 

1

‘ 4.. b.“
23.9., .

"mm?

when .1—

Ieg'v

9p

. a. 13.9»
. .. I!»

C-
‘ 92179-0 I-r‘zh

.l l 1...: 51 u 4 'P... .’ air “11%; priﬁtnu‘ at s L . H ‘ . . . ‘ , : .J . Jun 13:41.". A. fun]. 4 . .
. ., V .3 u \ . .7». u. A315: _ x 13312.155. ‘ . . ‘ , . ‘ . .
v! in... , .2 1.: r no. .3 1. 9....» .u .- h? w v‘ :3. . A 1.... u. l . . .. ‘

 

This is to certify that the
dissertation entitled

RELATIONSHIPS BETWEEN COASTAL WETLAND
HABITAT AND YOUNG OF THE YEAR FISH COMMUNITY
CHARACTERISTICS, DISTRIBUTION, AND GROWTH

presented by

SHARON SCHAPEL

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Fisheries and Wildlife and

 

 

l//Ecology, Evolutionary Biology, and Behavior

Clam/L

Major Professor’s Signature
4 Nor. was

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

_ LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 C'jCIRC/DateDuep65-p. 1 5

 

RELATIONSHIPS BETWEEN COASTAL WETLAND HABITAT
AND YOUNG OF THE YEAR FISH COMMUNITY CHARACTERISTICS.
DISTRIBUTION, AND GROWTH
By

Sharon Schapel

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
Ecology, Evolutionary Biology, and Behavior Graduate Program

2003

many

young

charac
two spz
charact
differen'
includ in.
ﬁsh grm
habitat (
oxygen.
perCh, a
eXDeIIm
different
I” Wetlar
Habitat (
abUnder

VeGEtatic

ABSTRACT
RELATIONSHIPS BETWEEN COASTAL WETLAND HABITAT
AND YOUNG OF THE YEAR FISH COMMUNITY CHARACTERISTICS,
DISTRIBUTION, AND GROWTH
By

Sharon A. Schapel

Although wetlands are important habitat for the early life history stages Of
many ﬁsh species, little is known about how wetland habitat Characteristics affect
young of the year (YOY) ﬁsh. This study examines the effect Of wetland habitat
Characteristics on YOY ﬁsh community composition, distribution, and growth at
two Spatial scales. At the macrohabitat scale, relationships between habitat
characteristics (eg. habitat complexity, wetland area, and percent cover of
different vegetation types) and Characteristics of the YOY ﬁsh community,
including ﬁsh abundance, species richness, diversity, species composition, and
ﬁsh growth, were examined. At the microhabitat scale, relationships between
habitat characteristics (eg. vegetation type and density, water depth, dissolved
oxygen, and turbidity) and the distribution Of YOY bluegill, banded killiﬁsh, yellow
perch, and spottail shiners were examined. Additionally, two enclosure
experiments were conducted to evaluate the growth Of YOY bluegill under
different microhabitat conditions. Young Of year ﬁsh were responding to Changes
in wetland habitat Characteristics at the macrohabitat and microhabitat scale.
Habitat characteristics that had the largest inﬂuence on YOY ﬁsh included the
abundance and type of wetland vegetation, and water depth. The amount Of

vegetation inﬂuenced species richness, diversity, community composition, and

ﬁsh gro
the pre
growth
wetlant
bluegill
oi the t
depth I
that sh
oomple
microt
correle
distribi
quality
both n
Smet
200p];
habita
type c

SUbITlI

ﬁsh growth at the macrohabitat scale. The type Of vegetation present inﬂuenced
the presence or absence of banded killiﬁsh, and the presence or absence and
growth Of bluegill at the microhabitat scale. Wetlands that had areas of deeper
wetland habitat had a higher diversity of YOY ﬁsh, and faster growth rates of
bluegill. Within wetland areas, the presence or absence Of YOY ﬁsh from three
of the four fish species examined was positively correlated with increasing water
depth within the 0 tO 100 cm range sampled. Other macrohabitat characteristics
that showed strong relationships with the YOY ﬁsh community included habitat
complexity and the presence of islands, sand bars, or gravel bars. At the
microhabitat scale, the presence or absence Of resident wetland species was
correlated with changes in the physical structure Of the habitat, while the
distribution of transient wetland species was correlated with changes in water
quality. Results Of the enclosure experiments indicated that YOY bluegill fed on
both macroinvertebrates and zooplankton when conﬁned to emergent vegetation,
submersed vegetation, or open water habitats; however they consumed more
zooplankton and less macroinvertebrates in open water than in vegetated
habitats. Bluegill growth was inﬂuenced by the density of YOY ﬁsh as well as the
type of vegetation present, growing faster at low densities, and in areas of

submersed vegetation.

l
guidanc
thank DI
me. l at
member
implemt

l
graduat.
and Ecc
Michiga
Laborati
the Micl

T
Undergn
KOVacs,
Aaron S
greatly;
Fish Lat
lots of tt

alld SUp

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Tom Coon for all of his time,
guidance, and patience with me over the past four years. I would also like to
thank Dr. Dan Hayes for the use of his lab, and for all the advice he has given
me. I also appreciate the help of Dr. Tom Burton and Dr. Mary Bremigan, who as
members Of my graduate committee, provided valuable insights into the design,
implementation, and analysis of my research.

I would also like to thank those organizations that provided funding my
graduate education and research, including: Cooperative Institute for Limnology
and Ecosystems Research, International Association of Great Lakes Research,
Michigan Sea Grant College Program, Great Lakes Environmental Research
Laboratory, Department of Fisheries and Wildlife, Michigan State University, and
the Michigan State University Distinguished Fellowship Program.

This research was conducted with the assistance of numerous
undergraduate and fellow graduate students, including Amy Morgan, Cory
Kovacs, Vern Moore, Kelly DeGrandChamp, Matt Klungle, Alfred Thomson,
Aaron Shultz, and Randy Schapel. Their help in the ﬁeld and laboratory is
greatly appreciated. I would also like to thank my fellow members of the MSU
Fish Lab, for their assistance and humor over the past four years. Finally, I send
lots Of thanks to my husband Randy Schapel for always being so encouraging

and supportive of my desire to “play with the ﬁshes.”

LIST OF
LIST OF
INTROD
CHAPTE

CHARAC
SAGINA'

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................. vi
LIST OF FIGURES ............................................................................................... ix
INTRODUCTION ................................................................................................... 1

CHAPTER 1: EFFECT OF COASTAL WETLAND MACROHABITAT
CHARACTERISTICS ON THE YOUNG OF THE YEAR FISH COMMUNITY IN
SAGINAW BAY, LAKE HURON

ABSTRACT ................................................................................................ 4
INTRODUCTION ........................................................................................ 6
METHODS ............................................................................................... 10
RESULTS ................................................................................................. 23
DISCUSSION ........................................................................................... 41
LITERATURE CITED ............................................................................... 55

CHAPTER 2: EFFECT OF COASTAL WETLAND MICROHABITAT
CHARACTERISTICS ON THE DISTRIBUTION OF YOUNG OF THE YEAR FISH
IN SAGINAW BAY, LAKE HURON

ABSTRACT .............................................................................................. 62
INTRODUCTION ...................................................................................... 64
METHODS ............................................................................................... 69
RESULTS ................................................................................................. 78
DISCUSSION ......................................................................................... 105
LITERATURE CITED ............................................................................. 116

CHAPTER 3: EFFECTS OF VEGETATION AND FISH DENSITY ON THE
GROWTH OF YOUNG OF THE YEAR BLUEGILL

ABSTRACT ............................................................................................ 123
INTRODUCTION .................................................................................... 125
METHODS ............................................................................................. 128
RESULTS ............................................................................................... 134
DISCUSSION ......................................................................................... 171
LITERATURE CITED ............................................................................. 181
CONCLUSIONS ................................................................................................ 186

Table 1.1

Table 1.2

Table 1.3

Table 1.4

Table 1.!

Table 1.1

Table 1.

Table 2.

Table 2.

Table 2_

Table 1.1

Table 1.2

Table 1.3

Table 1.4

Table 1.5

Table 1.6

Table 1.7

Table 2.1

Table 2.2

Table 2.3

LIST OF TABLES
Summary of ﬁsh collection methods used in 2000 and 2001 ......... 14

Summary Of overall ﬁsh catch, and number Of species Of young Of
year ﬁsh captured at each study site in 2000 and 2001 ................. 24

List of young Of the year ﬁsh species sampled indicating wetlands
where each was collected; where, A = AU Gres, B = Bayport, C =
Coggins, E = Essexville, F = Fish point, H = Hog Island, L =
Linwood, O = Oakhurst, P = Pinconning, R = Pine River, Q =
Quanicassee, S = Sand Point, K = Killarney, T = Thomas, W =
Wigwam. ........................................................................................ 25

General characteristics Of wetlands measured from handheld GPS,
1998 aerial photographs, maps, and visual Observations in the ﬁeld.
Numbers in the structure category represent the number of islands,
sand bars, and/or gravel bars if present, and their cumulative length
(m) ................................................................................................. 28

Summary Of transect data, indicating percentage cover of different
vegetation types (open water, emergent, submersed, and ﬂoating
leaf) within each wetland, habitat complexity, and maximum water
depth measured along transects. .................................................. 29

Mean growth rate (cm/week) Of young of year yellow perch, spottail
shiner, banded killiﬁsh, and bluegill over the week preceding
capture, in each Of the twelve wetland areas where ﬁsh were
captured during 2001. .................................................................... 36

Summary table indicating relationships between YOY ﬁsh
community characteristics, ﬁsh growth, and habitat Characteristics,
where + represent statistically signiﬁcant positive relationships, and
— represent statistically signiﬁcant negative relationships. ............. 43

Mean, standard deviation, and range of microhabitat conditions
sampled at Linwood, Bayport, and Pine River wetlands from July 1,
2001 to August 15,2001 ................................................................ 79

Number and percent of categorical microhabitat variable conditions
sampled at Linwood, Bayport, and Pine River wetlands from July 1,
2001 to August 15, 2001 ................................................................ 80

Estimated coefﬁcients, standard errors, and X2 tests for the logistic
regression model for bluegill habitat selection. .............................. 82

vi

Table 2.4

Table 2.5

Table 2.6

Table 2.7

Table 2.8

Table 29

Table 21

Table 21

Table 21;

Table 2.1;

Table 214

Table 2.4

Table 2.5

Table 2.6

Table 2.7

Table 2.8

Table 2.9

Table 2.10

Table 2.11

Table 2.12

Table 2.13

Table 2.14

Estimated Odds ratio point estimates and 95% Wald Conﬁdence
Limits for variables in the bluegill habitat selection logistic
regression model. .......................................................................... 85

Mean, standard deviation, and range of microhabitat conditions
sampled at Pine River, Linwood, Oakhurst, and Bayport wetlands
from May 15, 2001 to August 15, 2001 .......................................... 86

Number and percent of categorical microhabitat variable conditions
sampled at Pine River, Linwood, Oakhurst, and Bayport wetlands
from May 15, 2001 to August 15, 2001 .......................................... 88

Estimated coefﬁcients, standard errors, and X2 tests for the logistic
regression model for YOY banded killiﬁsh habitat selection .......... 89

Estimated Odds ratio point estimates and 95% Wald Conﬁdence
Limits for variables in the banded killiﬁsh habitat selection logistic
regression model ........................................................................... 92

Mean, standard deviation, minimum, maximum, median, and
quartiles of water depths (cm) sampled in each of the four wetland
areas .............................................................................................. 95

Estimated coefﬁcients, standard errors, and X2 tests for the logistic
regression model for yellow perch habitat selection ...................... 96

Estimated Odds ratio point estimates and 95% Wald Conﬁdence
Limits for variables in the yellow perch habitat selection logistic
regression model ........................................................................... 98

Estimated coefﬁcients, standard errors, and X2 tests for the logistic
regression model for spottail Shiner distribution ........................... 101

Estimated Odds ratio point estimates and 95% Wald Conﬁdence
Limits for variables in the spottail Shiner habitat distribution logistic
regression model ......................................................................... 103

Overall % concordance of logistic regression models and
microhabitat Characteristics that showed statistically signiﬁcant
relationships with the predicted presence or absence of YOY ﬁsh.
Numbers indicate regression coefﬁcients. For habitat and next
habitat type, letters indicate ﬁsh habitat preference, where S =
submersed vegetation, E = emergent vegetation, and O = open
water habitat ................................................................................ 106

vii

Table 3.1

Table 3.2

Table 3.3

Table 3,4

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Mean density (#/m3 1: standard deviation) of macroinvertebrates
from different taxonomic groupings sampled in experiment one
enclosures located in different habitat types. Where amp =
amphipoda, ara = aranea, col = coleoptera, com = collembola, dec
= decapoda, dip = diptera, eph = ephemeroptera, gas =
gastropoda, hem = hemiptera, hir = hirudinea, Odo = Odonata, and
tri = trichoptera ............................................................................. 142

Mean density (#/m3 :1: standard deviation) of macroinvertebrates
from different taxonomic groupings sampled in experiment two
enclosures located in different habitat types, with different ﬁsh
densities. Where col = collembola, deC = decapoda, dip = diptera,
eph = ephemeroptera, gas = gastropoda, hem = hemiptera, Odo =
Odonata, and tri = trichoptera ....................................................... 160

Mean number (i 2 standard error) Of organisms found within the
stomach Of YOY bluegill harvested from experiment two
enclosures. Cla = Cladocera, Cop = Copepod, Dip = Diptera, Eph
= Ephemeroptera, Hem = Hemiptera, Odo = Odonata ................ 167

Mean biomass (,ug t 2 xstandard error) Of organisms found within
the stomach of YOY bluegill harvested from experiment two
enclosures. Cla = Cladocera, Cop = Copepod, Dip = Diptera, Eph
= Ephemeroptera, Hem = Hemiptera, Odo = Odonata. ............... 168

viii

Figure 1.1

Figure 1.2

Figure 1.3

Figure 1.4

Figure 15

Figure 1.6

Figure 1]

Figlue 1.8

FIgUre 1.9

Fhwezt

FEWEZg

Figure 1.1

Figure 1.2

Figure 1.3

Figure 1.4

Figure 1.5

Figure 1.6

Figure 1.7

Figure 1.8

Figure 1.9

Figure 2.1

Figure 2.2

LIST OF FIGURES

Water levels (m IGLD 1985) recorded by NOAA Essexville Station
#9075035 in Saginaw Bay, Lake Huron in 2000 and 2001

compared to long term mean (1918-2000). ................................... 11
Map of approximate locations of study sites in Saginaw Bay, Lake

Huron. ............................................................................................ 13
Illustration of modiﬁed Breder trap used to catch YOY ﬁsh ............ 16

Design of 1 m2 throw trap used to sample YOY ﬁsh in wetlands in
late July thru September of 2000. .................................................. 17

Design of trapnet used to sample YOY ﬁsh in wetlands in 2001 ...19

Abundance Of YOY ﬁsh captured in wetlands without a structure
such as a sand bar, grave bar, or island, and wetlands with such a
structure ......................................................................................... 31

Graph of wetland PCOA scores for the ﬁrst two axes. Symbols

represent the ﬁrst letter in the name Of each wetland area sampled,
except for the Pine River wetland which is represented by the letter
R. ................................................................................................... 33

Relationship between A) mean growth rates (cm/week) Of YOY
yellow perch and amount of submersed vegetation cover and B)
mean growth rates (cm/week) Of perch and mean standard length
(cm) Of perch. Symbols represent the ﬁrst letter in the name Of
each wetland area sampled, except for the Pine River wetland
which is represented by the letter R ............................................... 37

Relationship between A) mean growth rate of YOY bluegill
(cm/week) and maximum water depth (cm) measured along a
transect and B) mean growth rate of YOY bluegill (cm/week) and
mean standard length of bluegill (cm). Symbols represent the ﬁrst
letter in the name of each wetland area sampled, except for the
Pine River wetland which is represented by the letter R. ............... 40

Map of approximate locations of study sites in Saginaw Bay, Lake
Huron. ............................................................................................ 70

Water levels (m IGLD 1985) recorded by NOAA Essexville Station
#9075035 in Saginaw Bay, Lake Huron in 2001 compared to long
term mean (1918-2000) ................................................................. 71

FgweZL

FgweZ!

FgmeZE

Fgwe26

FgmeZJ

Fbme31

Fewe32

Figure 3.3

Fewe34

Figme 35

Figure 3.7

Figure 3.8

Figure 2.3

Figure 2.4:

Figure 2.5:

Figure 2.6:

Figure 2.7:

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Design Of 1 m2 throw trap used to sample YOY ﬁsh in wetlands. ..73

Relationships between the predicted probability of presence Of YOY
bluegill and water depth, plant density, distance to next habitat, and
type of vegetation present. ............................................................ 83

Relationships between microhabitat Characteristics (wetland,
habitat type, next habitat type, and turbidity) and the predicted
probability of presence Of YOY banded killiﬁsh. ............................. 90

Relationships between the predicted probability Of presence of YOY
yellow perch and conductivity, water temperature, and water
depth .............................................................................................. 97

Relationships between the likelihood of presence Of YOY spottail
Shiner and water temperature, water depth, turbidity and dissolved
oxygen. ........................................................................................ 102

Water temperature (A) and dissolved oxygen concentration (B)
(mean i 2xstandard error) in experiment one enclosures located in
emergent vegetation, Open water and submersed vegetation ..... 135

Taxonomic composition (by density) of zooplankton during
experiment one. ........................................................................... 137

Density (mean :I: 2x standard errors) of A) Cladocerans and B)
copepods sampled in enclosures in different habitat types in
experiment one ............................................................................ 138

Biomass (mean pg/l 1 2x standard error) Of A) Cladocerans and B)
copepods sampled in enclosures in different habitat types during
experiment one ............................................................................ 140

Density of macroinvertebrates sampled in enclosures from different
habitats at the beginning and end of experiment one .................. 143

Biomass (mg/l) of macroinvertebrates at the beginning and end of
experiment one, sampled in enclosures located in emergent
vegetation, Open water, and submersed vegetation habitat ......... 145

Growth (mm/day) of YOY bluegill in experiment one enclosures
located in different habitat types .................................................. 146

Comparison Of YOY bluegill stomach weight (9) among different
habitat types in experiment one ................................................... 148

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

Figure 3.17

Figure 3.18

Figure 3.19

Density (mean # / stomach t 2xstandard error) (A) and biomass
(mean pg I stomach :I: 2 standard error) (B) of organisms dissected
from stomachs of bluegill in experiment one enclosures from
different habitats. Where AMP = amphipoda, DIP = diptera, CLA =
Cladocera, COP = copepod, EPH = ephemeroptera, HEM =
hemiptera, and ODO = Odonata ................................................... 149

Biomass Of A) zooplankton and B) macroinvertebrates dissected
from bluegill stomachs in experiment one enclosures located in
different habitats .......................................................................... 150

Water temperature (A) and dissolved oxygen concentration (B)
(mean i 2xstandard error) in experiment two enclosures located in
Open water and submersed vegetation ........................................ 152

Taxonomic composition (by density) of zooplankton during
experiment two ............................................................................ 153

Mean density (#/| i 2xstandard error) of Cladocerans in enclosures
located in A) open water and B) submersed vegetation during
experiment two ............................................................................ 1 55

Biomass (mean :t 2xstandard error g/l) of Cladocerans sampled in
enclosures stocked with different densities of bluegill in A) Open
water and B) submersed vegetation during experiment two ........ 156

Mean density (#/l t 2xstandard error) Of copepods in enclosures
located in A) open water and B) submersed vegetation during
experiment two ............................................................................ 157

Mean biomass (g/l :l: 2xstandard error) of copepods in enclosures
located in A) open water and B) submersed vegetation during
experiment two ............................................................................ 159

Density of macroinvertebrates (#lm3) at the beginning and end Of
experiment two sampled in open water and submersed vegetation
enclosures stocked with different densities of YOY bluegill. ........ 161

Mean biomass (g/l i 2xstandard error) of macroinvertebrates in
enclosures located in A) open water and B) submersed vegetation
during experiment two ................................................................. 163

Growth of YOY bluegill (mm/day) in experiment two enclosures

located in different habitats, and stocked with different densities of
ﬁsh ............................................................................................... 164

xi

Figure 3.20 Stomach weights (9) from harvested bluegill within experiment two

Figure 3.21

enclosures ................................................................................... 166
Biomass (pg/stomach) of A) zooplankton and B)

macroinvertebrates dissected from YOY bluegill stomachs in
experiment two enclosures located in different habitats. ............. 169

xii

COVI
component
role in shor
important it
Over 75% (
their life Cy:
habitation
provide qua
submersed
places to h.
larger DTECI;
maCIOphm
allows IOr a

INTRODUCTION

Covering approximately 1,209 kmz, coastal wetlands are an important
component Of the Great Lakes ecosystem. These wetlands play an important
role in shoreline stabilization, sedimentation, and nutrient dynamics. One Of the
important functions of Great Lakes coastal wetlands is providing habitat for ﬁsh.
Over 75% Of Great Lakes fish species utilize coastal wetlands for some portion Of
their life cycle. For many Of these species, coastal wetlands serve as important
habitat for early life history stages. There are several reasons why wetlands
provide quality habitat for young ﬁsh. Wetlands are areas with abundant
submersed and emergent vegetation. This vegetation provides small ﬁsh with
places to hide from potential predators, and also reduces the maneuverability of
larger predatory ﬁsh. Wetlands are also very productive habitats, with abundant
macrophyte, phytoplankton, and periphyton communities. This high productivity
allows for abundant populations of zooplankton and macroinvertebrates which
serve as food for many young ﬁsh.

Wetland losses have been considerable in all of the Great Lakes, with an
estimated 60-80% Of the Great Lakes wetlands having been lost as a result Of
agriculture, land ﬁlling, residential, and industrial development. In some regions,
up to 75% of coastal wetlands have been destroyed, and the remaining wetlands
in these areas are heavily degraded due to development pressures, nutrient

enrichment, watershed Changes, and pollution inputs. These activities have

 

altered the t
characteristi
alterations o
delerrnine tt

This 1
young of the
because the
their overwi

Past
uniform witl
is known at
diversities i
and QTOWth
SI'lOuld Cho
bECause gr
increase Iil
between tt
efﬁClency i
ChOose Th E
maXIleing

The
charaqens

Sun/“’3' of

altered the total amount Of habitat available to young ﬁsh as well as the
Characteristics Of the remaining habitat. In order to quantify the impact of habitat
alterations on ﬁsh populations, there is a need to identify habitat attributes that
determine the use Of Great Lakes wetlands by ﬁshes.

This study focuses on the use of Great Lakes coastal wetland habitat by
young Of the year (YOY) ﬁsh. Young Of the year ﬁsh were Chosen for study
because the growth rate Of ﬁsh in their ﬁrst year Of life is critical in determining
their ovewvinter survival and ultimately year class strength.

Past research has demonstrated that the distribution Of YOY ﬁsh is not
uniform within wetland areas or among different wetland environments. Yet, little
is known about which habitat Characteristics lead to higher abundances and
diversities Of young Of year ﬁsh, or how habitat inﬂuences species composition
and growth rates of YOY fish. Optimal foraging theory predicts that YOY ﬁsh
should Choose habitat patches that maximize energy intake per unit time,
because growth rates, particularly in the ﬁrst year, must be maximized to
increase likelihood Of survival to breeding age. But there is Often a trade-Off
between the increased protection from predation and decreased foraging
efﬁciency in areas of increased habitat structure. As a result, a ﬁsh should
Choose the environment that allows it to minimize its risk of predation while
maximizing foraging success.

The overall goal of this project is to assess how wetland habitat
Characteristics at two different spatial scales affect the distribution, growth, and

survival of YOY ﬁsh in the Laurentian Great Lakes. By increasing understanding

of which habitat Characteristics inﬂuence the distribution, abundance, diversity,
and growth Of young ﬁsh, managers will be better informed to make decisions
concerning the protection, restoration, or alteration of Great Lakes coastal habitat

and the Great Lakes ﬁsheries that depend on these habitats.

CHAPTER 1: EFFECT OF COASTAL WETLAND MACROHABITAT
CHARACTERISTICS ON THE YOUNG OF THE YEAR FISH
COMMUNITY IN SAGNINAW BAY, LAKE HURON
ABSTRACT
Wetland loss and degradation have been extensive throughout the Great
Lakes, impacting wetland habitat conditions and ﬁsh populations. The goal Of
this study was to examine relationships between wetland habitat and the young
Of year (YOY) ﬁsh community, including ﬁsh abundance, species richness,
diversity, and species composition. Additionally, relationships between habitat
Characteristics and the growth rates of YOY ﬁsh were examined. Study sites
included ﬁfteen wetland areas in Saginaw Bay, Lake Huron. At each site, habitat
Characteristics including location, percent cover Of vegetation types, maximum
water depth, habitat complexity, wetland area, and proximity to other
geographical features such as rivers and islands were recorded. Gear used to
sample the YOY ﬁsh community included seines, minnow traps, Breder traps,
throw traps, and trapnets. Habitat Characteristics that explained the greatest
variation in the YOY ﬁsh community included the percent cover of submersed
vegetation or open water, the presence Of islands and other land features, and
water depth. Percent cover Of submersed vegetation was positively correlated
with species richness and the growth of yellow perch and banded killiﬁsh, while
the amount of Open water had a negative relationship with species diversity, and
impacted species composition. The presence of sand bars, gravel bars or
islands was positively correlated with ﬁsh abundance, and species richness, and

impacted ﬁsh community composition. Greater species diversity and faster

individual growth rates of YOY bluegill were found in wetlands with areas of
deeper water. Regional factors such as distance to the outer Bay did not have
signiﬁcant relationships with ﬁsh abundance, species diversity, evenness or ﬁsh

growth; however they did impact ﬁsh community composition.

INTRODUCTION

Covering approximately 1,209 km2 (Herdendorf et al. 1981), coastal
wetlands are an important component of the Great Lakes ecosystem. These
wetlands play an important role in shoreline stabilization, sedimentation, and
system productivity (Jude and Pappas 1992). Great Lakes coastal wetlands also
serve as important habitat for various life stages of ﬁsh (Liston and Chubb 1985,
Chubb and Liston 1986, Jude and Pappas 1992, Leslie and Timmins 1994,
Minns et al. 1994, Brazner 1997, Brazner and Beals 1997). A high percentage Of
Great Lakes ﬁsh species (>75%) depend upon coastal wetlands for some part of
their life cycle (Stephenson 1990, Whillans 1992).

Wetland losses have been considerable in all of the Great Lakes, with an
estimated 60-80% Of the Great Lakes wetlands having been lost as a result of
agriculture, land ﬁlling, residential, and industrial development (Comer et al.
1995). In some regions, up to 75% of coastal wetlands have been destroyed,
and the remaining wetlands in these areas are heavily degraded due to
development pressures, nutrient enrichment, watershed Changes, and pollution
inputs, compromising their ecosystem value (Jude and Pappas 1992, Krieger et
al. 1992). For example, the alteration of habitat within coastal wetlands affects
the abundance and composition of ﬁsh populations that use these habitats (Poe
et al. 1986, Leslie and Timmins 1994, Minns et al. 1994, Brazner 1997, Brazner

and Beals 1997, Johnson et al. 1997).

Among the important but poorly deﬁned values Of coastal wetlands is their
role as habitat for the early life history stages of a number of ﬁsh species
(Jaworski and Raphael 1978). Coastal wetlands provide spawning habitat for
some species (e.g., Esox spp., Lepomis spp.) and habitat for early life history
stages Of these and other species (e.g., Perca ﬂavescens, Micropterus spp.)
(Chubb and Liston 1986, Herdendorf et al. 1981, Stephenson 1990, Petering and
Johnson 1991, Jude and Pappas 1992, Brazner et al. 1998). In order to evaluate
the impacts of wetland loss and degradation on ﬁshery resources throughout the
Great Lakes, more research is needed on the ﬁsh community structure within
coastal wetlands, and the function of wetlands as spawning areas and habitat for
early life history stages Of ﬁsh.

Potential beneﬁts of coastal wetlands for juvenile ﬁsh include the high
primary productivity Of these environments, reduced risk of predation, and rich
zooplankton and benthos food sources that wetlands provide (Jude and Pappas
1992). Macrophytes are a key component of the habitat Of Great Lakes coastal
wetlands, providing young of the year (YOY) ﬁsh with refuge from predation by
serving as protective cover that reduces the visibility of YOY ﬁsh and the
maneuverability Of adult piscivores (Savino and Stein 1982, 1989, Werner et al.
1983, Mittelbach 1986, Gotceitas and Colgan 1987, Werner and Hall 1988,
Gotceitas 1990a,b, Heck and Crowder 1991 ). Vegetation also plays a role in
determining the quality and quantity of invertebrates within wetlands (Turner

1988, Cardinale et al. 1997, Cardinale et al. 1998).

The abundance, growth, and survival of juvenile ﬁsh are not uniform within
a single wetland or among wetlands within the same lake or geographic area
(Chubb and Liston 1986, Bryan and Scamecchia 1992, Baltz et al. 1993, Brazner
1997, Brazner and Beals 1997, Weaver et al. 1997, Duffy and Baltz 1998,
Jackson and Jones 1999). Many ﬁshes segregate within and among habitats
along gradients of distance from shore, vegetation structure, temperature, and
depth (e.g., Keast and Harker 1977, Werner et al. 1977, Johnson and Stein
1979, Brandt et al. 1980, Keast 1984; Killgore et al. 1989, Bryan and
Scamecchia 1992, Michaletz 1998, Pierce et al. 2001). Understanding the
spatial distribution of organisms in ecosystems is of crucial importance for
understanding ecosystem functioning (Hayes et al. 1996). In order to determine
the impact of habitat alterations on ﬁsh populations, there is a need to identify
habitat attributes that affect the use of Great Lakes wetlands by ﬁshes (Brazner
and Beals 1997).

This study builds upon previous efforts to identify linkages among habitat
attributes and ﬁsh communities, by examining not only community characteristics
and species composition, but also relationships between wetland habitat and
YOY ﬁsh growth. Additionally, this study was carried out within a single aquatic
system, Saginaw Bay, Lake Huron, thereby minimizing potential effects of
differences in adult ﬁsh community Characteristics among lake systems.

The goal of this study was to examine relationships between habitat
characteristics Of wetland environments and the YOY ﬁsh community.

Speciﬁcally, relationships between wetland habitat characteristics and attributes

Of the YOY ﬁsh community such as ﬁsh abundance, species richness, species
diversity, and species evenness were examined. Additionally, I sought to identify
which habitat Characteristics may play an important role in determining the
species composition of the YOY ﬁsh community in different wetland areas.
Finally, because the growth rates of freshwater ﬁsh within their ﬁrst year of life
are of critical importance in determining their likelihood of winter survival and
Ultimately year-Class strength (Houde 1994), this study also examined the
inﬂuence of wetland habitat Characteristics on the success Of YOY ﬁsh from four

different species as measured by their growth rates.

Slur

Con
witi
al. ‘
slit
of ti
tovt
Pa;
exc

Shc

in I
201

aDl

are

Sm

win

METHODS

Study Site Description

My study sites were located within Saginaw Bay, Lake Huron, USA.
Compared to the rest of Lake Huron, Saginaw Bay is a eutrophic embayment
with high productivity (Jude and Pappas 1992, Skubinna et al. 1995, Cardinale et
al. 1998). The Saginaw River ﬂows into the southern portion Of the Bay, carrying
silt and nutrients, and thereby affecting water quality along the eastern portions
of the Bay. Conditions gradually improve from the mouth of the Saginaw River
toward Lake Huron, creating gradients in environmental conditions (Jude and
Pappas 1992). Several other smaller tributaries ﬂow into Saginaw Bay. With the
exception of urban and industrial landuses near Bay City on the southwestern
shore of the Bay, the surrounding land use is primarily agricultural.

Saginaw Bay is relatively shallow; eighty percent of its volume is less than
5.5 m deep (Beeton et al. 1967). Water levels within the Great Lakes are cyclical
in nature, with alternating periods Of high and low water (Geis 1979, Boutin
2000). Water levels in Saginaw Bay during the study years averaged
approximately 0.5 m below the long term mean (Figure 1.1). There are extensive
areas Of gravel and sand beds scattered throughout the Bay, as well as several
small islands. These structures may help to shelter the coastal wetlands from
damaging wind and wave energy as strong southwesterly and northeasterly

winds are common within Saginaw Bay (Schelske and Roth 1973).

10

 

Water Level (m)

 

 

 

177.0
—0— 2000
--I--2001
+Mean
176.54
A—‘K‘
’7‘:'....i- ...\ ,-
1760~ ,o-e’x -\... ~~~~~~ .
o.‘ ,A’ \
* .aw' ' \T‘o
l--""
175.5 E T 7 f m. T t I f I r

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Month

Figure 1.1: Water levels (m IGLD 1985) recorded by NOAA Essexville Station
#9075035 in Saginaw Bay, Lake Huron in 2000 and 2001 compared to long

term mean (1918-2000).

11

A great diversity Of coastal habitats exists within Saginaw Bay. There are
large and extensive wetlands with emergent vegetation communities dominated
primarily by Scirpus pungens and Typha angustifolia. Submersed plant beds are
dominated by Myriophyllum spicatum, Valisineria americana, Potamogeton spp.
and Chara globularis.

Fifteen different wetland areas located throughout the Bay were selected
as study sites and sampled in the summer of 2000 (Figure 1.2). Due to lower
water levels in 2001, only twelve of these wetlands were sampled in 2001 (Au
Gres, Essexville, and WigWam were not sampled). All of the wetlands sampled
were at least 1.5 hectares in size, and separated by at least 10 km. Seven of
these sites are located along the western shore Of Saginaw Bay, while eight Sites
are located along the eastern shore of Saginaw Bay. Study sites varied in
habitat Characteristics, with some wetland areas being rather small and
contained within distinct boundaries (e.g., Bayport), while other wetlands were a
portion of continuous vegetated habitat along the coast that stretched on for
many kilometers (e.g., Pinconning). The wetlands also varied in the presence of
a structure such as islands, sand bars, or gravel bars that may Offer protection

from wave action.

.Fish Collection

A variety of gear types was used to sample ﬁsh in 2000 (Table 1.1). As

the size of YOY ﬁsh increased and vegetation growth became more proliﬁc, it

12

 

 

WiWam

Pine River Au Gres %

Pinconning .

. Sand Point

O

Coggins Bayport
Hog Island
Linwood

' Fish Point

5 Thomas

Killarne

Essexville Oakhurst

Quanicassee

 

 

Figure 1.2: Map of approximate locations of study sites in Saginaw Bay,

Lake Huron.

13

 

Table 1.1: Summary Of ﬁsh collection methods used in 2000 and 2001.

 

 

#sWetlands T333352: Start Date End Date
ampled wetland

Seine 15 5 6/12/2000 6/28/2000
Minnow trap 1 5 20 7/6/2000 7/1 3/2000
Breder trap 1 5 10 7/25/2000 8/14/2000
Throw trap 1 5 10 7/25/2000 8/14/2000
Throw trap 12 1 0 9/1/2000 9/4/2000
Trapnet 12 4 7/23/2001 8/10/2001

 

14

wa
tak
site

ear

sei

tra

20

we
Ple
5n

b0

was necessary to switch sampling gears. The same number Of samples was
taken at each of the ﬁfteen wetlands with each gear type. For each gear type, all
sites were sampled within a three week period, for a total of ﬁve sample events in
each wetland over a 15-week period in 2000.

A 5 m beach seine (3mm mesh) was used in May 2000, and five 15 m
seine hauls were conducted at each site.

Minnow traps were used to sample ﬁsh in early June 2000. A total Of 20
traps were set overnight in each wetland in water less than 1 m deep. Ten of the
20 traps were baited with saltine crackers.

Ten modiﬁed Breder traps (Breder 1960) were set overnight in each
wetland in late July thru early August 2000. These traps were constructed of
Plexiglas, with dimensions Of 32 cm x 17 cm x 17 cm, and 46 cm long wings. A
5m long lead made of 5 mm mesh nylon netting with lead weights and ﬁshing
bobbers for ﬂoats was attached to each trap to increase catch (Figure 1.3).
Rebar placed through holes drilled through the trap was used to attach the lead
to the trap and fasten the other end into the sediment.

A modiﬁed throw trap (Figure 1.4) was used to sample ﬁsh within the
wetlands in late July thru September 2000. Throw traps have been shown to be
an effective method of sampling YOY ﬁsh in vegetated wetlands (Kushlan 1981,
Freeman et al. 1984, Dewey 1992, Baltz et al. 1993, Raposa and Oviatt 2000).
The throw trap used was designed with an angle iron base, weighted with lead,
to cause the trap to sink rapidly. A PVC (4 cm diameter) frame was used for the

t0p of the trap, which would ﬂoat on the water’s surface. Three mm mesh nylon

15

Overhead View

 

5m

V.'..'"V'V'"""..'."'.'...’V’..'.'."' .'
.r.o2c.4020to202¢20202920292024.1024? 202020202020202020302401024.3292. .02

   

D 0
'20.

17cm

 

 

32 cm

  

Side View

5m 17cm

          
   
  

 

 

 

 

Ill-III --__. ._-
yvvvvvvvvvvvvvvvvvvvvvvvv'vvvvvvvvvvvvvvv'vvvvvvvvvvy
QﬁQQﬂQﬁ{éﬁﬁégéQﬁégQQQQﬁﬁﬁﬁﬁgééﬁﬁéﬁéﬁQﬁégﬁﬁﬁﬁﬁe@ﬁﬁﬁﬁﬂ‘

0.9 0.0 0.0 o o o o o o o o o o 0.0 9.0 o o o o o o o o o o o o o o 9 0.0.0.0...9’. t...o.o.o.o.q
oo.ococo.oooooooooooooooooooooeooooooooooeoooooq
co 00 0000 o o 90 090 o
o ’0’... .0 o o o 0.9 0.0 o o 9.0 o o o

90 o 0009

   
    
 
     
     
  

 
 
   

 

  
    

o
co... 9
oooooo 09 0.00099
Qﬁ’ﬂﬁ qegepww.
00
.6

    

00 o
ﬁﬂﬁﬁﬁﬁﬁﬁ
ﬁﬁﬂﬂﬁﬁﬁﬁwb
900 see.
0’0... 0“ ’o’b’o’o’a’o

 
 
 
 

32 cm

Figure 1.3: Illustration of modiﬁed Breder trap used to catch YOY ﬁsh.

Buoyant PVC 1” diameter 1 m2 frame

 

 

 

 

 

 
 
       

/

3 mm mesh sides Of trap

 

 

 

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

'\

1 m2 bottom frame, constructed Of angle-iron and weighted with lead

—
r\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ \\

 

Figure 1.4: Design of 1 m2 throw trap used to sample YOY ﬁsh in wetlands in

late July thru September Of 2000.

17

netting was used for the sides of the throw trap. The trap sampled 1 m2 of
wetland habitat up tO 1 m in depth. The net was deployed by two people tossing
the frame into the water. Pihl and Rosenberg (1982) found that animals
generally did not exhibit an avoidance response to throw traps until moving
people were within 1.5 m, and our throws typically covered a distance Of three to
four meters. The trap was deployed ten times in each wetland habitat. After
deployment, the bottom Of the frame was pushed into the sediment. A 5 mm
mesh seine was used to remove fish from the trap. Seining continued until six
consecutive seine hauls resulted in no additional ﬁsh being captured.

Trapnets (Figure 1.5) were used to sample ﬁsh in late July and August Of
2001. Four traps were deployed overnight in each of the twelve wetland areas.
The traps consisted of a PVC frame measuring 100 cm x 50 cm x 35 cm. The
PVC was ﬁlled with a sand and gravel mix to cause the trap to sink and increase
rigidity. Mesh netting (3 mm) was used for the sides and funnel of the trap. The
funnel had a round opening of 11.5 cm. Wings constructed of 5 mm mesh
netting were attached to the traps. Each wing was 7.5 m in length, with lead
weights and plastic ﬂoats.

With all gear types, a stratiﬁed random sampling procedure was used to
select sample locations. Effort was made to sample the diversity of habitat
conditions available to YOY ﬁsh within a given wetland in proportion to the
availability of that habitat type. All ﬁsh collected were identified to species, and
counted. At each sample location, the standard length of ﬁve to ten individuals of

each species was recorded. A minimum often individuals per species were

18

Overhead View

 

 

 

 

 

50 cm
width
7.5 m long 100 cm length
Wing
Side View
/ 100 cm length

7.5 m long 11.5 cm funnel 35 cm
Wing opening height

 

 

 

/

Figure 1.5: Design of trapnet used to sample YOY ﬁsh in wetlands in 2001.

19

preserved in a 90% ethanol solution, for otolith analysis of growth rates. An
Optimas imaging system was used tO measure the total radius Of the otolith, and
the radius of the otolith before the last seven rings were deposited. Growth rates
of ﬁsh over the week preceding their capture were calculated according to

procedures for the body proportional hypothesis outlined in Francis (1990).

Habitat Measurements

The location Of each sample was recorded in the ﬁeld using a hand held
Garrnin GPS unit. The presence of gravel bars, sand bars, islands or other
structures that may serve to shelter the wetland from wave energy was also
recorded.

Five 200 m transects were established within each wetland to sample
characteristics Of the vegetation community. Random locations were selected
along the shore for the starting point of each transect, with the criteria that all
transects were located at least 20 m apart. Transects ran perpendicular from
shore into the wetland. Water depth was measured every ﬁve meters along each
transect. Additionally, the presence of Open water, submersed, emergent, or
ﬂoating leaf vegetation was recorded every ﬁve meters. These data were used
to determine the percent cover of Open water, submersed vegetation, emergent
vegetation, and ﬂoating leaf vegetation, habitat complexity, and the maximum
water depth along the transects for each wetland.

Black and white aerial photographs from 1998 were Obtained from the

Michigan DNR website. The MrSlD GeOViewer software program was used to

20

measure the area of each wetland and the distance from each wetland area to
the nearest river. These photographs were also used to measure the size of
islands, sand bars, or gravel bars adjacent to the wetlands sampled.

USGS topographic maps were used to determine the distance from each
wetland to the Saginaw River and to the outer Bay. The outer Bay was deﬁned
as water northeast of a line connecting AU Gres Point on the western bank Of
Saginaw Bay to Sand Point on the eastern bank of Saginaw Bay (Figure 1.2).
USGS maps were also used to categorize each wetland as on the eastern or
western bank of Saginaw Bay, with the Saginaw River being used as the dividing

line (Figure 1.2).

Data Analysis

Analysis Of Covariance (ANCOVA, fonrvard selection procedure, a = 0.05)
was used to relate ﬁsh abundance, species richness, species diversity, and
species evenness to habitat Characteristics of each wetland. Only the twelve
wetlands that were sampled in both 2000 and 2001 were used for this analysis.

Principal Coordinate Analysis (PCOA, City block distance measure) was
used to compare the YOY ﬁsh communities in each of the twelve wetlands
sampled in 2000 and 2001. A matrix was constructed containing information on
the presence / absence Of all species collected within the wetlands. Species that
were found in only one Of the twelve wetlands sampled were dropped from this
analysis. Jaccard’s similarity index was used tO measure similarity of the YOY

fish communities among wetlands. ANCOVA analysis was used to examine

21

relationships between the PCOA scores Of each wetland for the ﬁrst two PCOA
axes and habitat Characteristics of those wetlands.

The mean weekly growth rate of four ﬁsh species, yellow perch, Perca
ﬂavescens, spottail shiner, Notropis hudsonius, banded killiﬁsh, Fundulus
diaphanous, and bluegill, Lepomis macrochirus, over the week preceding their
capture was compared among wetlands. TO minimize the variability in growth
rates due to time, ﬁsh size, ﬁsh age, and year, only ﬁsh collected over the 3-week
sampling period in 2001 were used for this analysis. Approximately 20 to 30 ﬁsh
of each species per wetland were used in this analysis. Growth rates were
calculated according to the body proportional hypothesis model outlined in
Francis (1990). ANCOVA was used to examine relationships between mean ﬁsh
growth rates in different wetland sites and the habitat characteristics Of those

sites.

22

RESULTS
General Catch Characteristics

A total Of 3,800 young of the year ﬁsh was collected in the ﬁfteen wetland
areas throughout the study period. Fish sampling in 2000 resulted in a capture of
2,768 YOY ﬁsh, while 1,032 ﬁsh were collected from the twelve wetland areas
sampled in 2001 (Table 1.2). In 2000, total catch of ﬁsh ranged from a low Of 10
at WigWam to 684 at Bayport. In 2001, total YOY catch ranged from 8 at
Killamey to 396 at Linwood. The number of ﬁsh species caught at a single
wetland in 2000 or 2001 ranged from two to eighteen (Table 1.2).

Young of year ﬁsh from 36 different species were collected throughout the
two-year study period. WigWam had the lowest species richness, with only two
species collected (Table 1.3). The highest number of species was collected at
Linwood, where young of year ﬁsh from 21 species were collected. Common
carp (Cyprinus carpio ) was the only species found in all ﬁfteen wetland areas.
Other common ﬁsh species included quillback (Carpiodes cypn'nus), banded
killiﬁsh, bluegill, spottail shiner, largemouth bass (Micropterus salmoldes), and
yellow perch. Twelve ﬁsh species were collected in only one of the ﬁfteen

wetland areas (Table 1.3).

23

Table 1.2: Summary Of overall ﬁsh catch, and number Of species of young of

year ﬁsh captured at each study site in 2000 and 2001.

 

 

Total Fish Caught # Species Caught
2000 2001 2000 2001

Au Gres 382 - 5 -
WigWam 10 - 2 -
Pine River 20 82 7 6
Pinconning 195 82 4 2
Coggins 96 14 5 4
Linwood 160 396 18 1 1
Killamey 304 8 1 1 3
Essexville 105 - 7 -
Quanicassee 29 22 5 5
Oakhurst 132 28 10 5
Thomas 36 14 4 3
Fish Point 213 34 6 8
Hog Island 389 247 8 6
Bayport 684 55 13 3
Sand Point 13 50 3 6

 

24

 

x x x noomExQEaa
x x x x x x x cmccam :85
x x Low 80:98.
x 37.523 x005
x 558 .9520
BocEE >3an
x x x x x x x x x x x x x 55:: amazon
x x x Loren >552
x 93 52:0:
cmcm EmNNa
xxxxxxxxxxxxxxxemocoEEoo
855 :52?
x EEO exp.
x .983 233
x x 3.82.5".
5:28
cEson
x “385.3 .565
x 39:59 26=o>

X

X

X

X X
X

X

X

X
X

X
X
><><><><><
X

X
X

mmmn x02

383.5 £583
3:0:me £qu3
«58:00 32833
«538.6 mOSOOUBS
2:833 9:332
.Eowsxcmc 35$?an
3:3qu «Suntan
SEQ: mESmomsm
«5.63 xowm

Ezcmﬁoqmo mEomoLoQ
0.288 mzctho
$3.5qu mamctho
OSOQEEQ «.3330
.EowLoEEoo manwmemo
outtake monoEBmo
«.583 35.350

meg NEE

38393: mEEoE<
maﬁa: wEEoEc‘
«Saga OOEQOEE<

 

m m I n. ._. O O w v. .— 0 a 1 >> < :OEEOO

oEa: oEEEom

 

.Em>>m_>> u >> .mmEoE. u ._. .>oEm___v. u v_ 450d ccmm n m .oowmmoEmao n O :95. 9.5 u m .mcEcooEn.

u d .655me n 0 60035.. n ._ 652m. mo: u I .E_on_ :mE um. 6:58me u M 65800 n 0 .toa>mm u m 620

:< u < .9653 638:8 mm; comm 9oz; 8:363 @5865 uoEEmm moOoam cm: Loot» 9: Co mcao> Co 63 an... aim...

25

 

 

x o>o__m>> 829:3 5.8.8835
x 0:590 :83 «Bosommumwmm
305E
x $0595 350: mgmcquE
x x x x x :93 26:9» 25896: moLom
x LocEm O_E_E 3:839 @5982
555 oomSmoL «5:32 @3962
x x 555m 0:3 323.63: 3982
x x x x x x x x .82» .588 SE82: ”.5982
x x .055 085.83 GLASSES: mﬁoboz
x x 55% 289cm $28.:ch mEoEoz
x 555m 520m “Mum.“mewmz
3528. c0200 EEESEO oEoBoon
x mmmn 2E; mqowbco 0:902
x x x x x x x 32 EnoEomﬁ 8223mm 25390.5
x x 55% coEEoo @3358 9:33
x x x x x x x :63: 235985 e583
w m I G v. ._ n. 1 >> < :OEEOO oEn: oEucﬂom

 

.6353 3 2%»

26

General Wetland Characteristics

Seven wetlands were located along the western shore Of Saginaw Bay,
while eight were located on the eastern shore (Figure 1.2, Table 1.4). The size
of the wetlands ranged from 1.6 hectares at Coggins to 26.6 hectares at Hog
Island. The distance from the wetland to the nearest river ranged from 0 km at
Quanicassee to 8.0 km at Oakhurst. The distance from wetlands to the Saginaw
River ranged from 4.8 km (Killamey) to 41.8 km (Au Gres and Sand Point), while
the distance to the outer portion of Saginaw Bay ranged from 2.4 km (Au Gres) to
43.9 km (Quanicassee). Six Of the ﬁfteen wetlands sampled had one or more
structures such as sand bars, gravel bars, or islands that may have provided the
wetland area with some degree Of shelter from wave action. The cumulative
length of these islands, sand bars, or gravel bars ranged from 253 m to 2,173 m.

There was a broad range in the percent cover of open water within a
wetland, with values ranging from a low Of 10% at Pine River to 96% at
Essexville (Table 1.5). Percent cover by emergent vegetation ranged from 1%
(Killarney and Oakhurst) to 68% (Pinconning). The amount of cover provided by
submersed vegetation also had a broad range from 0% (Pinconning and Fish
Point) to 69% (Pine River). The amount of cover by ﬂoating leaf vegetation was
very low in all ﬁfteen wetlands sampled. Only two Of the wetlands (Linwood and
Au Gres) had any patches of ﬂoating leaf vegetation intersecting a transect.
Percent cover of ﬂoating leaf vegetation was 5% at Linwood and 22% at AU

Gres. All other wetlands sampled had 0% cover of ﬂoating leaf vegetation.

27

Table 1.4: General Characteristics Of wetlands measured from handheld GPS,
1998 aerial photographs, maps, and visual observations in the ﬁeld. Numbers in
the structure category represent the number of islands, sand bars, and/or gravel

bars if present, and their cumulative length (m).

 

 

GPS Shore Area km to km to km to
Coord. Of Bay (ha) Nearest Sag. Outer Structure
River River Bay
Au Gres ggffgx W 10.9 3.8 41.8 2.4 2, 378 m
WigWam 33:23,” W 5.6 1.1 39.9 10.5 1, 800 m
Pine River 3:: 3?” W 7.0 0.8 36.5 15.0 -
Pinconning g3: :52,” W 6.1 2.4 24.9 23.6 -
Coggins 3:: £35,: W 1.6 0.2 18.0 29.0 -
Linwood 33: 51ng W 8.1 0.1 14.5 34.6 1, 1914
Killamey 33: 32:” W 5.1 1.7 4.8 40.2 -
Essexville 33:25” E 8.8 5.9 6.7 41.3 -
Quanicassee :3: 45:9,,” E 6.1 0 16.4 43.9 -
Oakhurst 3:: 3785:, E 5.6 8.0 20.4 35.9 -
Thomas 83:33:11! E 6.5 4.9 26.0 28.1 -
Fish Point 33: 335i E 14.6 1.3 34.3 25.7 4, 551
Hog Island 3:: 33:” E 26.6 1.9 30.0 12.9 6, 2173
Bayport 33:321.,” E 2.9 3.8 37.5 7.5 2, 253
Sand Point 3;: 253,2, E 6.6 0.1 41.8 6.4 -

*

 

28

 

Ti

GE

tr:

Table 1.5: Summary of transect data, indicating percentage cover of different
vegetation types (open water, emergent, submersed, and ﬂoating leaf) within

each wetland, habitat complexity, and maximum water depth measured along

 

 

transects.
% Cover Max.
"/0 Cover °/o Cover % Cover _ .
Floating Habitat Water
Open Emer. Submer.
Leaf Comp. Depth
Water Plants Plants
Plants (cm)
Au Gres 33 3 42 22 29 15
WigWam 60 27 13 0 22 15
Pine River 10 21 69 0 12 97
Pinconning 32 68 O 0 20 25
Coggins 7O 9 21 0 27 25
Linwood 20 1 5 60 5 18 61
Killarney 89 1 10 0 21 35
Essexville 96 2 2 0 7 20
Quanicassee 20 62 18 0 38 30
Oakhurst 45 1 54 0 44 61
Thomas 88 2 20 0 23 35
Fish Point 36 64 0 0 24 35
Hog Island 34 34 32 0 41 33
Bayport 1 5 40 45 0 18 74
Sand Point 69 9 22 0 37 30

u

29

 

 

fi

Habitat complexity values ranged from 7 (Essexville) to 41 (Hog Island). The
maximum water depth along a transect ranged from 15 cm at Au Gres and

WigWam to 97 cm at Pine River.

Relating Fish Community Characteristics to Wetland Habitat

The abundance of YOY ﬁsh was signiﬁcantly greater (R2 = 0.677, F140 =
20.924, p = 0.001) at sites that had one or more structures such as sand bars,
gravel bars, or islands that may have provided the wetland with shelter from
wave action, compared to sites that lacked such a structure (Figure 1.6). The
average catch of fish in wetlands without a structure was 141 1 36 ﬁsh, while the
average catch of ﬁsh in wetlands with one or more of these structures was 545 i
106 ﬁsh.

Species richness within the twelve wetlands ranged from 6 to 22 species.
Results of the ANCOVA (R2 = 0.854) indicate that there are signiﬁcant positive
relationships between the presence of one or more islands, sand bars, or gravel
bars and species richness (Fm = 21.826, p = 0.002), and percent cover of
submersed vegetation (F13 = 17.335, p = 0.003). A negative relationship was
found between species richness and distance from the wetland to the Saginaw
River (F13 = 8.247, p = 0.021 ).

The Shannon-Weiner species diversity index (H') was used to calculate
the diversity of the YOY ﬁsh community sampled within each of the twelve
wetlands. Species diversity ranged from 0.240 at Pinconning to 1.424 at Pine

River. A positive relationship was found between species diversity and maximum

30

 

800 i 7

700 - T —
600 - -
500 - —
400 — -

 

 

 

 

 

300 - T -
200 - -
100 - -

:5

O l 1
Absent Present

Presence of Structure

 

Total YOY Fish Catch

 

 

 

 

 

 

 

Figure 1.6: Abundance of YOY ﬁsh captured in wetlands without a structure

such as a sand bar, gravel bar, or island, and wetlands with such a structure.

31

water depth along a transect (F13 = 64.385, p < 0.001 ), while a negative
relationship was found between species diversity and the amount of coverage of
open water (F19 = 14.617, p = 0.003). Overall model performance resulted in a
R2 = 0.952.

Species evenness for the twelve wetland areas ranged from 0.149 at
Pinconning to 0.682 at Thomas. Results of the ANCOVA (R2 = 0.959) indicate
that there are signiﬁcant positive relationships between species evenness and
maximum water depth along a transect (F13 = 52.118, p < 0.001) and percent

cover of open water (F13 = 32.904, p < 0.001).

Relating Fish Community Composition to Wetland Habitat

The ﬁrst two PCoA axes explained 79.7% of the variance in the species
presence/absence data, with the ﬁrst axis explaining 48.3% of the variance, and
the second axis explaining 31.4% of the variance. Visual inspection of the PCoA
plot (Figure 1.7) indicates that Coggins (C) and Quanicassee (Q) had very similar
species composition, with both wetlands having similar scores for axes one and
two. Sand Point (8) and Thomas (T) also had similar scores for the ﬁrst two
axes. Other wetlands such as Pine River (R) and Fish Point (F) had similar
scores for axis 1, but not for axis 2. Similarly, Linwood (L) and Coggins (C) had
similar scores for axis 2, but very different scores along axis 1.

Multiple linear regression analysis was used to relate the wetland scores

along axes one and two to wetland habitat characteristics. PCoA scores along

32

 

 

 

 

4 l l l l l l l i
F
3 _ _
2 — T —
N 1__ S _
(I)
.30 (30
8 K
o. '1‘ O —
-2 _ _
-3 _ P R _
_4 l l l l l I I l
-4 -3 -2 -1 0 1 2 3 4 5
PCoA axis 1

Figure 1.7: Graph of wetland PCoA scores for the ﬁrst two axes. Symbols
represent the ﬁrst letter in the name of each wetland area sampled, except for

the Pine River wetland which is represented by the letter R.

33

 

 

 

 

 

wetlands t;

wetlands

(F16 = 8.87.
located (F11:
12.625, p =
7.927, p = 0
0' more isla:
Shore of Sag
and that We:
that Were ”it
largemouth l

pumpkins ee(

axis 1 were related to characteristics of the wetland habitat (R2 = 0.737). There
were signiﬁcant relationships between PCoA axis 1 scores and the presence of
one or more islands, sand bars, or gravel bars (F13 = 17.447, p = 0.002), and
habitat complexity (F 1,9 = 5.188, p = 0.048). PCoA axis 1 scores were higher in
wetlands that had one or more islands, sand bars, or gravel bars, and in
wetlands with low habitat complexity. Species that were more common in
wetlands with high scores along PCoA axis 1 included: bluegill, green sunﬁsh,
pumpkinseed, yellow perch, brown bullhead, and spotﬁn shiner.

Habitat characteristics of the wetlands sampled were also able to explain
much of the variation in scores for PCoA axis 2 (R2 = 0.871 ). Scores for this axis
were related to the presence of one or more islands, sand bars, or gravel bars
(F15 = 8.879, p = 0.025), the shore of Saginaw Bay where the wetland was
located (F1;3 = 18.697, p = 0.005), the amount of cover of open water (F15 =
12.625, p = 0.012), and the distance from that wetland to the nearest river (Fm =
7.927, p = 0.031 ). In general, wetlands with high scores for PCoA axis 2 had one
or more islands, sand bars, or gravel bars, and were located on the eastern
shore of Saginaw Bay. Additionally, wetlands with large amounts of open water,
and that were located near rivers had higher scores for PCoA axis 2. Species
that were more common in wetlands with low scores along PCoA axis 2 included:
largemouth bath, common shiner, brown bullhead, white sucker, and

pumpkinseed.

34

 

 

 

h...——-—.———_.._._~_,__._.,.__M. _ _

Relating l

 

 

MM

Ycl
the three- ‘
perch in tl"I
cm/week a
growth rate

A 51

was found

 

the amount
perch grew
There was I
0.927) bet“
wetlands ar

(Figure 1.8)

S Ottail Shi

YOY
the three-m
the Week prl
Baypon t0 0

Relating Fish Growth to Wetland Habitat
Yellow Perch

YOY yellow perch were collected in seven different wetland areas during
the three-week sampling period in 2001. Mean individual growth rates of yellow
perch in the week preceding capture in these wetlands ranged from 0.45
cm/week at Thomas to 0.81 cm/week at Linwood (Table 1.6). Mean weekly
growth rate over all seven wetlands was 0.59 i 0.05 cm/week.

A signiﬁcant positive relationship (R2 = 0.792, F15 = 18.982, p = 0.007)
was found between mean growth rates of yellow perch in the seven wetlands and
the amount of submersed vegetation cover (Figure 1.8). In general, YOY yellow
perch grew better in areas that had higher amounts of submersed vegetation.
There was not a statistically significant relationship (R2 = 0.002, F15 = 0.009, p =
0.927) between mean weekly growth rate of YOY yellow perch within the seven
wetlands and mean standard length of YOY yellow perch from those wetlands

(Figure 1.8).

Spottail Shiner

YOY spottail shiners were collected in six different wetland areas during
the three-week sampling period in 2001. Mean growth rates of spottail shiners in
the week preceding capture in these wetlands ranged from 0.32 cm/week at
Bayport to 0.50 cm/week at Fish Point (Table 1.6). Mean weekly growth rate

over all six wetlands was 0.43 :l: 0.03 cm/week.

35

 

 

Table 1.6
shiner, be

the twelve

Wetland

 

Pine River

Pinconnin

 

Coggins
Linwood
Killamey
Quanicass
OakhUrst
Thomas
Fish pOint
H09 Island
B“Wort

Sand pOinl

Table 1.6: Mean growth rate (cm/week) of young of year yellow perch, spottail

shiner, banded killiﬁsh, and bluegill over the week preceding capture, in each of

the twelve wetland areas where ﬁsh were captured during 2001.

 

 

Yellow Spottail Banded ‘
Wetland Bluegill

Perch Shiner Killiﬁsh
Pine River 0.74 1 0.11 0.38 1 0.02 0.55 1 0.03 0.58 1 0.03
Pinconning - - 0.50 1 0.02 -
Coggins - - 0.49 1 0.03 -
Linwood 0.81 1 0.03 0.41 1 0.02 0.50 1 0.02 0.59 1 0.01
Killamey - - - 0.25 1 0.03
Quanicassee - - - _
Oakhurst 0.64 1 0.04 0.47 1 0.02 0.55 1 0.05 -
Thomas 0.45 1 0.04 - 0.47 1 0.03 0.35 1 0.01
Fish Point 0.46 1 0.04 0.50 1 0.07 0.42 1 0.07 0.27 1 0.02
Hog Island 0.50 1 0.02 - 0.44 1 0.02 0.24 1 0.01
Bayport 0.57 1 0.02 0.32 1 0.01 0.51 1 0.02 0.67 1 0.02
Sand Point - 0.48 1 0.05 0.48 1 0.03 0.30 1 0.02

 

36

A. YOY yellow perch growth vs. "/0 submersed vegetation

 

 

0.9 l l l l l l

0.8 — '- -
3:; R
E 0.7— -l
3
g 0
9 0.6— -
(D B

0.5 - H ‘

F T
l l l l 1 l

 

 

0.4
010 20 30 40 50 60 70
% Submersed Vegetation

B. YOY yellow perch growth vs. standard length of perch

 

 

 

0-9 l l l r

0.8— l— —
3‘; R
E 0.7— —
3
z o
g 0.6— ‘l
5 B

0.5— H -

F T
0.4 l l l l

 

1 2 3 4 5 6
Mean Fish Length (cm)

Figure 1.8: Relationship between A) mean growth rates (cm/week) of YOY
yellow perch and amount of submersed vegetation cover and B) mean growth
rates (cm/week) of perch and mean standard length (cm) of perch. Symbols
represent the ﬁrst letter in the name of each wetland area sampled, except for

the Pine River wetland which is represented by the letter R.

37

ANCOVA was used to relate spottail shiner growth rate to wetland habitat
characteristics, and was able to explain a large proportion of the variation in
mean growth rates (R2 = 0.976). A positive relationship was found between
mean growth rates and habitat complexity (FM = 63.612, p = 0.004), while a
signiﬁcant negative relationship was found between growth rate and maximum
water depth sampled along a transect (FM = 20.654, p = 0.020). There was not
a statistically signiﬁcant relationship between mean weekly growth rate of YOY
spottail shiners within the six wetlands and mean standard length of YOY spottail

shiners from those wetlands (R2 = 0.566, F14 = 5.223, p = 0.084).

Banded Killiﬁsh

Young of the year banded killiﬁsh were collected in ten of the twelve
wetland areas during the three-week sampling period in 2001. Mean growth
rates of banded killifish in the week preceding capture in these wetlands ranged
from 0.42 cm/week at Fish Point to 0.55 cm/week at Oakhurst and Pine River
(Table 1.6). Mean growth rate of YOY banded killiﬁsh over all ten wetlands was
0.49 1 0.01 cm/week.

ANCOVA was used to relate banded killiﬁsh growth to habitat
characteristics (R2 = 0.785). Results of the multiple linear regression indicate
that there is a signiﬁcant negative relationship between killiﬁsh growth rate and
the presence of one or more islands, sand bars, or gravel bars (F1; = 9.726, p =

0.017), and a signiﬁcant positive relationship with percent cover of submersed

38

 

 

 

 

vegetatio
areas tha

abundant

 

relationsh
of YOY be

0.137).

Bluegill
The

analyzed fc

 

wetland rar
(T able 1.6).

A sig
Was found t

areas Of 061
00421 F16 I
blueGill in th

bluegill from

vegetation (F1; = 17.591, p = 0.004). In general, banded killiﬁsh grew better in
areas that lacked islands, sand bars, or gravel bars, but where there was
abundant submersed vegetation cover. There was not a statistically significant
relationship between mean banded killiﬁsh growth rate and mean standard length
of YOY banded killiﬁsh within the ten wetland areas (R2 = 0.254, F13 = 2.726, p =

0.137).

Bluegill

The growth rates of YOY bluegill in the week preceding their capture was
analyzed for eight wetland areas. Mean weekly bluegill growth rates within a
wetland ranged from 0.24 cm/week at Hog Island to 0.67 cm/week at Bayport
(Table 1.6). Overall mean bluegill growth rate was 0.41 1 0.06 cm/week.

A signiﬁcant positive relationship (R2 = 0.756, F1; = 18.640, p = 0.005)
was found between mean growth rates of bluegill in the eight wetlands and the
maximum water depth measured along a transect within those wetlands (Figure
1.9). In general, YOY bluegill grew faster in wetland areas that had at least some
areas of deeper water. There was not a statistically signiﬁcant relationship (R2 =
0.042, F15 = 0.260, p = 0.628) between the mean weekly growth rate of YOY
bluegill in the different wetland areas and the mean standard length of YOY

bluegill from those wetlands (Figure 1.9).

39

A. Bluegill growth vs. water depth

 

0-7 r T i l i i l

.o .0
0'! O)
l l

r.
l 1

Growth (cm/wk)
.0
h
l
l

T

S '—1
F
PK

1 l l L l l l

0.2
20 30 40 50 60 70 80 90 100
Water Depth (cm)

9
on
i

 

 

 

 

B. Bluegill growth vs. st8n7dard length of bluegill

i l
0.6 — |_ —
0.5 - -

0.4 — -

Growth (cm/wk)

0.3

l
U)
l

 

 

 

0.2 l 1
1.5 2.0 2.5 3.0

Mean Fish Length (cm)

Figure 1.9: Relationship between A) mean growth rate of YOY bluegill (cm/wk)
and maximum water depth (cm) measured along a transect and B) mean growth
rate of YOY bluegill (cm/wk) and mean standard length of bluegill (cm) in eight
wetlands. Symbols represent the ﬁrst letter in the name of each wetland area

sampled, except for the Pine River wetland which is represented by the letter R.

40

DISCUSSION

Young of the Year Fish Community in Saginaw Bay, Lake Huron

This study sampled the YOY ﬁsh community in 2000 and 2001 in ﬁfteen
different wetlands areas around Saginaw Bay, Lake Huron. The young of year
ﬁsh community in these wetlands was numerically dominated by eight ﬁsh
species: banded killiﬁsh, quillback, common carp, spottail shiner, bluegill, green
sunﬁsh, goldfish, and yellow perch. The dominant ﬁsh species found within
Saginaw Bay are similar to those reported for other wetlands in Lake Huron
(Leslie and Timmins 1994) and other coastal wetlands throughout the Great
Lakes (e.g., Chubb and Liston 1986, Jude and Pappas 1992, Brazner 1997).

Overall, YOY ﬁsh from 36 species were captured in the coastal wetlands
of Saginaw Bay over 2000 and 2001. This value is within the range of species
richness (18-38 species) reported from other studies of coastal wetlands in Lake
Huron (e.g., Loftus 1982, Minns et al. 1994, Leslie and Timmins 1997). The
number of species captured within a single wetland during this study ranged from
2 to 18 species. This value is low compared to the range (15-41 species)
reported in other studies of Great Lakes coastal wetlands, many of which
examined both the adult and juvenile ﬁsh community (e.g., Chubb and Liston
1986, Poe et al. 1986, Randall et al. 1996, Brazner 1997, Johnson et al. 1997).

Although a total of 36 ﬁsh species was captured within the coastal

wetlands in Saginaw Bay, twelve of the ﬁsh species captured were found in only

41

one wetland area. An abundance of rare species was also documented in a
study of coastal wetland areas in Severn Sound, Lake Huron (Leslie and
Timmins 1994). Fish species considered to be uncommon in their study
outnumbered common species by more than three to one. However the authors
noted that it was likely that these rare species were not sampled in proportion to
their true abundance, which is probable in this study of YOY ﬁsh in Saginaw Bay

as well.

Inﬂuence of Habitat Characteristics on the YOY Fish Community

A wide range of YOY ﬁsh abundances, species richness, diversity,
community composition, and ﬁsh growth rates were observed among the
different wetlands sampled in Saginaw Bay. These results indicate that all
wetland environments are not equal in terms of their value for YOY ﬁsh. More
ﬁsh, greater numbers of species, different species, and faster growing ﬁsh were
collected from certain wetlands compared to others. Table 1.7 provides a
summary of the statistically signiﬁcant relationships between YOY ﬁsh
community metrics and measured wetland habitat characteristics included in the
ANCOVA models.

In general, characteristics of the macrophyte community within the
wetlands were important in determining YOY ﬁsh use of wetland areas, with
percent cover of open water or submersed vegetation having a statistically

signiﬁcant relationship to 60% of the metrics of the YOY ﬁsh community

42

 

___mo:_m
cmEEx bmucmm
..ocEm =3“on

cocoa 26=o>

 

Ecummw

I._.>>Omo
+ + + N £2
+ F m_x<
coEmanoo bisEEoo
mmocco>m
33520 $6on
1 + mmocgoi meooaw
+ oocmuczn< 5E
>mm Loam 52m coamuomo> 93025
.mmw .m . b_xo_anO .
o 905 265 mm 6932 82355 ho
a. 2 837.5 9 8:990 cocomoi

 

.maEmcozﬂo. gnome:

E8536 >__mo:m_§m E8059 1 cam 59:80:39 $358 E8536 2.833% “commie. + 2oz; .wozmtouoﬁmco

“Sim; new .5265 cm: .mozmtoaomcmco bEsEEoo cm: >O> comics 3320339 95865 2an meEsm K... 023%

43

measured. In addition to macrophyte cover, habitat characteristics relating to
hydrology and geomorphology (presence of one or more islands, sand bars, or
gravel bars, and maximum water depth measured along sample transects), had

large inﬂuences on the YOY ﬁsh community.

ﬁmatic Veqetation

Macrophytes are a key component of the habitat of Great Lakes coastal
wetlands, providing young of the year ﬁsh with refuge from predation by serving
as protective cover that reduces the visibility of YOY ﬁsh and the maneuverability
of adult piscivores (Savino and Stein 1982, 1989, Werner et al. 1983, Mittelbach
1986, Gotceitas and Colgan 1987, Werner and Hall 1988, Gotceitas 1990a,b,
Heck and Crowder 1991). Macrophytes also serve as important habitat for
macroinvertebrate and zooplankton food resources (Cardinale et al. 1998). The
amount of submersed vegetation and open water had impacts on 60% of the
YOY community metrics examined in this study.

There was a signiﬁcant relationship between the percent cover of
submersed vegetation and species richness in wetlands in this study. YOY ﬁsh
from more species were collected in wetland areas that had high amounts of
submersed vegetation cover. Species of YOY ﬁsh that were more commonly
present in wetlands with high amounts of submersed vegetation cover included:
largemouth bass, green sunﬁsh, pumpkinseed, bowﬁn, Iongnose gar, and brown
bullhead. Several other studies have documented increases in species richness

or diversity in wetlands with increased vegetation cover (Keast et al. 1978, Bryan

44

and Scamecchia 1992, Leslie and Timmins 1994, Minns et al. 1994, Randall et
al. 1996, Pierce et al. 2001).

Extremely high plant cover (70-80%), has been shown to result in a
decrease in ﬁsh species richness (Crowder and Cooper 1982, Wiley et al. 1984,
Killgore et al. 1989, Lillie and Budd 1992, Dibble et al. 1996). No decline in
species richness with increased vegetation cover was found in this study.
Previous studies on the detrimental impacts of high vegetation cover have
examined effects of plant cover on the entire ﬁsh community, whereas this study
focused only on YOY ﬁsh use of coastal wetlands. YOY ﬁsh are extremely
vulnerable to predators, and are more likely to remain associated with vegetation
because of the shelter they provide, regardless of potential decreases in foraging
efﬁciency. Larger ﬁsh face less of a predation risk, and therefore may move from
vegetated areas when macrophytes become too abundant.

Leslie and Timmins (1994) documented a higher diversity of YOY ﬁsh in
areas of Severn Sound, Lake Huron that had submersed vegetation compared to
areas located on exposed shore and dominated by open water habitat. Similarly,
this study found a signiﬁcant negative relationship between species diversity and
the amount of open water within a wetland. In general, wetlands with more open
water areas, or less vegetation, had lower diversities of YOY ﬁsh.

The structure that aquatic macrophytes provide has been shown to have a
dramatic inﬂuence on the abundance, distribution, and species composition of
ﬁsh in the littoral zone (Werner et al. 1977, Keast et al. 1978, Poe et al. 1986,

Heck and Crowder 1991, Leslie and Timmins 1994, Dibble et al. 1996, Randall et

45

al. 1996, Miranda and Pugh 1997, Weaver et al. 1997, Pierce et al. 2001).
These studies have demonstrated a general increase in ﬁsh abundance with
increasing vegetation cover. This study did not ﬁnd a relationship between
vegetation cover and ﬁsh abundance; however, a signiﬁcant relationship was
found between the amount of open water and species composition. Brown
bullhead, bowﬁn, white suckers and emerald shiners were more common in
wetlands with low amounts of open water.

Positive relationships were found between the amount of submersed
vegetation and the growth of yellow perch and banded killiﬁsh. These faster
mean weekly growth rates in wetlands with abundant submersed macrophytes
are likely a result of the fact that submersed plants serve as habitat for abundant

macroinvertebrate and zooplankton food resources (Cardinale et al. 1998).

Presence of Sand Bars, Gravel Bars or Islands

The presence of one or more habitat structures such as sand bars, gravel
bars or islands had a large impact on the YOY ﬁsh community, with this
characteristic appearing in 50% of the models constructed. The presence of one
or more of these structures was positively correlated to ﬁsh abundance and
species richness. More ﬁsh were collected, and more species were collected in
wetlands that had one or more of these structures than in wetlands that lacked
these structures. Additionally, this habitat characteristic impacted ﬁsh community
composition. Yellow perch, brown bullhead, spotﬁn shiners, green sunﬁsh, and

pumpkinseed were all more common in wetlands with one of these structures

46

than wetlands without. Finally, the growth of YOY banded killiﬁsh was slower in
wetlands with one or more of these structures.

Although this study did not measure wave or wind energy within the
different wetland areas, I hypothesize that sand bars, gravel bars and islands
serve to shelter the wetland areas from wave action. The dissipation of wind
energy through the presence of vegetation has been shown to result in changes
in water quality, turbidity, phytoplankton, epiphytic algae, zooplankton,
macroinvertebrates, and ﬁsh (Suzuki et al. 1995, Randall et al. 1996, Cardinale
et al. 1997, Cardinale et al. 1998). Several studies have documented decreases
in the cover of aquatic vegetation with increasing wave action. There was not a
statistically signiﬁcant decrease in the amount of open water in wetlands that had
one of these structures (F133 = 2.446, a: 0.142), however mean percent open
water in areas that lacked one of these structures was 58% while mean percent
open water in areas that had one of these structures was 33%.

Additionally, exposure to wave action has been shown to inﬂuence the
littoral ﬁsh community. This study documented large decreases in ﬁsh
abundance in wetlands that lacked a sand bar, gravel bar or island. In a study
comparing ﬁsh production in wetland areas in Lakes Ontario and Huron, Randall
et al. (1996) found that exposure to wind resulted in lower abundances of ﬁsh.
Randall et al. (1996) did not compare species richness or composition among the
wetland areas.

There was a considerable range in the number of islands, sand bars or

gravel bars located adjacent to the wetlands sampled, where only one structure

47‘

was present at WigWam and Linwood, while 6 structures were present at Hog
Island. There was also a large difference in the total length of these structures
among wetland areas, ranging from 253 m to 2,173 m. The differences in the
number and size of these structures likely affects the impact they have on wind
and wave energy, with single larger structures providing the highest degree of
shelter from wave and wind energy. Multiple smaller structures may provide
some protection from wave energy while allowing for increased mixing and

interactions with offshore areas of Saginaw Bay.

Water Depth

The maximum water depth identiﬁed along sampling transects also had a
large impact on the YOY fish community, being present in 40% of the models
developed. Species that were more commonly present in wetlands with greater
maximum water depths included: brown bullhead, bowﬁn, Iongnose gar, and
common shiner. Wetlands that had areas with deeper water levels (>60 cm) had
higher species diversity and species evenness. The areas of deeper water likely
provided the YOY ﬁsh with refuges of cooler water temperatures and higher
dissolved oxygen levels during warm summer days, resulting in more diverse ﬁsh
communities.

Maximum water depth was correlated to the growth of YOY ﬁsh of two of
the four species examined; bluegill grew faster in wetlands where there were
areas of deeper water, whereas spottail shiners grew faster in wetlands where

maximum water depths remained quite shallow. According to Leslie and

48

Timmins (1994), bluegill can be classiﬁed as phytophillic species, spending most
of their lives in close association with aquatic vegetation, whereas spottail
shiners are psammophils preferring sandy habitats. Because of their phytophillic
nature, bluegill are more likely to require areas of deeper water in close
association to the vegetation. They may have a higher dependence on nearby
deeper water refuges in times of low oxygen and extreme water temperatures.
Spottail shiners, on the other hand, may be more likely to temporarily leave
vegetated areas in search of better habitat conditions.

Water levels within the Great Lakes are cyclical in nature (Geis 1979,
Boutin 2000), and water levels within Saginaw Bay during the study period
averaged approximately 0.5 meters below the long term average. The
importance of having areas of deeper water is likely exaggerated during times of
low water levels. It is possible that during times of normal or high water levels,
habitat factors other than water depth would increase in importance in their
affects on the YOY ﬁsh community.

A review of the literature revealed few studies that have examined the role
of water depth in inﬂuencing the YOY ﬁsh community; with the exception of Keast
and Harker (1977) who found a steady decrease in ﬁsh abundance from an
average depth of 0.8 to 8 meters. Although not directly related to maximum
water depth, Benson and Magnuson (1992) found a change in species
composition with increasing depth gradient.

A few studies have compared ﬁsh abundance and diversity in relation to

water depth at larger spatial scales. For example, Chubb and Liston (1986)

49

found highest densities of larval ﬁsh in shallow (<1 meter) vegetated areas of
Pentwater Marsh, Lake Michigan. Similarly, Keast et al. (1978) found that
vegetated areas < 2 meters deep had more diverse and abundant ﬁsh
communities compared to deeper areas of submersed vegetation, and adjacent
open water areas In Lake Opinicon. Finally, Bryan and Scamecchia (1992)
documented a higher species richness and abundance of larval and juvenile ﬁsh
in wetlands areas with a mean water depth less than 1 meter compared to areas

with deeper water levels in Spirit Lake, Iowa.

Habitat Complexity

Several studies have documented increases in the species richness and
diversity of littoral ﬁsh communities with increasing habitat complexity or habitat
heterogeneity (e.g., Benson and Magnuson 1992, Tom and Magnuson 1982,
Eadie and Keast 1984, Leslie and Timmins 1994, Jenkins and Sutherland 1997,
Weaver et al. 1997). This study did not ﬁnd a relationship between habitat
complexity and species richness or species diversity.

Habitat complexity did have an inﬂuence on species composition. Brown
bullhead, bowﬁn, and white suckers were more common in wetlands with low
habitat heterogeneity, while gizzard shad and sand shiners were more common
in wetlands with high habitat heterogeneity. Weaver et al. (1997) also found
vegetation patchiness to be important in determining species assemblages, with
some species increasing in abundance with vegetation patchiness, and others

decreasing in abundance.

50

Habitat complexity was positively related to the growth of spottail shiners.
The faster growth rates with increased heterogeneity may be a result of the fact
that patchiness within vegetation provides prey ﬁshes with a habitat in which
shelter is in close proximity to open spaces that harbor rich zooplankton food
resources. This reduces the amount of time and energy these ﬁsh need to spend
moving from areas that provide them with protection from predators to feeding

areas.

Importance of Regional Factors

In examining patterns of ﬁsh community assemblages in coastal wetland
and beach habitat in Green Bay, Lake Michigan, Brazner and Beals (1997) found
that ﬁsh abundance, species richness, and community composition were greatly
inﬂuenced by geographical region within the Bay. They found distinct species
assemblages in wetlands from the lower, middle and upper portions of the Bay,
with regional factors being more important than surrounding development
pressure, habitat related factors (eg turbidity, conductivity, pH, macrophyte
cover), predation, or competition in determining the distribution of ﬁsh species
(Brazner and Beals 1997, Brazner 1997). Regional patterns were also seen in
the grouping of ﬁsh species utilizing coastal wetlands into assemblages in the
Canadian waters of the Great Lakes (Kelso and Minns 1996).

This study examined the impact of several regional factors (e.g., distance
from Saginaw River, distance to the outer bay, eastern or western shore of

Saginaw Bay) on the YOY ﬁsh community. Contrary to the results of Brazner

51

(1997) and Brazner and Beals (1997), the regional factors examined in this study
did not have signiﬁcant impacts on YOY ﬁsh abundance. Additionally, no
signiﬁcant relationships were found between these regional factors and species
diversity, species evenness, or ﬁsh growth. However, there was a signiﬁcant
relationship between species richness and distance to Saginaw River. Species
richness was higher in wetland areas located closer to Saginaw River, and
species richness decreased with increasing distance from Saginaw River.
Regional factors also appeared to have an impact on species composition, with
different ﬁsh communities associated with the eastern compared to the western
shore of Saginaw Bay. Gizzard shad, green sunﬁsh, sand shiners, and yellow
perch were more common in wetlands on the eastern bank, while white suckers,
rock bass, and bowﬁn were more common in wetlands on the western bank of

Saginaw Bay.

Future Research Needs

The presence of a structure such as an island, sand or gravel bar within a
wetland area had a large impact on YOY ﬁsh abundance, species richness, and
community composition. These structures may be important because they help
to shelter the wetland and the YOY ﬁsh community from damaging wave and
wind energy. However, more research is needed in determining why these
structures had such a positive effect on the YOY ﬁsh community, if they have a

signiﬁcant effect on wind and wave energy, and if they may help deter larger

52

predatory ﬁsh from entering the wetland environments, and thereby improve the
value of the wetland for YOY ﬁsh.

There was a high degree of similarity in the types of species that
dominated the YOY ﬁsh community in the wetlands sampled in Saginaw Bay, as
compared to the YOY ﬁsh community sampled in other wetlands throughout the
Great Lakes. Additionally, species richness in Saginaw Bay was within the range
of species richness identiﬁed for other Great Lakes’ coastal wetlands areas. As
a result of this similarity in species composition and richness, it appears that the
YOY ﬁsh community within Saginaw Bay can be considered representative of a
typical YOY ﬁsh community of Great Lakes wetlands. Further research is
necessary to determine if the patterns identiﬁed in this study between habitat
characteristics and the YOY ﬁsh community can be applied to other wetland
areas within the Great Lakes.

Many of the coastal wetland habitats along the Great Lakes are under
pressure from development, resulting in degraded water and habitat quality. In
order to understand how alterations in coastal wetland systems may inﬂuence
young of the year recruitment, there is a need to ﬁrst identify habitat
characteristics that may inﬂuence YOY community attributes, species
composition, and growth rates of YOY ﬁsh. This study identiﬁed several habitat
characteristics that can inﬂuence the YOY ﬁsh community. However, further
research is necessary to determine how these relationships may differ as a result
of changes in surrounding land use, development pressure, and human activity in

the vicinity of these wetland environments.

53

Finally, the success of YOY ﬁsh within wetland environments is inﬂuenced
by characteristics of the wetland as a whole, such as those investigated in this
study. However, the distribution of YOY ﬁsh is not uniform within a wetland, and
the success of YOY ﬁsh is likely also impacted by within wetland characteristics,
such as the types of vegetation habitat available, wetland productivity, and water
quality. Future research is needed on how these characteristics may inﬂuence

YOY ﬁsh recruitment.

54

LITERATURE CITED

Baltz, D.M., C. Rakocinski, and J.W. Fleeger. 1993. Microhabitat use by marsh-
edge ﬁshes in a Louisiana estuary. Environmental Biology of Fishes
36:109-126.

Beeton, A., S. Smith and F. Hooper. 1967. Physical limnology of Saginaw Bay,
Lake Huron. Technical Report No. 12. Great Lakes Fishery Commission,
Ann Arbor, MI.

Benson, B.J. and J.J. Magnuson. 1992. Spatial heterogeneity of littoral ﬁsh
assemblages in lakes: relation to species diversity and habitat structure.
Canadian Journal of Fisheries and Aquatic Sciences 49:1493-1500.

Boutin, C. 2000. Great Lakes Water Levels. NOAA Great Lakes Environmental
Research Laboratory, Ann Arbor, Michigan.

Brandt, S.B., J.J. Magnuston, and LB. Crowder. 1980. Thermal habitat
partitioning by ﬁshes in Lake Michigan. Canadian Journal of Fisheries and
Aquatic Sciences 37: 1 557-1 564.

Brazner, JG. 1997. Regional, habitat and human development inﬂuences on
coastal wetland and beach ﬁsh assemblages in Green Bay, Lake
Michigan. Journal of Great Lakes Research 23:36-51.

Brazner, J.C. and E.W. Beals. 1997. Patterns in ﬁsh assemblages from coastal
wetland and beach habitats in Green Bay, Lake Michigan: a multivariate
analysis of abiotic and biotic forcing factors. Canadian Journal of
Fisheries and Aquatic Sciences 54:1743-1761.

Brazner, J.C., D.K. Tanner, D.A. Jensen and A. Lemke. 1998. Relative
abundance and distribution of ruffe (Gymnocephalus cemuus) in a Lake
Superior Coastal Wetland Fish Assemblage. Journal of Great Lakes
Research 24:293-303.

Breder, C.M. Jr. 1960. Design for a fry trap. Zoologica 45:155-160.
Bryan, MD. and D.L. Scamecchia. 1992. Species richness, composition, and
abundance of ﬁsh larvae and juveniles inhabiting natural and developed

shorelines of a glacial Iowa lake. Environmental Biology of Fishes 35:329-
341.

55

Cardinale, B.J., T.M. Burton and V.J. Brady. 1997. The community dynamics of
epiphytic midge larvae across the pelagic-littoral interface: do animals
respond to changes in the abiotic environment. Canadian Journal of
Fisheries and Aquatic Science 54:1-9.

Cardinale, B.J., V.J. Brady and TM. Burton. 1998. Changes in the abundance
and diversity of coastal wetland fauna from the open water/macrophyte
edge towards shore. Wetlands Ecology and Management 6:59-68.

Chubb, S.L. and CR. Liston. 1986. Density and distribution of larval ﬁshes in
Pentwater Marsh, a coastal wetland on Lake Michigan. Journal of Great
Lakes Research 12:332-343.

Comer, A.J., D.A. Albert, H.A. Wells, B.L. Hart, S.B. Raab, D.L. Price, D.M.
Kashian, R.A. Corner, and D.W. Shuen. 1995. Michigan’s Native
Landscape, as Interpreted from General Land Ofﬁce Surveys, 1816-1856.
Michigan Natural Features Inventory Report to Water Division, US. EPA
and Wildlife Division, Michigan Department of Natural Resources,
Lansing, Michigan.

Crowder, LB. and WE. Cooper. 1982. Habitat structural complexity and the
interaction between bluegills and their prey. Ecology 63:1802-1813.

Dewey, MR. 1992. Effectiveness of a drop net, a pop net, and an electroﬁshing
frame for collecting quantitative samples of juvenile ﬁshes in vegetation.
North American Journal of Fisheries Management 12:808-813.

Dibble, E.D., K.J. Killgore, and S.L. Harrel. 1996. Assessment of ﬁsh-plant
interactions. American Fisheries Society Symposium 16:357-372.

Duffy, KC. and D.M. Baltz. 1998. Comparison of ﬁsh assemblages associated
with native and exotic submerged macrophytes in the Lake Pontchartrain
estuary, USA. Journal of Experimental Marine Biology and Ecology
223:199-221 .

Eadie, J. McA. and A. Keast. 1984. Resource heterogeneity and ﬁsh species
diversity in lakes. Canadian Journal of Zoology 62:1689-1695.

Francis, R.I.C. 1990. Back-calculation of ﬁsh length: a critical review. Journal of
Fish Biology 36:883-902.

Freeman, B.J., H.S. Greening, and JD. Oliver. 1984. Comparison of three
methods for sampling fishes and macroinvertebrates in a vegetated
freshwater wetland. Journal of Freshwater Ecology 2:603-609.

56

Geis, J.W. 1979. Shoreline processes affecting the distribution of wetland
habitats. Transactions of the North American Wildlife Natural Resources
Conference 44:529-542.

Gotceitas, V. and P. Colgan. 1987. Selection between densities of artiﬁcial
vegetation by young bluegills avoiding predation. Transactions of the
American Fisheries Society 116240-49.

Gotceitas, V. 1990a. Plant stem density as a cue in patch choice by foraging
juvenile bluegill sunﬁsh. Environmental Biology of Fishes 29:227-232.

Gotceitas, V. 1990b. Foraging and predator avoidance: a test of a patch choice
model with juvenile bluegill sunﬁsh. Oecologia 83:346-351.

Hayes, D.B. C.P. Fererri, W.W. Taylor. 1996. Linking ﬁsh habitat to their
population dynamics. Canadian Journal of Fisheries and Aquatic
Sciences 53:383-390.

Heck, KL, and LB. Crowder. 1991. Habitat structure and predator-prey
interactions in vegetated aquatic systems. pp282-299 IN S.S. Bell, E.D.
McCoy and HR. Mushinsky (eds.). Habitat structure: the physical
arrangement of objects in space. Chapman and Hall, New York.

Herdendorf, C.E., S.M. Hartley, and MD. Barnes (eds.). 1981. Fish and wildlife
resources of the Great Lakes coastal wetlands within the United States.
Vol.1: Overview. United States Fish and Wildlife Service, Ofﬁce of
Biological Services, Washington, DC, FWS/OBS-81/02-V3.

Houde, ED. 1994. Differences between marine and freshwater ﬁsh larvae:
implications for recruitment. ICES Journal of Marine Science 51:91-97.

Jackson, G. and GK. Jones. 1999. Spatial and temporal variation in nearshore
ﬁsh and macroinvertebrate assemblages from a temperate Australian
estuary over a decade. Marine Ecology Progress Series 182:253-268.

Jaworski E. and ON. Raphael. 1978. Fish, wildlife, and recreational values of
Michigan’s coastal wetlands. Rep. to Lands Resources Programs division,
Michigan Department of Natural Resources, Eastern Michigan University,
Ypsilanti, Michigan.

Jenkins, GP. and CR. Sutherland. 1997. The inﬂuence of habitat structure on
nearshore ﬁsh assemblages in a southern Australian embayment:
colonization and turnover rate of ﬁshes associated with artiﬁcial
macrophytes beds of varying physical structure. Journal of Experimental
Marine Biology and Ecology 218:103-125.

57

Johnson, D.L., W.E. Lynch, and T.W. Morrison. 1997. Fish communities in a
diked Lake Erie wetland and an adjacent undiked area. Weltands
17(1):43-54.

Johnson, D.L. and RA. Stein. 1979. Response of ﬁsh to habitat structure in
standing water. Special Publication NO. 6, North Central Division,
American Fisheries Society, Bethesda, MD.

Jude, DJ. and J. Pappas. 1992. Fish utilization of Great Lakes coastal
wetlands. Journal of Great Lakes Research 18:651-672.

Keast, A. and J. Harker. 1977. Fish distribution and benthic invertebrate
biomass relative to depth in an Ontario lake. Environmental Biology of
Fishes 2(3):235-240.

Keast, A., J. Harker, and D. Tumbull. 1978. Nearshore ﬁsh habitat utilization
and species associations in Lake Opinicon (Ontario, Canada).
Environmental Biology of Fishes 3(2):173-184.

Keast, A. 1984. The introduced aquatic macrophyte, Myriophyllum spicatum, as
habitat for ﬁsh and their invertebrate prey. Canadian Journal of Zoology
62:1289-1303.

Kelso, J.R.M. and OK. Minns. 1996. Is ﬁsh species richness at sites in the
Canadian Great Lakes the result of local or regional factors? Canadian
Journal of Fisheries and Aquatic Sciences 53(Suppl. 1):175-193.

Killgore, K.J., R.P. Morgan, and NB. Rybicki. 1989. Distribution and abundance
of ﬁshes associated with submersed aquatic plants in the Potomac River.
North American Journal of Fisheries Management 9:101-111.

Krieger, K.A., D.M. Klarer, R.T. Heath, and CE. Herdendorf. 1992. A call for
research on Great Lakes coastal wetlands. Journal of Great Lakes
Research 18:525-528.

Kushlan, J.A. 1981. Sampling characteristics of enclosure ﬁsh traps.
Transactions of the American Fisheries Society 110:557-562.

Leslie, J.K. and CA Timmins. 1994. Ecology of young-of-the-year ﬁshes in
Severn Sound, Lake Huron. Canadian Journal of Zoology 72:1887-1897.

Lillie, RA. and J. Budd. 1992. Habitat architecture of Myriophyllum spicatum L.

as an index to habitat quality for ﬁsh and macroinvertebrates. Journal of
Freshwater Ecology 7:1 13-125.

58

Liston, CR. and S.L. Chubb. 1985. Relationships of water level ﬂuctuations and
ﬁsh. pp 134-136 IN Prince, H.H. and FM. D’Itri (eds.). Coastal Wetlands.
Proceedings of the First Great Lakes Wetlands Colloquium. Lewis
Publishers, Inc., Chelsea, MI.

Loftus, DH. 1982. Larval ﬁsh sampling in Lake Huron, 1978. Ontario Fisheries
technical Report Series No. 2.

Michaletz, PH. 1998. Population characteristics of gizzard shad in Missouri
reservoirs and their relation to reservoir productivity, mean depth, and
sport ﬁsh growth. North American Journal of Fisheries Management
18:114-123.

Minns, C.K., V.W. Cairns, R.G. Randall and J.E. Moore. 1994. An index of biotic
integrity (lBl) for ﬁsh assemblages in the littoral zone of Great Lakes”
Areas of Concern. Canadian Journal of Fisheries and Aquatic Science
51 :1804-1822.

Miranda, LE. and LL. Pugh. 1997. Relationship between vegetation coverage
and abundance, size, and diet of juvenile largemouth bass during winter.
North American Journal of Fisheries Management 17:601-610.

Mittelbach, G.G. 1986. Predator-mediated habitat use: some consequences for
species interactions. Environmental Biology of Fishes 16:159-169.

Petering, R.W. and D.L. Johnson. 1991. Distribution of ﬁsh larvae among
artiﬁcial vegetation in a diked Lake Erie wetland. Wetlands 11:123-138.

Pierce, C.L., M.D. Sexton, M.E. Pelham, and JG. Larscheid. 2001. Short-terrn
variability and long-term change in the composition of the littoral zone ﬁsh
community in Spirit Lake, Iowa. American Midland Naturalist 1462290-
299.

Pihl, L. and R. Rosenberg. 1992. Production, abundance, and biomass of
mobile epibenthic marine fauna in shallow waters, western Sweden.
Journal of Experimental Marine Biology and Ecology 57:273-301.

Poe, T.P., C.O Hatcher, C.L. Brown, and D.W. Schloesser. 1986. Comparison
of species composition and richness of ﬁsh assemblages in altered and
unaltered littoral habitats. Journal of Freshwater Ecology 3(4):525-536.

Randall, R.G., C.K. Minns, V.W. Cairns, and J.E. Moore. 1996. The relationship
between an index of ﬁsh production and submerged macrophytes and
other habitat features at three littoral areas in the Great Lakes. Canadian
Journal of Fisheries and Aquatic Science 53(Supp1):35-44.

59

Raposa, KB. and CA. Oviatt. 2000. The inﬂuence of contiguous shoreline type,
distance from shore, and vegetation biomass on nekton community
structure in eelgrass beds. Estuaries 23(1):46-55.

Savino, J.F. and RA. Stein. 1982. Predator-prey interaction between largemouth
bass and bluegills as inﬂuenced by simulated, submersed vegetation.
Transactions of the American Fisheries Society 111:255-266.

Savino, J.F. and RA. Stein. 1989. Behavioral interactions between ﬁsh predators
and their prey: Effects of plant density. Animal Behaviour 37:311-321.

Schelske, C. and J. Roth. 1973. Limnological survey of Lakes Michigan,
Superior, Huron, and Erie. Publication No. 17. University of Michigan,
Great Lakes Research Division, Ann Arbor, MI.

Skubinna, J.P., T.G. Coon, and TR. Batterson. 1995. Increased abundance
and depth of submersed macrophytes in response to decreased turbidity
in Saginaw Bay, Lake Huron. Journal of Great Lakes Research
21 (4):476-488.

Stephenson, TD. 1990. Fish reproductive utilization of coastal marshes of Lake
Ontario near Toronto. Journal of Great Lakes Research 16:71-81.

Suzuki, N., S. Endoh, M. Kawashima, Y. Itakura, C.D. McNabb, F.M. mm, and
TR. Batterson. 1995. Discontinuity bar in a wetland of Lake Huron’s
Saginaw Bay. Journal of Freshwater Ecology 10:111-123.

Tonn, WM. and J.J. Magnuson. 1982. Patterns in the species composition and
richness of ﬁsh assemblages in northern Wisconsin lakes. Ecology
63:1149-1166.

Turner, RE. 1988. Secondary production in riparian wetlands. Transactions of
the 53rd North American Wildlife and Natural Resources Conference. 491-
501.

Weaver, M.J., J.J. Magnuson, M.K. Clayton. 1997. Distribution of littoral ﬁshes
in structurally complex macrophytes. Canadian Journal of Fisheries and
Aquatic Science 54:2277-2289.

Werner, E.E., D.J. Hall, D.R. Laughlin, D.J. Wagner, L.A. Wilsmann, and F.C.
Funk. 1977. Habitat partitioning in a freshwater ﬁsh community. Journal
Fisheries Research Board of Canada 34:360-370.

Werner, E.E., G.G. Mittelbach, D.J. Hall, and J.F. Gilliam. 1983. Experimental

tests of optimal habitat use in ﬁsh: the role of relative habitat proﬁtability.
Ecology 64:1525-1539.

60

Werner, E.E. and DJ. Hall. 1988. Ontogenetic habitat shifts in bluegill: the
foraging rate-predation risk trade-off. Ecology 69(5):1352-1366.

Whillans, TH. 1992. Assessing threats to ﬁshery values of Great Lakes
wetlands. In Wetlands of the Great Lakes: protection, restoration, policies
and status of the science. Proceedings of the International Wetland
Symposium, 16-19 May 1990. Edited by J. Kusler and R. Smardon.
Association of State Wetland Managers, lnc., Berne, N.Y. pp 156-165.

Wiley, M.J., R.W. Gorden, S.W. Waite and T. Powless. 1984. The relationship
between aquatic macrophytes and sport ﬁsh production in Illinois ponds: a
simple model. North American Journal of Fisheries Management 4:1 1 1-
119.

61

CHAPTER 2: EFFECT OF COASTAL WETLAND MICROHABITAT
CHARACTERISTICS ON THE DISTRUBUTION OF YOUNG
OF THE YEAR FISH IN SAGNINAW BAY, LAKE HURON
ABSTRACT

Although coastal wetlands are important habitat for early life history stages
of many ﬁsh species, little is known about how wetland habitat characteristics
affect the distribution of young of year (YOY) ﬁsh. This study examined the
relationship between wetland habitat characteristics and the distribution of four
ﬁsh species: bluegill, banded killiﬁsh, yellow perch, and spottail shiner. Changes
in habitat preferences as YOY bluegill and yellow perch increased in size and
comparisons in habitat preferences between resident and transient wetland
species were also examined. A 1-m2 throw trap was used to sample the
distribution of YOY ﬁsh within four coastal wetlands in Saginaw Bay, Lake Huron.
The number and size of YOY ﬁsh, microhabitat characteristics, and adjacent
habitat conditions were recorded for each sample. Logistic regression analysis
was used to model relationships between the distribution of YOY ﬁsh and
microhabitat characteristics. Water depth, vegetation type, and water
temperature were the most inﬂuential characteristics in predicting the distribution
of YOY ﬁsh. This study did not identify differences in the habitat preferences of
YOY bluegill or yellow perch of different size classes. A difference in the types of
microhabitat characteristics that were important to resident and transient YOY
ﬁsh was observed. The presence or absence of resident wetland species
(bluegill and banded killiﬁsh) was correlated with physical habitat structure such

as the type of vegetation present, while the presence or absence of transient

62

wetland species (yellow perch and spottail shiner) was correlated with variations

in water quality.

63

INTRODUCTION

Coastal wetlands play an important role in Great Lakes ecosystems, and
their ecosystem values have been altered and likely compromised by a variety of
human inﬂuences (Krieger et al. 1992). Among the important but poorly deﬁned
values of coastal wetlands is the spawning and early life history habitat that they
provide to Great Lakes ﬁshes (Jude and Pappas 1992). Coastal wetlands
provide juvenile ﬁsh with many beneﬁts including rich food resources and
protective cover. This cover reduces the visibility of small ﬁsh by providing many
places for these ﬁsh to hide from predators, and reduces the maneuverability of
larger ﬁsh (Savino and Stein 1982, 1989, Werner et al. 1983a,b, Mittelbach 1986,
Gotceitas and Colgan 1987, Werner and Hall 1988, Gotceitas 1990a,b, Heck and
Crowder 1991, Carr 1994). Coastal wetlands also provide young ﬁsh with rich
zooplankton and macroinvertebrate food sources (Rozas and Odum 1988,
Turner 1988, Cardinale et al. 1998).

Alteration of coastal habitats affects the abundance and composition of
ﬁsh populations that use these habitats (Poe et al. 1986, Leslie and Timmins
1994, Minns et al. 1994, Brazner1997, Brazner and Beals 1997, Johnson et al.
1997, Duffy and Baltz 1998). The destruction and degradation of coastal
wetlands throughout the Great Lakes over the past 200 years have coincided

with major changes in the ﬁsheries of the Great Lakes (Krieger et al. 1992).

64

In order to quantify the impact of habitat alterations on ﬁsh populations,
there is a need to identify habitat attributes that determine the use of Great Lakes
wetlands by ﬁshes. Understanding which habitat characteristics are important to
the success of young of the year (YOY) ﬁsh can lead to better informed decisions
on the management, protection, and restoration of coastal habitat and the Great
Lakes ﬁsheries that depend on these habitats.

Several authors have called for more information on the nursery function
of Great Lakes coastal wetlands (Herdendorf et al. 1986, Jude and Pappas
1992). This study focuses on the role of coastal wetland areas as early life
history habitat for juvenile fish rather than larval ﬁsh or spawning habitat for two
reasons. First, juveniles have the ability to select habitats, whereas habitats for
larval ﬁsh are based more on spawning habitat selection by adults and water
currents. Additionally, in an examination of potential impacts of habitat alteration
on northern pike, Esox lucius, recruitment in a Lake Ontario coastal wetland,
Casselman and Lewis (1996) concluded that the use of wetlands as early life
history habitat plays a more critical role than use of wetlands as spawning habitat
in determining year class strength.

Secondly, in freshwater systems, it has been suggested that the growth
rate of ﬁsh in their ﬁrst year of life is critical in determining their overwinter
survival and ultimately year class strength (Houde 1994). Fish recruitment can
be increased by enhancing the survival and growth of juvenile fish within their
eany life history habitats (Rothschild 1986, Baltz et al. 1993). However, little is

known about how environmental factors inﬂuence habitat selection by juvenile

65

ﬁsh. At the microhabitat level, the habitat preferences of an individual species
should be inﬂuenced by their ability to forage, grow, and avoid predation (Baltz et
al. 1993). Optimal foraging theory predicts that young ﬁsh should choose habitat
patches that maximize energy intake per unit time, because growth rates,
particularly in the ﬁrst year, must be maximized to increase likelihood of survival
to breeding age (Post and Prankevicius 1987, Tonn and Paszkowski 1987). But
there is a trade-off between the increased protection from predation and
decreased foraging efﬁciency in areas of increased habitat structure. As a result,
a ﬁsh should choose the environment that allows them to minimize the predation
risk/foraging rate ratio (Werner and Hall 1988, Gotceitas 1990b).

Much of the recent work that has increased our understanding of coastal
wetland inﬂuences on Great Lakes fish species has used artiﬁcial substrates
(e.g. Petering and Johnson 1991), was conducted at a coarse spatial scale of
resolution (e.g. Jude and Pappas 1992, Brazner 1997), or in a controlled
mesocosm setting (e.g. Dionne and Folt 1991). Each of these methods has
unique advantages and disadvantages. For example, the use of artiﬁcial
substrates has the advantage of being able to control vegetation characteristics
such as plant density and leaf surface area; however this may not present the
same biological situation to juvenile ﬁsh as natural vegetation due to changes in
food density between artiﬁcial and natural stands of vegetation. Using a coarse
spatial scale of resolution has been shown to be beneﬁcial in determining the
relative importance of wetland habitat compared to other littoral zone

environments; however, it lacks information on the speciﬁc attributes of wetlands

66

that attract ﬁsh. The use of mesocosms has been beneﬁcial in demonstrating
changes in the behavior or growth of ﬁsh with changes in vegetation density or
structure. However, the nature of mesocosms limits the diversity of habitat
patches available to an individual fish. This study built upon these previous
efforts by quantifying relationships between the presence or absence of juvenile
ﬁsh in the ﬁeld and microhabitat characteristics, where the microhabitat of an
individual ﬁsh is deﬁned as the place where that individual is located at a point in
time (Baltz 1990).

The goal of this study was to increase understanding of how wetland
microhabitat characteristics inﬂuence the distribution of juvenile ﬁsh. Speciﬁc
objectives of this study were to 1) determine which microhabitat characteristics
are related to the presence or absence of juvenile ﬁsh from four different species,
2) determine if habitat preferences change as YOY ﬁsh increase in size, and 3)
determine if the same habitat characteristics are important to resident and
transient wetland ﬁsh species. Distribution patterns of four ﬁsh species were
investigated: bluegill, Lepomis macrochirus, banded killiﬁsh, Fundulus
diaphanous, yellow perch, Perca ﬂavescens, and spottail shiner, Notropis
hudsonius. These species were chosen in part because they are abundant
within Saginaw Bay. These species also represent a combination of resident
wetland species (bluegill and banded killiﬁsh) and transient wetland species
(yellow perch and spottail shiner), where resident wetland species are those
species that spends the majority of their life cycle in close association with

wetland environments. Transient wetland species are those species that may

67

utilize wetlands as juvenile habitat, and intermittently throughout their lifecycle,
but may also be commonly found in other aquatic habitats. The four ﬁsh species
chosen for this study also represent a variety of feeding styles. Juvenile yellow
perch are active predators, preying on larger macroinvertebrates and larval fish
within wetland environments, banded killiﬁsh are surface feeding topminnows,
spottail shiners are primarily zooplanktivores, and juvenile bluegill feed on a

combination of macroinvertebrate and zooplankton food resources.

68

METHODS
Study Site Description

Four different wetland areas in Saginaw Bay, Lake Huron were selected
as study sites and sampled from May to August, 2001 (Figure 2.1). These
wetlands represent a subset of the ﬁfteen locations described in Chapter 1.
Compared to the rest of Lake Huron, Saginaw Bay is a eutrophic embayment
with high productivity (Jude and Pappas 1992, Skubinna et al. 1995, Cardinale et
al. 1998). Emergent vegetation communities within Saginaw Bay wetlands were
dominated primarily by Scimus pungens and Typha angustifolia. Submersed
plant beds were dominated by Myriophyllum spicatum, Vallisneria americana,
Potamogeton spp. and Chara globularis.

Saginaw Bay is relatively shallow with eighty percent of its volume less
than 5.5 m in depth (Beeton et al. 1967). Water levels within the Great Lakes are
cyclical in nature, with alternating periods of high and low water occurring over a
period of 7-10 years (Herdendorf 1992). Water levels in Saginaw Bay during the
study averaged approximately 0.5 m below the long term average (Figure 2.2).
Periods of low water level can have dramatic impacts on wetland environments,
causing vegetation dieback, erosion of wetlands, and lateral displacement of

vegetative zones of wetlands (Herdendorf 1992).

69

 

Pine River %

Baypon

Linwood .

b

. Oakhurst

 

 

 

 

 

 

 

 

Figure 2.1: Map of approximate locations of study sites in Saginaw Bay,

Lake Huron.

70

177.0

 

 

 

-*-2001
-+—Mean
is:
T; 176.51
(I)
_l
.9
(D
g I.
o.". ....‘-.. I
176.01 l," --.. ------- u
I"'-'.'.. . .
175.5 I I I j T T I I f I I
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Figure 2.2: Water levels (m IGLD 1985) recorded by NOAA Essexville Station

#9075035 in Saginaw Bay, Lake Huron in 2001 compared to long term mean
(1918-2000).

71

 

Fish Collection

A modiﬁed throw trap (Figure 2.3) was used to sample young of the year
fish within the wetlands. Throw traps have been shown to be an effective
method of sampling YOY ﬁsh in vegetated wetlands (Kushlan 1981, Freeman et
al. 1984, Dewey 1992, Baltz et al. 1993, Raposa and Oviatt 2000). The throw
trap used was designed with an angle iron base, weighted with lead, to cause the
trap to sink rapidly. A PVC (4 cm diameter) frame was used for the top of the
trap, which would ﬂoat on the water’s surface. Nylon netting (3 mm mesh) was
used for the sides of the throw trap. The trap sampled 1 m2 of wetland habitat up
to 1 m in depth. The net was deployed by two people tossing the frame into the
water. Pihl and Rosenberg (1982) found that animals generally did not exhibit an
avoidance response to throw traps until moving people were within 1.5 m, and
our throws typically covered a distance of three to four meters. The trap was
deployed a total of 360 times in each wetland between May 15, 2001 and August
15, 2001 for a total of 1,440 samples. Typically 30 samples were taken in each
wetland each week. After deployment, the bottom of the frame was pushed into
the sediment. A 5 mm mesh seine was used to remove ﬁsh from the trap.
Seining continued until six consecutive seine hauls resulted in no additional ﬁsh
being captured.

A stratiﬁed random sampling procedure was used to select sample
locations. Effort was made to sample the diversity of habitat types available to
YOY ﬁsh within a given wetland in proportion to the availability of that habitat

type. Habitats within the wetlands were classiﬁed as open water, emergent,

72

O.
00
0%

\

   

 

 

 

 

Buoyant PVC 1" diameter 1 m2 frame

 

        

  

     

          

 

V.‘..““ 0.. ‘ ...“."
aﬁﬂﬂﬁﬁgﬁbi?o uﬂ€§%%%%§§§%
00000000 .00 0.9. 0.0090.
voooooooo.ooo 90.00 00000
00090009.... 00000090900
voooooooooooooooooo. coooooooooooooooooooooooooooo
v0.0.0.9...9909033330000

v

f

€¢d§¢o ”I
v0.9.9.

00. 09
v...

0000. o
.0...

0000. o
v... .
O... O
.009

000 to
V.....

90090 009
.0900

090. o
v....

000.. 9
v0.0.

.0099 0
'00..

0.9.0 o
.0009

.0990 o
{6&68 o
0000

vvv
’

v
0

vvvv v
0

V

v
0
O
O
9
0
§
9

vv vvv
0%,. v” §8§ﬁ§
oo o so

0 co

90

0.

$9

00

                           

of 0

§’

0

O

    

 

                                                                     
                                                  

    

iron and weighted with lead

1 m2 bottom frame, constructed of angle-

in wetlands.

throw trap used to sample YOY ﬁsh

Design of 1 m2

Figure 2.3

73

submersed, or ﬂoating leaf vegetation. Five 200 m transects were established at
each wetland site. Habitat type was recorded every 5 meters along the
transects, and these data were used to calculate the proportion of a given habitat
type within a wetland. All ﬁsh collected were identiﬁed to species and counted.
At each sample location, the standard length of ﬁve to ten individuals of each

species was recorded.

Habitat Measurements

The location of each sample was recorded in the ﬁeld using a hand held
Garmin GPS unit. Water depth was measured using a hand held folding ruler.
Water depth was taken at the center of the throw trap, and measured to the

nearest cm.

The type of vegetation within each enclosure was recorded. Plants were
identiﬁed to species according to Fassett (1985). A count was made of the
number of plant stems within each throw trap sample as a measure of vegetation
density. For each microhabitat sampling location, the habitat was classiﬁed as
emergent, submersed, ﬂoating leaf vegetation, or open water based upon the
dominant vegetation. The percent vegetation cover of emergent, submersed,
ﬂoating leaf vegetation and open water within each throw trap sample was

visually estimated to the nearest 10%.

The nearest habitat type, different from the one where the trap was
located, was recorded for each sampling location. For example, if the trap was

thrown in submersed vegetation, I recorded which other type of habitat (open

74

water, emergent vegetation, or ﬂoating leaf vegetation) was closest to that
sample location. After determining what the next nearest habitat type was, the

distance to that habitat type was measured to the nearest meter and recorded.

Substrate at each sample location was classiﬁed as rock, sand, clay, or
organic muck based on the size of soil particles and depth of the organic layer. It
the organic layer was more than 4 cm deep, the soil was classiﬁed as organic

muck.

A YSI model 50b dissolved oxygen meter was used to measure water
temperature to the nearest 0.1°C, and dissolved oxygen to the nearest 0.01
mg/L. A YSI model 33 Salinity-Conductivity-Temperature meter was used to
measure conductivity to the nearest 1 uS/cm. Temperature, dissolved oxygen,
and conductivity measurements were taken at the center of the throw trap, at
approximately 2/3 of the water depth.

At each sample location a water sample was taken for later measurement
of water clarity. Water samples were collected in a 50 ml dark plastic bottle. The
bottles were placed in a dark, cool, container in the ﬁeld. A Monitek model 21

Nephelometer was used to measure turbidity to the nearest 1 NTU.

Data Analysis

Multiple logistic regression analysis (a = 0.05) was used to relate the
presence and absence of YOY ﬁsh of four ﬁsh species to microhabitat
characteristics according to procedures outlined in Allison (1999) and Hosmer

and Lemeshow (2000). Chi-square test statistics for the signiﬁcance of

75

microhabitat coefﬁcients were used in selecting variables to include within the
logistic regression models. Odds ratios were calculated to describe ﬁsh selection
for different habitat characteristics. An odds ratio including 1 indicates no
preference for the habitat characteristic, odds ratios greater than 1 indicate a
preference for the habitat characteristics, and odds ratios less than 1 indicate
selection against a habitat characteristic. All quantitative habitat characteristics
were standardized to have a mean of 0 and standard deviation of 1. To control
for seasonal effects on water temperature, water temperatures were
standardized based upon the mean measured water temperature on each
sampling date.

The four ﬁsh species investigated were: bluegill, banded killiﬁsh, yellow
perch, and spottail shiner. For the bluegill, only samples taken after July 1, 2001
were included in the analysis because YOY bluegill did not start appearing in
samples until after this date. Because YOY bluegill were captured in only one
sample from Oakhurst, data from this site were not used to construct the logistic
regression models. For the three other species, data from the entire summer
and all four wetland sites were used.

To examine if juvenile ﬁsh habitat preferences change as the ﬁsh increase
in size, YOY bluegill and yellow perch were divided into two size categories,
small (<25 mm) and large (225 mm). A break of 25 mm was chosen because it
has been demonstrated that shifts in feeding and habitat use occur around 25
mm for yellow perch (Wahl et al. 1993), bluegill (Bremigan and Stein 1994), as

well as several other ﬁsh species (Simonovic et al. 1999). Statistically signiﬁcant

76

interactions between size class and microhabitat characteristics were tested
using logistic regression modeling. If there was no statistically signiﬁcant change
in habitat use of ﬁsh between the different size classes, a model with pooled data
from both size classes was constructed.

Models were constructed examining relationships between YOY ﬁsh
presence or absence and habitat characteristics in each of the wetlands. When
differences in relationships were not found among wetlands, a pooled model was
used, combining data from all four wetlands sampled. All data analysis was done
using the SAS® System. Variance Inﬂation Factor (VlF) values were used to test
for multicollinearity among variables. Due to strong relationships between the
estimated percent cover of vegetation and plant stem density, only vegetation

density values were used in the models.

77

RESULTS

Bluegill

A total of 510 throw trap samples was taken between July 1, 2001 and
August 15, 2001, at three of the wetland areas, Pine River, Linwood, and Bayport
(Figure 2.1), for a total of 170 samples per wetland. A total of 392 YOY bluegill
were collected in 146 (29%) of the samples. Throw trap samples were taken
over a wide range of microhabitat conditions. Table 2.1 provides a summary of
the habitat conditions sampled for the quantitative variables (water depth, stem
density, water temperature, dissolved oxygen, conductivity, turbidity, and
distance to next habitat type), and Table 2.2 provides a summary for the
categorical variables (habitat type, substrate, and next habitat type).

YOY bluegill sampled were classiﬁed as small (< 25 mm in length) or large
(> 24 mm in length). Small YOY bluegill were captured in 87 (60%) of the
samples where bluegill were present. Large YOY bluegill were captured in 86
(59%) of the samples where bluegill were present. YOY bluegill from both size
classes were captured in 27 (18%) of the samples where YOY bluegill were
present. Size class was not a signiﬁcant predictor of the presence or absence of
YOY bluegill (X21 = 0.0712, p = 0.7896). Interactions between size class and
other habitat variables also were not signiﬁcant (e.g., size class and water depth,
X21 = 1.6282, p = 0.2019). Because interactions between microhabitat variables

and size class were not statistically signiﬁcant in the logistic regression model,

78

Table 2.1: Mean, standard deviation, and range of microhabitat conditions

sampled at Linwood, Bayport, and Pine River wetlands from July 1, 2001 to

August 15, 2001.

 

 

Variable Mean St. Dev. Minimum Maximum
Water depth (cm) 42.5 18.6 5 97
Stem density 26.8 19 0 103
Water temp (°C) 24.4 2.8 20.2 35.8
Dissolved Oxygen (mg/I) 8.3 2.6 3.12 12.54
Conductivity (uS/cm) 1005 658 280 1880
Turbidity (NTU) 17.9 20 2 200
Distance to next habitat (m) 17.6 37.5 1 410

 

79

Table 2.2: Number and percent of categorical microhabitat variable conditions
sampled at Linwood, Bayport, and Pine River wetlands from July 1, 2001 to

August 15, 2001.

 

 

Variable # of Samples % of Samples
Habitat
Emergent 118 23
Submersed 322 63
Open water 66 13
Floating leaf 4 1
Sediment
Rock 1 <1
Sand 1 <1
Clay 346 68
Organic 162 32
Next Habitat Type
Emergent 268 53
Submersed 135 26
Open water 94 18
Floating leaf 13 3

 

80

data for both size classes were pooled for further analysis of relationships
between YOY bluegill presence or absence and microhabitat characteristics.

Of the 146 samples where YOY bluegill were collected, 49 (34%) of these
samples were taken at Pine River, 45 (31%) of these samples were taken at
Linwood, and 52 (36%) of these samples were taken at Bayport. There were no
statistically signiﬁcant differences in the predicted probability of presence of YOY
bluegill among the wetland areas (X22 = 1.4181, p = 0.4921). Additionally,
interactions between wetland and microhabitat characteristics were not
signiﬁcant (e.g. water depth, x22 = 5.0349, p = 0.0807). Data from the three
wetlands were thus pooled for use in constructing logistic regression models
relating the presence or absence of YOY bluegill to microhabitat characteristics.

Logistic regression identified four microhabitat characteristics, water
depth, habitat type, stem density, and distance to next habitat type, that were
signiﬁcantly associated with the presence or absence of YOY bluegill (Table 2.3,
Figure 2.4). Water quality variables such as water temperature (X21 = 0.3143, p
= 0.5751 ), dissolved oxygen (x2, = 0.0166, p = 0.8975), conductivity (x2, =
0.2328, p = 0.6295), and turbidity (X25 = 0.7895, p = 0.3743) were not signiﬁcant
predictors of the presence or absence of YOY bluegill.

The resulting logistic regression model correctly predicted the presence or
absence of YOY bluegill 74.6% of the time. The overall model was statistically
signiﬁcant (Wald’s X25 = 81.4339, p < 0.0001 ). Regression coefﬁcients were
negative for emergent vegetation and distance to next habitat type, and positive

for water depth, ﬂoating leaf and submersed vegetation, and stem density.

81

Table 2.3: Estimated coefﬁcients, standard errors, and X2 tests for the logistic

regression model for bluegill habitat selection.

 

 

Variable o.r=. Coeff. Std. Err. x2 l=>x2
Intercept 1 -2.1876 0.3076 50.5840 <0.0001
Water depth 1 0.5209 0.0925 31 .7322 <0.0001
Habitat

Emergent 1 -0.2278 0.3603 0.3999 0.5271

Floating leaf 1 0.4470 0.8186 0.2981 0.5851

Submersed 1 0.5067 0.3291 9.8234 0.0017
Stem density 1 0.2482 0.1091 5.1771 0.0229
Distance to next habitat 1 -0.6846 0.2304 8.8284 0.0030

 

82

 

 

Predlcted Pmbablllty of
Pmen

 

Water Depth (cm)

1. :1

X X X
1 X ‘x

t".:‘§": * 34:1. 1"

l! x §* . x 1“in:
x“ T I ‘ 4: j: 1“ {£55
“1;?“ “Elli! 1,51%; &:§:§‘*¥

all

20 30 40
Plant penalty (mmdm’)

X

Prodlctod Probablllty of
men

 

 

 

 

 

h
0
E
3
3
2
0.
§ X
t *-
‘1
l
g 4 ‘*x*- . t x
60 90 120 150

Distance to Next Habltat (m)

 

0.6 l I l V

.o 9 .o
09 # U!
I 7 l
I L J

_o
N
I
I

a? a

x K e9 at
90° 09% ‘9 dt’b‘
61(9‘ W 90‘69 $9“
HAB I TAT

1

Predicted Probability of Presence
.0
I

 

 

 

.0
c

Figure 2.4: Relationships between the predicted probability of presence of YOY
bluegill and water depth, plant density, distance to next habitat, and type of

vegetation present.

83

Regression coefﬁcients for emergent and ﬂoating leaf vegetation habitat types
were not statistically signiﬁcant.

The odds ratio estimate for water depth is 1.684, indicating that YOY
bluegill are more likely to be located in deeper water (Table 2.4). Odds ratio
estimates for emergent and ﬂoating leaf vegetation did not indicate any
preference for these habitats over open water. However, the odds ratio estimate
for submersed vegetation is 3.430, indicating that YOY bluegill prefer submersed
vegetation habitat over open water. The odds ratio estimate for stem density
was 1.282, indicating that the likelihood of sampling YOY bluegill showed a slight
increase in more densely vegetated habitats. The odds ratio estimate for
distance to next habitat was 0.504, indicating that YOY bluegill are more likely to
be located near the edge of habitat patches.

The residual chi square test (X221 = 26.5138 p = 0.1875) and Hosmer
Lemeshow goodness of ﬁt test (X23 = 6.5096, p = 0.5903) indicate that the model
provided a reasonable ﬁt to the data, and that residuals from the model were not

associated with other (not included) explanatory variables.

Banded Killiﬁsh

A total of 1,440 throw trap samples was taken between May 15, 2001 and
August 15, 2001, at four wetland areas, Pine River, Linwood, Oakhurst, and
Bayport, for a total of 360 samples per wetland. A total of 1,808 YOY banded
killiﬁsh were collected in 414 (29%) of the samples. Throw trap samples were

taken over a wide range of microhabitat conditions. Table 2.5 provides a

84

Table 2.4: Estimated odds ratio point estimates and 95% Wald Conﬁdence

Limits for variables in the bluegill habitat selection logistic regression model.

 

 

Variable Point Estimate Minimum Maximum
Water depth 1.684 1.404 2.018
Habitat
Emerge. vs. open water 1.645 0.589 4.597
Floating vs. open water 3.231 0.331 31.564
Sub. vs. open water 3.430 1.340 8.777
Stem density 1.282 1.035 1.587
Distance to next habitat 0.504 0.321 0.792

 

85

Table 2.5: Mean, standard deviation, and range of microhabitat conditions

sampled at Pine River, Linwood, Oakhurst, and Bayport wetlands from May 15,

2001 to August 15, 2001.

 

 

Variable Mean St. Dev. Minimum Maximum
Water depth (cm) 33.8 17.4 3 97
Stem density 19.6 32.0 0 103
Water temp (°C) 22.3 4.8 12.5 35.8
Dissolved oxygen (mg/I) 7.6 2.8 0.41 12.54
Conductivity (uS/cm) 516 360 210 1880
Turbidity (NTU) 36.7 51.6 1 560
Distance to next habitat (m) 31.0 69.8 1 700

 

86

summary of the habitat conditions sampled for the quantitative variables (water
depth, stem density, water temperature, dissolved oxygen, conductivity, turbidity,
and distance to next habitat type), and Table 2.6 provides a summary for the
categorical variables (habitat type, substrate, and next habitat type).

Of the 414 samples where YOY banded killiﬁsh were collected, 115 (28%)
of these samples were taken at the Pine River, 174 (42%) of these samples were
conducted at Linwood, 57 (14%) of these samples were taken at Oakhurst, and 1
68 (16%) of these samples were taken at Bayport. There was a statistically .,
signiﬁcant relationship between which wetland the sample was collected and the
predicted presence of YOY banded killiﬁsh (X23 = 190.7959, p < 0.0001).
0.0004). Interactions between microhabitat characteristics and wetland were
tested, and were not statistically signiﬁcant (p values ranged from 0.3479 to
0.8667). Due to the large difference in the percent of samples containing YOY
banded killiﬁsh among the different wetland areas, I retained wetland where the
sample was taken as an explanatory factor for constructing logistic regression
models relating the presence or absence of YOY banded killiﬁsh to microhabitat
characteristics.

Logistic regression identiﬁed four characteristics, wetland, type of habitat,
next habitat type, and turbidity, which had statistically signiﬁcant relationships
with the presence or absence of YOY banded killiﬁsh (Table 2.7, Figure 2.5).
The resulting logistic regression model correctly predicted the presence or
absence of YOY banded killiﬁsh 81.8% of the time. The overall model was

statistically signiﬁcant (Wald’s X211 = 298.6320, p < 0.0001).

87

Table 2.6: Number and percent of categorical microhabitat variable conditions
sampled at Pine River, Linwood, Oakhurst, and Bayport wetlands from May 15,

2001 to August 15, 2001.

 

 

Variable # of Samples % of Samples
Habitat
Emergent 248 17.2
Submersed 658 45.7
Open water 525 36.5
Floating leaf 9 0.6
Sediment
Rock 2 0.1
Sand 328 22.8
Clay 784 54.4
Organic 326 22.6
Next Habitat Type
Emergent 645 44.8
Submersed 329 22.8
Open water 448 31.1
Floating leaf 18 1 .3

 

88

Table 2.7: Estimated coefﬁcients, standard errors, and X2 tests for the logistic

regression model for YOY banded killiﬁsh habitat selection.

 

 

Variable o.r=. Coeff. Std. Err. x2 l=>x2
Intercept 1 -0.41 10 0.3278 1 .5724 0.2099
VVeﬂand

Linwood 1 0.6148 0.1211 25.7672 <0.0001

Oakhurst 1 -0.1915 0.1665 1.3239 0.2499

Bayport 1 -0.7602 0.1365 31 .0151 <0.0001
Habitat

Emergent 1 0.0847 0.2398 0.1246 0.7241

Floating leaf 1 1.0175 0.5681 3.2082 0.0733

Submersed 1 0.5546 0.2065 7.2114 0.0072
Next Habitat Type

Emergent 1 -0.1834 0.2752 0.4440 0.5052

Floating leaf 1 1 .9845 0.7760 6.5398 0.0105

Submersed 1 -0.3934 0.2852 1 .9032 0.1677
Turbidity 1 -0.6874 0.1361 25.5250 <0.0001

 

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 8 I
5 5
g 0.8F — g 03— i _
o. o.
“6 r - .
g; 0.6— 1 _L _, g 0.6- Q A
a a t
e 0.4 ’- — e 0.4 '_ . -l
O. Q
l3 — E $
.9 0,2— - .2 0.2F ‘
E T 123 _t_
o. [L
0.0 o x o o 1‘
(e «19’ 0" o 9 ‘0‘
#90 ~11 6" Q-‘ .90 (\6 (.9 *9
9° ‘° 00* 9'6"” 06° (09 9969 06,0
Wetland Habitat
1.0
a, ' ..l— ' ' 1.0._ . j , , ,
0 a:
C 0
o c
(D 0.8— - a) .
0’ , —1
*- o.
o .,_
g. 0.6 "‘ — o , . —
E g 0.6 .7
.0 '1, - ".
.8 0-4 — - g . -.-‘~ .15,
8 °- f-j' :-
7§ 0.2— — E . .
“‘- .2 0,2 a
D. l T E, .21....
0.0 O. 1:. "._’,6‘ I
of” 60” 6‘” we 0.0 7M“? 11.4 . - 4 a i .4
,c° 60° 960° 0,00 0 100 200 300 400 500 600
Next Habitat Turbldlty (NT U)

Figure 2.5: Relationships between microhabitat characteristics (wetland, habitat
type, next habitat type, and turbidity) and the predicted probability of presence of

YOY banded killiﬁsh.

90

The odds ratio estimates for wetland were calculated based upon
comparisons to Pine River wetland (Table 2.8). The odds ratio estimate for
Linwood is 1.320, however the conﬁdence interval for this estimate contains 1,
indicating no difference in the likelihood of sampling YOY banded killiﬁsh at
Linwood compared to Pine River. The odds ratio estimates and conﬁdence
intervals for Oakhurst and Bayport are both less than 1, indicating that YOY
banded killiﬁsh are less likely to be present in samples from these wetlands than
at Pine River. Odds ratio estimates for habitat type were calculated based upon
comparisons to open water. The conﬁdence intervals for emergent, ﬂoating leaf,
and submersed vegetation are all greater than 1, indicating that banded killiﬁsh
are more likely to be sampled in vegetated habitats than open water. The odds
ratio estimates for next habitat type were also calculated based upon
comparisons to open water. All of the next habitat type odds ratio estimates and
conﬁdence intervals are greater than one, indicating that YOY banded killiﬁsh are
less likely to be captured near open water habitat than emergent, submersed, or
ﬂoating leaf vegetation. The odds ratio estimate for turbidity is 0.503 indicating
that YOY banded killiﬁsh are more likely to be present in areas with low turbidity.

The residual chi square test (X25 = 0.9518, p = 0.9170) and Hosmer
Lemeshow goodness of ﬁt test (X28 = 15.1020, p = 0.0572) indicate that the
model provides a reasonable ﬁt to the data, and that residuals from the model

were not associated with other (not included) explanatory variables.

91

Table 2.8: Estimated odds ratio point estimates and 95% Wald Conﬁdence

Limits for variables in the banded killiﬁsh habitat selection logistic regression

 

 

model.
Variable Point Estimate Minimum Maximum
Wetland
Oakhurst vs. Pine 0.590 0.370 0.939
Bayport vs. Pine 0.334 0.223 0.500
Linwood vs. Pine 1.320 0.928 1.878
Habitat
Emerge vs. open 5.706 3.454 9.426
Floating vs. open 14.502 3.214 65.432
Submersed vs. open 9.129 6.330 13.165
Next Habitat Type
Emerge. vs. open 3.402 2.225 5.202
Floating vs. open 29.734 3.808 232.192
Submersed vs. open 2.758 1.807 4.208
Turbidity 0.503 0.385 0.667

 

92

Yellow Perch

A total of 849 yellow perch was collected in 2001 from the 1440 samples,
over a wide range of microhabitat conditions (Tables 2.5 and 2.6). Yellow perch
were collected in 348 or 24% of the samples. YOY yellow perch were divided
into two size categories. Large YOY yellow perch (> 24 mm in length) were
present in 178 or 51% of the samples where yellow perch were collected. Small
YOY yellow perch (< 25 mm in length) were present in 194 or 56% of the
samples where yellow perch were collected. Fish from both size classes were
found in 24 (7%) of the samples where yellow perch were collected. There was
not a statistically signiﬁcant relationship between size class and the predicted
presence or absence of YOY yellow perch (X21 = 0.0498, p = 0.8234).
Interactions between size class and other microhabitat variables also were not
signiﬁcant (p values ranged from 0.1258 to 0.8624). Therefore, YOY yellow
perch from both size classes were pooled for further analysis of relationships
between yellow perch presence or absence and microhabitat characteristics.

Of the 348 samples where YOY yellow perch were collected, 58 (17%) of
these samples were taken at Pine River, 54 (16%) of these samples were
conducted at Linwood, 76 (22%) of these samples were taken at Oakhurst, and
160 (46%) of these samples were taken at Bayport. There was not a statistically
signiﬁcant relationship between which wetland the sample was taken in and the
predicted presence of YOY yellow perch (X23 = 5.2905, p = 0.1517). There was
a statistically signiﬁcant interaction between wetland and water depth (X23 =

23.9679, p < 0.0001). However, there was a signiﬁcant difference in the mean

93

water depth and range of water depths available for ﬁsh among the three habitats
(F3,1435 = 116.550, p < 0.0001, Table 2.9). Interactions between wetland and the
remaining microhabitat characteristics were not signiﬁcant (p values ranged from
0.0552, to 0.9526). Due to the lack of statistically signiﬁcant relationships, data
from all four wetlands were pooled for use in constructing logistic regression
models relating the presence or absence of YOY yellow perch to microhabitat
characteristics.

Logistic regression identiﬁed three microhabitat characteristics, water
depth, water temperature, and conductivity, which had statistically signiﬁcant
relationships with the presence or absence of YOY yellow perch (Table 2.10,
Figure 2.6). The resulting logistic regression model correctly predicted the
presence or absence of YOY yellow perch 75.5% of the time. The overall model
was statistically signiﬁcant (Wald’s X23 = 25.8195, p < 0.0001). Regression
coefﬁcients were negative for water temperature, and positive for water depth
and conductivity. Microhabitat variables selected to describe physical habitat
conditions such as habitat type (X23 = 0.4170, p = 0.9367), plant stem density
(x2, = 0.3851, p = 0.5349), next habitat type (X2. = 0.3147, p = 0.9572), and
distance to next habitat patch (X21 = 3.1993, p = 0.0737) were not signiﬁcant
predictors of the presence or absence of YOY yellow perch.

The odds ratio estimate for water depth is 1.670, indicating that YOY
yellow perch are more likely to be located in deeper water (Table 2.11). The
odds ratio estimate for water temperature is 0.440, indicating that the probability

of presence of YOY yellow perch increases with cooler water temperatures. The

94

Table 2.9: Mean, standard deviation, minimum, maximum, median, and

quartiles of water depths (cm) sampled in each of the four wetland areas.

 

 

Wetland Mean St. Dev. Min. Max. Median 1st Quar. 3rd Quar.
Pine River 47.8 20.7 4 97 46 32 64
LinWOOd 30.2 1 1 .1 4 55 32 20 38
Oakhurst 28.3 12.2 3 61 27 19 37
Bayport 30.3 17.4 3 74 29 15 43

 

95

Table 2.10: Estimated coefficients, standard errors, and X2 tests for the logistic

regression model for yellow perch habitat selection.

 

 

Variable o.r=. Coeff. Std. Err. x2 P>X2

Intercept 1 -3.7307 0.3786 97.0818 <0.0001
Water depth 1 0.5130 0.2017 6.4671 0.0110
Water temperature 1 -0.8217 0.3540 5.3888 0.0203
Conductivity 1 0.7739 0.2151 12.9464 0.0003

 

96

0.35 -

Predicted Probability of
Presence
9
a

 

 

Predicted Probability of
Presence

 

0.4 Water Temperature (°C)
0.35 1 x
0.3 x?‘

0.25 X x X
0.2 § *

0.15
0.1

0.05

Predicted Probability of
Presence

 

 

Water Depth (cm)

Figure 2.6: Relationships between the predicted probability of presence of YOY

yellow perch and conductivity, water temperature, and water depth.

97

Table 2.11: Estimated odds ratio point estimates and 95% Wald Conﬁdence

Limits for variables in the yellow perch habitat selection logistic regression model.

 

 

Variable Point Estimate Minimum Maximum
Water depth 1 .670 1 .125 2.480
Water temperature 0.440 0.220 0.880
Conductivity 2.168 1 .422 3.305

 

98

odds ratio estimate for conductivity is 2.168 indicating that YOY yellow perch
were more likely to be present in areas with high water conductivity.

The residual chi square test (X211 = 8.8394, p = 0.6367) and Hosmer
Lemeshow goodness of ﬁt test (X28 = 4.7285, p = 0.7862) indicate that the model
provided a reasonable ﬁt to the data, and that residuals from the model were not

associated with other (not included) explanatory variables.

Spottail Shiner

A total of 819 YOY spottail shiners was collected in 2001 from 1440
samples, over a wide range of microhabitat conditions (Tables 2.5 and 2.6).
Spottail shiners were collected in 277 or 19% of the samples. Of the 277
samples where YOY spottail shiners were collected, 100 (36%) of these samples
were taken at Bayport, 40 (15%) of these samples were conducted at Linwood,
93 (34%) of these samples were taken at Oakhurst, and 44 (16%) of these
samples were taken at Pine River. There was a statistically signiﬁcant
relationship between which wetland the sample was taken in and the predicted
presence of YOY spottail shiner (X23 = 13.0233, p = 0.0046). Interactions
between wetland and microhabitat characteristics were not signiﬁcant (p-values
ranged from 0.3816 to 0.6681 ), with the exception of wetland*water depth (X23 =
13.0233, p = 0.0046). However, there was a signiﬁcant difference in the ranges
of water depths sampled among the four wetland areas (F3,1435 = 116.550, p <

0.0001, Table 2.9).

99

Logistic regression identiﬁed ﬁve microhabitat characteristics, water
temperature, water depth, turbidity, wetland, and dissolved oxygen, which had
statistically signiﬁcant relationships with the presence or absence of YOY spottail
shiners (Table 2.12, Figure 2.7). The resulting logistic regression model correctly
predicted the presence or absence of small YOY spottail shiners 78.4% of the
time. The overall model was statistically signiﬁcant (Wald’s X27 = 94.0427, p <
0.0001). Regression coefﬁcients were negative for turbidity, and positive for
water temperature, water depth, and dissolved oxygen. The only coefﬁcient
within the wetland variable that was statistically signiﬁcant was for Bayport.
Microhabitat variables selected to describe physical habitat conditions such as
habitat type (X23 = 0.3488, p = 0.9506), plant stem density (X21 = 2.7237, p =
0.0989), next habitat type (X23 = 2.2077, p = 0.5304), and distance to next
habitat patch (X21 = 0.0075, p = 0.9308) were not signiﬁcant predictors of the
presence or absence of YOY spottail shiner.

The odds ratio estimate for water temperature is 2.283, indicating that
YOY spottail shiner are more likely to be found in areas with warmer water
temperatures (Table 2.13). The odds ratio estimate for water depth is 1.527,
indicating that YOY spottail shiner are more likely to be present in areas of
deeper water. The odds ratio estimate for turbidity is -0.5054, indicating that the
likelihood of presence of YOY spottail shiner decreases with increased turbidity.
The odds ratios for wetland were calculated based upon comparisons to the Pine
River wetland. The conﬁdence interval for the odds ratio estimate for Linwood

includes 1, indicating no difference in the likelihood of sampling YOY spottail

100

Table 2.12: Estimated coefﬁcients, standard errors, and X2 tests for the logistic

regression model for spottail shiner distribution.

 

 

Variable o.r=. Coeff. Std. Err. x2 l=>x2
Intercept 1 -1.7633 0.1 147 236.2869 <0.0001
Water temperature 1 0.8253 0.1233 44.7905 <0.0001
Water depth 1 0.4230 0.1096 14.9115 0.0001
Turbidity 1 05054 0.1562 10.4669 0.0012
VVeﬂand

Linwood 1 -0.3278 0.1973 2.7617 0.0965

Oakhurst 1 0.2045 0.1745 1.3746 0.2410

Bayport 1 0.7240 0.1827 15.7057 <0.0001
Dissolved oxygen 1 0.3200 0.0990 10.4413 0.0012

 

101

 

 

0.8 — n

0.6— -- ' i -

 

Predicted Probability of Presence
Predicted Probability of Presence

 

 

. 0,0 :52».-. ﬁt“ I "it I L l
. “ ' .... ' 0 10 20 30 40 50 60 70 8090100
10 20 30 o 40
Water De th cm
WaterTemperature( C) p ( )

l | I I 1'0 l l l I l |

 

 

 

 

 

 

 

0.8 — -

0.6— . . .-. —

Predicted Probability of Presence
Predicted Probability of Presence

 

 

 

   

0 100 200 300 400 500 600 0 2 4 6 8 12 14
Turbidity (NT U) Dissolved Oxygen (mg/I)

Figure 2.7: Relationships between the likelihood of presence of YOY spottail

shiner and water temperature, water depth, turbidity and dissolved oxygen.

102

Table 2.13: Estimated odds ratio point estimates and 95% Wald Conﬁdence

Limits for variables in the spottail Shiner habitat distribution logistic regression

 

 

model.
Variable Point Estimate Minimum Maximum
Water temperature 2.283 1 .793 2.907
Water depth 1.527 1.232 1.892
Turbidity 0.603 0.444 0.891
VVeﬂand
Linwood vs. Pine River 1.314 0.691 2.499
Oakhurst vs. Pine River 2.237 1.203 4.160
Bayport vs. Pine River 3.761 1.987 7.116
Dissolved oxygen 1.377 1.134 1.672

 

103

shiners at Linwood compared to Pine River wetland. However, the odds ratio
estimate for Bayport (3.761) and Oakhurst (2.237),and their conﬁdence intervals
are both greater than 1, indicating that there is a higher probability of sampling
YOY spottail shiners if the sample is taken at Bayport or Oakhurst compared to
Pine River wetland. The odds ratio estimate for dissolved oxygen is 1.377,
indicating that the probability of presence of YOY spottail shiners increases with
increased oxygen concentrations.

The residual chi square test (X211 = 9.9719, p = 0.5329) and Hosmer
Lemeshow goodness of ﬁt test (X28 = 6.5496, p = 0.5859) indicate that the model
provided a reasonable ﬁt to the data, and that residuals from the model were not

associated with other (not included) explanatory variables.

104

DISCUSSION

Influence of microhabitat characteristics on YOY ﬁsh distribution

This study examined the spatial distribution of young of the year bluegill,
banded killiﬁsh, yellow perch, and spottail shiners. These Species were not
uniformly distributed within the wetland areas investigated, and there were
predictable patterns in their distribution. Logistic regression was used to model
the presence or absence of YOY ﬁsh based upon microhabitat characteristics,
with the models correctly predicting the presence or absence of ﬁsh 74.6% to
81.8% of the time. With the exception of substrate type, all of the microhabitat
variables measured were associated with the presence or absence of YOY ﬁsh
for at least one of the species investigated (Table 2.14).

Water depth was the most inﬂuential microhabitat characteristic, showing
a statistically signiﬁcant relationship with the presence or absence of YOY ﬁsh
from three of the four species investigated (Table 2.14). In all three cases, YOY
ﬁsh had a higher likelihood of presence in areas with deeper water levels.
Habitat type, water temperature, and turbidity were also important, and were
signiﬁcant variables in 50% of the models. The other microhabitat characteristics.
measured, plant density, distance to next habitat type, next habitat type,
dissolved oxygen, and conductivity, showed Signiﬁcant relationships with the

presence or absence of YOY ﬁsh from one of the Species investigated.

105

 

6.0- and 8.0 Ned E: 555.8

Kc mm o- 33 mmn no.8 >
8.0- 066 06.6 6.5 55.26
8.0 8.0- 8.0 omxw SR ___oo=_m

 

E81 83
. exoz 2 53...:
8:665 682

36:60 83 28:00
Em_n_ #95me .x. .

c6995 dEo... Sumo

559:5 6:00 .85 66.3 .263

 

.2356; Loam; :80 n O “Em .cozﬂomg EomcoEo u
m. .8398? 686E936 .i. w 982, 6085.65 6536; 6.... 9865 26:6. .83 6.59m; 68: new “956; Lon. .mEQoEmS
86856.. 2865 EonEsz .6: >O> ..o 8:88 Lo 8885 8669a 65 .53 6968366. E8556 2.836366

66266 55 62686996 EznmchE “Em m_o_ooE 86869 2660. ho 8:628c8 e\c __Eo>O "Emu 63m...

106

WateLEJepth

Water depth showed a statistically signiﬁcant relationship with the
presence or absence of YOY ﬁsh from three of the four species investigated
(Table 2.14). In all three cases, YOY ﬁsh were more likely to be sampled from
deeper water habitats (7O - 100 cm). Water depth also had a large impact on
the young of year ﬁsh community at the macrohabitat scale (see chapter one)
showing positive relationships with species diversity, species evenness, and
growth of YOY bluegill.

Deeper water may be important to young of the year ﬁsh for several
reasons. First of all, deeper water may provide young ﬁsh with a refuge from
warm water temperatures during hot summer months. Measured water
temperatures ranged as high as 35.8 °C during the study period. These very
warm water temperatures may be detrimental to ﬁsh growth, and as a result
during warm days, YOY ﬁsh may seek out pockets of deeper water to escape
high temperatures. Secondly, areas of deep water may provide young ﬁsh with a
refuge from avian predation. Avian predators, including herons, egrets, and
terns, were frequently seen throughout Saginaw Bay. These birds commonly
feed upon ﬁsh. As a result of the presence of these birds, young ﬁsh may select
deeper water habitat to avoid predation.

The data collected as part of this study showed an increase in likelihood of
presence of YOY ﬁsh with increased water depth; however, the use of a throw
trap limited sampling to areas with water levels less than 100 cm. Several other

studies have indicated that there is an increase in the density and diversity of

107

YOY ﬁsh in wetland areas less than 1 to 2 m in depth (Keast et al. 1978, Chubb
and Liston 1986, Bryan and Scamecchia 1992). Data here complement these
studies by indicating that while YOY ﬁsh may be more abundant in shallow
wetland areas (< 1-2 m), they may be selecting for deeper habitats within this
range.

Water levels in the Great Lakes are cyclical in nature (Herdendorf 1992),
and during the study period water levels were approximately 0.5 m below the
long term mean. It is possible that during times of low water levels in Saginaw
Bay, the importance of water depth to YOY ﬁsh may be exaggerated. Saginaw
Bay has a shallow littoral slope. As a result of water levels being 0.5 m below the
long term mean, there were few areas within the wetlands l sampled where there
was habitat with water depths greater than 0.5 m, adjacent to vegetation.
However, if low water levels within Saginaw Bay continue, over time the
submersed vegetation community will spread into areas further off-shore where

there is deeper water.

Habitat Type

The presence of vegetation structure was important in determining the
presence or absence of YOY ﬁsh from two of the four species investigated (Table
2.14). Young of the year bluegill and banded killiﬁsh selected vegetated habitats
over areas of open water. The importance of vegetation in providing habitat for
young ﬁsh has been documented in several studies (e.g. Chubb and Liston 1986,

Conrow et al. 1990, Petering and Johnson 1991, Leslie and Timmins 1994,

108

Minns et al. 1997, Brazner and Beals 1997, Duffy and Baltz 1998). Vegetated
habitats provide beneﬁts over open water such as increased protection from
predation (e.g. Werner et al. 1983b, Gotceitas and Colgan 1987, Gotceitas
1990b, Heck and Crowder 1991). Vegetation cover was also important to young
of year ﬁsh at the t the macrohabitat scale. The percent cover of submersed
vegetation showed positive relationships with species richness, while the amount
of open water showed negative relationships with species diversity, and
inﬂuenced species composition (see chapter 1).

While this study adds support to the general ﬁnding that the presence of
vegetation structure is important to young ﬁsh, this study also suggests that the
type of vegetation present may also inﬂuence the presence or absence of some
species of young ﬁsh. Bluegill were more likely to be collected from submersed
vegetation habitats over areas of emergent vegetation and open water, while
banded killiﬁsh preferred both types of vegetated habitats over open water. In
addition to inﬂuencing the presence or absence of YOY ﬁsh, the abundance of
vegetation may also affect their growth rates. The growth of banded killiﬁsh was
positively related to the amount of submersed vegetation cover at the
macrohabitat scale (see chapter 1).

The type of vegetation present may be important to young ﬁsh because
the foraging efﬁciency of individual ﬁsh may decline with changes in the
vegetation community structure and increased habitat complexity (Winﬁeld 1986,
Diehl 1988, Gotceitas 1990b, Dionne and Folt 1991). However, the protection

that vegetation provides to young ﬁsh from predation will increase with vegetation

109

complexity (Werner et al. 1983a,b, Lillie and Budd 1992, Jacobsen and Berg
1998). Crowder and Cooper (1982) suggested that the value of aquatic plants as
habitat for ﬁsh varies among plant species based on the density and architecture
of the species. While few studies have been conducted documenting the
distribution of young ﬁsh in the wild in response to different plant communities,
Petering and Johnson (1991) used artiﬁcial substrates to demonstrate a
preference of larval ﬁsh to ﬂoating leaf and emergent vegetation over submersed
vegetation. This study suggests that YOY ﬁsh from different species may prefer
different types of vegetation structure, based upon the habitat needs of each
species.

The presence of vegetation may not be important to all species of YOY
ﬁsh. This study found no relationship between the presence or absence of YOY
yellow perch and spottail shiner and aquatic vegetation. In a study by Fisher et
al. (1999), they also found no relationship between the abundance of juvenile
yellow perch and the presence of emergent or submersed aquatic vegetation.
However, at the macrohabitat scale, I did ﬁnd a positive relationship between
growth rates of yellow perch and the amount of submersed vegetation (see

Chapter 1).

Water Temperattﬂ
Water temperature was important in describing the distribution of yellow
perch and spottail shiners. In their ﬁrst year of life, ﬁsh need to grow rapidly to

increase their likelihood of overwinter survival (Post and Prankevicius 1987, Tonn

110

and Paszkowski 1987). Spottail shiners may select warmer temperatures
because they can grow faster as water temperatures increase.

Young of the year yellow perch selected habitats with cooler water
temperatures. Compared to spottail shiners, yellow perch can be considered
cool water species. Tidwell et al. (1999) raised juvenile yellow perch in tanks
with water temperatures of 20°C, 24°C, and 28°C. They found that yellow perch
grew signiﬁcantly better in water temperatures of 24°C compared to 28°C, and
survival rates were higher for perch 24°C tanks compared to 28°C. Measured
water temperatures across the duration of this study ranged from 125°C to
358°C, and were 28°C or warmer 14.6% of the time. Growth rates of YOY
yellow perch in Saginaw Bay may be hindered as water temperatures
approached 35.8 °C, and as a result they selected habitats with cooler water
temperatures.

Fisher et al. (1999) showed that the abundance of juvenile yellow perch in
South Dakota lakes was negatively correlated to water temperatures. Results of
this study also indicate that water temperatures are important to the distribution
of YOY yellow perch, with perch preferring cooler water temperatures.

While there are few studies that have examined the effects of temperature
on the distribution of YOY ﬁsh, there are several studies that compare yearly
ﬂuctuations in temperature to spawning success and larval ﬁsh densities (e.g.
Chubb and Liston 1986). Results of this study, and those by Fisher et al. (1999)
and Darf and Powell (1997), indicate that within a year, water temperatures

within the early life history habitat are associated with YOY ﬁsh distribution, and

111

therefore may inﬂuence their likelihood of survival. Further investigation of the
impacts of water temperature on ﬁsh distribution and abundance is necessary to

determine its potential importance to YOY ﬁsh growth, and probability of survival.

Turbidity

The presence or absence of two of the four ﬁsh species investigated was
related to turbidity. Both banded killiﬁsh and spottail shiner were more likely to
be sampled in areas with higher water clarity. Several other studies have also
documented the importance of turbidity with regards to ﬁsh distribution in the
Great Lakes (Jude and Pappas 1992, Petering and Johnson 1991, Brazner and
Beals 1997). Increased turbidity may inhibit the ability of YOY ﬁsh to locate and
capture prey. As a result, YOY banded killiﬁsh and spottail shiners may select

areas with improved water clarity.

Other Microhabitat Characteristics

Although water depth, habitat type, water temperature, and turbidity were
the most common habitat characteristics to inﬂuence the distribution of YOY ﬁsh,
other microhabitat characteristics such as plant density, conductivity, distance to
next habitat, and dissolved oxygen also inﬂuenced YOY ﬁsh presence or
absence (Table 2.14). For example, YOY bluegill selected habitats with
increased plant density, which may be a result of increased protection from
predators in more densely vegetated habitat (Savino and Stein 1989, Gotceitas

19903, Lillie and Budd 1992). Therefore, it may be important to evaluate

112

additional microhabitat characteristics to determine their potential impacts on the
distribution, growth, and survival of YOY ﬁsh. Which microhabitat characteristics

are important to ﬁsh distribution will vary among species.

Ontogenetic Shifts in Habitat Use

Many ﬁsh species undergo ontogenetic shifts in habitat use as they
develop (e.g. Werner and Hall 1988, Post and McQueen 1988, Wahl et al. 1993,
Eggleston 1995, Simonovic et al. 1999, Svanbaeck and Ekloev 2002). These
shifts in habitat use can inﬂuence ﬁsh growth rates (Ruzycki and Wurtsbaugh
1999, Dahlgren and Eggleston 2000), and typically occur in response to changes
in diet (VanderKooy et al. 2000, Svanbaeck and Ekloey 2002) or predation risk
(Werner and Hall 1988, Sillett and Foster 2000, Dahlgren and Eggleston 2000).
Although other authors have demonstrated ontogenetic shifts in habitat use by
bluegill and yellow perch (Post and McQueen 1988, Werner and Hall 1988, Wahl
et al. 1993), results of our logistic regression modeling did not indicate
differences in the distribution of smaller (<25 mm) and larger (>24 mm) ﬁsh from
these species in response to the microhabitat characteristics measured.
However, the previously mentioned studies on bluegill and yellow perch were
conducted at larger temporal and spatial scales; documenting broad habitat shifts
(e.g. offshore vs. inshore), over longer time spans (larval to juvenile to adult

stages) and greater size differences.

113

Resident vs. Transient Wetland Species

This study examined the influence of microhabitat characteristics on the
distribution of YOY ﬁsh from two resident wetland species (bluegill and banded
killiﬁsh) and two transient wetland species (yellow perch and spottail shiner). A
summary of the results of the logistic regression modeling is presented in Table
2.14. These results indicate a difference in the types of microhabitat
characteristics that were important to YOY ﬁsh from both groups. In general, the
resident wetland species were responding to microhabitat characteristics that
describe the physical structure of the habitat (e.g. habitat type, plant density).
Contrary to this, the presence or absence of transient wetland species was not
correlated to microhabitat characteristics used to describe physical habitat
conditions. No statistically signiﬁcant relationships were found between the
presence or absence of these ﬁsh and microhabitat characteristics such as
habitat type, next habitat type, and vegetation density. Instead, these ﬁsh were
associated with variation in water quality parameters such as water temperature,
dissolved oxygen, conductivity, and turbidity. These results suggest a signiﬁcant
difference in the way that YOY ﬁsh from resident and transient wetland species
respond to conditions within wetlands. Resident wetland species appear to be
more restricted in their distribution to the type of habitat structure within wetlands,
while the presence or absence of transient wetland species was not related to
the microhabitat characteristics measured here to describe physical habitat

conditions. Further research is needed to determine if this pattern applies to a

114

broader suite of transient and resident wetland species, and to determine what

implications this may have for wetland and ﬁsheries management.

115

LITERATURE CITED

Allison, PD. 1999. Logistic Regression Using the SAS System: Theory and
Application. SAS Institute, Inc., Cary, NC.

Baltz, D.M. 1990. Autecology. pp. 585-607 IN C.B. Schreck and PB. Moyle
(eds.). Methods for Fish Biology. American Fisheries Society, Bethesda,
MD.

Baltz, D.M., C. Rakocinski, and J.W. Fleeger. 1993. Microhabitat use by marsh-
edge ﬁshes in a Louisiana estuary. Environmental Biology of Fishes
36:109-126.

Beeton, A., S. Smith and F. Hooper. 1967. Physical limnology of Saginaw Bay,
Lake Huron. Technical Report No. 12. Great Lakes Fishery Commission,
Ann Arbor, MI.

Brazner, J.C. 1997. Regional, habitat and human development inﬂuences on
coastal wetland and beach ﬁsh assemblages in Green Bay, Lake
Michigan. Journal of Great Lakes Research 23:36-51.

Brazner, J.C. and E.W. Beals. 1997. Patterns in ﬁsh assemblages from coastal
wetland and beach habitats in Green Bay, Lake Michigan: a multivariate
analysis of abiotic and biotic forcing factors. Canadian Journal of
Fisheries and Aquatic Sciences 54:1743-1761.

Bremigan, MT. and RA. Stein. 1994. Gape-dependent larval foraging and
zooplankton size: Implications for fish recruitment across systems.
Canadian Journal of Fisheries and Aquatic Sciences 51 :913-922.

Bryan, MD. and D.L. Scamecchia. 1992. Species richness, composition, and
abundance of ﬁsh larvae and juveniles inhabiting natural and developed
shorelines of a glacial Iowa lake. Environmental Biology of Fishes 35:329-
341.

Cardinale, B.J., V.J. Brady and TM. Burton. 1998. Changes in the abundance

and diversity of coastal wetland fauna from the open water/macrophyte
edge towards shore. Wetlands Ecology and Management 6:59-68.

116

Carr, M.H. 1994. Predicting recruitment of temperate reef ﬁshes in response to
changes in macrophyte density caused by disturbance. pp 255-269 IN
D.J. Stouder, K.L. Fresh, R.J. Feller (eds.). Theory and Application in Fish
Feeding Ecology. University of South Carolina Press, Columbia, SC.

Casselman, J. M., and CA. Lewis. 1996. Habitat requirements of northern pike
(Esox lucius). Canadian Journal of Fisheries and Aquatic Science 53:161-
174.

Chubb, S.L. and CR. Liston. 1986. Density and distribution of larval ﬁshes in
Pentwater Marsh, a coastal wetland on Lake Michigan. Journal of Great
Lakes Research 12:332-343.

Conrow, R., A.V. Zale, and R.W. Gregory. 1990. Distributions and abundances
of early life stages of ﬁshes in a Florida lake dominated by aquatic
macrophytes. Transactions of the American Fisheries Society 1192521-
528.

Crowder, LB. and WE. Cooper. 1982. Habitat structural complexity and the
interaction between bluegills and their prey. Ecology 63:1802-1813.

Dahlgren, GP. and DB. Eggleston. 2000. Ecological processes underlying
ontogenetic habitat shifts in a coral reef ﬁsh. Ecology 81 :2227-2240.

Darf, BA. and J.C. Powell. 1997. Distribution, abundance, and habitat
characteristics of juvenile tautog (Tautoga onitis, Family Labridae) in
Narragansett Bay, Rhode Island, 1988-1992. Estuaries 20:589-600.

Dewey, MR. 1992. Effectiveness of a drop net, a pop net, and an electroﬁshing
frame for collecting quantitative samples of juvenile ﬁshes in vegetation.
North American Journal of Fisheries Management 12:808-813.

Diehl, S. 1988. Foraging efﬁciency of three freshwater ﬁshes: effects of
structural complexity and light. Oikos 53:207-213.

Dionne, M. and CL. Folt. 1991. An experimental analysis of macrophytes
growth forms as ﬁsh foraging habitat. Canadian Journal of Fisheries and
Aquatic Sciences 48:123-131.

117

Duffy, KC. and D.M. Baltz. 1998. Comparison of ﬁsh assemblages associated
with native and exotic submerged macrophytes in the Lake Pontchartrain
estuary, USA. Journal of Experimental Marine Biology and Ecology
223:199-221 .

Eggleston, DB. 1995. Recruitment in Nassau grouper Epinephelus stn’atus:
post-settlement abundance, microhabitat features, and ontogenetic habitat
shifts. Marine Ecology Progress Sweries 124:9-22.

Fassett, NC, 1985. A Manual of Aquatic Plants. The University of Wisconsin
Press, Madison, WI.

Fisher, S.J., C.R. Pyle, and D.W. Willis. 1999. Habitat use by age-0 yellow perch
in two South Dakota glacial lakes. Ecology of Freshwater Fish 8:85-93.

Freeman, B.J., H.S. Greening, and JD. Oliver. 1984. Comparison of three
methods for sampling ﬁshes and macroinvertebrates in a vegetated
freshwater wetland. Journal of Freshwater Ecology 2:603-609.

Gotceitas, V. and P. Colgan. 1987. Selection between densities of artiﬁcial
vegetation by young bluegills avoiding predation. Transactions of the
American Fisheries Society 116:40-49.

Gotceitas, V. 1990a. Plant stem density as a cue in patch choice by foraging
juvenile bluegill sunﬁsh. Environmental Biology of Fishes 29:227-232.

Gotceitas, V. 1990b. Foraging and predator avoidance: a test of a patch choice
model with juvenile bluegill sunﬁsh. Oecologia 83:346-351.

Heck, K.L., and LB. Crowder. 1991. Habitat structure and predator-prey
interactions in vegetated aquatic systems. pp282-299 IN S.S. Bell, E.D.
McCoy and HR. Mushinsky (eds.). Habitat structure: the physical
arrangement of objects in space. Chapman and Hall, New York.

Herdendorf , CE. 1992. Lake Erie Coastal Wetlands: an overview. Journal of
Great Lakes Research 18:533-551.

Herdendorf, C.E., C.N. Raphael, and E. Jaworski. 1986. The ecology of the
coastal marshes of western Lake Erie: a community proﬁle. US Fish and
Wildlife Service Biological Report 85.

Hosmer, D.W. and S. Lemeshow. 2000. Applied Logistic Regression, 2nd
edition. John Wiley 8 Sons, Inc., New York, NY.

118

Houde, ED. 1994. Differences between marine and freshwater ﬁsh larvae:
implications for recruitment. ICES Journal of Marine Science 51:91-97.

Jacobsen, L. and S. Berg. 1998. Diel variation in habitat use by planktivores in
ﬁeld enclosure experiments: the effect of submerged macrophytes and
predation. Journal of Fish Biology 53:1207-1219.

Johnson, D.L., W.E. Lynch, and T.W. Morrison. 1997. Fish communities in a
diked Lake Erie wetland and an adjacent undiked area. Weltands
17(1):43-54.

Jude, DJ and J. Pappas. 1992. Fish utilization of Great Lakes coastal
wetlands. Journal of Great Lakes Research 182651-672.

Keast, A., J. Harker, and D. Tumbull. 1978. Nearshore ﬁsh habitat utilization
and species associations in Lake Opinicon (Ontario, Canada).
Environmental Biology of Fishes 3(2):173-184.

Krieger, K.A., D.M. Klarer, R.T. Heath, and CE. Herdendorf. 1992. A call for
. research on Great Lakes coastal wetlands. Journal of Great Lakes
Research 18:525-528.

Kushlan, J.A. 1981. Sampling characteristics of enclosure ﬁsh traps.
Transactions of the American Fisheries Society 110:557-562.

Leslie, J.K. and CA. Timmins. 1994. Ecology of young-of-the-year ﬁshes in
Severn Sound, Lake Huron. Canadian Journal of Zoology 72:1887-1897.

Lillie, RA. and J. Budd. 1992. Habitat architecture of Myriophyllum spicatum L.
as an index to habitat quality for ﬁsh and macroinvertebrates. Journal of
Freshwater Ecology 7:113-125.

Liston, CR, and S.L. Chubb. 1986. Relationships of water level ﬂuctuations and
ﬁsh. pp 121-140 IN Prince, H.H. and PM. D’ltri (eds.). Coastal Wetlands.
Lewis Publishers, Inc., Chelsea, Ml.

Minns, C.K., V.W. Cairns, R.G. Randall and J.E. Moore. 1994. An index of biotic
integrity (IBI) for ﬁsh assemblages in the littoral zone of Great Lakes’
Areas of Concern. Canadian Journal of Fisheries and Aquatic Science
51 :1804-1822.

Mittelbach, G.G. 1986. Predator-mediated habitat use: some consequences for
species interactions. Environmental Biology of Fishes 16:159-169.

119

Petering, R.W. and D.L. Johnson. 1991. Distribution of ﬁsh larvae among
artiﬁcial vegetation in a diked Lake Erie wetland. Wetlands 11:123-138.

Pihl, L. and R. Rosenberg. 1992. Production, abundance, and biomass of
mobile epibenthic marine fauna in shallow waters, western Sweden.
Journal of Experimental Marine Biology and Ecology 57:273-301.

Poe, T.P., C.O Hatcher, C.L. Brown, and D.W. Schloesser. 1986. Comparison
of species composition and richness of ﬁsh assemblages in altered and
unaltered littoral habitats. Journal of Freshwater Ecology 3(4):525-536.

Post, JR, and DJ McQueen. 1988. Ontogenetic changes in the distribution of
larval and juvenile yellow perch (Perca ﬂavescens ): A response to prey
or predators? Canadian Journal of Fisheries and Aquatic Sciences 45:
1820-1826.

Post, JR, and AB. Prankevicius. 1987. Size-selective mortality of young-of-
the-year yellow perch (Perca ﬂavescens): evidence from otolith
microstructure. Canadian Journal of Fisheries and Aquatic Sciences
44:1840-1847.

Raposa, KB. and CA. Oviatt. 2000. The inﬂuence of contiguous shoreline type,
distance from shore, and vegetation biomass on nekton community
structure in eelgrass beds. Estuaries 23(1):46-55.

Rothschild, B.J. 1986. Dynamics of marine ﬁsh populations. Harvard University
Press, Cambridge, MA 277pp.

Rozas, LP. and WE. Odum. 1988. Occupation of submerged aquatic
vegetation by ﬁshes: testing the roles of food and refuge. Oecologia
77:101-108.

Ruzycki, JR. and WA. Wurtsbaugh. 1999. Ontogenetic habitat shifts of juvenile
Bear Lake sculpin. Transactions of the American Fisheries Society
128:1201-1212.

Savino, J.F. and RA. Stein. 1982. Predator-prey interaction between largemouth
bass and bluegills as inﬂuenced by simulated, submersed vegetation.
Transactions of the American Fisheries Society 111:255—266.

Savino, J.F. and RA. Stein. 1989. Behavioral interactions between ﬁsh predators
and their prey: Effects of plant density. Animal Behaviour 37:311-321.

120

Sillett, KB. and SA. Foster. 2000. Ontogenetic niche shifts in two populations
of juvenile threespine stickleback, Gasterosteus aculeatus, that differ in
pelvic spine morphology. Oikos 91:468-476.

Simonovic, P.D., P. Garner, E.A. Eastwood, V. Kovac, and G.H. Copp. 1999.
Correspondence between ontogenetic shifts in morphology and habitat
use in minnow Phoxinus phoxinus. Environmental Biology of Fishes
56:1 17-128.

Skubinna, J.P., T.G. Coon, and TR. Batterson. 1995. Increased abundance
and depth of submersed macrophytes in response to decreased turbidity
in Saginaw Bay, Lake Huron. Journal of Great Lakes Research
21 (4):476-488.

Svanbaeck, R. and P. Ekloev. 2002. Effects of habitat and food resources on
morphology and ontogenetic growth trajectories in perch. Oecologia
131 :61-70.

Tidwell, J.H., S.D. Coyle, J. Evans, C. Weibel, J. McKinney, K. Dodson, and H.
Jones. 1999. Effect of culture temperature on growth, survival, and
biochemical composition of yellow perch Perca ﬂavescens. Journal of the
World Aquaculture Society 30(3):324-330.

Tonn, WM. and CA. Paszkowski. 1987. Habitat use of the central mudminnow
(Umbra limb and yellow perch (Parca ﬂavercens) in Umbra—Perca
assemblages: the role of competition, predation, and abiotic environment.
Canadian Journal of Zoology 65:862-870.

Turner, RE. 1988. Secondary production in riparian wetlands. Transactions of
the 53rd North American Wildlife and Natural Resources Conference. 491-
501.

VanderKooy, D.E., C.F. Rakocinski, and R.W. Heard. 2000. Trophic
relationships of three sunﬁshes (Lepomis spp.) in an estuarine bayou.
Estuaries 23:621-632.

Wahl, C.M., E.L. Mills, W.N. McFarland, and J.S. DeGisi. 1993. Ontogenetic
changes in prey selection and visual acuity of the yellow perch, Perca
ﬂavescens. Canadian Journal of Fisheries and Aquatic Sciences 50:743-
749.

Werner, E.E. and DJ. Hall. 1988. Ontogenetic habitat shifts in bluegill: the
foraging rate-predation risk trade-off. Ecology 69(5):1352-1366.

121

Werner, E.E., G.G. Mittelbach, D.J. Hall, and J.F. Gilliam. 1983a. Experimental
tests of optimal habitat use in ﬁsh: the role of relative habitat proﬁtability.
Ecology 64:1525-1539.

Werner, E.E., J.F. Gilliam, D.J. Hall, and G.G. Mittelbach. 1983b. An
experimental test of the effects of predation risk on habitat use in ﬁsh.
Ecology 64:1540-1548.

Winﬁeld, I.J. 1986. The inﬂuence of simulated aquatic macrophytes on the

zooplankton consumption rate of juvenile roach, Rutilus rutilus, and perch,
Perca ﬂuviatilis. Journal of Fish Biology 29:37-48.

122

CHAPTER 3: EFFECTS OF VEGETATION AND FISH DENSITY ON THE
GROWTH OF YOUNG OF THE YEAR BLUEGILL

ABSTRACT

Although wetland environments are important habitats for juvenile bluegill,
young of the year (YOY) bluegill are not uniformly distributed within wetlands,
and often inhabit vegetated areas in preference to open water. The goal of this
study is to determine if YOY bluegill habitat preferences inﬂuence ﬁsh success as
measured by their growth rates. Additionally, effects of ﬁsh density and
macroinvertebrate and zooplankton resource levels on bluegill growth are
discussed. Two enclosure experiments were conducted to examine bluegill
growth rates under different habitat conditions. The ﬁrst experiment compared
bluegill growth rates in emergent vegetation, submersed vegetation, and open
water habitats. The second experiment evaluated bluegill growth rates at high
and low ﬁsh densities in submersed vegetation and open water. Water quality
within the enclosures was monitored daily, zooplankton populations were
monitored weekly, and macroinvertebrates were sampled at the beginning and
end of the experiment. Fish growth and diet were assessed at the end of the
experiment. Few differences in zooplankton abundance were found among
habitat types. Macroinvertebrates were more abundant in vegetated areas than
open water. Bluegill fed on both macroinvertebrates and zooplankton in all
habitats; however, they consumed more zooplankton biomass and less
macroinvertebrate biomass in open water than in vegetated habitats in
experiment one. Fish grew slower in emergent habitat than submersed

vegetation or open water in experiment one. In experiment two, habitat type had

123

no impact on growth rates at low ﬁsh density, however at high densities YOY
bluegill grew better in submersed vegetation than in open water. In both
submersed vegetation and open water habitats, bluegill grew faster at lower
densities. Bluegill also had faster growth rates in experiment one where food

resources were more abundant.

124

INTRODUCTION

Wetland environments serve as important eany life history habitats for
many ﬁsh species, including bluegill (Lepomis macrochirus). Potential beneﬁts of
wetland areas for juvenile bluegill include the high primary productivity of these
environments, reduced risk of predation, and the rich zooplankton and benthos
food source that wetlands provide (e.g. Mittelbach 1984, Rozas and Odum 1988,
Turner 1988, Werner and Hall 1988, Schramm and Jirka 1989, Jude and Pappas
1992, Cardinale et al. 1998). However, the abundance and distribution of young
of the year (YOY) bluegill are not uniform within wetlands (see Chapter 2) or
among wetland areas (see Chapter 1). The structure that aquatic macrophytes
provide inﬂuences the distribution of YOY bluegill. In general, they are more
likely to be found in submersed vegetation habitats compared to areas of
emergent vegetation or open water (see Chapter 2). The goal of this study is to
determine if YOY bluegill habitat preferences inﬂuence individual growth rate.

The habitat preferences of a species should be inﬂuenced by the ﬁsh‘s
ability to forage, grow, and avoid predation. Optimal foraging theory predicts that
young ﬁsh should choose habitat patches that maximize energy intake per unit
time, because growth rates, particularly in the ﬁrst year, must be maximized to
survive to breeding age (Post and Prankevicius 1987, Tonn and Paszkowski
1987, Houde 1994). Studies have shown that the foraging efﬁciency of individual

ﬁsh may decline with changes in the vegetation community structure and

125

increased habitat complexity (Winﬁeld 1986, Diehl 1988, Gotceitas 1990a,
Dionne and Folt 1991), but the protection that vegetation provides to young ﬁsh
from predation increases with vegetation complexity (Crowder and Cooper 1982,
Werner et al. 1983b, Lillie and Budd 1992, Jacobsen and Berg 1998). As a
result, a ﬁsh should choose the environment that allows it to minimize the

predation risk/foraging rate ratio (Gotceitas 1990b).

Previous studies on bluegill have shown that bluegill switch habitats and E

rm

diet in response to food availability, predation risk, competition, and experience in
a habitat type (Wemer and Hall 1976, 1977, 1979, Mittelbach 1981, 1984,
Werner et al. 1981, 1983a). There are tradeoffs in the energetic return of feeding
between open water and vegetated habitats. Bluegill feeding on zooplankton in
open water environments spend little time searching and handling prey, but the
small prey size makes each item’s overall proﬁtability low. In vegetated
environments, bluegill feeding on macroinvertebrates spend longer periods in
search for prey, due to the increased habitat structure, and in handling prey, due
to their increased size, both of which act to lower prey proﬁtability (Mittelbach
1981 ). Adult bluegill have been shown to select prey to optimize their energy
intake (Werner and Hall 1974), and switch habitats from vegetated areas to open
water as the relative proﬁtability of the different habitats change (Mittelbach
1981). In general, adults will grow faster in areas of vegetated habitat, but move
to open water in response to competition (Werner and Hall 1977, 1979). Smaller
bluegill, however, may remain in vegetated habitats where they feed less

optimally and are subject to more intense competition. Foraging in open water

126

may be more proﬁtable, but the increased protection from predation in vegetated
habitats outweighs the beneﬁt of open water feeding (Mittelbach 1981, 1984,
1988, Werner et al. 1983a, Werner and Hall 1988, Gotceitas 1990b). However,
in areas with high populations of zooplanktivorous ﬁsh and low abundances of

zooplankton, juvenile bluegill may grow faster in vegetated areas compared to

4..

areas of open water (Belk 1993).

‘1

While there have been several studies examining bluegill use of different

I“;
. A .A

habitats, most of these studies have focused on shifts in diet and energy
consumption between habitats. Few studies have examined growth rates of
bluegill conﬁned to different habitats, and of these, all focused on larger ﬁsh (age
one and older) and evaluated the impacts of competition on growth rates rather
than potential effects of habitat type. Fish growth in their ﬁrst year of life is critical
in determining the length of time they spend vulnerable to predators, the over
winter survival rate, and year class strength (Mittelbach 1984, Houde 1994). The
goal of this study is to examine the impacts of selecting different habitat types on
the growth rates of YOY bluegill. Speciﬁc objectives of this study are to: 1)
compare the growth of YOY bluegill in different habitat types, 2) compare the
abundance and types of food resources in different habitats, 3) compare the diet
of YOY bluegill in different habitats, and 4) determine if ﬁsh density inﬂuences

growth rates of YOY bluegill in different habitats.

127

METHODS

Study Site Description

One wetland and one pond located at the Experimental Lakes Area at
Michigan State University, Ingham County, Michigan were selected as study
sites. The ﬁrst experiment took place in a 0.40 hectare wetland. This wetland
had a maximum water depth of 1.3 m. The plant community within the wetland
was dominated by cattails (Typha spp.), Eurasian water milfoil (Myriophyllum
spicatum), and common wateniveed (Elodea canadensis). The only ﬁsh captured
within the wetland were fathead minnows (Pimephales promelas). Due to low
oxygen concentrations and declining water levels in the wetland, the second
experiment was conducted in a larger pond at the Experimental Lakes Area.
Pond 4 is approximately 5 hectares in size, with a maximum water depth of 2.4
m, and average depth of 1.8 m. Vegetation within the pond is dominated by
common reed (Phragmites australis), Eurasian water milfoil (Myriophyllum
spicatum), and common waterweed (Elodea canadensis). The ﬁsh community
within the pond is dominated by sunﬁsh (Lepomis spp.), brook sticklebacks

(Culaea inconstans), and fathead minnows.
Experiment 1

A total of 24 enclosures were used in experiment 1. Each square

enclosure measured 1 m on a side, and was constructed with a polyvinyl chloride

128

r'r-a-rr—f

pipe frame (2.5 cm in diameter) covered with 500 pm mesh made of plastic
sidewalk underliner. The enclosures were ﬁxed into the substrate with stakes at
least 10 cm deep. The bottom of the sides of the enclosures were weighted with
chain, and buried in the sediment to prevent ﬁsh from escaping. The tops of the
enclosures were covered with orange plastic construction fencing to prevent
avian predation. The enclosures were placed within the wetlands so that eight
enclosures were located in emergent vegetation, eight were located in
submersed vegetation, and the remaining eight enclosures were located in open
water habitat. Water depths where enclosures were placed ranged from 28 to 64
cm, with a mean of 45 cm. Two enclosures in each habitat type were randomly
selected to serve as controls and were not stocked with ﬁsh. The remaining
enclosures were stocked with YOY bluegill (size ranged from 17 to 23 mm,
standard length) captured from a local lake on July 23 - July 25, 2002. Before
stocking of ﬁsh, each enclosure was extensively seined to remove any ﬁsh that
may have been inadvertently captured while the enclosures were installed.

A YSI model 55 dissolved oxygen meter was used to sample water
temperature and dissolved oxygen concentrations in the enclosures daily, before
10 am. Readings were taken at approximately two-thirds of the water depth.

A modiﬁed Gerking sampler (Gerking 1957) was used to sample

macroinvertebrates in the enclosures at the beginning and end of the experiment.

Each sample covered 491 cm2, and water depth at the time of each sample was
measured for later calculation of water volume and macroinvertebrate density.

The Gerking sampler was placed over a random patch of open water, emergent,

129

.1

1"

or submersed vegetation within the enclosure, and quickly pushed down to the
sediment surface. Vegetation, if present, was cut at the substrate surface, and
the bottom of the sampler was closed. The sampler was then inverted and the
sample was ﬁltered through a 253 pm mesh nylon netting. Macroinvertebrates
were picked from the samples and preserved in a 90% ethanol solution. The
plant material was placed into sealed plastic bags and frozen for later analysis.
In the lab, plant samples were allowed to thaw, washed, and examined under a
microscope for associated macroinvertebrates, which were hand picked from the
plant material. All macroinvertebrates from the ethanol sample in the ﬁeld and
associated plant material were then identiﬁed to order, and measured for body
length and head capsule width using a dissecting scope and digitizing tablet.
Plants were identiﬁed to species according to Fassett (1985), and dried in an
oven at 80°C for 24 hours and weighed to the nearest 0.001 g to estimate
macrophyte biomass. A subsample of 20 macroinvertebrates from each order
were measured for body length and head capsule width, and then dried in an
oven at 105°C for four hours (Smock 1980). After drying, the macroinvertebrates
were weighed to the nearest 0.0001 gram. Data from these macroinvertebrates
were used to construct regression relationships between head capsule width and
dry weight. Head capsule width was measured for all macroinvertebrates
captured, and used to calculate macroinvertebrate biomass within the enclosures
from the regression relationships. Macroinvertebrates from the order gastropoda
were not included in calculations of overall macroinvertebrate biomass because

they did not represent a food resource for the YOY bluegill.

130

’77--

A 76 mm diameter tube sampler was used to sample zooplankton within
the entire water column. Zooplankton were sampled at the beginning and end of
the experiment, as well as weekly throughout the study. Two samples with the
tube sampler were taken during each sampling event. Water depth was
recorded for each sample for later calculation of water volume and zooplankton
density. The samples were combined and ﬁltered through an 80 um ﬁlter, and
preserved in 20 ml of a 90% ethanol solution. A 1 ml sedgewick rafter counting
cell was used for counting zooplankton and estimating zooplankton density.
Zooplankton within the counting cell were examined under a dissecting scope at
40x power, and classiﬁed as Daphnia spp., Bosmina spp., Chydorus spp.,
Ceriodaphnia spp., or calanoid, cyclopoid, or nauplii copepods according to
Balcer et al. (1984). Rotifers and other species of small zooplankton were not
counted. One ml aliquots were subsampled until a minimum of 100 zooplankton
were counted. Body length of the ﬁrst 20 zooplankton within a classiﬁcation
group was measured using a digitizing tablet according to procedures outlined in
Culver et al. (1985). Once 100 zooplankton were counted the remainder of the
counting cell was examined, and subsampling for that sample was discontinued.
Body length-weight regression relationships from Culver et al. (1985) were used
to calculate zooplankton biomass.

At the end of the experiment, August 19, 2002, seining was conducted to
remove the stocked bluegill from the enclosures. The standard length of all ﬁsh
was recorded, and ﬁsh were weighed to the nearest 0.0001 gram. Fish were

preserved in a 90% ethanol solution. In the lab, the sagittal otolith was dissected

131

and used to measure growth rates. An Optimas imaging system was used to
measure the total radius of the otolith, and the radius of the otolith before the last
seven, fourteen, and twenty-one rings were deposited. Growth rates of ﬁsh were
calculated according to procedures for the body proportional hypothesis outlined
in Francis (1990).

Stomachs were removed from the YOY bluegill for examination of diet.
Dissected stomachs were weighed to the nearest 0.00001 gram, and contents of
the stomachs were identiﬁed under a dissecting scope. Macroinvertebrates were
identiﬁed to order, and head capsule width was measured with a digitizing tablet.
Zooplankton were classiﬁed as Cladocerans or copepods, and body length of

zooplankton was measured using the digitizing tablet.

Experiment 2

A total of 18 enclosures were installed along the littoral edge of Pond 4, in
areas of submersed vegetation and open water (9 enclosures each). Design of
the enclosures was the same as for those used in experiment 1. The mean
water depth where enclosures were installed was 49.8 cm, and ranged from 39
to 61 cm. Before stocking of ﬁsh, each enclosure was extensively seined to
remove any ﬁsh that may have been inadvertently captured while the enclosures
were installed. Enclosures were stocked with YOY bluegill (size ranged from 21
to 28 mm, standard length) from a local lake on August 21, 2002. Three
randomly assigned enclosures in open water and submersed vegetation were

stocked with O, 5, or 15 YOY bluegill.

132

Following the sampling procedures detailed for experiment one, dissolved
oxygen and temperature readings were taken each morning. Macroinvertebrates
were sampled using a modiﬁed Gerking sampler at the beginning and end of the
experiment. Zooplankton samples were taken with a tube sampler weekly. All
macroinvertebrate and zooplankton samples were processed as described for
experiment one.

Fish were harvested from the enclosures after 23 days on September 12,
2002. All ﬁsh were preserved in 90% ethanol solution. Standard length of all ﬁsh
was measured to the nearest mm, and all ﬁsh were weighed to the nearest
0.0001 g. The diet and growth rates of all ﬁsh were sampled according to

procedures outlined in experiment 1.

Data Analysis

Repeated Measures Analysis of Variance (a = 0.05) was used to analyze
differences in water temperature, dissolved oxygen, zooplankton and
macroinvertebrate density, and biomass among the habitat types and ﬁsh
densities over time (Littell et al. 1991). Analysis of Variance (a = 0.05) was used
to analyze differences in mean ﬁsh growth rates, stomach weights, and diet

composition. All data analysis was done using the SAS system.

133

RESULTS

Experiment 1

Muskrat activity and low oxygen levels within the wetland resulted in the
destruction of enclosures or loss of ﬁsh in three of the eight enclosures located in
emergent vegetation, and two of the eight enclosures located in open water and
submersed vegetation. As a result, all comparisons between enclosures with ﬁsh
and those without ﬁsh were made by comparing only enclosures where bluegill
were captured at the end of the experiment to enclosures that were not stocked

with YOY bluegill.

WATER QUALITY

Water temperatures varied throughout the duration of experiment one,
ranging from 173°C to 259°C, with an average water temperature of 227°C
(Figure 3.1). Water temperatures were similar among habitat types on any given
day (F221 = 2.09, p = 0.1492). Dissolved oxygen levels also varied daily
throughout the experiment (Figure 3.1), ranging from 0.14 mg/I to 12.25 mg/I,
with an average of 3.05 mg/I. Differences in dissolved oxygen concentrations
among the habitat types were observed (F221 = 29.13, p < 0.0001), with highest
concentrations in open water habitats, ranging from 0.47 to 9.91 mg/I, with an
average of 3.76 mg/l. Submersed vegetation had the next highest oxygen

concentrations, ranging from 0.49 to 10.31 mg/l, with an average of

134

 

A. Water Temperature

 

 

 

 

 

26
8 25 -
0., 24
93 l
g 23 -
a 22 .
2' 21 q
'1) 20 1 - -O- -ernergent
a3 19 i —I—openwater
H
g 13 l —A— submersed
17 T T T T T T T T T j T T T T T T T T T T T T T
N “a <34 9.5 0.33.33.99
Day
B. Dissolved Oxygen
10
e - -0- -errergent
g) 8 _ —l—openwater
: —t— submersed
8: 6
>. 1 ll
6 _ l
'o 4 '1 | g
3
8 2 - ‘ 5:}
,9 ‘ I {J- i
D
0 l T T T T T f T T

 

 

 

~2<o«:<\<beee¢~e

Day

Figure 3.1: Water temperature (A) and dissolved oxygen concentration (B)
(mean :t 2xstandard error) in experiment one enclosures located in emergent

vegetation, open water and submersed vegetation.

135

3.08 mg/l. Enclosures located in emergent vegetation had the lowest oxygen

concentrations, ranging from 0.14 to 12.25 mg/I with an average of 2.31 mg/I.

ZOOPLANKTON

In general, the zooplankton communities were dominated by copepods
(Figure 3.2). Nauplii made up the highest percentage by density of the
zooplankton community, averaging 38.8%, while cyclopoid copepods averaged
21 .2%, and calanoid copepods averaged 7.6%. The Cladocerans were
dominated by Chydorus spp. (12.6%), followed by Ceriodaphnia spp. (9.6%),
Daphnia spp. (7.7%) and Bosmina spp. (2.5%). There was little difference in the
percent contribution of each zooplankton taxa to overall zooplankton density over
the course of the experiment (Figure 3.2).

Cladoceran density varied over time (F516 = 5.82, p = 0.0030) ranging
from 0.02 to 1.53 Cladocerans per cm3 of water, with an average of 0.32
Cladocerans per cm3 of water (Figure 3.3). The pattern of variation in Cladoceran
density over time was not the same among all three habitat types (F220 = 4.04, p
= 0.0336), and there were differences in overall Cladoceran density among
habitat types (F1032 = 2.19, p = 0.0456). Submersed vegetation had higher
densities than emergent (t = 2.58, p = 0.0174 ) or open water habitat (t = 2.48, p
= 0.0219). There was no difference in Cladoceran density among emergent

vegetation and open water (t = 0.11, p = 0.9164). The pattern of variation

136

100% -

80% I Nauplii
Cyclopoid

[:1 Calanoid

[II Daphnia
Chydorus

El Ceriodaphnia
l Bosmina

o:
o
o\°

40%

% Composition

20%

 

 

 

0%
start week 1 week 2 week 3 week 4 end

Time

Figure 3.2: Taxonomic composition (by density) of zooplankton during

experiment one.

137

A. Cladocerans

 

1.2
- + - emergent ..
1.0 ~ —l—open water
—A— submersed
0.8 2

 

 

 

 

 

 

 

 

 

 

 

 

0.2 2
0.0 T I l I T
start week1 week2 week 3 week4 end
Time
B. Copepods
2.5
- + - emergent
20 _j .- —I-—open water
A —A— submersed
”E
o 2 9 .
s 1-5 ~. ,I '~-.
3 I ‘ ~ . ' “' T ‘ \
._ t, c ' / Q
g 1 0 ~ {~ § _ ~ . I / N
8 K \ / \ ‘
. \ __ ’
0.5 ~
0.0 I I I I I
start week1 week2 week3 week4 end
Time

Figure 3.3: Density (mean :I: 2x standard errors) of A) cladocerans and B)

copepods sampled in enclosures in different habitat types in experiment one.

138

 

in cladoceran density over time was similar between enclosures where ﬁsh were
present and absent (F5,15 = 0.90, p = 0.5038), and overall cladoceran density did
not differ between enclosures with and without ﬁsh (F120 = 0.00, p = 0.9638).

The biomass of cladocerans (Figure 3.4), remained fairly constant over
the duration of the experiment (F545 = 1.86, p = 0.1583). The variation in
cladoceran biomass over time was similar among habitats (F1032 = 1.14, p =
0.3639), and a similar amount of cladoceran biomass was found in all habitat
types (F220 = 2.47, p = 0.11031). No differences were found in how cladoceran
biomass varied over time in enclosures with ﬁsh compared to those without (F515
= 1.92, p = 0.1468), or in the overall amount of cladoceran biomass in enclosures
with or without ﬁsh (F120 = 0.02, p = 0.8779).

Copepod density varied over time (F546 = 7.09, p = 0.0011), ranging from
0.107 to 2.83 copepods per ml of water, with an average of 0.882 copepods per
ml of water (Figure 3.3). Although the pattern of variation in copepod density
was similar among habitat types (F1032 = 1.49, p = 0.1876), there were
differences in overall copepod density among habitats (F220 = 4.42, p = 0.0258).
With a mean of 1.134 copepods per ml of water, emergent vegetation had higher
densities of copepods than open water (mean of 0.649) or submersed vegetation
(mean of 0.870). Similar densities (F120 = 1.03, p = 0.3234), and patterns of
variation in copepod density over time (F546 = 0.93, p = 0.4893) were found in

enclosures with and without YOY bluegill.

139

A. Cladocerans

 

 

 

 

 

 

 

 

 

3.0 ~
—-o—emergent
2-5 i - + — open water .-
3) 20 ﬂ - - -A- - - submersed
(D
‘8
E 1.5 -
.9
cm 1'0 - xxx,
0.5 I
0.0 I
start week 1 week 2 week 3 week 4 end
B. Copepods Time
0.20 —
-- -e- - - emergent __ __
—I—open water
g 0-15 f -+—submersed
93
(D
8
E 0.10 —
.9
m
0.05 -
0.00 I
start week 1 week 2 week 3 week 4 end
Time

Figure 3.4: Biomass (mean pgll :t 2x standard error) of A) cladocerans and B)

copepods sampled in enclosures in different habitat types during experiment one.

140

Copepod biomass (Figure 3.4) varied over the duration of experiment one

(F515 = 9.00, p = 0.0003), in a similar pattern among the different habitat types

(F 1032 = 0.70, p = 0.7189). However, with a mean of 0.078 :I: 0.070 g/l, there was
a greater biomass of copepods (F220 = 7.06, p = 0.0048) sampled in emergent
vegetation than in open water (0.038 :t 0.027 g/l) or submersed vegetation (0.046
.t 0.024 g/l). Similar patterns in variation of copepod biomass over time (F515 =
0.62, p = 0.6862) were found in enclosures with and without ﬁsh. Additionally, no
differences were found in copepod biomass between enclosures with and without

ﬁsh (F120 = 0.01, p = 0.9354).

MACROINVERTEBRATES

Macroinvertebrates from twelve different taxa were sampled from
experiment one enclosures (Table 3.1). Overall, macroinvertebrates were
dominated by gastropods and hemipterans; however, ephemeroptera and
odonates were also abundant. Several taxa were only found in the emergent
vegetation habitat type, including amphipoda and aranea. There was only one
taxon, coleoptera, which was present in open water or submersed vegetation
habitat and not in emergent vegetation.

Macroinvertebrate densities were similar at the beginning and end of the
experiment (F120 = 0.26, p = 0.6140), averaging 440 macroinvertebrates/m3 at
the start of experiment one, and 403 macroinvertebrates/m3 at the end of
experiment one (Figure 3.5). There were no differences in the pattern of change

in macroinvertebrate density from the start to the end of the experiment in

141

“if ‘

 

SHAWN No 30m swam: mmﬂwo #0 $63 vamN them 0 o o :6 o o www.mEnaw
F ram mmﬁmm N Pam mm FHmNF mo“ 5 no Exam 0 v 36 o FNHm o o 553 chO

vamN \Imﬁmr NV ENE. No 2:. F _. mwNHmmN mhﬂmv wNHoF 3am FNHm o ooﬁnm OVHNF #:09me

 

E 80 LE Em: 80 Eu ea 86 £8 .8 ea. 62

 

6530205 n E cam .8968 u one .8525:
u .2 6.29:5; u Em; .mooaobmmm u 93 .mLcEBcEccac n cam .9636 u QB 66886“. n own .m_onEc=oo u Eco
6.63028 u Be .838 u 9m .mcanaEm u aEm can; .39: Exam; E2956 E 6980. $58.05 95 EmEtcaxm E

umaEmm 35955 2:55me E96t6 E0: m2m5¢t¢>c_o.omE Co EcumSou Emucﬁm H NEE bacon cows. 3.” min...

142

 

1 500

 

 

 

 

 

 

 

HABITAT] "T
D emergent E
* submersed s
«7‘ _ ‘ open water 0 E
g 1000 E S 1
$5 0 "TI;
t S T I
‘7) ‘1‘: ——L .0
E 500— I —
a _ I
O l ' .2 ' ' are:
beginning end
TIME

Figure 3.5: Density of macroinvertebrates sampled in enclosures from different

habitats at the beginning and end of experiment one.

143

enclosures located in different habitats (F220 = 0.16, p = 0.8550), or in enclosures

with or without YOY bluegill (F120 = 0.15, p = 0.7034). However, with a mean of

610 macroinvertebrates/m3, there was a higher density of macroinvertebrates

(F220 = 5.06, p = 0.0166) in emergent vegetation habitat then in open water (252

macroinvertebrates/m3) or submersed vegetation (408 macroinvertebrates/m3).

There was not a difference in the density of macroinvertebrates in enclosures

with YOY bluegill compared to those without YOY bluegill (F120 = 1.57, p = g

0.2243). L.
Macroinvertebrate biomass ranged from a mean of 0.814 :I: 0.311 mg /l at

the start of the experiment, to 0.283 mg/I at the end of the experiment (Figure

3.6); however, this difference was not statistically signiﬁcant (F120 = 4.33, p =

0.0506). Additionally, the pattern of macroinvertebrate biomass measured at the

beginning to the end of the experiment was similar in each of the habitats (F220 =

1.70, p = 0.2072) and in enclosures with and without YOY bluegill (F120 = 1.37, p

= 0.2549). Similar amounts of macroinvertebrate biomass were measured in

enclosures from all habitat types (F220 = 1.69, p = 0.2105), and enclosures with

or without ﬁsh (F120 = 0.78, p = 0.3872).

FISH GROWTH

The mean growth rate of ﬁsh within an enclosure ranged from 0.347
mm/day to 0.700 mm/day with an overall mean of 0.523 :I: 0.103 mm/day (Figure
3.7). With a mean growth rate of 0.271 mm/day, bluegill from enclosures with

emergent vegetation had slower growth rates (F28 = 11.046, p = 0.005) than

144

 

6 .
Time 7 I ..'

59 beginning b g
end e

.h
T
l

BIOMASS (mg/l)
O)
l
l

N
l
]——-«
J

 

 

 

 

 

e b e e
1- & _ —
b . L
o 2:13 ..-I Til
9x 5 at
3th (0696 069‘
66‘ 96° ch
HA B | TAT

Figure 3.6: Biomass (mg/l) of macroinvertebrates at the beginning and end of
experiment one, sampled in enclosures located in emergent vegetation, open

water, and submersed vegetation habitat.

145

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.8 I I I
’5? o 7 — I —
"E
E
v 0.6- T —
E
l
O ._ l _
m 0.5
(D
If.
I :- -
0.3 I I l
K t
900 90 «1&9
3106‘ 6606\0 OQGO
HABITAT

Figure 3.7: Growth (mm/day) of YOY bluegill in experiment one enclosures

located in different habitat types.

146

7‘EF1

 

bluegill from enclosures located in open water (mean = 0.410 mm/day) or

submersed vegetation habitats (mean = 0.393 mm/day).

FISH DIET

Weights of YOY bluegill stomachs ranged from a low of 0.02304 9 to
0.11811 9 (Figure 3.8). Stomach weight was similar among ﬁsh harvested from
enclosures located in different habitats (F23 = 0.117, p = 0.891).

The diet of bluegill in all habitat types was numerically dominated by
zooplankton, both cladocerans and copepods (Figure 3.9). However, the main
contributors to biomass within the stomachs of bluegill were macroinvertebrates,
with high amounts of diptera and ephemeroptera biomass consumed (Figure
3.9). To compare bluegill diet among habitat types, prey items were grouped as
macroinvertebrates and zooplankton (Figure 3.10). There was a signiﬁcant
difference in the biomass of zooplankton within the stomachs of ﬁsh from open
water enclosures compared to other habitat types (F23 = 4.750, p = 0.044). With
a mean zooplankton biomass of 81 pg per stomach, there was a greater amount
of zooplankton biomass dissected from the stomach’s of bluegill from open water
habitat than from emergent or submersed vegetation (means = 25 and 22 pg per
stomach respectively). With a mean of 171 pg per stomach, there was less
macroinvertebrate biomass (F23 = 5.236, p = 0.035) in the stomachs of bluegill
from open water enclosures than from submersed or emergent enclosures

(means were 537 and 769 pg per stomach respectively).

147

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.08 .. I I
A T
C)
'—
35 0.07— I —
LTJ
E
I
2
E 0.06- l -
O .l.
'5 ..
0.05 l l l
O 3‘
90° 60 ’b‘
0&6‘ 960(06‘ Q60
HABITAT

Figure 3.8: Comparison of YOY bluegill stomach weight (9) among different

habitat types in experiment one.

148

A. Density

 

 

 

 

120
A 100 I emergent
5 -
8 open water
g 80 CI submersed
‘05
g 60 I
2*
7, 40
C
a)
D 20 " I

0 Ti";- M

AMP DIP CLA COP EPH HEM ODO
Taxa

B. Biomass

1 600

 

1400 ~ T I emergent
1200 - El open water
1000 I D submersed
800 ~
600 4
400 I
200 -
0 _J_é . JILL , L

AMP DIP CLA COP EPH HEM ODO

Taxa

 

Biomass (pg/stomach)

 

 

 

 

 

Figure 3.9: Density (mean # / stomach i 2xstandard error) (A) and biomass
(mean pg / stomach :t 2 standard error) (B) of organisms dissected from
stomachs of bluegill in experiment one enclosures from different habitats. Where
AMP = amphipoda, DIP = diptera, CLA = Cladocera, COP = copepod, EPH =

ephemeroptera, HEM = hemiptera, and ODO = odonata.

149

inn-.91
m

f...

A. Zooplankton

BIOMASS (pg/stomach)

B. Macroinvertebrates

BIOMASS (pg/stomach)

Figure 3.10: Biomass

from bluegill stomachs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 50 I I I
100— —
50 — I I I —
O I T 1
a 0‘
(90° 0‘90 $7"
0‘00 996‘ 0900
HABITAT
1200 r l I
1000 '- I ‘4
800 — 'l' ‘
600 — “
400 - l "
200 _ P
E3
0 l l l
\ 6 ‘
«(0‘96“ «(0‘90 «9‘6
0 e)9 096
HABITAT

of A) zooplankton and B) macroinvertebrates dissected

in experiment one enclosures located in different habitats.

150

Experiment 2
WATER QUALITY

Water temperatures varied throughout the duration of experiment two,
ranging from a low of 21 .8°C to a high of 259°C, with an average water
temperature of 23.4°C (Figure 3.11). However, on any given day during the
experiment, the water temperature in submersed vegetation and open water
enclosures was similar (Fm; = 0.02, p = 0.8945). Dissolved oxygen
concentrations also varied daily throughout the experiment (Figure 3.11), ranging
from a low of 2.43 mg/I to a high of 10.11 mg/I, with an average of 6.15 mg/l.
Mean oxygen concentrations were similar (Fm; = 4.29, p = 0.0549) in open
water (6.41 mg/l) and submersed vegetation habitats (5.89 mg/I) throughout the

study.

ZOOPLANKTON

In general, the zooplankton communities within the enclosures were
dominated by copepods (Figure 3.12). Nauplii had the highest density in all
habitats and levels of ﬁsh density, followed by cyclopoid and calanoid copepods.
The cladoceran community was dominated by Daphnia spp., Chydorus spp., and
Bosmina spp., with very few Ceriodaphnia spp (Figure 3.12). There was little
difference in the composition or relative contribution of different zooplankton taxa

to the total density of zooplankton over the course of the experiment.

151

 

 

A. Water Temperature

 

27
+ openwater
—a— submersed
26 ~
25 ~

Water Temp (°C)
'0
&

 

 

 

23~
22«
21 T T T T T T T T T T T T T T T T T T 7 T T
12345678910111213141516171819202122
Day

B. Dissolved Oxygen

 

12
+openwater

1° 1 —.— submersed
.8; a-
P /
C) V-
a: \ \

I \ ‘ /

g 6‘ , I \ 4 I \
3 / x ’ ' I
§ ‘ I ‘ ’
.9 4 ‘" I
O \ /

2 . i

o T T T T T T T f T ﬂ 1 T 1 T T I T T T 7 I

 

 

 

12 3 4 5 6 7 8 910111213141516171819202122
Day

Figure 3.11: Water temperature (A) and dissolved oxygen concentration (B)
(mean :t: 2xstandard error) in experiment two enclosures located in open water

and submersed vegetation.

152

 

 

'5. 80 ~
g I Nauplii
D 3 III] Daphnia spp.
E 60 ‘ :3 Cyclopoid
E I El Chydorus
T: 40 . III] Ceriodaphnia
é ‘ El Calanoid
- l I Bosmina
g 20 ~
o\
0 ,,

   

1 8 15
Day of Experiment

Figure 3.12: Taxonomic composition (by density) of zooplankton during

experiment two.

153

Cladoceran density varied little over the duration of experiment two
(Figure 3.13, F312 = 1.85, p = 0.1918). The pattern in variation of cladoceran
density over time was similar in submersed vegetation and open water (F312 =
0.61, p = 0.6228), and among different ﬁsh densities (F312 = 0.36, p = 0.7844).
Additionally, habitat type did not affect the pattern of variation in cladoceran
density among different densities of stocked bluegill (F1,14 = 4.19, p = 0.0599).
Little variation in cladoceran density was found between habitat types (F144 =
1.38, p = 0.2593), and ﬁsh densities (F114 = 0.32, p = 0.5808).

The biomass of cladocerans within the enclosures also varied over time
(Figure 3.14, F343 = 3.81, p = 0.0370). The pattern in change of cladoceran
biomass over time was similar between open water and submersed vegetation
(F113 = 0.70, p = 0.5677), and overall biomass was similar between the two
habitats (Fm 5 = 0.08, p = 0.7767). Changes in cladoceran biomass over time
varied among ﬁsh densities (F343 = 4.04, p = 0.0311), although overall biomass
was similar among enclosures stocked with 0, 5, or 15 YOY bluegill (F1_15 = 0.55,
p = 0.4716).

Copepod density was fairly constant during experiment two (Figure 3.15,
(F312 = 0.86, p = 0.4881), with both habitats (F3_12 = 1.28, p = 0.3243) and all
levels of ﬁsh density (F342 = 0.70, p = 0.5725) showing similar patterns in
variation in copepod density over time. Additionally, habitat type did not affect
the pattern of variation in copepod density among different densities of stocked

bluegill (F1,14 = 0.66, p = 0.4287). There were similar densities of copepods in

154

A. Open Water
70 —
50 «

404
30~

Density (#ll)

20«

 

10~

 

 

 

1 8 15 22
Day of Experiment

7o—
60«
50«
401
30«

Density (#ll)

204
1o-

 

 

 

0

 

Day of Experiment

Figure 3.13: Mean density (#II :I: 2xstandard error) of cladocerans in enclosures

located in A) open water and B) submersed vegetation during experiment two.

155

A. Open water

400 —
350 .
300 —
V 250 -
200 ~
150 —
100 ~
50 —
0

9")

Biomass

 

 

 

Day of Experiment
B. Submersed Vegetation

400 - ——+——O
350 ~ “r __.__5
300“ ---A---15
.. 250 -
200 ~
150 ~
100 4
sol

0

9")

 

Biomass

 

 

 

 

 

Day of Experiment

Figure 3.14: Biomass (mean :t 2xstandard error g/l) of cladocerans sampled in
enclosures stocked with different densities of bluegill in A) open water and B)

submersed vegetation during experiment two.

156

A. Open Water

200 ~

Denslty (#Il)

 

 

 

 

 

1 8 15
Day of Experiment

B. Submersed Vegetation

 

200- —-o—0
—-I—--5
150~ ---A---15
i 120 -
5‘
2
q, 80—
D
40 -
0 I

 

 

 

Day of Experiment

Figure 3.15: Mean density (#/l :l: 2xstandard error) of copepods in enclosures

located in A) open water and B) submersed vegetation during experiment two.

157

enclosures located in different habitats (F144 = 0.05, p = 0.8234) and with
different densities of bluegill (F1_14 = 0.06, p = 0.8126).

The biomass of copepods varied little over the duration of experiment two
(Figure 3.16, F343 = 0.59, p = 0.6349). Patterns in variation of copepod biomass
were similar among the different habitat types (F343 = 1.34, p = 0.3055) and
stocked densities of ﬁsh (F313 = 0.42, p = 0.7412). Copepod biomass was similar
in submersed vegetation and open water (F1,15 = 0.13, p = 0.7228), and in
enclosures stocked with different densities of YOY bluegill (F145 = 0.64, p =

0.4347).

MACROINVERTEBRATES

Macroinvertebrates from eight different taxa were sampled from
experiment two enclosures (Table 3.2). Macroinvertebrates were dominated by
trichoptera; however, odonata and ephemeroptera were also abundant. Two
taxa, gastropoda and hemiptera, were sampled only in enclosures that were not
stocked with ﬁsh.

Macroinvertebrate density (Figure 3.17) was similar at the beginning and
end of experiment two (means = 70 t 131 and 56 :l: 102 macroinvertebrates / m3
respectively, F1,“ = 0.15, p = 0.7014). The pattern of change in
macroinvertebrate density over time was the same for both habitat types (F 1,14 =
0.03, p = 0.8544), all levels of ﬁsh density (F214 = 0.69, p = 0.5173), and among

habitats and ﬁsh densities (F212 = 0.62, p = 0.5529). With a mean of

158

A. Open Water

145

 

+0
124 "'"5 r
10«
€85
.2 6‘
to
4-4
2‘1

 

 

 

Day of Experiment

B. Submersed Vegetation

14~

.3 .3
m C N
l l l

Biomass (9!!)
O)

 

 

 

1 8 1 5 22
Day of Experiment

Figure 3.16: Mean biomass (g/l :t 2xstandard error) of copepods in enclosures

located in A) open water and B) submersed vegetation during experiment two.

159

 

 

 

53m :3 o o we“: 0 8an on“: 2.26

mumﬁoﬁ 8an o o Ram om“: o 2: 225

8:? Snow the on “8 mug o o o oasm
o mm“? o o 83m o o o 3 580

om H NF 0 o o o o o o m .58

mm + E o o o o o o o o 580
E 80 Em... 28 new ea 8o _oo

 

653055 u E van .9958 u one .mEﬁEmc n Em; .muoaozmmm n wmm .mEEBmEmﬁm u cam 6536 n a6 .8833
n can 6.2.82.8 u .8 29.3 .8223 cm: EoEEU 5.3 .89: “9.3m... E9930 c. “3380. $58.05 05 EeEtoaxm c.

BEES wmcazoa o_EocoxS EEmEn EB. $559..chme v6 326va Emncﬂw H ”ER... bacon :85. ”Na min...

160

450 1
400 q - --Start,OPen
End,Open
—-x—- Start, Submersed
_)(—. End, Submersed

350 ~
300 —
250 a
200 1
150 4

_L

O

O
L

 

DENSITY (#lm3)

 

 

 

Fish Density

Figure 3.17: Density of macroinvertebrates (#lm3) at the beginning and end of
experiment two sampled in open water and submersed vegetation enclosures

stocked with different densities of YOY bluegill.

161

7 macroinvertebrates / m3, 3 lower density of macroinvertebrates was sampled in
enclosures located in open water than in submersed vegetation (mean = 115
macroinvertebrates / m3, F114 = 8.39, p = 0.0117). Macroinvertebrate densities
were similar in enclosures stocked with 0, 5, or 15 YOY bluegill (F214 = 0.67, p =
0.5265).

Macroinvertebrate biomass (Figure 3.18) was similar at the beginning
(mean = 12.3 1 21.7 g/l) and end (mean = 2.5 1 4.2 g/l) of experiment two (F135 =
3.74, p = 0.0722). Similar patterns in variation of macroinvertebrate biomass
over time were found in both habitat types (F115 = 1.70, p = 0.2114) and all levels
of ﬁsh density (F145 = 0.27, p = 0.6100). No differences were found in
macroinvertebrate biomass among different habitat types (Fug, = 3.62, p =

0.0765), or enclosures with different densities of ﬁsh (F1_15 = 0.76, p = 0.3984).

FISH GROWTH

All stocked ﬁsh were recovered from the enclosures, with the exception of
one of the open water enclosures where only 13 of the 15 stocked ﬁsh were
recovered. Mean ﬁsh growth rates of enclosures ranged from 0.269 to 0.410
mm/day, with a mean of 0.327 mm/day (Figure 3.19). Slower growth rates were
observed in enclosures stocked with higher densities of YOY bluegill (F 1,9 = 8.91,
p = 0.0153). No differences were observed in overall mean growth rates of YOY
bluegill among the two habitat types (F19 = 0.26, p = 0.6210); however, at a
density of 15 ﬁsh per square meter, YOY bluegill grew better in submersed

vegetation habitat compared to open water habitat (FM = 8.15, p = 0.0462).

162

A. Open Water

80 1
Fish Densi

70 — ty 1
,,. 0

60 — 5 —

40E -

 

30' 10 “
20— : —

BIOMASS (g/l)

10’ 05i 0510

0— —— —-——-—-—~-» ﬂ

 

 

 

-10 1 1
beginning end

TIME

B. Submersed Vegetation

 

 

 

 

 

80 1 I
70 _ I Fish DensntyA
0 1 E1 0
6° “ 5? 5 ‘
s 50 — L 15 —-
532
2 30 - ——; -
Q I 0
m 20 ~ I 5 —
10— 17.10 [F 10 —
I I
0 _ . _ ﬁ
_10 l l
beginning end

TIME

Figure 3.18: Mean biomass (g/l 1 2xstandard error) of macroinvertebrates in
enclosures located in A) open water and B) submersed vegetation during

experiment two.

163

 

0.45

Fish Density I
[3 5 5
15 5 ..

b— —

.O
.p.
O

 

 

 

 

 

 

 

 

 

 

 

GROWTH (mm/day)
O
00
01
l
l

 

 

 

15
0.30 — a
0.25 J 1
open water submersed
HABITAT

Figure 3.19: Growth of YOY bluegill (mm/day) in experiment two enclosures

located in different habitats, and stocked with different densities of ﬁsh.

164

FISH DIET

Weights of bluegill stomachs ranged from 0.00347 g to 0.09146 9 (Figure
3.20). There were no differences observed in the way stomach weight varied
with ﬁsh density between open water and submersed vegetation (F13 = 2.26, p =
0.1708). Bluegill stomach weights were similar in both habitats (F1’9 = 0.16, p =
0.6991 ), and ﬁsh densities (F13, = 0.30, p = 0.5945).

The diet of bluegill in all habitat types was numerically dominated by
zooplankton, both cladocerans and copepods (Table 3.3). Macroinvertebrates
from the order Diptera were also abundant within the bluegill stomachs, with
means raging from 5 to 12 dipterans per stomach.

The main contributor to biomass within the stomachs of bluegill was
macroinvertebrates from the order Diptera (Table 3.4). Cladocerans and
copepod zooplankton, and macroinvertebrates from the order Odonata also
made up high proportions of the biomass of items within bluegill stomachs, when
compared to Ephemeroptera and Hemiptera.

To compare bluegill diet among habitat types, prey items were classiﬁed
as macroinvertebrates and zooplankton. Mean zooplankton biomass in bluegill
from different enclosures ranged from 3.8 to 53 pg per stomach (Figure 3.21).
No interaction between habitat and ﬁsh density in predicting the biomass of
zooplankton within the bluegill stomachs was observed (Fm = 0.14, p = 0.7196).
Additionally, no differences in zooplankton biomass in the stomachs of bluegill
harvested from enclosures located in different habitat types (F 1,9 = 0.12, p =

0.7359) or different ﬁsh densities (F19 = 0.19, p = 0.6743) were observed.

165

 

.0
o
c»

 

 

 

 

 

 

 

 

 

l l
Fish Density

3;, 0.05— D i l5 -
+- 5
I 15
(5'3 004— 5 1
g 0.03— 7 —
e I r n  ‘ I ‘
a) 0.02— -—

0.01 I '

 

 

 

Open water Submersed
HABITAT

Figure 3.20: Stomach weights (g) from harvested bluegill within experiment two

enclosures.

166

Table 3.3: Mean number (1 2 standard error) of organisms found within the

stomach of YOY bluegill harvested from experiment two enclosures. Cla =

Cladocera, Cop = Copepod, Dip = Diptera, Eph = Ephemeroptera, Hem =

Hemiptera, Odo = Odonata.

 

 

Cla Cop Dip Eph Hem Odo
Open Water. 5 3239* 7:3: ‘lé’lJ “6361* 1593* Off;
Open Wat... 15 6535* 3'31; 4589* ”6933* ”£33" “1333*
s. 7.5:: 3.1:; 6.5:; 0.9:: 1.9;:
Submersed. 15 25f? 6585* 8539" 357133" 1591’ 3" ‘58:

 

167

Table 3.4: Mean biomass (pg 1 2 xstandard error) of organisms found within the

stomach of YOY bluegill harvested from experiment two enclosures. Cla =

Cladocera, Cop = Copepod, Dip = Diptera, Eph = Ephemeroptera, Hem =

Hemiptera, Odo = Odonata.

 

 

Cla Cop Dip Eph Hem Odo
055" W55 5 4553301 15945301 13322; 348.311 15330" 41652151
Open Water. 15 81%? 13512; 851933? 7355135 21363981 33.3%"?
SUDmersed' 5 1119152566: 71859;}; 192707314: 7522361: 0 130%:
.5 1.9.52.5 13.5: 1.53.1.5 7:32; 1.5:: 1.8.1;

 

168

«1

I

 

A. Zooplankton

 

 

 

 

 

 

 

 

60 l I
Fish Density 15
A 50 — 1 5 5 V w -
‘g 15 :
g 40 ~ * 5 ~
g . 15
3 30 — ~
0)
«‘8
E 20 — —
.9 —— -
m 7
10 - T 1 —
O I L I
open water submersed
Habitat
B. Macroinvertebrates
4000 1 I
Fish Density
A :1 5
.c 5 15
8 3000 — 15 l -
E
9.
£
0) __.
3: 2000 — —
a , '
<0 5 -' “ '
g 15 g .
5 1000— g . " —
O l I

 

 

 

open water submersed

Habitat

Figure 3.21: Biomass (pg/stomach) of A) zooplankton and B)
macroinvertebrates dissected from YOY bluegill stomachs in experiment two

enclosures located in different habitats.

169

Mean macroinvertebrate biomass in bluegill stomachs ranged from 612 to
3520 pg per enclosure (Figure 3.21). No interaction between habitat and ﬁsh
density was observed (Fm = 2.54, p = 0.1498). Additionally, there were no
differences in the amount of macroinvertebrate biomass in the stomachs of
bluegill from different habitat types (F19 = 0.01, p = 0.9215) or different ﬁsh

densities (F13 = 0.37, p = 0.5565).

170

‘1

 

DISCUSSION
YOY Bluegill Growth and Habitat

In experiment one, the growth rates of YOY bluegill stocked into
enclosures located in emergent vegetation were slower than the growth rates of
bluegill in enclosures located in open water or submersed vegetation. However,
dissolved oxygen levels were signiﬁcantly lower in emergent vegetation
enclosures, with levels often falling below 3.0 mg/l. These low oxygen
concentrations may have caused the decline in growth rate observed in YOY
bluegill within these enclosures. Low oxygen concentrations in emergent
vegetation compared to open water habitats have been documented in other
studies (e.g., Rose and Crumpton 1996). More research is needed to determine
if growth rates of YOY bluegill would be slower in areas of emergent vegetation
that did not experience similarly low oxygen concentrations.

Bluegill in Saginaw Bay, Lake Huron were more likely to be sampled in
areas with submersed vegetation than in areas of emergent vegetation or open
water (see Chapter 2). One objective of this study was to examine if a habitat
preference for submersed vegetation may lead to faster growth rates of bluegill.
This study provides some evidence that submersed vegetation may have a
positive impact on YOY bluegill growth rates. While no differences in bluegill

growth rates were found between open water and submersed vegetation in

171

experiment one, or at low densities in experiment two, at high densities bluegill
located in submersed vegetation grew faster than those located in open water.

One possible explanation for the difference in growth rates between
habitat types is that there are differences in food resource levels among
emergent vegetation, submersed vegetation, and open water habitats. I found
some differences in food resource levels among the different habitats in
experiment one and two. The macroinvertebrate community in both experiments
was dominated by ephemeropterans, gastropods, hemipterans, trichopterans,
and odonates, which is consistent with macroinvertebrate communities reported
in other studies of wetland and pond areas (e.g. Crowder and Cooper 1982,
Schramm and Jirka 1989, Belk 1993). Macroinvertebrate food resources may
have been greater in vegetated habitats than in open water in both experiments.
Higher densities of macroinvertebrates were measured in enclosures located in
emergent vegetation compared to open water and submersed vegetation in
experiment one. In experiment two, higher densities of macroinvertebrates were
found in submersed vegetation than in open water. However, no differences in
macroinvertebrate biomass among the habitat types were observed in either
experiment. The pattern of higher densities of macroinvertebrates in vegetated
areas than in open water has been documented in other studies of wetland
systems (e.g. Schramm et al. 1987, Belk 1993). Crowder and Cooper (1982)
also documented increases in macroinvertebrate biomass with macrophyte

density.

172

While there is some evidence of higher macroinvertebrate food resources
in the submersed vegetation enclosures compared to open water, there is less
evidence for differences in zooplankton levels among habitat types. Copepod
density and biomass did not vary among habitat types in either experiment.
Cladoceran biomass also did not vary among habitats; however, in experiment
one there was a signiﬁcantly greater density of cladocerans in enclosures located
in submersed vegetation compared to emergent vegetation and open water.
Belk (1993) also documented higher abundances of zooplankton in vegetated
areas compared to open water in a South Carolina reservoir.

One other possible explanation for the differences in growth rates between
submersed vegetation and open water habitats in experiment two is that there
are differences in the feeding behavior and diet of bluegill between the two
habitats. YOY bluegill in this study fed on a combination of zooplankton and
macroinvertebrate food resources. In both experiments, bluegills fed on more
cladocerans than copepods, even though copepods were found in higher
densities and biomass levels. Bluegill also fed on macroinvertebrate food
resources, with dipterans, ephemeropterans, and odonates making up large
proportions of the biomass of food items within their stomachs. Similar types of
macroinvertebrate prey have been found in the diet of older bluegill in other
studies (Crowder and Cooper 1982, Mittelbach 1984, Mittelbach 1988, Schramm
and Jirka 1989).

Diet analysis of bluegill in experiment one enclosures indicated that YOY

bluegill fed on more macroinvertebrates in submersed and emergent vegetation

173

habitats that in open water. Bluegill also fed on more zooplankton in open water
enclosures than in the other habitat types. This pattern of feeding on
macroinvertebrates in vegetated areas and zooplankton in areas of open water is
consistent with results of other studies (Werner and Hall 1977, 1988, Mittelbach
1981, 1984, Crowder and Cooper 1982, Schramm and Jirka 1989); however, this
pattern was not observed in experiment two enclosures. In experiment two there
were no signiﬁcant differences in the zooplankton or macroinvertebrate biomass
from the stomachs of bluegill from enclosures located in open water compared to
submersed vegetation. Under low resource conditions like those in experiment
two, bluegill may switch from feeding on zooplankton to macroinvertebrates, or
from macroinvertebrates to zooplankton depending upon resource levels,
allowing them to maximize energy intake, at any given time (Mittelbach 1981,
Werner 1983, Schramm and Jirka 1989). While studying age one and older
bluegill, Werner (1983) demonstrated that to maximize energy intake, ﬁsh should
feed on zooplankton when zooplankton populations are high, but then switch to
macroinvertebrates as zooplankton populations decrease. Since resources of
both macroinvertebrates and zooplankton were low in experiment two enclosures
compared to experiment one enclosures, bluegill may have been feeding on both
types of food resources rather than specializing on one resource.

Although I did not identify differences in the biomass of zooplankton and
macroinvertebrates in bluegill stomachs from different habitats in experiment two,
diet was only sampled at the end of the experiment, and may not be

representative of the feeding behavior of bluegill throughout the experiment. If

174

the bluegill were actually feeding more on zooplankton in open water and
macroinvertebrates in submersed vegetation, there may have been differences in
the energy YOY bluegill derive from those different food sources. Bluegill
feeding on zooplankton in open water environments spend little time searching
and handling prey, but the small prey size makes each item’s overall proﬁtability
low (Mittelbach 1981). Bluegill feeding on macroinvertebrates in vegetated
habitats may have to spend more time searching for and handling prey; however
due to the increased prey size, they capture greater biomass per feeding event
(Mittelbach 1981). In addition to capturing greater biomass food items in
vegetated habitats, the overall caloric content per gram dry weight is typically
higher for macroinvertebrates belonging to the orders diptera and odonata (the
most common prey items consumed in this study) compared to cladoceran and
copepod zooplankton (Cummins and Wuycheck 1971 ). These differences in
energy consumption and expenditure could impact bluegill growth rates, leading
to faster growth in submersed vegetation.

While I found that at high densities, bluegill grew faster in submersed
vegetation than in open water, other studies have suggested that the opposite
pattern may occur. Bluegill may feed more optimally in open water
environments, resulting in faster growth rates of bluegill in open water habitat
compared to vegetated areas (e.g. Mittelbach 1981, Werner and Hall 1988).
However these estimates of prey proﬁtability were calculated using large bodied
zooplankton such as Daphnia spp. in artiﬁcial tank environments (Mittlebach

1981). I found low abundances of daphnids in experiment two enclosures. The

175

low abundance of daphnids is expected, as all enclosures (open water and
submersed vegetation) were located in the shallow littoral zone, and daphnids
are typically more abundant in the pelagic zone compared to the littoral edge
(Gliwicz and Rykowska 1992, Hulsmann et al. 1999). The lack of large species
of zooplankton may be one reason why YOY bluegill in this study grew slower in

open water enclosures compared to enclosures located in submersed vegetation.

Fish Density and Bluegill Growth Rates

Fish density can have a negative effect on the growth rates of YOY
bluegill. Growth rates of bluegill in the second experiment were signiﬁcantly
slower in enclosures stocked with 15 ﬁsh compared to those stocked with 5 ﬁsh.
This result is not entirely surprising, as intraspeciﬁc competition has been shown
to have negative impacts on the growth rates of older bluegill (Mittelbach 1988,
Osenberg et al. 1988, Belk 1993); however this is the ﬁrst study to document
negative impacts of high ﬁsh densities on the growth of small YOY bluegill.

Fish density can have a negative impact on growth rates through
exploitative competition. If there is exploitive intraspeciﬁc competition among the
bluegills used in this study, then a decline in the density and biomass of
zooplankton and macroinvertebrates over time, and a stronger decline with
increased ﬁsh density would be expected. Other studies have shown that bluegill
can have signiﬁcant negative impacts on cladoceran density (Turner and
Mittelbach 1990), macroinvertebrate biomass (Crowder and Cooper 1982),

macroinvertebrate size (Crowder and Cooper 1982, Mittelbach 1988) and

176

macroinvertebrate density (Mittelbach 1988); however these studies involved the
use of older bluegill and were longer in duration. Diet analysis of bluegill in this
study did not indicate any signiﬁcant differences in the biomass of zooplankton or
macroinvertebrates harvested from the stomachs of bluegill in enclosures
stocked with different densities of ﬁsh. However, measurements of zooplankton
abundance and biomass over time show some evidence of resource depletion at
high ﬁsh densities. While there were no signiﬁcant differences in the abundance
or biomass of zooplankton from the start date to the end date of the experiment,
there were differences in the pattern of zooplankton variation over time among
the different ﬁsh densities. There was an increase in cladoceran abundance and
biomass in open water enclosures stocked with 0 or 5 ﬁsh measured on day 8 of
the experiment. This increase in zooplankton may have been a result of release
from predation pressure from existing populations of fathead minnows, sunﬁsh,
and sticklebacks within the pond. There was not an increase in density or
biomass of cladocerans observed in enclosures stocked with 15 ﬁsh. Stocking
15 bluegill into enclosures may have maintained sufﬁcient predation pressure on
zooplankton to prevent the resource from growing. The increase in zooplankton
resources on day eight of the experiment was not observed in enclosures placed
in submersed vegetation; however zooplankton populations have been shown to
persist under ﬁsh predation in vegetated areas because macrophytes can serve

as a refuge from ﬁsh predation (Stansﬁeld et al. 1997).

177

Growth Rates of Bluegill in Experiments 1 and 2

There was a large difference in mean growth rates of YOY bluegill
between the two experiments. In experiment one, growth rates of YOY bluegill
averaged approximately 0.523 1 0.103 mm/day, which is within the range (0.343
to 0.957 mm/day) of mean growth rates of YOY bluegill sampled in wetland areas
in Saginaw Bay, Lake Huron in 2001 (see chapter one). Growth of YOY bluegill
was considerably slower in experiment two, averaging approximately 0.327 1
0.042 mm/day.

The slower growth rates of YOY bluegill in experiment two may be a result
of the considerably lower food resources available in experiment two enclosures
compared to experiment one. Experiment one was conducted in a wetland with
abundant populations of zooplankton. Mean cladoceran density in experiment
one enclosures was 300 organisms/l, and mean copepod density was 882
organisms/I. The densities of cladocerans and copepods measured in
experiment one are within the range of zooplankton densities reported in other
wetland environments (Belk 1993, Sandilands and Hann 1996, Stansﬁeld et al.
1997). The density of zooplankton was much lower in experiment two
enclosures, with a mean cladoceran density of 8 organisms/l and mean copepod
density of 81 organisms/l. With a mean density of 422 macroinvertebrates/m3,
enclosures in experiment one also had higher densities of macroinvertebrates
compared to experiment two enclosures where mean density was 61

macroinvertebrates/m3.

178

Growth rates of bluegill have been shown to be correlated with prey
availability, with slower growth rates in areas with lower densities of large
macroinvertebrates (Mittelbach 1988). This study did not show differences in the
biomass or density of macroinvertebrates and zooplankton between enclosures
stocked with ﬁsh and those without ﬁsh at the end of the experiment one, but
there were differences in cladoceran biomass in experiment two. This indicates
that food resources may have been limiting in experiment two. Therefore,
managers concerned with bluegill growth may need to monitor zooplankton and
macroinvertebrate community dynamics to ensure abundant food supplies for

YOY ﬁsh.

Future Research

This study demonstrated that ﬁsh density has an impact on the individual
growth rates of YOY bluegill. Therefore managers should be concerned with the
density of YOY bluegill if they desire to maximize growth rates. Although lower
ﬁsh densities may produce fewer numbers of ﬁsh, the ﬁsh should have faster
growth rates. It is not yet known how large of an impact bluegill density may
have on growth rates of small YOY bluegill. Additionally, this study only
examined three levels of bluegill density, 0, 5, and 15 ﬁsh per square meter.
While this study indicated that there is a relationship between bluegill growth rate
and ﬁsh density, the nature of this relationship is not known, and requires further

study.

179

Results of the second experiment indicated that at high ﬁsh densities,
YOY bluegill grow faster in submersed vegetation than in open water. This study
did not identify the mechanism driving this pattern. There are several possible
explanations for why small YOY bluegill may grow better in submersed
vegetation than in open water. One potential explanation is that there are
differences in the feeding behavior and diet of small YOY bluegill between
submersed vegetation and open water habitat. Although, this study did not ﬁnd
any signiﬁcant differences in the diet of YOY bluegill between the two habitats in
experiment two, the diet of bluegill was only sampled once at the end of the
experiment. More research is needed on the potential feeding behavior and diet
of small YOY bluegill in these habitats to determine if differences in growth rates

are due to feeding.

180

LITERATURE CITED

Balcer, M.D., N.L. Korda, and SJ Dodson. 1984. Zooplankton of the Great
Lakes: A Guide to the Identiﬁcation and Ecology of the Common
Crustacean Species. The University of Wisconsin Press, Madison, WI.

Belk, MC. 1993. Growth and mortality of juvenile sunﬁshes (Lepomis sp.) under
heavy predation. Ecology of Freshwater Fish 2:91-98.

Cardinale, B.J., V.J. Brady and TM. Burton. 1998. Changes in the abundance
and diversity of coastal wetland fauna from the open water/macrophyte
edge towards shore. Wetlands Ecology and Management 6:59-68.

Crowder, LB. and WE. Cooper. 1982. Habitat structural complexity and the
interaction between bluegills and their prey. Ecology 63:1802-1813.

Culver, D.A., M.M. Boucherle, D.J. Bean, and J.W. Fletcher. 1985. Biomass of
freshwater crustacean zooplankton from length-weight regressions.
Canadian Journal of Fisheries and Aquatic Sciences 42:1380-1390.

Cummins, K.W. and J.C. Wuycheck. 1971. Caloric equivalents for investigations
in ecological energetics. Mitteilungen - lntemationale Vereinigung fur
Theoretische und Angewandte Limnologie 18:1-

Diehl, S. 1988. Foraging efﬁciency of three freshwater ﬁshes: effects of
structural complexity and light. Oikos 53:207-213.

Dionne, M. and C.L. Folt. 1991. An experimental analysis of macrophytes
growth forms as ﬁsh foraging habitat. Canadian Journal of Fisheries and
Aquatic Sciences 48:123-131.

Fassett, NO, 1985. A Manual of Aquatic Plants. The University of Wisconsin
Press, Madison, WI.

Francis, R.l.C. 1990. Back-calculation of ﬁsh length: a critical review. Journal of
Fish Biology 36:883-902.

181

Gerking, SD. 1957. A method of sampling the littoral macrofauna and its
application. Ecology 38:219-226.

Gliwicz, 2M. and A. Rykowska. 1992. “Shore avoidance” in zooplankton: a
predator-induced behaviour or predator-induced mortality? Journal of
Plankton Research 14: 1331-1342.

Gotceitas, V. 1990a. Plant stem density as a cue in patch choice by foraging
juvenile bluegill sunﬁsh. Environmental Biology of Fishes 29:227-232.

Gotceitas, V. 1990b. Foraging and predator avoidance: a test of a patch choice
model with juvenile bluegill sunﬁsh. Oecologia 83:346-351.

Houde, ED. 1994. Differences between marine and freshwater ﬁsh larvae:
implications for recruitment. ICES Journal of Marine Science 51:91-97.

Hulsmann, S., T. Mehner, S. Worischka, and M. Plewa. 1999. Is the difference
in population dynamics of Daphnia galeata in the littoral and pelagic areas
of a long-term biomanipulated reservoir affected by age-0 ﬁsh predation?
Hydrobiologia 408/409z57-63.

Jacobsen, L. and S. Berg. 1998. Diel variation in habitat use by planktivores in
ﬁeld enclosure experiments: the effect of submerged macrophytes and
predation. Journal of Fish Biology 53:1207-1219.

Jude, DJ. and J. Pappas. 1992. Fish utilization of Great Lakes coastal
wetlands. Journal of Great Lakes Research 18:651-672.

Lillie, RA. and J. Budd. 1992. Habitat architecture of Myriophyllum spicatum L.
as an index to habitat quality for ﬁsh and macroinvertebrates. Journal of
Freshwater Ecology 7:1 13-125.

Littell, R.C., R.J. Freund, and PC. Spector. 1991. SAS System for Linear
Models, Third Edition. SAS Institute Inc., Cary, NC.

Mittelbach, G.G. 1981. Foraging efﬁciency and body size: a study of optimal diet
and habitat use by bluegills. Ecology 62:1370-1386.

182

 

Mittelbach, G.G. 1984. Predation and resource partitioning in two sunﬁshes
(Centrarchidae). Ecology 65:499-513.

Mittelbach, G.G. 1988. Competition among refuging sunﬁshes and effects of
ﬁsh density on littoral zone invertebrates. Ecology 69:614-623.

Osenberg,. C.W., E.E. Werner, G.G. Mittelbach, and DJ. Hall. 1988. Growth
patterns in bluegill (Lepomis macrochirus) and pumpkinseed (L. gibbosus)
sunﬁsh: environmental variation and the importance of ontogenetic niche
shifts. Canadian Journal of Fisheries and Aquatic Sciences 45:17-26.

Post, JR, and AB. Prankevicius. 1987. Size-selective mortality of young-of-
the-year yellow perch (Perca ﬂavescens): evidence from otolith
microstructure. Canadian Journal of Fisheries and Aquatic Sciences
44: 1 840-1 847.

Rose, C. and WG. Crumpton. 1996. Effects of emergent macrophytes on
dissolved oxygen dynamics in a prairie pothole wetland. Wetlands
16:495-502.

Rozas, LP. and WE. Odum. 1988. Occupation of submerged aquatic
vegetation by ﬁshes: testing the roles of food and refuge. Oecologia
77:101-108.

Sandilands, K.A. and B.J. Hann. 1996. Effect of ﬁsh and submersed
macrophytes on the abundance of zooplankton in a prairie wetland. UFS
(Delta Marsh) Annual Report 31:58-62.

Schramm, H.L. Jr., and K.J. Jirka. 1989. Epiphytic macroinvertebrates as a food
resource for bluegills in Florida lakes. Transactions of the American
Fisheries Society 88:416-426.

Schramm, H.L. Jr., K.J. Jirka, and M.V. Hoyer. 1987. Epiphytic
macroinvertebrates on dominant macrophytes in two central Florida lakes.
Journal of Freshwater Ecology 4:151-161.

Smock, LA. 1980. Relationships between body-size and biomass of aquatic
insects. Freshwater Biology 10:375-383.

183

 

. v» ..

I... . . .
T .
....

. 3.15.5145: .
. .19 z...

 

Stansﬁeld, J.H., M.R. Perrow, L.D. Tench, A.J.D. Jowitt and A.A.L. Taylor. 1997.
Submerged macrophytes as refuges for grazing Cladocera against ﬁsh
predation: observations on seasonal changes in relation to macrophyte
cover and predation pressure. Hydrobiologia 342/343z229-240.

Tonn, WM. and CA. Paszkowski. 1987. Habitat use of the central mudminnow
(Umbra limi) and yellow perch (Perca ﬂavescens) in Umbra-Perca
assemblages: the role of competition, predation, and abiotic environment.
Canadian Journal of Zoology 65:862-870.

Turner, AM. and G.G. Mittelbach. 1990. Predator avoidance and community
structure: Interactions among piscivores, planktivores, and plankton.
Ecology 71: 2241 -2254.

Turner, RE. 1988. Secondary production in riparian wetlands. Transactions of
the 53rd North American Wildlife and Natural Resources Conference. 491-
501.

Werner, E.E. and DJ. Hall. 1974. Optimal foraging and the size selection of
prey by the bluegill sunﬁsh (Lepomis macrochirus). Ecology 5521042-
1052.

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

Werner, E.E. and DJ. Hall. 1977. Competition and habitat shift in two sunﬁshes
(Centrarchidae). Ecology 58:869-876.

Werner, E.E. and DJ. Hall. 1979. Foraging efﬁciency and habitat switching in
competing sunﬁshes. Ecology 60:256-264.

Werner, E.E. and DJ. Hall. 1988. Ontogenetic habitat shifts in bluegill: the
foraging rate-predation risk trade-off. Ecology 69:1352-1366.

Werner, E.E., G.G. Mittelbach, and DJ. Hall. 1981. The role of foraging
proﬁtability and experience in habitat use by the bluegill sunﬁsh. Ecology
62:1 16-125.

184

 

Werner, E.E., J.F. Gilliam, D.J. Hall, and G.G. Mittelbach. 19833. An
experimental test of the effects of predation risk on hab itat use in ﬁsh.
Ecology 64:1540-1548.

Werner, E.E., G.G. Mittelbach, D.J. Hall, and J.F. Gilliam. 1983b. Experimental
tests of optimal habitat use in ﬁsh: the role of relative habitat proﬁtability.
Ecology 64:1525-1539.

Winﬁeld, I.J. 1986. The inﬂuence of simulated aquatic macrophytes on the
zooplankton consumption rate ofjuvenile roach, Rutilus rutilus, and perch,
Perca ﬂuviatilis. Journal of Fish Biology 29:37-48.

185

 

CONCLUSIONS

Results of this research suggest that not all wetlands are of equal value to
young of the year ﬁsh. Along the coastal wetlands of Saginaw Bay, Lake Huron,
the distribution and growth of YOY ﬁsh varies both within wetland environments
and among wetland areas. The YOY ﬁsh community responded to both changes
in the physical stmcture of the habitat (e.g. type of vegetation present, water
depth), as well as changes in water quality (e.g. dissolved oxygen
concentrations, and turbidity). Wetland habitat characteristics inﬂuenced the
abundance, diversity, species composition, and distribution of YOY ﬁsh. Wetland
habitat also has the potential to inﬂuence the growth of YOY ﬁsh at the
macrohabitat and microhabitat scale. Because growth rates are critical in
determining the probability of over winter survival and year class strength, small
changes in the characteristics of wetland environments that may inﬂuence YOY
ﬁsh growth rates may have large implications on ﬁsh population dynamics in the
Great Lakes.

Due to the large impact that wetland habitat characteristics can have on
the YOY ﬁsh community, not all wetland habitats are of equal value to YOY ﬁsh.
By understanding which habitat characteristics may increase the success of YOY
ﬁsh, we can make better decisions on which type of wetlands to protect for Great
Lakes ﬁsheries resources. For example, this study suggests that water depth

had a large impact on the YOY ﬁsh community, with the growth, presence, and

186

 

 

 

5.. 7

4 . . ..a..,0.w.....
...5.....1 v...

 

diversity of YOY ﬁsh all increasing in wetlands with areas of deeper water within
the 0 — 100 cm range sampled. Water levels within the Great Lakes are cyclical
in nature, and were approximately 0.5 meters below the long term mean during
this study period. As a result of the low water levels, within the wetlands
sampled there were few wetland areas with water depths greater than 60cm that
also had abundant vegetation growth. There are several reasons why having
access to deeper water habitats adjacent to vegetation may be beneﬁcial to
young ﬁsh, including providing refuges with cooler water temperatures, higher
oxygen concentrations, and reduced risk of avian predation. Protecting wetland
areas with greater littoral slopes would help insure that during periods of low
water levels in the Great Lakes, YOY ﬁsh still have access to both areas of
deeper water and abundant wetland vegetation.

Another habitat characteristic that had a large impact on the YOY ﬁsh
community at both the macrohabitat and microhabitat scale is the amount of
vegetation cover within a wetland. Vegetation cover inﬂuenced species richness,
diversity, community composition, and ﬁsh growth at the macrohabitat scale,
while the type of vegetation present inﬂuenced the distribution of banded killiﬁsh,
and the distribution and growth of bluegill at the microhabitat scale. Higher
abundances of submersed vegetation may be particularly advantageous to the
growth of YOY ﬁsh. In addition to the amount of vegetation present, YOY ﬁsh
may prefer patchy vegetation communities, and wetland environments with high
habitat heterogeneity. At the macrohabitat scale habitat complexity inﬂuenced

which species of YOY ﬁsh were found within a wetland as well as the growth of

187

 

.E.?.._.§..s ...P... .m........._.........:.ﬁ.r.9.2.55.1...1;.1...

r

 

YOY spottail shiner. At the microhabitat scale, YOY bluegill showed a
preference for patchy vegetation communities, preferring to be located along the
edge of habitat patches where they may have easy access to more than one
habitat type. As a result of the importance of wetland vegetation to the YOY ﬁsh
community, Great Lakes ﬁsheries managers may wish to protect wetlands with
patchy, diverse, and abundant vegetation communities.

This study examined the YOY ﬁsh community in ﬁfteen different wetland
areas located throughout Saginaw Bay, Lake Huron. Six of these wetland areas
had islands, sand bars, or gravel bars adjacent to the wetland. While the number
and size of these structures varied among the wetland areas, these structures
may help to shelter wetland areas from damaging wind and wave action. The
presence of these structures had a large impact on the YOY ﬁsh community, with
increased abundances of YOY ﬁsh, and higher species richness with one of
these structures compared to those without. These presence of these structures
also impacted species composition. Yellow perch, brown bullhead, spotﬁn
shiner, green sunﬁsh, and pumpkinseed were all more commonly found in
wetlands with one or more of these structures than in wetlands without. Due to
the importance of these structures to the YOY ﬁsh community, managers may
wish to focus on wetlands with these types of structures when managing coastal
habitats for YOY ﬁsh.

The shore of Saginaw Bay where a wetland area was located appeared to
have an impact on species composition, with different ﬁsh communities

associated with the eastern compared to the western shore of Saginaw Bay.

188

 

 

Gizzard shad, green sunﬁsh, sand shiners, and yellow perch were more
commonly present in wetlands on the eastern bank, while white suckers, rock
bass, and bowﬁn were more commonly present in wetlands on the western bank
of Saginaw Bay. Due to the change in species composition observed between
wetlands on the eastern versus western shore of Saginaw Bay, it may be
important to protect and manage wetland areas on both shores for the beneﬁt of
YOY ﬁsh.

Water quality parameters such as turbidity and dissolved oxygen may also
be important to YOY ﬁsh. Both banded killiﬁsh and spottail shiner were more
likely to be sampled in areas with higher water clarity. Increased turbidity may
inhibit the ability of YOY ﬁsh to locate and capture prey, and as a result, YOY
banded killiﬁsh and spottail shiners may select areas with increased water clarity.
Dissolved oxygen concentrations inﬂuenced the distribution of spottail shiners
within wetland areas. Additionally, results of the enclosure experiment indicated
that the growth of YOY bluegill was slower in areas of emergent vegetation.
Although the enclosures with emergent vegetation had larger macroinvertebrate
populations and abundant zooplankton food resources, dissolved oxygen
concentrations were signiﬁcantly lower in these enclosures, and may have had a
negative impact on YOY bluegill growth. The distribution of transient wetland
species, or those species that are likely to move in and out of wetland areas, was
largely impacted by changes in water quality compared to the importance of
changes in the physical structure of wetland habitat. Therefore, managers may

need to monitor and manage water quality characteristics such as turbidity and

189

dissolved oxygen within wetland areas to maximize their value to and use by
YOY ﬁsh.

Finally, the density of YOY ﬁsh can have a negative effect on their growth
rates. Growth rates of bluegill in the second enclosure experiment were
signiﬁcantly slower in enclosures stocked with 15 ﬁsh compared to those stocked
with 5 ﬁsh. Fish density can have a negative impact on growth rates through
exploitative competition. While diet analysis of bluegill in this study did not
indicate any signiﬁcant differences in the biomass of zooplankton or
macroinvertebrates harvested from the stomachs of bluegill in enclosures
stocked with different densities of ﬁsh, measurements of zooplankton abundance
and biomass over time show some evidence of resource depletion at high ﬁsh
densities. Because ﬁsh density has an impact on the individual growth rates of
YOY bluegill, managers should be concerned with the density of YOY bluegill if
they desire to maximize growth rates. Although lower ﬁsh densities may produce

fewer numbers of ﬁsh, the ﬁsh should have faster growth rates.

190

 

 

   

iiiiiiiiiiijiiiiiijiiiiiiiijiii