50"" -
«'1‘ 1382 0
g b 240 D 12°C , i
z 360 o -
B 040 "' 80 cm q
.3. l3 0 240cm
n
x , ,
3’
0
C rs/r a: 30° Cm
— d
I: 0 20 ./ a
E ”/’D r- " b a
‘3 “r4I‘IIM: ° " C)
/"‘ 5" I". Q
000 Zia L . I #1 g. a L . - L .‘L e I L e 1_
' 0.0 5.0 10.0 . 15.0 200 25.0
Water Velocity (cmfiec)
€0.03 w B
o 0
E0030 Section I G i
8 SuaanIC]
3002:? o "'
i:
3210201 I
3 o
O
>
$0.0I 5 r u I
.é " IIIII|._21
g n ‘: l'l
'3 "ll'
0 O~OIO ' 4
8 I
5') a
0.0075 .
0 so 120 I80 240 300

Distance Distance (cm)

Figure 10.--a. Relationships between water velocity and drift energy
at five distances from the upper end of Section I
during the allopatry phase of the brown-trout-vs.-
coho-salmon experiment.

b. Slope of the drift-vs.-velocity relationship as a
function of distance from upper ends of Sections I
and II during the same experiment. Bars show 95%
sonfidence intervals for slopes, transformed to natural
ogs.

46

TABLE 3.--Relati0nships between slopes of drift-energy-vs.-water-
velocity regressions and distance downstream from the
food source.

 

 

 

 

 

 

Stream Drift
Experiment section ~ equation
1. Profit vs. Growth
Coho salmon Upstream S=0.1002 e'0°0236 D
Brown trout Downstream S=0.0546 e'o'0207 D
Drift without fish Upstream S=0.02572 e'o'002338 D
Downstream S=0.02598 e'o'005103 D
2. Brook trout vs. Brown troutb
Sympatry Upstream S=0.05185 e-0°006076 D
Downstream S=0.06237 e'0'007885 D
Allopatry
BIOW" trout Upstream s=o.o1551 e-o.o1190 D
Brook trout Downstream S=0.02480 e'o'DODDDD D
3. Brook trout vs. Coho salmon
Sympatry . Upstream S=0.03501 e'o-0003560 D
Downstream S=0.03004 e'0'002370 D
Allopatry f -
Brook trout Upstream S=0.03298 e'0'0007465 D
Coho Salmon Downstream S=0.02458 e'o'003053 D
4. Brown trout vs. Coho salmon
Sympatry Upstream S=0.02222 e'o'003185 D
Downstream S=0.01645 e'o’003946 D
Allopatry
C°h° salmon Upstream S=0.03304 e'0'004095 D
Brown trout Downstream S=0.01539 e'o'001003 D

 

a Drift equation of general form: Slope = b.e-c-D1stance downstream(cm)

b All drift equations for this experiment divided by four (see text).

47

When the potential profits at fish positions were calculated
for each day, the cost of swimming reduced the potential energy from
the drift by a relatively small amount, usually 10-25%. Therefore,
it made little difference whether the equations for rainbow trout or
coho salmon (equations (6) or (7)) were used to calculate the cost
of swimming and potential profit. When both were used to calculate
the mean potential profit for coho salmon (Figure 9a), the threshold
values for the two relationships only differed by about 0.5 cal/hr.
In subsequent experiments, the coho metabolic equation was always
used to calculate the cost of swimming for coho, and the rainbow
trout equation was used for the brook and brown trout.

During this experiment, only three coho salmon and two brown
trout grew; the rest of the fish lost weight. However, fish were
fed only 10-11% of their wet weight, and both species captured
virtually all of the Daphnia introduced to their respective sections.
In two-hour drift samples, less than three Daphnia were captured in
each of four nets positioned 300 cm below the upper end of the brown
trout section, and none were caught in the coho section at the 300-cm
mark.

One brown trout held a position in a crevasse in the stream
bed of the upper riffle (45-60 cm) for most of the experiment, and
was assigned very high potential profits each day because the swift
currents overhead (25-35 cm/sec) and the upstream position contri-
buted to a high estimate for drift energy. However, due to the swift

currents, this fish was able to capture only a small proportion of

48

the drift passing by, and so was excluded as an outlier from the
relationship (Figure 9b).

In this and subsequent experiments, individual fish occasion-
ally swam to other parts of the stream for short periods to use
atypical positions. These outliers were detected from abnormally
high or low potential profit values for one day relative to other
days, and were excluded by the method of Grubbs and Beck (1972).

When one, or rarely two, daily potential profit measurements were
excluded for a fish, the-mean profit was recalculated using the remain-
ing measurements.

Although both coho salmon and brown trout were fed similar
fractions of their wet weight, the relationships show that coho were
more efficient in capturing this energy and converting it to growth
(Figure 9). In addition, coho salmon began growing at a lower thresh-
old of potential profit than did brown trout. The three parameters
needed to calculate the Michaelis-Menten relationship (equation (15)):
maximum specific growth rate (umax), half-saturation constant (KR),
and threshold value (T), are shown in Table 4 for each experiment.

The results of these preliminary experiments on brown trout
and coho salmon in allopatry may be used to predict which species
would grow larger in sympatry if interspecific behavior were ignored.
Thus, at equally profitable positions in the stream, coho would
quickly grow larger than brown trout. I speculate that the larger
size of coho might then confer advantages in interspecific agonistic

bouts and allow coho to dominate advantageous stream positions.

TABLE 4.--Parameters for Michaelis-Menten relationships of specific growth rate as a function of
potential profit for juvenile salmonids.

 

Parameters in Mchaelis-Menten quation

 

 

 

 

 

 

xhnum specific Half: ’Threshold
growth rate ("max) saturation potential
Experiment constant (KR) profit (T)
1. Potential Profit vs. Growth Rate
Coho salmon
Coho metabolism 0.0192 3.3 0.0
Rainbow metabolism 0.0180 1.9 -0.5
Brown trout 0.2590 99.8 1.5
2. Brook Trout vs. Brawn Trouta
Sympatry
Brook trout
Upstream Section u - -0.01049 + 0.000334 P
Downstream Section u - -0.01020 + 0.000495 P
Brown trout ' ‘
Upstream Section D 8 -0.00885 + 0.000108 P
Downstream Section u - -0.00983 + 0.000277 P
Allopatry
Brook trout 0.0229 6.1 1.4
Brown trout 0.0162 13.0 2.0
3. Brook Trout vs. Coho Salmon
Sympatry
Brook troutb 0.0365 18.5 7.7
Coho salmon no relationship was fit
Allopatry
Brook trout 0.0348 11.6 5.5
Coho salmon 0.0500 3.75 2.5
4. Brown Trout vs. Coho Salmon
Sympatry ‘
Brown trout 0.0171 22.4 7.5
Coho salmon 0.0293 4.3 1.0
A1 1 opa try
Brown trout 0.0132 5.75 4.0
Coho salmon 0.0524 8.6 1.1

 

a Linear regressions were fit because relationships were poor (see text).
b No relationship was fit because of inadequate acclimation (see text).

c Equation fit by inspection.

50

Competition Experiments

Coho salmon were clearly the superior competitor when pitted
against either brook trout or brown trout of equal size in the stream
aquarium. Brook trout were dominant over brown trout of equal size.
These conclusions are evident when the relationships of potential
profit vs. specific growth rate, the downstream distances of fish
positions, and the behavior of fish in sympatry are compared with

those in allopatry.

Brook Trout vs. Brown Trout

All fish lost weight in sympatry (Figure 11a), but brook
trout lost weight at a lesser rate than brown trout, when the stream
sections are considered separately. The weight loss is not surpris-
ing, considering that fish were fed only 9% of their body weight per
day (Table 2), were acclimated only five days to the food and the
stream aquarium, and were probably too large to forage efficiently
on drifting Daphnia. To adjust for the selectivity by trout for
large Daphnia throughout this experiment, I reduced the potential
drift energy relationships to 25% of their original values. Because
the data for sympatry are variable, I simply fit straight lines to
relate specific growth rate to mean potential profit (Figure 11a).

During allopatry, two brook trout grew, and another main-
tained its weight (Figure 11b), but all brown trout lost weight,
even though all fish should haVe been acclimated by this time. One
brown trout was excluded as an outlier, but did not occupy an atypi-

cal position. I conclude that in the stream aquarium, brook trout

51

14

 

 

 
   

 

   

 

 

 

 

 

 

 

 

00040 b t ‘ ' fi f ' U V ' F I I, v r
1 A. Sympatry 1
o 030:- Brook Trout H j
.'. Brown Trout 0— -0 I
1 1
0.020 :- Upstream 0 j
: Downstream 0 1
0.010} 1
“a : 1
v 0 000 7"
2: ' 4_— i: i
v 1' ‘
r - W. .
b b - In--- 6" -- :
5.0.010 PP . 0 $0 ‘
13 . J
3 4
cl
L 1 1 l 1 1; l 1 L 1 l j _
500200 2 4 s a 10 12 14 |6 IS 20 227’ 27 29
O
3 0.040 . . - e . - . . 2 g r r
a: r .
5 : B. Allopatry :
g 0.030 '- 1
o I .
.2 : :
’5 0.020 :- 1
03
a. ' II
(D P :1
(JJJICIP 1
P
I 3
0.0007
I . I
-0.010:, (3 1
_o.ozc . l _ l L l 4 L L L l l 1 ‘

 

 

 

L
-2 0 2 4 6 8 10 l2 I4 16 18 20 22
Mean Potential Profit (cal/hr)

Figure ll.--Specific growth rate as a function of mean potential profit
for brook and brown trout in sympatry (a) and allopatry (b).
One brown trout in allopatry was excluded as an outlier (*).

52

of this size were more efficient at foraging on drifting Daphnia
and converting it to growth than were brown trout.

Because acclimation and feeding presented problems during
this experiment, the distribution of downstream distances of fish
positions, and qualitative behavioral observations provide better
evidence that brook trout were the dominant competitor. In sympatry,
the dominant brook trout maintained positions upstream of brown trout
and actively drove them downstream in both sections (Figure 12a and
12b). The black portions of the bars representing brook trout show
the positions of the dominant brook trout in each section during the
ten days of sympatry. Brook trout held positions that were signifi-
cantly further upstream than those of brown trout in both sections
during sympatry: the mean distances were 243 cm vs. 289 cm in the
upstream section (p<.00l) and 172 cm vs. 224 cm in the downstream
section (p<.001). The frequencies of fish positions in allopatry
(Figures 12c and 12d) are shown at half scale for easy comparison
with the sympatry distributions.

In allopatry, brown trout shifted to more upstream positions
(mean distance 204 cm) when released from competition with brook
trout (Figure l2c). The dominant brown trout chased other fish
from the upstream 130 cm of stream, which was 36% 0f the total sec-
tion. One dominant brook trout drove all others downstream onto the
lower riffle (mean distance 277 cm) during allopatry (Figure l2d),
reserving 76% of the stream section for itself.

In sympatry, the four fish holding positions furthest

upstream in each section were brook trout. Two brook trout

53

RIFFLE 9901. RIFFLE ,
A. Sympatry Section!

 

Brown Trout 0
Brock Trout a

   

20 *' Mean distance

Brook Trout I72c1n(nn70)
. Brown Trout 224cm(n870)

  

  

10L .

 

      

333W“Mimi?!“iiiHiiifiiiiiiiiiiiliéiEiiiiliilmiiiii?

11112112311111

 

..

  

 

 

 

 

B Sympatry Section}!

  
  

      

 

 

 

 

20......" distance ‘
Brook Trout 243cm (11870)
- Brown Trout 2890111 (0370) i
.
E '0' 2.1: : ‘
> 32 :21 '
3 - is: a 1
a V 33 E
g ..: ‘ 35 E
o a a a , . :' :
.- 0 r2; .. "I f‘
2 c. Allopatry Section I Brown Trout
a .
.n 301-. ,
:5, Mean distance 204cm (nsi40i
z

 

 

 

 

 

Oh—————=hJI_£lJH’J1 [1 [1]] i] I] [IJ;_[L:

    
   

   
    

    
 

 

 

 

 

D. Allopatry Sectiann Brook Trout .__
30F 3 E -
Mean distance 277cm (nal40) E 5:2
20.. E .
10- E E g

  

0 so 120' 180" 240 300 - sea
Distance from Upstream End (cm)

Figure lZ.--Distribution of positions held by brook and brown trout
during sympatry (a and b) and allopatry (c and d).
Black portions of bars are positions of dominant fish.

54

controlled the area from 40-l50 cm in Section I, and one brook trout
in Section II drove all others downstream below 240 cm. There
appeared to be no difference in the patterns of agonistic behavior
of brook trout and brown trout, but brown trout were more easily
intimidated by brook trout of equal or slightly larger size.

In this and subsequent experiments, brown trout that were
forced into positions in fast water often applied the leading edges
of their pectoral fins to the stream bed to hold themselves on the
bottom with little energy expenditure. Brook trout were also
observed doing this occasionally, and it is reported that Atlantic
salmon use this technique (Gibson l977, Kalleberg 1958). Coho salmon
never rested on the bottom. Whenever trout were seen resting on the
stream bed during the brook-vs.-brown and the brook-vs.-coho experi-
ments, they were assigned a daily potential profit of zero because
I assumed they were not foraging. During the brown-vs.-coh0 experi-
ment, brown trout were assigned a focal point velocity of zero
because I assumed that they required little energy to maintain the
position, and were given a maximum velocity measured only to the
distance from the focal point that they were observed to forage.

In natural populations, male brook trout can mature as early
as age 0 and females as early as age I (Jensen l97l), but brown trout
do not mature sexually until one or more years later for each sex
(McFadden and Cooper 1964). Because the fish used in this experiment
were hatched during January l980 but were not used in the experiment

until February 198l, I suspected that the dominant brook trout may

'55

have been sexually mature males. However, when dominant brook and
brown trout were dissected after the experiment, none were sexually

mature.

Brook Trout vs. Coho Salmon

In sympatry with brook trout, the specific growth rates of
all coho salmon were positive or zero, but were unrelated to mean
potential profit (Figure l3a). It appears that coho were not suffi-
ciently acclimated to the food or the stream aquarium, although all
but four fish gained weight. Only one brook trout grew in sympatry,
but the relationship between specific growth rate and potential
profit is similar to that for fish grown in allopatry (Figures 13a
and l3b).

In allopatry, most coho salmon grew, and converted potential
profit to growth more efficiently than brook trout (Figure l3b).
Only three brook trout maintained their weight or grew in allopatry.
Moreover, brook trout required a higher threshold of potential profit
to grow than did coho salmon (Figure 13b). Ten brook trout dis-
appeared into the gravel during allopatry and were never recovered.
All of these fish were healthy, but held unfavorable stream positions
and had negative mean potential profits for the daily positions
measured before they disappeared. One coho salmon disappeared on
the last day of allopatry and was never found.

In sympatry, most brook trout held positions in the lower
half of each stream section (Figures l4a and 14b), although one

subordinate brook trout consistently occupied the upper riffle in

 

 

 

 

 

 

 

 

 

 

131>3c:_ r. .7 . .2 . - . .0. , , ,. ,1

, A Sympatry ,

I A 00110 3.41.1101! A—AZ

002° :’ A A BROOK TROUT 121-433

E A A A A A .

1

cn01cat- .3: 1

a I’ A 'i

E A A ’5’ :

g 0.000er ‘

\ D I 1

.. : a 59’s: j

2:. .00'0 a I] ‘

'3 El ['3 :

3 -0.020 0,15 -

c , % "

2 all? 3 1

c- J

'3 «0133c1 c1 -

a: 1 a a1 a 1 1 1 1 L 1 1 1 1 l
.3

3 0.030 T f I I V r r i I U I r ‘

o '_' B. Allopatry ,
t-

o . I 2

2 0.020;- 1

’5 t a,” :

a .. 1’ .1

a) 0.0. OP 1’ q

: '3 i

I I’ 2

0.000 1:1 ‘

E B I, B 1

. I q

. I d

«0010— I ‘3 -

I l 13 I

i- " s1

'0.020i:’ a, 1

D ' d

P I "i

-0030 ' 1 ' a1 9 1 1 4 J 1 1 L 1 1 J

0 2. 14 £5 8 iCl l2 ht I6 "3 2C) 22 1&4

Mean Potential Profit (cal/hr)

Figure l3.--Specific growth rate as a function of mean potential profit
for brook trout and coho salmon in sympatry (a) and allo-
< patty (b).

RIFF P
A. Sympatry Section I

_ Mean distance
Coho Salmon i5icm (n=80)

Brook Trout 2290m (n=80)

RIEFLfi__,

COHO SALMON
BROOK TROUT fl -

 
   
 
  
       

 

20

     
  

i brook trout
position in gravel

   

    

-.. ..
I.:..ODI
see essus

     
 
 

 

'- -
I 0.0:...
II

n

  

.:::::1:::1:1:;:::::::::ii

9.x»§\\\\\\\\\\\\\\\\\\\\\\\\\\\\ °
(«ZZZ/((4
' i 3 i i: 8.3

 

\\\\\\\\\\\\\\\\\
Z{l{.//.({./../{{{.//..

§§§§\\\\\‘

 

EEEEEEEEEHEEEEEE

\\\\\\\V

 

In.
I"

B. Sympatry Sect on II

_ Mean distance 9 brook trout
Coho Salmon i430m (n=80) positions in gravel
Brook Trout 23ch(n=Bd

'5
T

 

.\\\\\\\\\\\\\\\\\\\§

\.\\\\\\\\\\\\\
.\\\\\\\\\\\\\\‘
.\\\\\\\\\\\\\

 

m\\\\\\\\\\‘
\\\\\\\\\\‘

§\\\\\\\\\\
\\\\\\\\\\\‘
:1111ziiiiiiiiiiiii:
3;;;:;;flitit::::;::
iiiiiiiiiiiiiéiii

1131111111

0

C. Allopatry SectionI Brook Trout

30t- Meon distance _,
IBch (nelsd

Number of Observations
N

O

I

45 brook trout
positions in gravel

O
I

    

 

o E: m ,,
‘0”. D. A iiopotry SOCT'OD I

 
   

     

   
   

Coho Salmon .1

Mean distance
30- IBTcm (n=l60)

 

 

.\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\l

 

 

 

 

'% , ‘
20'- g g ..
/ /
% é
/ /
é %
i0 " a g -
% f
4 ¢
0 Q 5 . M
O 60 iZO iBO 240 300 360

Distance from Upstream End (cm)
Figure l4.--Distribution of positions held by brook trout and coho salmon
during sympatry (a and b) and allopatry (c and d). Black
portions of bars are positions of dominant fish.

58

the upper section. In the downstream section, the largest brook
trout was the dominant fish, shown as the black portion of the bars
in Figure l4b, and competed with the largest coho for the most
advantageous position at the head of the pool.

Positions held by coho salmon in sympatry were significantly
further upstream than those held by brook trout in Section I (mean
distances were lSl vs. 229 vm, p<.OOl) and Section II (143 vs. 238 cm,
p<.OOl). A number of brook trout burrowed into the gravel in sympatry
and allopatry to escape competitors (Figure l4).

In allopatry, brook trout shifted to positions more evenly
distributed throughout the stream section (Figure 140). The dominant
fish occupied positions in the upper riffle and upper pool from 20-
180 cm, and drove away other fish from upstream positions, apparently
attempting to defend the upper 50% of the section. The dominant coho
salmon held a position at the head of the characteristic aggregation
at mid-pool (Figure 14d). This fish also drove away all other coho
from upstream positions, but tolerated subordinate positions that
were just downstream of its own.

In general behavioral observations, coho salmon were clearly
the dominant competitor in sympatry. They appeared to be more effi-
cient at foraging than brook trout, and more persistent at maintaining
stream positions in the face of agonistic bouts--in short, coho could
eat and defend positions at the same time. In contrast, brook trout
often retreated downstream after a few nips from a coho or larger

brook trout, and often did not feed for one or more minutes after

59

an agonistic bout. Brook trout also appeared to win agonistic bouts
only against coho of smaller size.

In measurements of agonistic acts, brook trout interacted
more often with coho salmon during sympatry than among themselves
in Section II (p<.05), but not in Section I. Brook trout initiated
and received about 1.5-5.0 times more agonistic acts to and from
coho than to and from themselves (Table 5). Coho salmon showed just
the opposite pattern, initiating and receiving about 2-3 times more
agonistic acts to and from brook trout than coho (Table 5), a signi-
ficantly greater amount (Sections I and II, both p<.025). These
data indicate that, in sympatry, coho exhibited more agonistic
behavior than brook trout, and reflect that not much agonism was
required by coho to drive brook trout away. Conversely, brook trout
spent more time and energy fighting with coho than among themselves
in sympatry.

The measurements of agonistic acts in allopatry are not com-
parable to those in sympatry, because more than twice as many of
either species were present in allopatry than in sympatry. Thus, it
might be expected that agonistic acts would be about twice as fre-
quent in allopatry than sympatry. However, in allopatry brook trout
interacted about 3-l3 times more frequently than in sympatry, while
coho salmon interacted slightly less frequently in allopatry than in

sympatry (Table 5).

6C)

TABLE 5.--Sunmary of agonistic behavior among trout and coho salmon. Percents of all two-
minute observations where any agonism was observed are shown, with actual nunbers
of observations where agonism was observed in parentheses.

 

 

 

Trout agonistic acts Coho agonistic acts
From To From ‘To a Fran To From To
Experiment trout trout coho coho n coho coho trout trout

 

Brook trout vs. Coho salggg,

5W3 try

 

Section I 14 ll 23 14 35b 50 31 22 i4
(5) (4) (8) (5) (18) (11) (8) (5)
Section II 3 5 l7 17 40b 43 36 25 14
(l) (2) (7) (7) (19) (15) (11) (6)
Allopatry 4o 36 50b 33 27
(20) (18) (25) (21)
Brown trout vs. Coho Salmon
Acciimetionc 48 54 25b 26 26
(12) (15) (9) (9)
Sympatry b
Section I 3 0 l8 15 33 27 33 13 20
(l) (0) (5) (5) (11) (T3) (5) (8)
Section II 10 7 10 5 4o 33 27 TO 7
(4) (3) (4) (2) (13) (11) (4) (3)
Allopatry 42b

17 24 39 36
(7) (10) (31) (29)

 

a timer of two-minute observations.
b Fish hiding in gravel were excluded from totals.
c Data are frail 6 days of pro-experiment allopatry.

61

Brown Trout vs. Coho Salmon

All coho salmon grew when in sympatry with brown trout, were
more efficient at growing on this food source than brown trout, and
required a lower threshold of potential profit for growth (Figure
l5a). All but one brown trout lost weight in sympatry. The rela-
tionships between specific growth rate and mean potential profit
Should be accurate for both sympatry and allopatry, because all brown
trout and coho salmon were acclimated to the stream aquarium and to
foraging on drifting Daphnia for 13.5 days in allopatry prior to
this experiment (Table 2). Three brown trout died during the two-
day rest period, and one was not found after allopatry, reducing the
number from 16 to 13 during allopatry and to 12 at the end of the
experiment.

All coho salmon grew in allopatry (Figure le), grew at
similar specific rates for a given potential profit as they did in
sympatry, and had a similar threshold of potential profit. In
allopatry, coho again showed higher specific growth rates and a
lower threshold than brown trout. Brown trout grew at higher rates
in allopatry than in sympatry for a given level of potential profit,
and had a lower threshold of potential profit for growth (Figure 15
and Table 4), although only five of twelve fish grew or maintained
their weight. This change in brown trout growth rates after release
from competition with coho salmon indicates that, in the presence
of coho salmon, brown trout were either unable to forage on the
available drifting Daphnia as efficiently, or were unable to convert

as much energy to growth as when they were alone. One coho salmon

62

r f e f r I fii f I r V V V

   

 

  
  

A. Sympatry
COHO SALMON A
BROWN TROUT G

 

 

 

 

 

 

 

 

 

1 e 1 L l 1 J

O 2 4 6 8 i0 l2 i4 l6 l8 202224 26
Mean Potential Profit (cal/hr)

Figure lS.--Specific growth rate as a function of mean potential profit
for brown trout and coho salmon in sympatry (a) and allo-
patry (b). One coho salmon in allopatry was excluded as
an outlier (*).

L J l

 

63

that held a position on the upper riffle during allopatry was excluded
as an outlier for the same reasons given for a similar fish in the
experiment of specific growth rate as a function of potential profit.

Coho actively drove brown trout from the pool onto the
riffles in both sections during sympatry (Figures 16a and 16b). Most
brown trout were forced into positions in the lower pool and the
lower riffle, but two occupied the upper riffle in Section I, which
caused the mean distances of coho and brown trout to be similar there
(181 vs. 191 cm, p>.25). In Section II, coho salmon positions were
significantly further upstream than brown trout (149 vs. 260 cm,
p<.OOl), but one trout hid in a crevasse at 140-160 cm in the middle
of the dominant coho's territory in the upper pool.

During allopatry, coho salmon showed the typical aggregation
in mid-pool (Figure 16c), although the dominant fish in this experi-
ment usually occupied a position on the upper riffle. Brown trout
used positions along the entire stream section (Figure 16d), but by
the end of the allopatry portion of the experiment, one fish defended
the stream from 40-180 cm, which was 39% of the total 360-cm section,
and three brown trout controlled the stream from 40-270 cm. Dominant
brown trout were Similar to brook trout in reserving more space for
themselves in allopatry than did coho salmon.

Brown trout appeared to win agonistic bouts only when coho
were smaller, which was not often because fish were sorted into two
groups before sympatry and were all nearly equal in size. I also
observed that brown trout were less persistent than coho in maintain-

ing stream positions in the face of agonistic attacks. In

64
RIFFLE POOL RIFFLE

 

  
  
   
 

 

 

 

 

 

 

 

[
2° A. Sympatry SectionI Coho Salmon
'5 Brown Trout C]
Mean distancei
C h ' 18 :64
10 o 0 tom (n ) 9brovm trout ..
Brown-isicm (n ‘54) positions in gravel
5 - ..
0---£I:I--[]-JJH—-lfl
B Sympatry Section 1
I51" Mean distance '
Coho -i49cm (n=64) ' mar" 3'01" 1
i0- 3'“?sz pan on in grave
Q
a 5 ' § ‘
3 §
3 a
1; k
E C. Allopatry SectionI
3 30 #- Coho Soimon .
O
46 Mean distance
._ 20’ I74cm (n=128l ‘
2
E
3
2
o I
D. Allopatry Section II
30- Brown Trout .
i2 brown trout
M90" distance positions in gravel
20- 2llcm (n=l28) ‘
10'— .
OMHnmfll—Inn Wflnall
0 60 i20 lBO 240 300 360

Distance from Upstream End (cm)

Figure l6.--Distribution of positions held by brown trout and coho salmon
during sympatry (a and b) and allopatry (c and d). Black
portions of bars are positions of dominant fish.

65

measurements of agonistic acts, brown trout initiated or received

5-6 times more agonistic acts to or from coho salmon than brown trout
in Section I during sympatry, a significantly greater amount (p<.025).
But in Section II, brown trout interacted with both species with
about equal frequency. Coho salmon interacted about l.5-4 times less
frequently with brown trout than among themselves in sympatry (Table
5), which was a significant difference (Section I, p<.lO; Section II,
p<.01).

As in the brook-trout-vs.-coho salmon experiment, the fre-
quency of intraspecific agonistic acts Should be expected to approxi-
mately double in allopatry due to the increased number of fish of
the same Species. However, brown trout were 2-9 times more aggres-
sive among themselves when alone than when coho were present, especi-
ally during the acclimation period. Conversely, coho were about as
aggressive when brown trout were present than when alone. These
trends in agonistic behavior are the same as those measured during

the experiment on brook trout vs. coho salmon.

General Stream Positions and Behavior

Throughout all experiments, certain stream positions and
behavior patterns were common to each group of fish tested in the
stream aquariun. The dominant fish in each section typically occu-
pied the most upstream position at the head of the pool where the
water became deeper and water velocity decreased (see Figure 2,
Section II). This fish often held a position a few centimeters

above the bottom about 10 cm from the outside stream wall, and

66

moved to faster and deeper water near this outside wall to catch
drifting food.

The subordinate fish in each section were arrayed in a
hierarchy downstream from the dominant, and were spaced closer to
each other with each successively lower member in the hierarchy,
with a group of fish positioned 180-220 cm from the upstream end.

One subordinate fish often occupied a position behind a protruding
stone on the upper riffle, if tolerated by the dominant who often
attacked all other fish trying to gain upstream positions.

The more subordinate fish in the hierarchy used positions
on the lower riffle or in shallow water along the bar in the pool.
Among coho salmon, there was often one opportunistic subordinate
fish that took a position in shallow water on the bar at the head
of the pool, but swam 30-40 cm to the deep fast water next to the
far wall to capture drifting food. Subordinate brook and brown trout
often burrowed in the gravel, presumably to escape their competitors,
but coho simply retreated to shallow, quiet water when intimidated,

and were never seen to burrow into the gravel.

Significance of Competition Experiments

From the experiments of competition in the stream aquarium,
I conclude that coho salmon were superior competitors over either
brook or brown trout of equal size at all food levels, and when no
visual concealment was available. Coho salmon drove trout from
~ advantageous positions in the stream aquariun, Which were those posi-

tions in the upstream portion of the pool. Coho also grew faster

67

at all levels of potential profit than trout, and continued to grow
at a lower threshold of potential profit than trout. Therefore, even
if brook or brown trout could have gained positions as profitable as
those of coho, the coho would quickly have grown larger, and larger
body Size would soon have ensured their dominance over trout.

Only a few other investigators have studied interactions
among juveniles of coho salmon and either brook or brown trout.
Gibson (1977) measured agonistic behavior of coho salmon that were
nearly smolt size (>120 cm) and brook trout of equal size, in a
stream aquariun devoid of cover. He found that brook trout displaced
coho more often than vice versa. However, the smolt transformation
may have reduced coho aggression, some brook trout may have been
sexually mature, and the coho length distribution he reported
appeared to be slightly smaller than that of trout. Although these
results contradict my findings, Gibson (1977) did note that coho
and brook trout both preferred positions in pools, that coho salmon
tended to aggregate in the pool, and that coho attacked each other
more frequently than they attacked trout in sympatry, all of which
coincided with my findings.

Taube (1975) calculated population estimates of brown trout
in two sections of the Platte River, Michigan, one of the original
streams where coho salmon were introduced, before and after the
salmon were allowed to spawn there. He found a Significant decrease
in the numbers of age-O brown trout when salmon were present for

both stream sections, but concluded that the decrease had little

68

long-term effect on this year class of trout because of their greater
compensatory survival to older ages.

On the basis of eight years of juvenile salmonid population
estimates in five Lake Superior tributaries, Stauffer (1977) con-
cluded that numbers of juvenile brook and brown trout were lower
when age-O coho were abundant, and suggested that juvenile coho may
depress trout numbers. The negative correlations between coho salmon
and trout abundance from these two field studies support my conclu-
sions that juveniles of coho salmon are superior competitors over
those of brook and brown trout.

My stream aquarium studies also show that juvenile brook
trout were superior competitors over equal-sized brown trout, gaining
more of the advantageous upstream positions and actively displacing
brown trout downstream. Kjellberg (1969) observed that brook trout
excluded brown trout from small lake inlets in Sweden, and eventually
crowded them out. He attributed this to competition for territories
between juvenile brook and brown trout during their first sunner of
life, and stated that brook trout, in general, show considerably
more aggressive territorial behavior than brown trout. However, on
the basis of shifts in adult brook trout positions after brown trout
were removed from a section of a Michigan stream, Fausch and White
(1981) concluded that brown trout larger than 150 mm were superior
competitors and excluded equal-sized brook trout from advantageous
positions.

In contrast to the assemblages of exotic salmonids discussed

above, studies of interactions among salmonids that evolved together

69

suggest that these species partition stream resources in several

ways to avoid competition. Research on interactions between steel-
head trout and coho salmon (Hartman 1965), cutthroat trout (Salmg_
213151) and coho salmon (Glova and Mason 1977), and Atlantic salmon
and brook trout (Gibson 1966) show that the former species of each
set occupies the riffles and the latter the pools during the sunner.
This is often termed interactive segregation, meaning that the segre-
gation occurs as a result of behavioral interactions, and is not
genetically fixed. Consequently, one species often shifts to a
different microhabitat if the other is removed.

Everest and Chapman (1972) propose that sympatric steelhead
trout and chinook salmon avoid competition by different timing of
fry emergence. Because the salmon are fall spawners and the young
emerge earlier than the spring-spawning steelhead, the young salmon
move from the stream margins into faster and deeper water before the
steelhead young emerge. This mechanism is probably important to
segregate all sympatric fall and spring spawning salmonids during
the early part of their first summer of life, but by early fall, the
size distributions of all salmonids are about equal (cf. Hartman
1965).

The measurements of agonistic behavior during the laboratory
competition experiments reveal two main patterns during sympatry
between coho salmon and brook or brown trout: (l) trout often fought
with coho more frequently than among themselves, and (2) salmon
always fought more frequently among themselves than with trout

(Table 5). I suspect that this reflects that coho drove trout into

70

scattered, unfavorable positions by occasional attacks, but that once
the trout were in unfavorable positions they interacted little with
other trout. In contrast, coho always aggregated in the pools during
both sympatry and allopatry, which probably resulted in higher fre-
quency of agonism among coho than was directed towards trout.

The frequency of agonistic acts among brook or brown trout
in allopatry should have been roughly double than in sympatry, but
was often higher, especially during acclimation of brown trout. In
each experiment, I observed that, after 9-10 days in allopatry, one
dominant trout defended a large area in the upstream part of the pool
where food was abundant, and that the three most dominant trout
usually defended a majority of the stream area. These more dominant
trout often travelled 50-75 cm to attack other trout. I suspect that
brook and brown trout defended large territories because the stream
aquarium offered no visual isolation other than that afforded by
distance along the stream bend. In contrast, the majority of coho
salmon remained closely aggregated in the pool during all allopatric
experiments, and the dominant fish defended a moderate area of the

stream, but tolerated subordinates within about 20 cm downstream.

Relationships Between Specific Growth Rate and Potential Profit

I chose the relationships between specific growth rate and
potential profit, shown in Figure 17, as the best curves for juveniles
of coho salmon, brook and brown trout on the basis of length of
acclimation to the food and the stream aquarium, and accuracy of

drift measurement. The relationships measured in allopatry account

71

 

E v v v r r 1 I f I r I

0.040} A. Allopatry

   
 
 
  

Coho Salmon

1
0.030-
: Coho
. Salmon ,’

0.020

fvrv

0.010"-

 

9’
<3
<3
<3

I t

 

t
1 ,’ Brown Trout vsCoho Salmon Experiment
-0.0i 0 :

,’ Brook Trout vs. Coho Salmon Experiment ----

ji'IT'V

 

.Qozo L I l l l L l 1 l 1 l 1

 

0.040 B. Sympatry

0.030

0.020 Coho Salmon

Specific Growth Rate in Weight (i/doy)

0.010

Brook Troqt I Brown TTOUT

I'Tt'r'jI'UfTU'IrIr'II

 
 

 

 

    

0.000

-0.0|0

UT'IIII

 

 

 

p I
'0.020 1 L 1 1 1 1 1 1 1 1 1 1
0 2 4 6 6 10 12 14 16 16 20 22 24
Mean Potential Profit (call hr)

Figure l7.--General relationships between specific growth rate and
mean potential profit for coho salmon, brook and brown
trout in allopatry (a) and sympatry (b).

72

for differences in the basic physiology and intraspecific behavior
among the three species. Those measured in sympatry account for
interspecific behavior and any changes in intraspecific behavior due
to sympatry, as well as basic physiology. Valid comparisons can
only be made within each experiment, but the differences between
sympatry and allopatry curves for a species should be caused mainly
by interspecific competition.

When sympatry is compared to allopatry for each species
within an experiment, the curves were lower in sympatry (Figure 17).
Although no measure is available about the statistical significance
of these decreases, inspection of the original data (Figures 13 and
l5) shows that the brown trout curve appears to be the only one that
clearly changed from allopatry to sympatry. This change was mainly
due to an increase in the threshold value for potential profit of
about 3.5 cal/hr. This general flattening of the brown trout curve,
and the increase in the threshold potential profit may have occurred
for two reasons: (l) brown trout spent more time and energy on
agonism during sympatry than allopatry, or (2) brown trout foraged
less efficiently on drifting Daphnia during sympatry than allopatry.
Because the measurements of agonistic behavior (Table 5) generally
refute the first reason, I suspect that brown trout may have been
intimidated by coho, and did not forage as efficiently in sympatry
as when alone.

I suspect that a more common effect of interspecific compe-
tition on these relationships would be to move individuals of the

subordinate species to the left and down the curves, as a result of

73

their being forced into poorer positions. This appears to have
occurred among brook trout when competing with coho salmon, because
the average of mean potential profit for brook trout was signifi-
cantly less during sympatry than allopatry (p<.05), as was the aver-
age specific growth rate (p<.025). This comparison cannot be made
for the brown trout vs. coho salmon experiment because the food
level was changed significantly between sympatry and allopatry
(Table 2).

Laboratory and Field Specific Growth Rates

One way to compare the suitability of the stream aquarium
and the drifting Daphnia for growth of juvenile salmonids, with the
conditions in natural streams is to compare laboratory specific
growth rates to those in the field for fish of equal size. The rela-
tionships between mean weight and specific growth rate for each
species are shown for all streams in Figure 18. The dashed lines
are negative exponential equations fit to all points to show the
average decrease in growth rate as mean weight increases. Although
coho are always larger (Figures 7 and 8), all species grew at similar
rates for equal weight. The three species grew at high rates after
emergence in early spring, but their specific growth rates decreased
to less than 0.020 per day during the first summer of life (Figure
l8).

The three negative exponential curves from Figure l8 are
shown in Figure l9 along with the highest growth rates for individual

coho salmon, brook and brown trout grown in sympatry and allopatry

74

 

0060

 

04330)

 

 

 

 

 

 

lOINDO
1:0136C)
or
‘2’
.‘g'. B. Brown Trout
15
53CM330
O
15
m - —
5o.ooo : : .1 e r s w
3 Study Streams
ET Lowelheek lC)
0 Sand Creek El
.2 0.090 Bigelow Creek 0
3.; Pine Creek A
to ifleeniflver <7
,3- C. Brook Trout Egypt Creek 0

Snufllineek E>

0'0 30 b P 3 0.05439 .‘Q37555' NW

 

0.000 1 1 1 L 1
00 20 ' 40 60 80 IOO I20

Mean Weight (g)

 

 

JI

Figure l8.--Specific growth rate of juvenile salmonids in eight Lake
Michigan tributaries as a function of mean weight. Dashed
lines are negative exponential equations fit to all data

for each species.

75

 

 

 

 

 

 

@0950 r r . r . . r r f .

g Laboratory Measurement:

1': . Brook Trout in streams Sympatry o .

:9 Allopatry Cl

0

3 o04o l: .

5 Brown Trout in stream: 3'00“ 7'0“? O

2 l 0 Brown Trout o

1: r .

a: Coho Coho Salmon 0

5 Salmon

3 0.0m l- in stream: 4

‘ II

CD

‘, -

E F \ .

2; 3°

0') 0.000 ‘ 1 l l L 1 L L 1 m
100 21) 4!) E“) 81) lCMD

Mean Weight (9)

Figure 19.--Comparison of salmonid specific growth rates as a func-
tion of mean weight in Lake Michigan tributaries with
the highest rates for individual fish in laboratory
experiments.

76

in the stream aquarium. The highest laboratory specific growth rates
for coho salmon are close to the field rates, indicating that condi-
tions in the stream aquarium adequately simulated those in Great
Lakes tributaries, at least for coho. However, the highest labora-
tory specific growth rates for trout fell far short of those measured
in natural streams.

It is evident that brook and brown trout in the stream
aquarium lacked some critical resource for growth; probably either
food or cover. Trout may require a different food type, possibly
foraging more on benthic invertebrates than those carried in the
drift. However, Wagner (1975) found that age-0 brown trout and coho
salmon foraged mainly on midge (Tendipedidae) and blackfly larvae
(Simuliidae), which I suspect were carried to their positions in
the drift. Moreover, Chapman (l966) proposed that salmonids, in
general, can not subsist on the benthos living within their territory,
and that benthos must move to be detected by foraging salmonids.

But if juvenile brook and brown trout food requirements and ability
to feed on drift are similar to those of coho, they should have
grown at field rates in the stream aquarium, all other things being
equal.

Another critical resource that probably limited brook and
brown trout growth in the stream aquarium was cover. Within the
confines of the stream sections, there was virtually no cover afford-
ing visual isolation, and during allopatry dominant trout drove away

all other trout that they could see. The time and energy required

77

for dominant trout to defend large territories detracted from energy
for growth. Mortensen (l977) found that natural mortality of age-0
brown trout, Corrected for density-dependent mortality, was higher
in Danish streams where weeds and wood debris were cleaned out than
in control streams. In contrast, coho salmon were not oriented to
cover in laboratory studies (Hoar l958, Glova and Mason 1977) nor
were they observed to use cover in the streams I sampled unless
frightened, whereas brook and brown trout were most often associated
with cover in these streams.

These relationships between specific growth rates and poten-
tial profit, and the need for visual isolation fit with what can be
surmised about the evolution of these fishes. Neave (1958) presents
evidence that the genus Oncorhynchus evolved from an ancestral anadro-
mous §almg, Coho salmon are considered to be the most primitive of
the Oncorhynchus, but have a short stream residence compared to
steelhead trout (Salmo gairdneri), which is thought to resemble the
ancestral line from which the Pacific salmon developed (Neave l958).
Natural selection appears to have favored less stream residence and
earlier smolting for the four more advanced salmon as well; pink and
chum salmon (Q, 3233) smolt soon after emergence, chinook salmon
smolt after a few months, and sockeye salmon smolt soon after emer-
gence but have a more complex life history involving growth in a
freshwater lake before ocean residence.

During the evolution of coho salmon, it is reasonable to

suspect that natural selection would favor larger smolts. Studies

78

of the return of adults from hatchery smolt releases reveal that
larger coho smolts produce both larger adult fish and a higher pro-
portion of early returning male fish, called "jacks" (Bilton 1978,
Hager and Noble 1976). The survival rate of smolts cannot be easily
determined because the salmon returning are only those that escape
the commercial fishery. Selection pressures favoring larger coho
smolts should favor maximum growth rates during juvenile stages,
much like those I measured in the stream aquarium. However, I sus-
pect that little selective advantage would be conferred to coho
juveniles reserving large areas of the stream to ensure suitable
cover or a future food supply, because coho are not oriented to
cover and most of the energy for coho growth and reproduction comes
from the ocean.

Brook and brown trout evolved a life history of residence
in streams of the northern hemisphere, and originally were generally
Holarctic in distribution (MacCrimmon and Campbell l969, MacCrimmon
and Marshall 1968). Although the relationships between specific
growth rate and potential profit measured in the laboratory evidently
do not apply to natural streams, it is clear that these trout must
extract enough energy from streams for growth and successful repro-
duction. To do this requires reserving enough space to provide an
adequate food supply, but also a space that affords cover to ensure
concealment from predators and competitors, and refuge from high
streamflow and winter ice. Therefore, in the absence of cover, it

is not surprising that brook and brown trout attempted to reserve

79

large areas of a stream aquariun and did not grow as fast as they

did in natural streams.

Interactions AmongJuvenile Salmonids in Great Lakes Tributaries

In Great Lakes tributaries, different ages and sizes of
juvenile salmonids occupied different areas of the stream, according
to my electrofishing and snorkeling observations. Coho salmon
emerged earlier and were larger than brook or brown trout on average,
although some equal-sized age-0 coho and trout were always present.
Newly emerged trout occupied the silt flats in early spring and were
often associated with cover afforded by wood debris or aquatic vege-
tation. Coho were found in shallow open areas after emergence, but
quickly moved to faster and deeper water as they grew.

During the summer, brook and brown trout juveniles were still
associated with cover, but had moved to the faster and deeper water
at the edge of the main channel. Coho salmon were found in groups
in the deeper water at the channel edge and in the main channel dur-
ing this time of year. Thus age-0 coho salmon and brook and brown
trout may partition space resources in streams and avoid direct
competition for this resource, while both probably exploit a common
invertebrate drift resource. Age-I and older trout also occupied
the main channel and were often associated with cover provided by
logs found there. Therefore, if coho juveniles grow as large as
the smallest age-I trout during their first sunner of life, they may
compete directly with these trout for food and space, because both

occupy the main stream channel.

80

Combining the results of the laboratory experiments, where
lack of visual isolation probably reduced trout growth, and the field
observations of cover use by juvenile trout, I further speculate that
visual isolation provided by cover in natural streams may be impor-
tant to the existence of juvenile brook and brown trout, especially
when faced with competition from coho salmon. In stream areas with
little cover, age-0 coho may severely inhibit the growth and survival
of age-0 trout. In laboratory tests of interspecific competition
among equal-sized fish, coho dominated brook and brown trout. In
natural streams, coho will have an even greater competitive advantage
because of their greater size. Moreover, juvenile salmon populations
often far outnumber those of brook or brown trout because adult
female salmon produce large nunbers of eggs relative to smaller
trout. With these considerations, I would expect juvenile coho
salmon populations to thrive, and brook and brown trout populations
to subsist with difficulty in Great Lakes tributaries where coho

salmon reproduce.

CONCLUSIONS

Juvenile coho salmon emerged 2-3 weeks earlier in the spring and
were always 6-20-mm larger than either brook or brown trout when
sympatric in Lake Michigan tributaries. Brook and brown trout
emerged at similar times and grew at about equal rates in both
sympatric and allopatric populations.

The specific growth rates of juvenile salmonids are predictable
functions of the potential net energy gain or "potential profit“
measured at their stream positions, and may be described by
Michaelis-Menten or Monod equations. Dominant individuals held
positions with the highest potential profit of any fish in the
stream section, and grew at the highest specific rates. This
supports the hypothesis that salmonids compete for stream posi-
tions that maximize net energy gain.

In laboratory tests of competition juvenile coho salmon dominated
juvenile brook and brown trout, excluding them from advantageous
positions in the stream aquarium. Coho salmon also grew more
efficiently than either brook or brown trout in both sympatry
and allopatry, and required a lower threshold of potential profit
to maintain growth than did the trout.

Juveniles of brook trout dominated those of brown trout and
excluded them from advantageous stream positions. During allo-

patry in each experiment, a few dominant trout always defended

Bl

82

large areas of the stream-from subordinates, probably because of
the lack of cover affording visual isolation in the stream
aquariun.

In natural streams where age-0 coho are significantly larger
than age-0 brook and brown trout on average, coho are expected
to exclude brook and brown trout from profitable areas of the
stream, and to thereby reduce their growth and survival. The
competitive disadvantage of brook and brown trout may be reduced
somewhat if coho and trout partition space resources in the
stream along depth and velocity gradients, or if large amounts

of cover affording visual isolation are present.

LITERATURE CITED

LITERATURE CITED

Bilton, H. T. l978. Returns of adult coho salmon in relation to
mean size and time at release of juveniles. Fish. Mar. Serv.
Tech. Rept. No. 832. 73 p.

Chapman, 0. W. 1966. Food and space as regulators of salmonid popu-
lations in streams. Amer. Nat. 100: 345-357.

Connell, J. H. 1975. Some mechanisms producing structure in natural
communities: a model and evidence from field experiments.
pp. 460-490 jn_M. L. Cody and J. M. Diamond, eds. Ecology
and Evolution of Communities. Harvard Univ. Press, Cambridge,
Mass.

Diamond, J. M. l978. Niche shifts and the rediscovery of inter-
specific competition. Amer. Sci. 66: 322-331.

Drummond, R. A. and W. F. Dawson. 1970. An inexpensive method for
simulating diel patterns of lighting in the laboratory.
Trans. Amer. Fish. Soc. 99: 434-435.

Emery, L. 1981. Range extension of pink salmon (Oncorh nchus
gorbuscha) into the lower Great Lakes. Fjsheries 3(2): 7-10.

Everest, F. H. l967. Midget Bentzel current Speed tubes for ecolo-
gical investigations. Limnol. Oceanogr. l2: l79-180.

Everest, F. H. and D. W. Chapman. 1972. Habitat selection and
spatial interaction by juvenile chinook salmon and steelhead
grout in two Idaho streams. Jour. Fish. Res. Board Can.

9: 9l-l00.

Fausch, K. D. and R. J. White. 1981. Competition between brook
trout (Salvelinus fontinalis) and brown trout (Salmo trutta)
for positibns in a Michigan stream. Can. Jour. Fish. Aquat.
Sci. 38: l220-1227.

Gibson, R. J. 1966. Some factors influencing the distributions of

brook trout and young Atlantic salmon. J. Fish. Res. Board
Can. 23: l977-l980.

83

84

Gibson, R. J. 1977. Behavioral interactions between juvenile coho
salmon (Oncorhynchus kisutch), and juvenile Atlantic salmon
(Salmo saTar) and’brook trout (Salvelinus fontinalis). Int.
Counc. Explor. Sea. CM 1977/M:23 manuscript. 18 p.

Gill, J. L. 1978. Design and Analysis of Experiments in the Animal
and Medical Sciences. Vol. 1. Iowa State Univ. Press,
Ames. 409 pp.

Glova, G. J. and J. 0. Mason. 1977. Interactions for food and space
between sympatric populations of underyearling coho salmon
and coastal cutthroat trout in a stream simulator during
sunner. Fish. Mar. Serv. Manuscr. Rept. No. 1428. 36 p.

Grubbs, F. E. and G. Beck. 1972. Extension of sample sizes and
percentage points for significance tests of outlying observa-
tions. Technometrics 14: 847-854.

Gruenfeld, G. 1977. A trojan horse? Atlantic Salmon Jour. 3: 30-
31.

Hager, R. C. and R. E. Noble. 1976. Relation of size at release of
hatchery-reared coho salmon to age, size, and sex composition
of returning adults. Prog. Fish Cult. 38: 144-147.

Hartman, G. F. 1965. The role of behavior in the ecology and inter-
action of underyearling coho salmon (Oncorhynchus kisutch)
and steelhead trout (Salmo gairdneri). Jour. *Fish. TRes.

Board Can. 22: 1035-108l.

 

Hildebrand, S. G. 1971. The effect of coho spawning on the benthic
invertebrates of the Platte River, Benzie County, Michigan.
Trans. Amer. Fish. Soc. 100: 61-68.

Hoar, W. S. 1958. The evolution of migratory behavior among juvenile
salmon of the genus Oncorhynchus. Jour. Fish. Res. Board
Can. 15: 391-428.

 

Horton, R. E. 1945. Erosional developments of streams and their
drainage basins: hydrophysical approach to quantitative
morphology. Geol. Soc. Amer. Bull. 56: 275-370.

Jenkins, T. M, Jr. 1969. Social structure, position choice and
microdistribution of two trout species (Salmo trutta and
Salmo airdneri resident in mountain streams. An. Beh.
ono. 2: 53-123.

Jensen, A. L. 1971. Response of brook trout (Salvelinus fontinalis)
populations to a fishery. Jour. Fish. Res.“Board Can. #28:
458-460.

85

Kalleberg, H. 1958. Observations in a small stream tank of terri-
toriality and competiton in juvenile salmon and trout (Salmo
salar L. and S, trutta L.). Rept. Inst. Freshwater Res.
Drottningholm 39: 55-98.

Kjellberg, G. 1969. Some data on the brook char (Nggra data om
backrbdingen) Translated from Swedish. Information from the
Freshwater Lab. Drottningholm No. 4. 12 p. mimeo.

Kwain, W. and A. H. Lawrie. 1981. Pink salmon in the Great Lakes.
Fisheries 6(2): 2-6.

LaBarbera, M. and S. Vogel. 1976. An inexpensive thermistor flow-
meter for aquatic biology. Limnol. Oceanogr. 21: 750-756.

Lampert, W. and U. Schober. 1980. The importance of "threshold"
food concentrations. pp. 264-267 jg_Evolution and Ecology
of Zooplankton Communities. Spec. Symposium Vol. 3. Amer.
Soc. Limnol. Oceanogr. W. C. Kerfoot, ed.

Latta, W. C. 1974. A history of the introduction of fishes into
Michigan. pp. 83-96 jg_Michigan Fisheries Centennial Rept.
1873-1973. Mich. Dept. Nat. Resour. Fish. Manage. Rept.
No. 6.

Leopold, L. a. and w. B. Langbein. 1966. River meanders. Sci.
Amer. 214: 50-70.

MacCrimmon, H. R. and T. L. Marshall. 1968. World distribution of
brown trout, Salmo trutta. Jour. Fish. Res. Board Can.
25: 2527-2548.

MacCrimmon, H. R. and J. S. Campbell. 1969. World distribution of
brook trout, Salvelinus fontinalis. Jour. Fish. Res. Board
Can. 26: 1699-1725.

 

McFadden, J. T. and E. L. Cooper. 1964. Population dynamics of
brown trout in different environments. Physiol. Zool. 37:
355-363.

Mortensen, E. 1977. Density-dependent mortality of trout fry (Salmo
trutta L.) and its relationship to the management of sma
streams. J. Fish Biol. 11: 613-617.

Neave, F. 1958. The origin and speciation of Oncorhynchus. Fish.
Res. Board Can. Studies No. 551, p. 25-39.

Peck, J. W. 1970. Straying and reproduction of coho salmon,
Oncorhynchus kisutch, planted in a Lake Superior tributary.
Trans. Amer. Fish. Soc. 99: 591-595.

86

Sale, P. F. 1979. Habitat partitioning and competition in fish
communities. Pages 323-331 jg_R. H. Stroud and H. Clepper,
eds. Predator-Prey Systems in Fisheries Management. Sport
Fishing Inst., Wash. 0. C.

Solomon, 0. J. 1979. Coho salmon in north-west Europe: possible
effects on native salmonids. U.K. Min. Agric. Fish. Food.,
Dir. Fish. Res. Lowestoft Lab. Leaflet. No. 49. 21 pp.

Spotte, S. H. 1979. Fish and Invertebrate Culture: water manage-
ment in closed systems. .Second ed. John Wiley and Sons,
New York. 145 p.

Stauffer, T. M. 1977. Numbers of juvenile salmonids produced in
five Lake Superior tributaries and the effect of juvenile
coho salmon on their nunbers, 1967-1974. Mich. Dept. Nat.
Res. Fish. Res. Rept. No. 1846. 29 p.

Stewart, 0. J. MS. Towards a general model of salmonid metabolism.
1981 draft of manuscript.

Taube, C. M. 1975. Abundance, growth, biomass, and interrelation-
ship of trout and coho salmon in the Platte River. Mich.
Dept. Nat. Res. Fish. Res. Rept. No. 1830. 82 p.

Tilman, D. 1981. Tests of resource competition theory using four
species of Lake Michigan algae. Ecology 62: 802-815.

Wagner, W. C. 1975. Food habits of coexisting juvenile coho salmon,
brown trout and rainbow trout in Platte River, 1967 and 1972.
Mich. Dept. Nat. Res. Fish. Res. Rept. No. 1831. 14 p.

Wankowski, J. W. J. 1981. Behavioural aspects of predation by
juvenile Atlantic salmon (Salmo salar L.) on particulate,
drifting prey. Anim. Behav. 29: 557-571.

Warren, C. E. 1971. Biology and Water Pollution Control. W. B.
Saunders. Philadelphia. 434 p.

Werner, E. E. and D. J. Hall. 1976. Niche shifts in sunfishes:
experimental evidence and significance. Science 191: 404-
406.

Young, T. C. and D. L. King. 1980. Interacting limits to algal
growth: light, phosphorus, and carbon dioxide availability.
Water Res. 14: 409-412.

APPENDIX

87

.om_ um mucmamcou sswcnwppzom soc» cmumpao—eu
.mucopm>w=cmp—PPE u .cmsm

 

 

 

 

 

 

 

 

No.~ o.m~ mo.~ No. 1 mm.~ mmp mm.m cop mp Pm\m—\~

mF.N “.mm up.~ Po. Po. Fo.~ omp ¢N.m mop mp pw\-\m

cospem ocoo .m> usosu exogm .e

mm.~ m.oe em.~ po. - ee.¢ cop up.w “pp m— pm\e\m

mm.~ ~.om mm.~ Po. No. mm.~ owp em.m mup «P Fm\Np\m

cospmm ogou .m> “sock xoogm .m

NF.¢ m.mm ¢_.e mo. Po. cm.m mmm o~.m New mp pm\e\m

Fe.e «.mm . cm.¢ No. Fo. pm.m mew mo.m NFN m— mepp\~

«sock exocm .m> ozone xoocm .N

eo.~ o.ve oo.~ Po. Po. m¢.m Amp mo.m cop mp om\¢~\op

B.N 9mm cod 8. E. mm.” amp 3d «2 3. 8R}:

oweoee Pewoeoooe .m> ooem guzoeu oeeeooem ._

AF\Poeev AFMFoeav sope_P mz. N2. M2. Amoueov ze Amoueuv Auv ooee

ncoagmo co toxpm :2 oz oz mmmc xupcwp new» opanm
owcmmco:_ was; m.oos 1wcmzl toxp< cope:

Peach Agog so Lou_p\mso copuugucmocoo

.acoswgqum

comm mo use use m=_:cwmmn on» no samgmacm seesaw esp cw Loam: mo mumummgmgomgmsu quwsosull.P< u4m<h

88

TABLE A2.--Mean dry weight and percent ash of frozen Daphnia fed
per 3 h during each experiment. SEM are s own 1n paren-

 

 

 

 

 

 

theses
Dry weight
Dry Ash-free Percent fed per
Experiment weight (g) dry weight (g) ash n day (9)
1. Growth vs. Profita
Coho salmon 0.070 0.065 7.5 5 0.280
(0.00309) (0.00286) (0.711)
Brown trout 0.043 0.039 8.7 5 0.172
(0.00430) (0.00396) (0.410)
2. Brook trout vs. Brown troutb
Sympatry and 0.0827 0.0752 9.1 13 0.331
Allopatry .(0.00272) (0.00250) (0.577)
3. Brook trout vs. Coho salmonb
Sympatry 0.0681 0.0625 8.2 8 0.272
(0.00373) (0.00339) (0.158)
Allopatry 0.0786 .0710 9.5 9 0.314
(0.00224) (0.00160) (0.811)
4. Brown trout vs. Coho salmonb
Sympatry
Upstream 0.0748 0.0669 10.6 4 0.299
(0.00837) (0.00733) (0.850)
Downstream 0.1048 0.0955 9.0 4 0.419
(0.00973) (0.00929) (1.533)
Allopatry 0.1046 0.0972 7.1 8 0.418
(0.00281) (0.00281) (0.379)

 

a Weighed on balance accurate to i .001 g.
b Weighed on balance accurate to i .0001 g.

191 CIa

Brook Trout
n

C!6

[ength
(11111)
37.0

11

ER

CI6

“131”

89
Love Creek

Sand Creek

Brown Trout

Length”§52’

Gun) CIa n

 

0.1111

0.0.0.00.0.0.flw0.nm

CI

Coho Salmon
Wei? t
(9

Sample size and half-widths of 95% confidence intervals are shown for each date.

 

Length'957

Date (mm) CIa n

 

 

TABLE A3.--Mean length and weight of juvenile salmonids in eight Lake Michigan tributaries during 1979.

211246339

9
on].
.00

b

b

b
1.14

.80

0.09
0 70
1.93
3.17
3. 23
2.87
3.91

3
.5
.3

211246339

60 6
I O 0
00096006

50.55J0.~I.4

9.30” .0.3°u8
245 7767

1

1.5

b

52.0

7wnsmuaamu

981mfinwflmfl

95 3 7 777
26 6“. 6

savanna”

7n. nuuumu

36911800420
0 . e . e

44440434
0 D C C
namnmuwn
65253344
44444441
5.066788%:

t Creek

53.16le
235L534...

5379055
mmmmafl.
235788“

42576

079
632
.b . . .
.l. 1121'

97 3
67 9
.l.‘ .I.
.00 b
0
7 m
C I O
.18 1|
.1 .I.
.11 .I
0.0 b
00 0
I O 0
77 2
09 .1
e1 .1
4.164455
4.1211fi2.’
. c . . -
5567889

law Creek

22.148

92
O I

£50.75
0" .Qua'm
78
38762
2 .13
112780
5.1.177

O 0
0.11100

444.

3 4
8 4.
8 6.
6 6
2 6

2
1
3

2 9
4 9
5.9
3 2
3 2

7533‘
O O C I O
nnumm
64455
24441
erase

Pine Creek

7768225
1.1.322

539.1276

0 e 0
1533355

90.38371d4.
fiNWWflw7

fifiwwwflfl
.1805 37
1127s 33

one I o. 0
0000000

0“.9.6 8
AU..I.32334.

0.11

Smith Creek

Green River
5 rin brook Creek

7-10
8-09

6 36
.l 33
«awn
O O O

00.1.“

m 33
34.5.0.
mmuu
4455

7555

Lama
44
nan

O I O O
000.1

awn“

24.5.].

flwflw
9704
abamnmam

“8J0.

7&9

8-29
9-19

 

b Confidence intervals not shown for samples of three of fewer fish.

a Half-width of 95% confidence interval.

90

 

 

SECTION I

 

Upstream 2.5 cm Velocity

 

 

FLOW

 

 

Figure Ala.--Water velocities (cm/sec) in Section I (upstream) 2.5

cm below water surface.
at points shown.

Hater velocities were measured

 

91

 

 

SECTION I Upstream 7.5cn1 Velocity

01'

 

14
'9'
F LOW

 

 

Figure Alb.--Nater velocities (cm/sec) in Section I 7.5 cm below water
surface. water velocities were measured at points shown.

92

 

 

SECTION I Upstream l2.5 cm Velocity

 

 

FLOW

 

 

Figure Alc.--Nater velocities (cm/sec) in Section I l2.5 cm below
water surface. Hater velocities were measured at points
shown.

93

 

 

SECTION II Downstream 2.5 cm Velocity

    
 

gravel bar

 

Figure A2a.--Water velocities (cm/sec) in Section II (downstream)
2.5 cm below water surface. Water velocities were
measured at points shown.

 

94

 

 

SECTION II Downstream 7.5cm Velocity

 

 

Figure A2b.--Hater velocities (cm/sec) in Section II 7.5 cm below
water surface. Hater velocities were measured at
points shown.

 

95

 

 

SECTION II

Downstream

I2.0 om Velocity

 

 

 

FWJOVU

 

 

 

 

 

Figure A2c.--Nater velocities (cm/sec) in Section II l2.0 cm below

water surface.

shown.

Water velocities were measured at points

 

96

 

 

SECTION I Upstream Light

I/.52£5 C)

d4 4'50

j ”\I -( §\..

 

 

.\ _
. 5 4)" 4038 now
Sim/Ix

 

Figure A3.--Light (uE/mZ/sec) at surface of Section I. Light was
measured l cm above water surface at sampling point
shown. The positions of mercury vapor (open circles)
and incandescent lamps (filled circles) are shown.

 

97

 

 

SECTION II Downstream Light

 

 

 

 

 

 

Figure 4.--Light (uE/mzlsec).at surface of Section II. Light was
measured 1 cm above water surface at sampling points
shown. The positions of mercury vapor (open circles)
and incandescent lamps (filled circles) are shown.