 

 

5 14v.

.. A:

 

. xxx... 4 .imhuim.

‘ an “ii

- _ f. 12.4 in”...
xﬁwxémwﬁ

u
:4 .
. F
A

r if {Mm—m
, ..- . i I
r3. (a. .
. .3

_
A?
x... Na.

41

_ Link! rt. .

. 1. ,5? 5
612%.? .

.JMVWSWMWV
fwﬁ

.
in u...
«I... .13...
. ;. .. '9
:s. nu»: I!“ ,
.9:an 3.531!

....A.u..v
... .

a»

E... n.

., iii .

2 LIBRARY
2 (7 08 Michigan State
University

This is to certify that the
dissertation entitled

EVALUATING COLLECTION, REARING, AND STOCKING
METHODS FOR LAKE STURGEON (ACIPENSER
FULVESCENS) RESTORATION PROGRAMS IN THE
GREAT LAKES

presented by

James Andrew Crossman

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Fisheries and Wildlife

 

 

Mrs/(Lg

 

Major Professor’s Signature

s-n-os

 

Date

MS U is an affirmative-action, equal-opportunity employer

 

 

_ - —.---l-t-I-o-I-I-l-u---a-o-C-I-I—l-l-n-0-o-t-I-I-o-l-I-I-I-l-u-o_u-v-l-c-o-d-c-

u.-.c|--—:---—--v—o-.—-.-._..~

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/DateDue indd

 

EVALUATING COLLECTION, REARING, AND STOCKING METHODS FOR
LAKE STURGEON (A CIPENSER F UL VESCENS) RESTORATION PROGRAMS IN
THE GREAT LAKES

By

James Andrew Crossman

A DISSERTATION

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

DOCTOR OF PHILOSHOPHY

Fisheries and Wildlife

2008

ABSTRACT

EVALUATING COLLECTION, REARING, AND STOCKING METHODS FOR
LAKE STURGEON (ACIPENSER FULVESCENS) RESTORATION PROGRAMS IN
THE GREAT LAKES

By
James Andrew Crossman

Lake sturgeon (Acipenserfulvescens) are native ﬁshes of the Great Lakes that
have been numerically depressed throughout their range due to anthropogenic factors
such as over-harvesting, degradation in water quality and spawning habitat and loss of
connectivity due to impoundments. Management and conservation strategies have been
implemented throughout the Great Lakes that focus on restoring remnant populations
through supplementation of hatchery produced juveniles. Alternative management
prescriptions have been advocated, largely in the absence of data, focusing on the best
way to collect gametes/larvae, the use of different hatchery rearing environments, and the
appropriate size or age of ﬁsh to stock. I determined the effects of different gamete/larval
collection methods, rearing environments, and stocking strategies on juvenile lake
sturgeon growth, survival, and levels of genetic diversity. I used three methods to collect
lake sturgeon progeny, 1) direct gamete takes from spawning adults, 2) collection of
naturally produced eggs from the stream substrate, and 3) collection of larvae dispersing
downstream ﬁ'om adult spawning grounds. Lake sturgeon offspring were reared in two
environments, half in a streamside hatchery constructed on the natal Upper Black River

and half at a traditional hatchery in southern Michigan. I also examined how hatchery

rearing environment and age affect survival and movement following release. Progeny
from each hatchery were marked and released at 8, 13, and 17 weeks of age. Overwinter
mortality was quantiﬁed by surgically implanting 20 juvenile lake sturgeon ﬁom each
hatchery with ultrasonic transmitters at 6 months of age and releasing them in late fall.
Finally, I investigated the effects of four predators on survival, behavior, and habitat
preference of three age classes of juvenile lake sturgeon reared in two hatchery
environments. I found that collecting dispersing larvae downstream captured the highest
amount of genetic diversity present in the adult breeding population. Streamside rearing
and the addition of habitat complexity within rearing tanks improved survival and
maintained natural variation in the size of lake sturgeon offspring. Signiﬁcantly higher
rates of recapture were realized for progeny reared in the streamside hatchery compared
to the traditional hatchery at 8 and 13 weeks of age. Recapture rates and dispersal
distances were signiﬁcantly higher for 17 week old ﬁsh compared to releases of earlier
ages. Large body size was negatively correlated with timing of movements across all
ages. Overwinter survival of juvenile lake sturgeon was 40% irrespective of rearing
location. Predation by crayﬁsh is likely to be a signiﬁcant source of mortality for juvenile
lake sturgeon up to three months of age. Results demonstrate that non-lethal indirect
effects potentially contribute to higher rates of mortality between alternate predator types.
Work described in thesis provides a framework for evaluating alternative strategies for
managers designing conservation programs for lake sturgeon and other long-lived
iteroparous species. Results indicate that supplementation protocols for lake sturgeon
should be developed on a site speciﬁc basis and demonstrate the importance of hatchery

rearing environment.

ACKNOWLEDGEMENTS

First and most importantly, I thank Christin Davis, whose love and support made
this work possible. I especially want to thank my parents, Wendy and Brion, my sister
Laura, and my grandparents Alex and Joan who have always encouraged me to follow
my dreams. I’m also grateful to Steve and Carol Davis for their support during this stage
of my life.

My time spent in the Department of Fisheries and Wildlife at Michigan State
University included many memorable moments and I wish to thank those individuals that
made a difference in my life and in my research. First, I thank my supervisor Dr. Kim
Scribner whose direction and guidance made my research a success. Secondly, I thank
my fellow lab members Patrick Forsythe, Kristi Filcek, Scot Libants, Laura McKinnon,
Jeanette Mcguire, Kristin Bott, Katy Caliﬁ‘, and Yen Duong. They provided help in the
lab, were bored by countless practice talks, provided help in the ﬁeld, and made the lab
an enjoyable place to work. I thank my committee members Dr. Gary Mittelbach, Dr.
Daniel Hayes, Dr. Don Garling, and Dr. Edward Baker. Their comments help to directly
improve the research included in this dissertation. Several outstanding technicians (Steve
Warner, Holly Wellard, Aaron Orhn, and Rachel VanI-Iome) and Jon Bivens helped make
my research a success and I thank them for putting up with my antics night and day. I
would also like to express my gratitude to the wonderful people residing on Crist Isle
Drive in Cheboygan County Michigan who provided support, meals, tools, and random
acts of kindness. I am grateﬁil to Sturgeon for Tomorrow, particularly Brenda and Gil
Archambo, whose devotion to lake sturgeon goes unmatched. Finally, the Michigan

Department of Natural Resources (MIDNR) provided many resources and man hours

iv

during my ﬁeld work and I would like to particularly thank the individuals at the Gaylord
ﬁsheries ofﬁce, Oden State Fish Hatchery, Marquette Fish Hatchery, and Wolf Lake
State Fish Hatchery.

Thanks to all the funding agencies that made this work possible including, the
Great Lakes Fisheries Trust, the MIDNR, the Michigan Agricultural Station, Presque Isle

Electric and Gas, and Sturgeon for Tomorrow.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES ........................................................................................................... xi
INTRODUCTION ............................. f .................................................................................. 1
CHAPTER 1

EVALUATION OF GAMETE AND LARVAL COLLECTION METHODS FOR LAKE
STURGEON; EFFECTS ON GENETIC DIVERSITY

Introduction .............................................................................................................. 4
Methods ................................................................................................................... 8
Study Site and Hatchery Rearing ................................................................. 8
Gamete/Larval Collection Methods ............................................................. 9
Rearing Methods and Environments .......................................................... 13
Genetic Sample Collection ........................................................................ 14
Genetic Analyses ....................................................................................... 14
Statistical Analyses .................................................................................... 15
Results .................................................................................................................... 21
Collection Methods .................................................................................... 21
Rearing Environment ................................................................................. 25
Adult Characteristics .................................................................................. 26
Discussion .............................................................................................................. 28
Conclusions ............................................................................................................ 40
CHAPTER 2

COMPLEXITY OF REARIN G ENVIRONMENT AF FECTS SURVIVAL AND
GROWTH OF LAKE STURGEON EMBRYOS, LARVAE AND JUVENILES

Introduction ............................................................................................................ 42
Methods ................................................................................................................. 45
Study Population ........................................................................................ 45
Hatchery Rearing Environment ................................................................. 46
Progeny Collection .................................................................................... 47
Experimental Design .................................................................................. 50
Egg Stage ................................................................................................... 50
Larval Stage ............................................................................................... 52
Juvenile Stage ............................................................................................ 56
Results .................................................................................................................... 60
Egg Stage ................................................................................................... 6O
Larval Stage ............................................................................................... 63
Juvenile Stage ............................................................................................ 65
Discussion .............................................................................................................. 68

Conclusions ............................................................................................................ 77

CHAPTER 3
HATCHERY REARIN G ENVIRONMENT AND AGE AFFECT SURVIVAL AND
MOVEMENTS OF STOCKED JUVENILE LAKE STURGEON

Introduction ............................................................................................................ 79
Methods ................................................................................................................. 82
Study Site ................................................................................................... 82
Rearing Environment ................................................................................. 83
Collection of Progeny ................................................................................ 83
Fish Stocking and Assessment ................................................................... 85
Statistical Analysis ..................................................................................... 87
Results .................................................................................................................... 89
Variation in Recapture Rates ..................................................................... 89
Variation in Size ........................................................................................ 91
Evidence of Mortality ................................................................................ 93
Discussion .............................................................................................................. 93
Conclusions ............................................................................................................ 98
CHAPTER 4

OVERWINTER SURVIVAL OF STOCKED AGE-O LAKE STURGEON REARED IN
NATAL AND NON-NATAL ENVIRONMENTS

Introduction .......................................................................................................... 100
Methods ............................................................................................................... 103
Study Site ................................................................................................. 103
Rearing Environment ............................................................................... 103
Transmitter Implantation and Monitoring ............................................... 104
Data Analysis ........................................................................................... 106
Results .................................................................................................................. 106
Discussion ............................................................................................................ 108
Conclusions .......................................................................................................... 1 1 1
CHAPTER 5

DIRECT AND INDIRECT EFFECTS OF PREDATION ON SURVIVAL AND
HABITAT USE OF JUVENILE LAKE STURGEON

Introduction .......................................................................................................... 112
Methods ............................................................................................................... 114
Experimental Design ................................................................................ 1 15
Tests of Indirect Effects of Predation Risk on Substrate Use .................. 116
Direct Effects of Predation on Age-Speciﬁc Survival ............................. 117
Alternate Species Trials ........................................................................... 118
Results .................................................................................................................. 119
Tests of Indirect Effects of Predation Risk on Substrate Use .................. 119
Direct Effects of Predation on Age-Speciﬁc Survival ............................. 120
Alternate Species Trials ........................................................................... 121
Discussion ............................................................................................................ 122
Conclusions .......................................................................................................... 126

vii

FINAL CONCLUSIONS ................................................................................................. 127

APPENDIX I
TABLES AND FIGURES ................................................................................... 130
LITERATURE CITED .................................................................................................... 160

viii

LIST OF TABLES

Table 1. Comparisons of sample size (N) and summary measures of genetic diversity
including coancestry (0) and effective population size for individuals sampled at different
developmental stages using each of three methods of gamete/larval collection over each
of three years. Effective population size estimates include calculations using the number

of males and females (N e), male and female reproductive variance (Nev), and coancestry

(Nee). Individuals were raised in one of two hatchery environments (streamside or
traditional) ........................................................................................................................ 131

Table 2. Comparisons of sample size (N) and summary measures of genetic diversity
including coancestry (0) and effective p0pulation size for offspring collected as naturally
produced eggs in each of two years. Effective population size estimates include

calculations using the number of males and females (N 6), male and female reproductive

variance (N CV), and coancestry (Nee). Individuals were raised in one of two hatchery
environments (streamside or traditional) ......................................................................... 132

Table 3. Comparisons of sample size (N) and summary measures of genetic diversity
including coancestry (9) and effective population size for offspring collecting as
dispersing larvae in each of three years. Effective population size estimates include

calculations using the number of males and females (N 6), male and female reproductive

variance (N (W), and coancestry (Ne6)- Individuals were raised in one of two hatchery
environments (streamside or traditional) ......................................................................... 133

Table 4. Reproductive success of males and females based on known pedigree
information for offspring produced from directed gamete takes and reared in two
hatchery environments (Streamside and Traditional) during each of 3 years. Variables
represent the mean i lSD ................................................................................................ 134

Table 5. Reproductive success of males and females based on genetically determined
pedigree information for offspring collected using two collection methods and reared in
two hatchery environments (Streamside and Traditional) during each of 3 years.
Variables represent the mean i lSD ................................................................................ 135

Table 6. Estimates of mean relatedness (rxy) of parent pairs for offspring collected using
three different gamete/larval collection methods and the adult spawning population (SP)
during 2005-2007. Collection methods included direct gamete takes (DGT), naturally

produced eggs (NPE) and dispersing larvae (DL). Mean rxy (:I: ISE) is presented along
the diagonal. Uncalculated values are represented with — ............................................. 136

ix

Table 7. Physical conditions of the Upper Black River during stocking events of three
different age classes of juvenile lake sturgeon. Physical variables including water depth
(m) and velocity (In/sec) are included for each downstream assessment site during each
release age ........................................................................................................................ 137

Table 8. Numbers released and total length (cm, mean i ISE) corresponding to age,
rearing environment, and gamete/juvenile collection method for three experimental
releases of juvenile lake sturgeon into the Upper Black River, Michigan. Letters (A, B, C)
correspond to statistically greater proportions of ﬁsh recaptured between the different
collection methods within each release age ..................................................................... 138

Table 9. The number of observations (Obs), average inter-observational distance traveled
(Mean 2}: 1 SD), and average inter-observational distance traveled (Mean :1: 1 SD) ﬁom
the release site for juvenile lake sturgeon surgically implanted with ultrasonic
transmitters and released into Black Lake, Michigan. Other characteristics of the
experimental ﬁsh represented include: individual ﬁsh ID number, hatchery rearing

location (Lee: S: streamside; T: traditional), total length (TL) and weight (W) .............. 139

LIST OF FIGURES

Figure 1. The Black Lake study site in Michigan, showing the adult spawning grounds,
the collection sites for naturally produced eggs C“), and the collection site for dispersing
larvae ................................................................................................................................ 140

Figure 2. Total number of adult lake Sturgeon captured by day on the spawning grounds
of the Upper Black River from 2005 to 2007 .................................................................. 141

Figure 3. Frequency distributions of inter-individual relatedness values (rxy) values for
adults producing offspring from A) each of three different gamete/larval collection
methods (Naturally Produced Eggs (NPE), Directed Gamete Takes (DGT), Dispersing
Larvae (DL)) and B) the total adult spawning population (SP) over three years (2005-
2007) ................................................................................................................................ 142

Figure 4. Estimates of inter-individual relatedness (rxy) values (Mean i ISE) between

adult female lake sturgeon concurrently breeding within distinct spawning zones in the
Upper Black River during each of three years (2005-2007). Numbers refer to the total
number of pairwise comparisons used to calculate the mean .......................................... 143

Figure 5. Estimates of inter-individual relatedness (rxy) (mean :I: ISE) between adult

female lake sturgeon concurrently breeding within distinct spawning runs in the Upper
Black River during each of three years (2005-2007). Numbers refer to the total number of
pairwise comparisons used to calculate the mean ........................................................... 144

Figure 6. Overall egg survival (mean d: ISE) to hatch for females incubated in two
different hatchery environments across two different years. Different sets of females were
used between years .......................................................................................................... 145

Figure 7. The effect of temperature on egg incubation time for eggs reared at a streamside
hatchery in 2005 and 2006 ............................................................................................... 146

Figure 8. Daily proportions of dead eggs infected by microbes (mean :t 18E) for 2
different lots of eggs (Early (E) and Late (L)) reared in streamside and traditional
hatcheries ......................................................................................................................... 147

Figure 9. Daily survival rates for lake sturgeon eggs incubating for a different number of
days until hatch ................................................................................................................ 148

Figure 10. Total numbers of larvae hatched (mean :t ISD) ﬁom females in two different

hatchery environments in A) 2005, and B) 2006. Different sets of females were used
between years ................................................................................................................... 149

xi

Figure 11. The effects of four treatments and time since hatch (days) on the change in
larval lake sturgeon total length (A) and yolk-sac utilization (B). Treatments correspond
to tanks that include refuge (cover), refuge and resources (cover-food), totally open tank
(open), and an open tank with resources (open-food) ..................................................... 150

Figure 12. Differences in total length (A) and yolk-sac utilization (B) over time among
larval lake sturgeon from six females reared at a streamside hatchery in 2006 .............. 151

Figure 13. Daily survival of juvenile lake sturgeon collected using three gamete/larval
collection methods (dispersing larvae (DL), naturally produced eggs (N PE), and direct
gamete takes (DGT)) reared in two different hatcheries (streamside and traditional) in
2005 ................................................................................................................................. 152

Figure 14. Growth of juvenile lake sturgeon reared in three different levels of habitat
complexity. Habitat complexity consisted of completely open, ﬁlled half way with large
rocks to simulate cover, and ﬁnally tanks that were completely covered ....................... 153

Figure 15. The Black Lake study site in Michigan, showing the release location and
sampling sites (14) for release assessments in the Upper Black River. Site 4 also
corresponds to the larval sampling site ............................................................................ 154

Figure 16. The proportion of juvenile lake sturgeon recaptured at downstream assessment
sites (Sites 14; Fig 1) that were reared in different hatchery environments. Ages 8, l3,
and 17 weeks are represented within a downstream interval respectively reading left to
right. Asterisks represent signiﬁcantly greater proportions of recaptured ﬁsh that were
reared at the streamside hatchery versus the traditional hatchery. Letters (A, B, C)
correspond to statistically greater proportions of ﬁsh recaptured at the different release
ages .................................................................................................................................. 1 55

Figure 17. The proportion of juvenile lake sturgeon captured at consecutive hours
following release into the Upper Black River at 8, 13, and 17 weeks of age .................. 156

Figure 18. Total length (means i- ] SE) of recaptured juvenile lake sturgeon over time
following release at 8, 13, and 17 weeks of age into the Upper Black River .................. 157

Figure 19. The effect of two predator classes (Fish, Crayﬁsh) on the proportion of
juvenile lake sturgeon observed on sandy substrate or that were swimming in the water
column. Gravel and large rock substrates were also available choices in this experiment
but are not included due to insigniﬁcant choice for these two habitat types. Predators
included rock bass (Ambloplites rupestrz's), smallrnouth bass (Micropterus dolomieue),
northern pike (Esox lucius) and the rusty crayﬁsh (Orconectes rusticus). Absent and
present refer to trials conducted with and without predators ........................................... 158

Figure 20. Survival (Mean :t 1 SE) of juvenile lake sturgeon and an alternate prey species

exposed to two predator species at two ages (smallmouth bass, Micropterus dolomieue,
and crayﬁsh, Orconectes rusticus) ................................................................................... 159

xii

INTRODUCTION

Lake sturgeon (Acipenserfulvescens) are a native ﬁsh of the Great Lakes that
have undergone signiﬁcant reductions in population abundance over the past century
resulting from over-harvest, habitat loss and degradation, and loss of habitat connectivity
(Holey et al. 2000). Currently, it is estimated that lake sturgeon are present at
approximately 1% of historical levels (Hay-Chmielewski and Whelan 1997). Lake
sturgeon restoration and protection has become a high priority for all state, federal, and
tribal management agencies throughout the Great Lakes. The lake sturgeon receives
protection to varying degrees throughout their range at the state and provincial levels,
listed as threatened, endangered, or of special concern (Welsh 2004). Management and
conservation strategies have been implemented throughout the Great Lakes that focus on
restoring remnant populations through supplementation of hatchery produced juveniles
(Holey et a1. 2000). Unfortunately, there are numerous impediments that complicate
current lake sturgeon restoration and recovery efforts, including the presence of dams on
rivers that historically supported spawning runs, habitat alterations, low spawner
abundance, and the species’ unique life history characteristics (late age at maturity,
infrequent spawning, and low recruitment rate). Lake sturgeon exhibit delayed sexual
maturation, ranging between 15 to 25 years (Houston 1987) which greatly complicate
efforts to document effects of hatchery supplementation on recruitment into the adult
breeding population. Once lake sturgeon mature, spawning occurs intermittently with

males spawning every one to two years and females spawning less ﬁequently, every three

to seven years depending on age (Auer 1999; Beamish et a1. 1996; F orsythe, unpublished
data).

Hatchery programs have been widely advocated as a viable means to enhance
declining wild ﬁsh populations, but are receiving increased scrutiny because of the
potential negative impacts of captive reared individuals to natural populations (Ford
2002; Araki et al. 2007a; Miller et a1. 2004). Criticisms have been raised based on
empirical evaluations of the relative effects of different management options including;
hatchery rearing environments (Berejikian et a1. 2000), age of release into the natural
environment (Paragamian and Kingery 1992), and gamete, juvenile, or broodstock
collection and maintenance protocols (Flag and Nash 1999). Currently, most scientiﬁc
data on hatchery ﬁsh are based on literature from a limited number of species that have
very different ecologies than lake sturgeon. Data on the complexities and inter-
dependencies among components of natural ecosystems have become more clearly
deﬁned for well studied ﬁsh such as salmonids (Fraser 2008). In light of recent ﬁndings,
empirical and experimentally-based evaluations of hatchery rearing and stocking
programs are needed to develop and incorporate new data based on the biology of lake
sturgeon into scientiﬁcally-based management programs.

Research is needed directed at determining the most effective culture and
stocking techniques that will repatriate lake sturgeon and also ensure the persistence of
remnant populations existing around the Great Lakes. My dissertation research
empirically determined the effects of different gamete/larval collection methods,
hatchery rearing environments, and stocking strategies commonly used by management

agencies on- juvenile lake sturgeon growth, survival, movements, and levels of genetic

diversity. In my ﬁrst chapter I investigate the effects of different gamete/larval collection
methods on levels of genetic diversity using modern molecular approaches. In the second
chapter I quantiﬁed and compared growth and survival rates of juvenile lake sturgeon
acquired using three different gamete/larval collection methods and reared in two
different hatchery environments. The hatchery environments I evaluated included a
streamside hatchery constructed on the natal river and a traditional hatchery environment
located on a non-natal water source. In chapter 3, I quantiﬁed survival and movements of
juvenile lake sturgeon reared in two different hatchery environments and released at
three different ages. In chapter 4, I quantiﬁed overwinter survival of age-0 lake sturgeon
using ultrasonic telemetry. In addition I examined overwinter survivorship between ﬁsh
reared in two different hatchery environments. In my ﬁfth chapter, I investigated the
effects of four different predators on survival, behavior, and habitat preference of three
age classes of juvenile lake sturgeon reared in two different hatchery enviromnents. I
discuss the signiﬁcance of my results in terms of designing and managing lake sturgeon

restoration programs throughout the Great Lakes.

Chapter 1
EVALUATION OF GAMETE AND LARVAL COLLECTION METHODS FOR
LAKE STURGEON; EFFECTS ON GENETIC DIVERSITY

INTRODUCTION

Hatchery supplementation has been employed as a management prescription for
the last century in order to augment existing stocks (Cowx 1994; Waples 1999),
mitigation for habitat destruction (Waples 1991), enhance ﬁshing opportunities through
introduction of desirable sport ﬁsh (Peck et al. 1999; Heidinger 1999), and to conserve
threatened populations (Waples 1991; Ryman 1991). Traditional hatcheries have
developed efﬁcient processes to maximize progeny production. Over the past decade,
conservation hatchery programs are increasingly being used to facilitate the recovery of
threatened or endangered species (Ryman et a1. 1995; Brown and Day 2002). The role of
a conservation hatchery is to assist in the recovery of population by minimizing the
probability that the genetic and ecological impacts of supplemental progeny on the wild
stock will be negative (Flagg and Nash 1999).

Due to the large numbers of ﬁsh produced in hatcheries and subsequently stocked
into natural environments, there is great concern about the potential impacts of hatchery
ﬁsh on native populations (Waples 1991; Utter 1998; Waples 1999; Lynch and O’Hely
2001). Increasing evidence for negative impacts of hatchery stocks on wild populations
has motivated research regarding the efﬁcacy of hatchery use for conservation and
management (W aples 1991; Ryman 1991; Hilbom 1992; Ryman et a1. 1995; Waples

1999; Hutchings and Fraser 2008). Recent data indicates that hatchery produced ﬁsh have

lower ﬁtness than wild ﬁsh (Araki et al. 2007a; Miller et a1. 2004) and have demonstrated
lower ﬁtness following release into the wild (Ford 2002; Araki et al. 2007b; Araki et a1.
2008). Concerns regarding the potential interactions of hatchery and wild populations
have led to evaluations of hatchery practices of maintenance of levels of genetic diversity
(Allendorf 1993). Genetic diversity is important for the long-term population viability by
preserving greater potential for adaptation to environmental change (Ford 2002).

The success of hatchery restoration programs is typically measured by total
numbers of progeny available for stocking or survival up to or following release (Brown
and Day 2002). As an alternate measure, successful conservation hatchery programs
should focus on maximizing the genetic diversity present in the adult breeding
population (Ryman 1991), and when possible, attempting to recreate natural reproductive
features of the species reproductive ecology by incorporating information that reﬂects
how variation in reproduction is apportioned naturally (e. g. locations or timing of
reproduction). There is increasing evidence that different methods of gamete take and
current hatchery practices can negatively impact conservation and management programs
(Bartron 2003; Page et a1. 2005). Different management options including how gametes,
juveniles, or broodstock are collected and maintained prior to release back into the
natural system (F lagg and Nash 1999) have been cited as important factors to retention
of levels of genetic diversity in hatchery progeny. Mating strategies have evolved in
natural systems that maximize ﬁtness and maintain genetic diversity. Reduced levels of
genetic diversity in captive environments have been attributed to the use of a small
number of breeders (Allendorf and Phelps 1980; Ryman 1991), as well as the differential

selection pressures of the captive environment (Hutchings and Fraser 2008). For

hatchery programs focused on species that do not use captive broodstocks there are many
factors such as population size, limited access to spawning adults, unique life history
characteristics (i.e. intermittent spawning, delayed maturity, skewed sex ratios), or
knowledge of species reproductive ecology that increase the probability that offspring
reared in hatcheries will represent a disproportionally small subset of the genetic
diversity represented in the adult breeding population.

Lake sturgeon (Acipenserﬁrlvescens), a native ﬁsh of the Great Lakes, has been
numerically depressed throughout their range due to anthropogenic factors (Hay-
Chmielewski and Whelan 1997). Management and conservation strategies have been
implemented throughout the Great Lakes that focus on restoring remnant populations
through supplementation of hatchery produced juveniles (Holey et al. 2000). Successful
use of hatchery supplementation for sturgeon conservation is dependent upon the degree
of domestication during rearing, adaptation to natural conditions following release, and
the extent hatchery-reared juveniles retain levels of genetic diversity present in the adult
population (Ireland et al. 2002; Secor et al. 2002).

There has been recent interest in the use of streamside hatcheries as a means of
minimizing domestication and facilitate imprinting (Holtgren et al. 2007). Streamside
hatcheries use water taken directly from natal rivers providing natural daily and seasonal
ﬂuctuation in water quality (i.e., temperature, oxygen, turbidity) and chemistry (i.e.,
dissolved organics, pheromones, predator cues etc). Unfortunately, despite advancements
in rearing technologies, there is still a critical need for studies evaluating the efﬁcacy of
available gamete/larval collection methods, how progeny collected from different

methods perform in different hatchery environments, and relative rates of loss of genetic

diversity.

Most scientiﬁc data on hatchery ﬁsh are based on literature ﬁom a limited number
of species that have very different ecologies than long-lived iteroparous species,
including lake sturgeon. Even for widely studied groups such as salmonids (Fraser 2008),
considerable effort over many years and by many investigators has been required to
accumulate information to formulate scientiﬁcally defensible hatchery-based restoration
programs. As data have become available, the complexities and inter-dependencies
among components of natural ecosystems have become more clearly deﬁned. Agency
logistical and ﬁnancial constrains must also be considered. It is necessary to develop and
incorporate new science that is based on empirical data on the biology of lake sturgeon
into management programs. Data from lake sturgeon collected in this study will allow
development of a scientiﬁc ﬁamework to be used to evaluate past supplementation
activities and predict outcomes of future management actions.

Collection of progeny for lake sturgeon hatchery and restoration programs
throughout the Great Lakes is limited to a few methods. Captive broodstock are not used
for lake sturgeon conservation due to logistical constraints imposed by the species large
body size, low abundance of wild populations (Holey et al. 2000), and genetic
differentiation between populations (DeHann et al. 2007). Without a captive broodstock
program collection of progeny for stocking is entirely reliant on gametes or larvae from
the wild. In order to successfully design conservation programs for this and other
species, comparisons are needed between available collection methods and different
hatchery rearing environments. Long-term success of these programs that rely on

supplementation will require that managers minimize relatedness between offspring

released and maximize the genetic diversity represented by available spawning adults.
The objectives of this study were to 1) use demographic and genetic variables to provide
a means of comparisons for three different methods of gamete/larval collection for lake
sturgeon, 2) determine the effect of two different hatchery environments (streamside and
traditional) on levels of genetic diversity within and among the different collection
methods, and 3) add to current knowledge of lake sturgeon reproductive ecology that can
be used to critically evaluate restoration programs that rely on hatcheries. This research
provides much needed guidance for managers involved in collecting and rearing lake
sturgeon progeny for restoration efforts. Results can also be used to improve the

effectiveness of lake sturgeon collection and culture techniques.

METHODS

Study Site and Hatchery Rearing

Research was conducted during each of three years (2005 to 2007) on the Upper
Black River in Michigan (Fig 1). The Upper Black River (see details in Smith and Baker
2005; Smith and King 2005) is a fourth order stream located in the northeastern corner
Michigan’s lower peninsula. The hydrology of the Upper Black River provides a unique
opportunity to enumerate a large proportion of the adults reproducing each year
(Forsythe, unpublished data), collect gametes from the stream substrate (Forsythe,
unpublished data), and to collect dispersing larvae (Smith and King 2005). The adult
reproductive season is comprised of different spawning runs. Adults migrate upstream to
spawning grounds, reproduce over several days and depart. Sequential groups comprised

of different subsets of adults arrive over a period ranging from 18-42 days. Timing of

8
m
/

female lake sturgeon reproduction is highly repeatable across multiple years (F orsythe,
unpublished data). Reproduction occurs in wadeable sections of the river which have
been delineated as distinct spawning zones based on a long-term data set collected from
this population (Fig. 1; F orsythe, unpublished data). Adults reproducing earlier in the
season spawn primarily in upstream spawning zones (Zones 1-3, Fig. 1) while ﬁsh
reproducing later frequent more downstream spawning zones (Zones 4-6, Fig. 1).
Detailed knowledge of the reproductive behavior of lake sturgeon in this system allows
for examination of genetic differences among different subsets of adults on both temporal
and spatial scales. Irnportantly, long-term ecological data provide a measure of the extent
of reproductive isolation between different spawning groups. We used capture-mark-
recapture and observational data to delineate different spawning aggregations in each of
three years. Spawning groups were assigned based on several consecutive days of zero
captures or a combination of capture data and observational data of spawning activity.
This included documenting whether females were actively releasing eggs or documenting

eggs deposited on the stream substrate following reproductive.

Gamete/Larval Collection Methods

Restoration of lake sturgeon or populations of any long-lived iteroparous ﬁsh
species requires a long and sustained program focused on preserving genetic diversity
and local phenotypic and life history adaptations. Different methods of collection focus
on different life history stages (eggs or larvae) capture diversity that can be quantiﬁed
based on empirically measurable variables including inter-individual relatedness,

heterozygosity, and numbers of contributing adults. Typically, lake sturgeon

supplemental progeny are collected through direct removal and fertilization of gametes
ﬁ'om spawning adults. Traditionally, fertilizations have been conducted with a relatively
few females mated with several males (Folz et al. 1983; Anderson 1984; Ceskleba et al.
1985; Pyatskowit et al. 2001) due to logistical challenges of river hydrology and reduced
population sizes which limit access to spawning adults (Holey et al. 2000). A second,
and less intrusive, collection method involves collecting larvae dispersing downstream
ﬁom the spawning grounds (Auer and Baker 2002; Smith and King 2005). This method
may prove to be closer to an unbiased representation of the relative reproductive success
of all spawning adults but has yet to be genetically assessed. A third potential method is
collecting eggs that have been naturally fertilized and deposited on the stream substrate
immediately downstream of spawning sites. This method has been used mainly to verify
successful reproduction (McCabe et al. 1990; Caswell et al. 2004). Recent work has
focused on the importance of environmental covariates including features of the stream
substrate on egg deposition and survival and has documented success collecting large
numbers of eggs (F orsythe, unpublished data). The effectiveness of collecting eggs in
this manner depends on speciﬁc knowledge regarding adult spawning site locations, but
represents a viable collection method in certain systems. Our design involved collections
of lake sturgeon gametes or larvae over a three-year period. Data allow for comparisons
between methods within and across years as well as examining the effects of rearing
environment. We quantiﬁed effort for each collection method by the number of
personnel hours needed by day over the entire sampling period each year.

I) Directed Gamete Takes (DGT): Adult lake sturgeon were captured on the

spawning grounds using large hand-held dip nets. We attempted to collect gametes from

10

all individuals. Eggs were removed by hand-stripping females captured while in the act of
spawning. Hand stripping involved applying pressure from the anterior section of the
abdomen to the posterior in order to extrude eggs from the uro-genital opening. Eggs
were kept in ovarian ﬂuid and placed in plastic bags and maintained at ambient river
temperatures. Milt was removed from individual males by applying pressure anterior to
the uro-genital opening, allowing the mil to pool, and using a 30ml syringe to collect it.
Milt was immediately placed on ice. Fertilizations were conducted within 12 hours of egg
collection. We used a partial factorial mating design for lake sturgeon where eggs from
each female were divided into equal lots and crossed with a maximum of two males
(depending on the available amount of eggs for each female) creating paternal half-sib
family groups of approximate equal size. Individual males were never used twice and
fertilizations occurred independently to avoid issues of sperm competition (Campton
2004). This mating design also increased the effective number of breeders compared to
single pair crosses (Busack and Knudsen 2007) and provided a quantitative genetic
framework to test for variation between males and females while accounting for
interactions between females (Vandeputte et al. 2004). We chose not to employ a full
factorial mating design, all males with all females, despite reported positive effects on
effective population size (Fiumera et al. 2004). Full factorial designs increase inter-
individual relatedness (Miller and Kapuscinski 2003.) which may pose a threat to
populations of lake sturgeon that are numerically depressed. Half of the fertilized eggs
ﬁom each paternal half-sib family were transported to a traditional hatchery and half

were maintained at a streamside hatchery for incubation.

ll

2) Dispersing Larvae (DL): Sampling for larval lake sturgeon larvae passively
dispersing downstream has been conducted in several studies to quantify recruitment and
chronology and duration of dispersal (Auer and Baker 2002; Smith and King 2005) and
represents a viable collection strategy. Systematic larval drift sampling was conducted
during a 5-hour period, beginning at dusk (2100) and ending in the early morning (0300).
The sampling location was located approximately 2km downstream ﬁom all spawning
areas on the Upper Black River (Fig. 1) allowing for collection of offspring ﬁom all
spawning groups. Sampling began seven days after the ﬁrst observation of reproduction.
We deployed ﬁve D-ﬁ'amed drift nets across the stream channel in equal intervals to
capture dispersing larval sturgeon. Nets were checked hourly and the larvae within each
net were enumerated. This design was replicated nightly each year for the entire drifting
period. Sampling was terminated after adult spawning ceased and zero larval captures
were observed for consecutive nights. Larval lake sturgeon captured each evening were
reared at the streamside hatchery until the end of the drifting period. The larvae were then
divided equally between the two hatchery rearing environments.

3) Naturally Produced Eggs (NPE): Systematic kick-net sampling was conducted
below observed spawning aggregations (Fig. 1) to collect eggs that were naturally
fertilized and deposited in the stream substrate. Transects were conducted across the
stream at 1 meter intervals. At each interval we conducted kick-net sampling for 10
seconds. Transects continued downstream in intervals of 5 meters until no eggs were
collected in consecutive transects. The number of sampled versus non-sampled locations
varied across years. Eggs were enumerated upon arrival at the streamside hatchery. Eggs

from this collection method were incubated and hatched at the streamside rearing

l2

environment because of pathogen concerns of the traditional hatchery. Larvae were then

divided equally between the two hatchery rearing environments.

Rearing Methods and Environments .

Lake sturgeon progeny from three different gamete/larval collection methods
were reared in two different hatchery environments. Half of the eggs or offspring from
each collection method were reared in natal water at a streamside hatchery constructed on
the Upper Black River (Fig. 1). The remaining eggs and larvae from each group were
reared in a non-natal water source, a state hatchery in southern Michigan (Michigan
Department of Natural Resources, Wolf Lake State Hatchery), representing a traditional
rearing environment. The streamside hatchery represented more natural rearing

conditions with temperature proﬁles mimicking that of the natal river. Ground water at

the traditional hatchery was heated to a constant 20°C.

Eggs were incubated and hatched using heath trays at both rearing environments.
Larvae produced from direct gamete takes were enumerated by paternal half sib family
groups at hatch and then reared separately by family for the duration within each
hatchery. Larvae and juveniles were reared in round tanks (diameter = 1.2m) that Were
divided to include two replicates of each of the three collection methods. This tank design
was replicated ten times at the streamside hatchery and six times at the traditional
environment. Daily husbandry activities, including cleaning and feeding, were consistent
between the two hatcheries. Fish ﬁom the three collection methods were reared until fall

ﬁngerling size (>20cm) and released into the Upper Black River.

13

Genetic Sample Collection

Genetic samples were collected for use in genetic analyses described below.
Dorsal ﬁn clips were taken from each adult captured during the enumeration of the
spawning run. Adult ﬁn clip samples (were dried and stored in individual envelopes at
ambient temperature. Juvenile mortalities were preserved separate by collection method,
hatchery environment, and year in 95% non-denatured ethanol. Prior to release, ﬁn clips
were taken from all surviving juveniles and stored in 95% non-denatured ethanol by

collection method, rearing environment, and year.

Genetic Analyses

A subset (5%) of juvenile lake sturgeon tissue samples were randomly chosen
from the naturally produced and larval drift collection methods. Samples were
proportional with respect to surviving/dead juveniles, the two hatchery environments, and
year of collection.

DNA was extracted ﬁom all adults and juvenile tissue samples using QIAGEN
DNeasy® kits (QIAGEN Inc.) according to the manufacturers’ protocol. DNA was
quantiﬁed using a Beckman DU® 7400 spectrophotometer and diluted to a constant
20ng/ul for use in Polymerase Chain Reactions (PCR). Individuals were genotyped at 12
disomically inherited microsatellite loci including LS68 (May et al. 1997), Aﬁ¢G68b
(McQuown et al. 2002), Spl 120 (McQuown et a1. 2000), on27 (King et al. 2001),
Aqu9, Aqu63, Aqu74, AﬁzGl 12, AﬁrGS6, AﬁrG160, Aﬁ4G195, and AﬁrGZO4 (Welsh et
al. 2003). PCRs were conducted in 25 pl volumes containing 100ng DNA, 10X PCR

Buffer (1M Tris-HCl, 1M MgClz, 1M KCl, 10% gelatin, 10% NP-40, 10% Triton-X),

l4

2mM of each dNTP, lOpmol of forward and reverse primer and 0.5111 Taq polymerase.
PCR conditions for each loci followed protocols outlined in the primary literature. PCR
products were run on 6% denaturing polyacrylarnide gels and visualized on a Hitachi

FMBIO II scanner. Allele sizes were scored using commercially available size standards

(MapMarkerTM, BioVentures Inc.) and based on standard samples of known genotype.

To minimize error, all genotypes were independently scored by two experienced lab
personnel and veriﬁed again after data were entered into electronic databases.
Furthermore, we re-ran a random subset (10%) of samples to determine an empirical per-

allele error rate for use in statistical analyses.

Statistical Analyses

Coancestry and Parentage: In closed populations such as Black Lake or in
populations restored primarily on the basis of hatchery-production, mean population
coancestry can be a direct measure of potential for future inbreeding. We calculated mean
coancestries (0) for offspring produced ﬁ'om fertilizations of eggs ﬁom known paternal
half sib families. Data on total number males and females spawned, the manner in which
eggs and rrrilt were crossed, and the total number of eggs fertilized and the subsequent
number of larvae at hatch were used to calculate 0. Mean 0 were estimated at both the
, egg and larval stage for both hatchery environments. Estimates were also calculated for
later juvenile stages at the streamside hatchery. Calculations of 0 were derived as
described by Chesser (1991). The number of individuals in a family (b), number of
families (11) and total numbers of adults spawned (N) was used to estimate mean 0 across

all half sib families as:

15

i132 — bl.
6 = i =1
4(N2 — N)

For offspring obtained ﬁom larval drift sampling or as naturally produced eggs
we estimated 0 among offspring using pedigree information derived from genetic
determination of parentage. Genetic determination of parentage was conducted separately
for each collection method across two sampling years (2006 and 2007). Parentage was
assigned using the likelihood-based approach in CERVUS Version 3.0 (Kalinowski et a1.
2007). Estimates of allele ﬁ'equencies, mean number of alleles per locus, and observed
and expected heterozygosities were calculated by CERVUS. Chi-squared goodness of ﬁt
tests were performed for each locus using CERVUS to test for deviations ﬁ'om Hardy
Weinberg expectations. The program CERVUS was also used to calculate the overall
probability of exclusion for the 12 loci. The simulation module within CERVUS was
used to generate critical values of likelihood ratios so that parentage could be assigned at
a given level of statistical confidence. Simulations were conducted with the rate of typing
error set at 0.02% which was empirically derived from our blind 10% genotyping error
check described above. Because we did not sample all possible parents during the annual
spawning runs we used the program PASOS 1.0 (Duchesne et al. 2005) to estimate the
proportion of parents sampled during the adult spawning run. Offspring were allocated to
either one or both parents of known sex in PASOS with the maximum number of offsets
between a parental and an offspring allele set conservatively to 0. The resulting
proportions for male and females were entered into the simulation settings in CERVUS.

Further analysis parameters for each simulation were as follows: 10,000 replication

16

cycles, the number of both female and male candidates (unique by year), and 99.54% of
loci typed which was empirically determined. Output from the simulation run was then
used in the parentage assignment module of CERVUS to assign offspring to a candidate
mother and father using the default stringent and relaxed conﬁdence intervals of 95% and
80% respectively. Paternity was assigned ﬁ'om the CERVUS output following two
criteria, a) the assignment of both parents was statistically signiﬁcant at either 80 or 95%
conﬁdence, or b) total exclusion between the offspring genotype and the genotype of both
parents. Offspring parent combinations meeting these criteria were then used in
calculation of mean 9 and in further analyses described below.

Reproductive Success: We calculated several measures of reproductive success
based on either known or genetically determined pedigree information ﬁom each
collection method. For directed gamete takes we calculated the mean (i 1 SD) egg
volume collected per female across each year and the number of larvae that hatched
across half-sib families. For all collection methods we calculated the number of mates for
both males and females. We used an analysis of variance (ANOVA) to test for
differences in the number of eggs obtained across females and sampling years. We used a
two-way ANOVA to test for signiﬁcant differences in numbers of larvae at hatch across
paternal half sib family groups in the two different hatchery environments. For offspring
collected as dispersing larvae or as naturally produced eggs we estimated the mean and
variance in male and female reproductive success (number of offspring per individual)
and mean numbers of mates for both males and females for each collection method across
each hatchery environment based on the genetically determined pedigree. Finally, we

examined whether females that we directly collected gametes from spawned naturally.

l7

This was to address concerns that direct gamete collection from spawning females may
cause undo levels of stress and increase the probability that they will abandon spawning.
We used parentage information estimated for the naturally produced eggs and dispersing
larvae in 2006 and 2007 to examine the percentage of females that we directly removed
gametes from that were represented in these samples. To test for differences in the
number of genetically assigned offspring per male and female between naturally
produced eggs and dispersing larvae within 2006 and 2007 we used a Fisher’s exact test.
We compared the number of offspring for all individual adults that spawned in the
particular year of question. We were interested if offspring ﬁom one adult of a given sex
were represented more by one collection method than the other. We calculated individual
heterozygosities from genotypes for naturally produced and larval drift offspring. Mean
differences in individual heterozygosities between collection methods within a year were
assessed using a Wald Test.

Eﬂective Population Size: We calculated three measures of effective population

size for each collection method, and for each hatchery rearing environment within each

collection method. Effective population sizes (Ne) based solely on the numbers of males

(Nm) and females (N!) used were determined following Wright (1931) as:

41vme
e=(Nm+Nf)

 

Secondly, we calculated a more fully parameterized measure of effective population size

using the effective numbers of males (Nem) and females (Nef) for each collection method

18

(Lande and Barrowclough 1987) utilizing our empirical estimates of male and female

reproductive variance as:

 

 

N _ (Nmkm-l)
em 02
k,.+ 1% —1
m
and
.N = (Nykf-d)
ef 02
kf+ kfk —1
f

respectively, where k is the mean number of progeny preduced by either the males (km)
or female (kf) for each collection method, and 02 is the variance in the number of progeny

for males or females. The variance population size (Nev) for each treatment was then

calculated (Lande and Barrowclough 1987) as:

 

Finally, we calculated coancestral effective population size (Neg, Chesser et al. 1993) for

each collection method. Neg is the number of breeding individuals consistent with

observed accrual of gene correlations over generations and was calculated as:

1
N = —
66 2A0

l9

where the change in mean 0 was calculated for any estimated made for the same
groupings over multiple time periods.

Relatedness: Relatedness coefficients (rxy; Queller and Goodnight 1989) were
calculated using the program Kinship 1.3.1 (Goodnight 2000). The estimates of

relatedness between individuals calculated as rxy differ from those calculated by

coancestry because coancestry calculates the probability of alleles being identical by

descent while rxy compares the degree of relatedness between individuals as compared to

the average intra-population relatedness. Allele frequencies were estimated using adult

genotypes from the Black Lake population. We calculated the rxy value for all parental

pairs for each collection methods and calculated the distribution of rxy values for the each

spawning population for each year. Forsythe (unpublished data) found repeatability for

spawning time to be high in female lake sturgeon so we calculated the distribution of rxy

for adult female lake sturgeon spawning in distinct spawning runs (temporal variation)
and within distinct spawning zones (spatial variation; Fig. 1). To test for mean
differences in relatedness made among assigned parents we used a permutation test
appropriate for similarity data. This analysis has been used previously for genetic
relatedness data (Lehmann et al. 1992; Ratnayeke et al. 2002; Wayne et al. 1991). Using
a SAS-based program (A. Saxton, University of Tennessee, unpublished software), the
observed difference between the means of the two groups was ﬁrst computed and then
data from the two groups were pooled. Pooled data was randomly subsampled 1000

times. Each Subsampling calculated the difference between the observed and theoretical

20

mean calculated ﬁom the pooled data and the one-tailed probability of the two means

being different was tested. We used a Mann-Whitney U test to test for differences in the

distribution of rxy values between females within different spawning zones and spawning

runs. Tests were conducted between spawning zones and runs within a year and between
the same spawning zones and runs across years with the null hypothesis being that the
samples were drawn from a single population.

F -Statistics: Estimates of PST (a measure of standardized variance in allele
ﬁ'equency) were calculated between collections of naturally produced eggs and dispersing
larvae using FSTAT version 2.9.3 (Goudet 2001 ). Per locus estimates of PST were tested
for signiﬁcance by jackkniﬁng across populations and the overall estimate of F51 between
the collection methods was tested for signiﬁcance by jackkniﬁng across loci. Finally, to
further address the issue of genetic differentiation between spawning runs we calculated

per locus estimates of PST for different spawning runs of adults within each year.

RESULTS

Collection Methods

We collected large numbers of eggs and larvae using all three collection methods
(Tables 1, 2, 3) with the highest totals realized using the DGT method. Numbers were
fairly consistent for both DGT and NPE across each year. However, larger variation was
observed in the DL samples among years despite relatively consistent adult spawning
numbers (n = 154, 234, 208 for 2005, 2006, and 2007, respectively; Fig. 2).

Microsatellite loci used showed moderate levels of polymorphism with 2 to 9

alleles observed per locus. Genotypic frequencies at all 12 loci did not deviate from

2l

Hardy-Weinberg expectations. Mean number of alleles per locus and observed
heterozygosity were 55/055 and 542/057 for 2006 and 2007 DL samples and 55/056
and 525/057 for NPE in 2006 and 2007. There were no signiﬁcant differences in
individual observed heterozygosities between naturally produced or larval drift samples
for either 2006 (df = 248, t = 1.97, P = 0.12) or 2007 (df= 122, t = 1.98, P = 0.85). We
had higher overall parentage assignment for the DL method compared to the NPE method
in both years. Assignment of offspring to both parents was 68% and 74% for the DL
method in 2006 and 2007 respectively while NPE had lower assignment rates of 57% and
21% in the same years.

We determined levels of coancestry (0) for lake sturgeon progeny obtained using
each of three collection methods as an indication of gene correlations among progeny
with known (DGT) or assigned (N PE and DL) parentage. For each year, estimates of 0
were consistently highest for DGT (Table 1) and decreased sequentially in NPE (Table 2)
and DL (Table 3) methods. For example, estimates in 2005 were >0.02 for individuals
collected using DGT compared to individuals collected using NPE in the same year

(0.014). Estimates for individuals collected using DL in 2005 were lowest (0.004).

Three estimates of effective population size (Ne, Nev, and Neg) were highest for
the DL (Table 3) collection method compared to both the DGT (Table 1) and NPE (Table
2) methods. This trend was consistent across both years. For example, in 2007 the
estimates of Nev were 38.6 for individuals collected using DGT, 59 for NPE individuals,
and a much higher estimate for individuals collected using the DL method (181.3). The

number of contributing adults was higher for offspring from the DL samples (Tables 5)

compared offspring collected using DGT (Table 4) and NPE (Table 5) and therefore

22

directly inﬂuenced the estimate of Ne, which is based entirely on numbers of males and

females. We documented higher variation in male and female reproductive success based

on individuals in the DGT method (Table 4) which lead to estimates of Nev being more
similar to the simpliﬁed estimate of Ne within years. Estimates of Neg were lower than. Ne
or Nev estimates for DGT offspring (Table 1) but were higher for both NPE (Table 2) and

DL (Table 3) samples than Ne estimates and varied with respect to Nev estimates.

Reproductive success is often measured by the number of offspring produced per
male and female. The number of offspring per male and female was much higher for
DGT method (Table 4) compared to samples obtained using NPE and DL (Table 5), as
anticipated given the large numbers of gametes collected from a relatively small number

of females. We documented signiﬁcant differences in the mean egg volumes collected

from females across years (ANOVA, F 2,40 = 4.44, P = 0.02). A ﬁsher’s exact test

revealed no signiﬁcant differences in the number of offspring per individual adult male or
female between the NPE and DL methods in either 2006 (Male: P = 0.15; Female: P =
0.95) or 2007 (Male: P = 0.25; Female: P = 0.77). There was no trend in the number of
mates per male or female based on collections using NPE or DL methods based on
genetic determination of parentage. The highest number of mates for either sex was for
several females in the 2006 and 2007 DL samples that were estimated to have
successfully reproduced with 6 males. The mean number of mates per male and female
was highest in the 2007 DL samples (Table 5) with an estimated 2.2 and 1.5 mates per

individual females and males respectively.

23

Permutation tests quantifying mean differences in distributions of relatedness

(rxy) values between collection methods within each sampling year revealed that the

offspring collected from the DGT method in 2006 had a lower mean rxy value than both

offspring collected using the DL method (P=0.038; Table. 6; Fig. 3) and the spawning

population (P=0.042; Table. 6; Fig. 3). There were no other signiﬁcant differences in the

distribution of rxy values between offspring collected using different collection methods

in each of the three years (P>0.05; Table. 6; Fig 3). Though not statistically signiﬁcant,

the estimates of rxy for DGT offspring in 2005 and 2007 were higher than all other

collection methods. This is comparable to the results found with the pedigree estimates of
0 between the three groups.

Multilocus estimates of PST between individuals collected using the NPE and DL
methods within a year were low signiﬁcant indicating that offspring represented different
subsets of adults or different adult contributions. The overall PST value between
individuals collected using the NPE and DL methods in 2006 was 0.019 (95% CI 0.009-
0.029). In 2007, the estimate of PST was 0.016 (95% CI 0.008- 0.024) between offspring
collected using the NPE and DL collection methods suggesting that different subsets of
the pool of available offspring were sampled using the different methods.

Effort expended to collect samples using each collection method was consistent
across years. However, effort differed substantially among collection methods. Directed
gamete takes ﬁom spawning adults required approximately 24 personnel hours per day
for the duration of spawning. In addition, fertilizations of paternal half sib family groups

required approximately 1 hour per female. Collecting eggs using the NPE method ﬁom

24

the stream substrate took approximately 30 personnel hours total for each year. NPE
samples were collected ﬁ'om two different spawning areas in each of the three years (Fig.
1). DL sampling consisted of 10 personnel hours per day for the entire duration of drift.
Drift duration was 26, 31, and 36 days for 2005 to 2007 respectively. These estimates of
effort for the three collection methods reﬂect collection only and do not include
personnel required to perform daily husbandry duties (egg incubation, feeding,
mortalities, water quality etc) in the two hatchery environments (See details in Crossman

2008, chapter 2).

Rearing Environment
Proportions of larval hatching ﬁom paternal half-sib family groups in the DGT
method was signiﬁcantly higher at the streamside hatchery in 2006 compared to the

traditional hatchery environment in the same year as well being higher than hatch in 2005

(ANOVA, F 3,42 = 3.39, P = 0.03). For offspring collected using the DGT method we

estimated consistently higher 0 at the traditional hatchery compared to the streamside
hatchery during both 2005 and 2006. Estimating the change in 0 at different life stages is
an important variable in supplemental progeny reared in hatcheries. We documented an
increase in 6 between the egg stage and larval stage for offspring collected using the
DGT method across all years and both hatchery environments (Table 1). Furthermore, we
documented the same increasing trend in 0 across four different stages at the streamside
hatchery environment over 2 years (Table 1).

The number of adults contributing to offspring within the different hatchery

environments was similar in the controlled DGT crosses (Table 4) but varied

25

considerably in the NPE and DL collection methods (Table 5). This is further reﬂected in

the estimates of Ne between the rearing environments. In both 2005 and 2006 all three

estimates of effective population size for DGT were higher at the streamside hatchery
compared to the traditional hatchery (Table 1). In 2006 all three measures of effective

population sizes were higher at the streamside hatchery for NPE samples. In the DL

samples during 2006 estimates of Ne and Neg were higher at the traditional hatchery

compared to the streamside hatchery. However, taking measures of reproductive variance

into account revealed a higher estimate of Nev for DL samples at the streamside hatchery

compared to the traditional hatchery in the same year.

Adult Characteristics

Duration of spawning varied among years. Spawning adults were captured over
periods of 23, 37, and 40 days during 2005, 2006, and 2007, respectively (Fig. 2). Sex
ratios (number of males per female) of the adult spawning population were 2.38, 2.55,
and 2.28 respectively for 2005, 2006, and 2007. The number of distinct spawning groups
differed among years with 2, 4, and 3 distinct reproductive events in 2005-2007,
respectively (Fig. 2). We sampled a high proportion of the adult spawning population in
2006 and 2007. PASOS estimates of the proportion of spawning adults sampled during
the adult spawning run for both NPE and DL samples within each year were higher for
DL samples (2006 Males (M):0.91, Females (F):0.93; 2007 M:0.93, F :0.96) compared to

NPE samples (2006 M:0.86, F:0.89; 2007 M:0.97, F:0.76).

26

 

Individual heterozygosities of adults did not differ signiﬁcantly between 2006 and
2007 (df = 440, t = 1.97, P=0.2) reﬂecting the large numbers of spawning adults. Pooled
individual heterozygosities from individuals collected using each of the collection
methods were not signiﬁcantly different ﬁom pooled adult heterozygosities (df = 814, t =
1.96, P = 0.82; mean of 0.55). However, adults do not all spawn at the same time and in
the same location. This aspect of the adult reproductive behavior has implications when

sampling based on DGT and NPE is conducted. We found signiﬁcant differences in mean

rxy values among adults from different spawning zones in 2007 (Fig. 4), with adults

spawning in zone 2 having a signiﬁcantly higher mean rxy value when compared adults

spawning in zones 4 and 5 (Mann-Whitney U test, P < 0.001 respectively). There were no

other signiﬁcant differences in mean rxy values between zones within the two other

sampling years. Tests for differences in mean rxy values between adults spawning in

different spawning zones across different sampling years resulted in signiﬁcant difference

between zone 2 in 2006 and 2007 (Mann-Whitney U test, P = 0.0.039). For spawning

runs within a year, we found signiﬁcant differences in mean rxy values only within 2007

(Fig. 5). Adults spawning in the second run had signiﬁcantly higher mean rxy estimates

than adults spawning in both the ﬁrst and third run during 2007 (Mann-Whitney U test, P

< 0.001 and 0.004 respectively). Between years, adults spawning in the second run in

2007 also had a signiﬁcantly higher mean rxy than both 2005 and 2006 (Fig. 5, Mann-

Whitney U test, P = 0.026 and <0.001 respectively).

27

We estimated PST to determine if there was a difference in the variance in allele
frequency between adults spawning at different times within a year. In 2005 there was no
signiﬁcant difference between the two spawning runs observed during that year (FST =
0.001, CI: -0.024, 0.026; Fig. 3). In 2006 the ﬁrst run of spawning adults was not
signiﬁcantly different from the second group (FST = ~0.008, CI: -0.010, -0.006). In 2007
the ﬁrst group of spawning adults was signiﬁcantly different from the last group of
spawning adults (FST = 0.012, CI: 0.002, 0.022) which implies that managers should use
caution when sampling ﬁom a limited portion of the total spawning season.

Parentage data from NPE and DL offspring identiﬁed that 67% and 60% of adult
females that we directly removed gametes from reproduced naturally in 2006 and 2007
respectively. Direct egg takes do not detract from a female’s ability to spawn naturally.
Thus, females contributing gametes based of direct gamete takes do contribute to natural

recruitment and mate with different males then used for hatchery crosses.

DISCUSSION

Hatchery pro grams require empirical evaluations of the efﬁcacy of different
collection methods for progeny that are to be reared and released into the natural
environment. Lake sturgeon restoration programs throughout the species range, including
the Great Lakes, have been initiated (Holey et al. 2000) despite considerable uncertainty
regarding the effects gamete/larval collection methods have on levels of genetic diversity
observed in supplemental juveniles. If conservation hatchery programs are to be
emphasized in lake sturgeon recovery planning, then alternative progeny collection

methods should be evaluated empirically so the beneﬁts of alternative management

28

actions can be quantitatively assessed. The relative beneﬁts and risks associated with
each collection method provide general guidelines that have direct bearing on future

restoration efforts for lake sturgeon and other long-lived iteroparous ﬁsh species.

Collection Methods Affect Measures of Genetic Diversity in Progeny

Over the past few decades there has been considerable concern expressed over
artiﬁcially increasing wild population abundance and effects on long-term genetic
variation (Tringali and Bert 1998). Molecular methods have been used in captive
broodstock programs to determine pedigrees in an effort to minimize levels of coancestry
between adults used in artiﬁcial crosses (Ballou and Lacy 1995; Wang 2004). However,
in programs that use limited wild broodstock of unknown pedigree there is a need to
determine the degree of genetic diversity that can be retained using collection measures
that focus on different life history stages. Comparing measures of genetic diversity of
progeny obtained using three collection methods allows for quantiﬁcation of levels of
genetic diversity captured conditional on numbers and the reproductive ecology of
available adults, and provides a framework to assess which method might be most
effective in retaining genetic diversity present in the wild population. Each method
evaluated provided large numbers of progeny for use in hatchery programs and
supplemental stocking.

Collections of gametes directly from spawning adults provided the highest initial
abundance of progeny (Table 1). However, offspring did not reﬂect the genetic diversity
of the wild population as did collections of naturally produced eggs or dispersing larvae.

We estimated coancestry, or levels of gene correlation, for individuals acquired using

29

each collection method to identify the method(s) that minimized levels of coancestry and
therefore future levels of inbreeding. Estimates of coancestry were higher for offspring
ﬁom directed gamete takes compared to both naturally produced eggs and dispersing
larvae (Table 1). The mechanism driving differences can be attributed to several factors.
First, the total number of adults contributing to offspring increased from directed gamete
takes to naturally produced eggs to dispersing larvae samples (Table 2). Reproductive
variance was high between females in the hatchery environment. Furthermore, hatchery
environments typically increase both fertilization success and survival at critical life
history stages (eggs and larvae) when compared to survival in the wild (Secor and Houde
1998). Though the number of adult females that we successfully spawned was quite high
relative to other lake sturgeon studies throughout the great lakes (Folz et a1. 1983;
Anderson 1984; Ceskleba et al. 1985; Pyatskowit et al. 2001), adult numbers were
comparably lower relative to the numbers of adults that provided offspring captured
during the period of larval dispersal. Naturally produced eggs were taken over relatively
small spatial scales from substrates associated with locations of two spawning groups
each year. Therefore, offspring had fewer contributing adults when compared to the
larval drift samples. Greater effort could result in reproductive contributions from more
adults. Irnportantly, naturally produced eggs and dispersing larvae samples were analyzed
ﬁ'om tissue samples collected during the juvenile stage. Therefore, these estimates of 0
refer to juveniles at older ages than those estimates produced for DGT offspring at both
the egg and larval stages. These estimates could have increased during the rearing process
if different families had different competitive abilities (and survival) in the hatchery

environment. Even though we observed no difference in offspring heterozygosities

30

between the three collection methods, both the NPE and DL methods resulted in lower
mean coancestry and should considered for populations where a limited number of adults
are available for gamete collection.

Parentage assignment based on molecular markers is an effective approach to
explore mating systems of species that are otherwise hard to study (Avise 2004). We used
a combination of programs to assign parentage to sampled offspring from the naturally
produced eggs and dispersing larvae methods. Allocation of offspring to candidate
parents with statistical conﬁdence can be difﬁcult in open systems where the proportion
of sampled adults is unknown. Lower assignment rates for NPE offspring was a product
of the proportion of adults sampled estimated using PASOS. The estimate of the
proportion of adults sampled was more sensitive for the NPE offspring because we were
sampling eggs ﬁom a subset of the spawning females each year or spawning run.
Unsampled females associated with spawning groups depositing NPE’s for our
collections could result in over estimation of the proportion of unsampled females (e. g.
only 76% of females sampled in 2007). This in turn limited our ability to assign NPE
offspring, particularly in 2007. The proportion of sampled adults was likely closer to the
true value for the dispersing larvae (>90% of both sexes each year) as we sampled
offspring ﬁom the entire reproductive period. Estimates of coancestry are contingent on
parentage assignments. However, despite inconsistency in assignment rates for offspring
from the NPE and DL collection methods, the rank order of our results remained
consistent across years

Effective population size is a predictive measure of generational changes in allele

frequency, heterozygosity, and inbreeding that has been widely used in ﬁsheries

31

management when supplementing wild populations with large numbers of hatchery
produced progeny (Ryman and Laikre 1991). We documented differences in three
measures of effective population size of lake sturgeon progeny collected using three
collection methods. Estimates were considerably higher for the DL samples compared to
both the NPE and DGT samples. Estimates of effective population size are inﬂuenced by
aspects of a species biology. Greater numbers of adults contributed to individuals
collected using the DL method in each year. The most fully parameterized estimate of
variance effective size which incorporates male and female reproductive variance is
likely the most reliable indicator for hatchery programs. The importance of keeping
family groupings separate through the rearing process so that individual contributions can
be determined ﬁom each pairing of adults has been emphasized in previous work
(Allendorf 1993; Fraser 2008). Logistically, this may be challenging due to high numbers
of breeders, and limited tank space. However, managers should recognize that variance
will be high and projections of effective population size based just on numbers of males
and females will lead to overestimation.

Estimates of PST are represent a measure of genetic differentiation among groups
(Whitlock and McCauley 1999), and were used as a measure of deviance in parental
contributions of offspring collected using the three different methods. Multilocus
estimates of F ST between offspring obtained from collection methods NPE and DL were
signiﬁcantly different from zero indicating that different subsets of adults were
contributing to offspring collected using each method and the proportional contribution
of adults varied. Genetic differentiation between the two groups of offspring indicates

methods of gamete larval collection should sample from many spawning areas (adequate

32

spatial coverage) or spawning runs (adequate temporal coverage) to increase probabilities
of retention of levels of genetic diversity present in the breeding adult population.

In populations of seasonally-breeding species, individuals breeding at similar
times will be more likely to mate with one another than with individuals breeding at
different times. Temporal assortative mating can be further reinforced if habitats used for
reproduction also vary throughout the reproductive season. If timing of reproduction is
heritable, and offspring share similar tendencies, multi-modal distributions of spawning
times can develop as observed in Black River lake sturgeon. Multiple spawning episodes
associated with different habitats tied to seasonal variation in thermal regimes have been
documented in lake sturgeon (Auer 1996). If selective regimes imposed by temporally
and spatially variable environmental conditions affect offspring traits and ﬁtness,
covariation between adult spawning time and juvenile traits can develop (Hendry and
Day 2005). Differences in allele ﬁequency between adults spawning at the same time and
in different locations should be considered and gametes collected from adults from
multiple locations and times to maximize retention of levels of genetic diversity
represented in neutral genetic markers and adaptive phenotypic traits.

Collecting gametes directly ﬁom spawning adults for use in sturgeon hatchery
conservation programs may provide the numerical abundance required. However, there
are several important aspects to consider. Current and past practices have been to remove
gametes from a female, fertilize with two or more males, and transfer and incubate eggs
in large mixed groups to limit the number of incubation vessels and therefore reduce
maintenance. Duration of egg takes around the Great Lakes has traditionally been

dependant upon a number of factors including the number of spawners present, and the

33

total amount of eggs required. Hatchery limitations (including space and personnel) may
dictate the number of gametes collected. Sturgeon are highly fecund (Bearnesderfer and
Farr 1997) allowing a relatively few number of adults to contribute a high proportion of
the supplemental progeny if absolute gamete number is the sole criteria. In our
experience, collection of milt form a large proportion of males is possible and tagging
ensures that the same male is not reused twice. We recommend that programs collecting
gametes form spawning adults sample in a spatial and temporally comprehensive manner.
If logistical constraints limit the ability to monitor families through the entire hatchery
rearing process then equalizing family contributions at the egg stage and if possible at

hatch should be an important consideration for conservation programs (Allendorf 1993).

Rearing Environment Aﬂects Measures of Genetic Diversity

Hatchery rearing environments have been shown to result in supplemental
progeny that are both genetically and biologically inferior to their wild counterparts
(Waples 1991; Ford 2002; Araki et al. 2007a; Miller et a1. 2004). We compared two
rearing environments, a traditional hatchery and a more novel streamside hatchery. We
found that estimates of 0 for DGT offspring were higher at the traditional hatchery
compared to the streamside. Equal numbers of eggs from each paternal half-sib family
group were represented at both hatcheries. However, higher proportions of eggs hatched
at the streamside hatchery among family groups compared to the traditional hatchery. We
documented an increase in 0 for DGT offspring at the time of hatch relative to estimates
based on the relative number of eggs fertilized. Consistently higher 0 was documented

for the traditional hatchery environment across years, resulting from differential survival

34

among family groups during the embryonic stage (Crossman chapter 2). Coancestry
increased over time at the streamside hatchery resulting over 4 sampling periods in each
of two years. Again, differential survival of larvae and juveniles among family groups
throughout the rearing process lead to increases in 0 (Crossman chapter 2). Studies on
other species (Herbinger et al. 1995; Unwin et al. 2003) have indicated family group
accounts for a large proportion of the variation in survival during early life stages.
Differences between hatchery rearing environments could inﬂuence survival as different
genotypes could be favored in different environments. We documented high variation in
reproductive success in adults used to produce offspring in the DGT collection method
within both hatcheries across both years. Finally, assignment rates within hatchery
rearing environments varied by collection method and location but our overall results

remained consistent across years.

Adult Reproductive Ecology

Examining the reproductive ecology of species such as lake sturgeon, which
broadcast gametes into the open water with no post-ovulatory parental care, is
challenging because of difﬁculties in designing sampling protocols for relevant groups of
offspring. Molecular-genetic markers have for the examination of reproductive patterns
in ﬁshes that utilized different reproductive tactics including broadcast spawners (Avise
et a1. 1992; DeWoody and Avise 2001; Avise et al. 2002). The ability to combine
molecular markers with ﬁeld data allows for the examination of aspects of a species
reproductive ecology that are otherwise uncertain (DeWoody and Avise 2001).

Information of the relative number of contributing adults, the number of mates, and the

35

number of offspring produced are variables that have eluded sturgeon ecologists. We
were able to use molecular methods to examine these variables using two different
methods of collection that represent offspring produced through natural reproductive
behaviors. Though the number of mates was controlled for DGT in the hatchery
environment, estimates of the number of mates per male and female among all three
collection methods were not widely different (Table 2). Adult males and females
contributing to the 2007 DL samples had the highest number of mates. Total natural
recruitment to the larval stage was much lower in this year compared to the other two
years. Offspring that were part of families that had very low initial abundance may have
been missed during sampling due to low representation or in the selection of genetic
samples to analyze. In the DL samples we documented individual females mating with as
many as six different males. The largest number of mates for one individual male was 3.
DeHaan (2003) found through genetic determination of parentage with a larger sample
size that females and males mated successfully with as many as 11 and 5 different
individuals, respectfully. Female lake sturgeon spawning in the Upper Black River have
been observed to be surrounded by several males (Forsythe, unpublished data), a
behavior that has been noted in other systems (Bruch and Binkowski 2002; Kempinger
1988). At the time females release their eggs, several males compete to position
themselves close to females to increase the likelihood of fertilizing the eggs (Bruch and
Binkowski 2002). However, the relative reproductive success of individual males has
been difﬁcult to assess. We used molecular markers to document fundamental aspects of

this species reproductive ecology and our ﬁndings complement behavioral observations

36

of reproduction recorded in the wild. Furthermore, this represents one of the ﬁrst attempts
to examine lake sturgeon reproductive behavior using a molecular approach.

Sampling methods that represent spatial and temporal aspects of a species
reproductive ecology will help to preserve adaptations that may have evolved to different
spawning enviromnents (Hendry and Day 2005). Heritability for either spawning location
or spawning time is a component of diversity that has been relatively unconsidered as of
yet in sturgeon. F orsythe (unpublished data) documented the timing of spawning to be
highly repeatable for individual female lake sturgeon. Other species have been shown to
have an increased heritability for both spawning time and location which has lead to
adaptive divergence, even within a population (Hendry and Day 2005). If spawning time
is a heritable trait in lake sturgeon then collecting gametes from only one subsection of
the entire spawning season may compromise future population adaptive potential. We
documented differences in estimates of relatedness among adult female lake sturgeon
within years across spawning locations and spawning runs. Furthermore, we documented
these differences in relatedness across years. These results indicate that adequate
sampling of offspring from adults ﬁom the entire spawning season will increase diversity
and decrease mean relatedness between progeny collected. Future studies that seek to
preserve the genetic diversity of lake sturgeon populations should design sampling
protocols accordingly.

Information regarding sex biased dispersal of adult lake sturgeon between
spawning locations and times is difﬁcult to estimate given the logistics of following
individuals. It has been documented that approximately 15% of adult males and 0% of

adult females spawn in multiple runs within a year in the Black Lake system (Forsythe,

37

unpublished data). Estimates of F ST were different for adults spawning at different
locations and at different times ﬁrrther reﬂecting low rates of dispersal between groups.
For lake sturgeon, dispersal on the population scale is very low due to the species
philopatric nature and the spatial genetic structuring of populations throughout the Great
Lakes (DeHaan et al. 2006). Further knowledge on dispersal rates between spawning
groups would be important to improve our understanding of the reproductive ecology of

lake sturgeon and to help deﬁne sampling strategies for progeny.

Handling Does Not A ffect Natural Reproduction

Handling-induced stress can have wide ranging effects on ﬁsh including nest
abandonment (i.e., several species of bass (Suski et al. 2003; Hanson et al. 2007; Siepker
et al. 2007) and interruption of spawning rrrigration due to incidental net capture or
tagging (i.e., sturgeon species including stellate (Acipenser stellatus, Kynard et a1. 2002),
shortnose (Acipenser brevirostrum, Moser and Ross 1995), white (Acipenser
transmontanus, Schaffter 1997), and green sturgeon (Acipenser medirostris, Benson et al.
2007). Observations of interrupted upstream migration of sturgeons postulated abandoned
of spawning due to stress incurred during capture and tagging. The majority of these
observations were been made based on tagging and telemetry data. It is still unknown
whether females that are subjected to partial gamete removal, either through hand
stripping or induced ovulation, continue on to spawn naturally after incurring this stress.
We documented through genetic determination of parentage that a large number of the
females used for direct removal of gametes successﬁrlly spawned in the wild either

before or after our interruption. Our estimate should be viewed as an underestimate given

that we only genotyped 5% of offspring samples. Results are encouraging for gamete
removal from actively spawning females. High fecundity in this species allows for a

relatively small amount of eggs to be taken for hatchery rearing.

Applicability of Results to Other Systems

Application of methods evaluated in our study in other systems will require
knowledge regarding the timing, duration, and locations of spawning activity, which will
vary among systems. Sturgeon typically spawn in large, deep, fast ﬂowing rivers that
may limit the choice of progeny collection methods. The Black Lake system is a unique
system because of access to a large number of spawning adults during the entire
reproductive season. Other systems may not be as conducive for collecting gametes and
might expect to expend even greater effort to do so. The number of eggs we obtained
through direct gamete takes represents a maximum as we attempted to collect gametes
ﬁ'om all spawning females, some on multiple occasions. Hand stripping was only
effective for females caught while releasing gametes, which might not be an efﬁcient
technique in larger systems where adults are caught by methods other than dip nets. In
this situation, adults would need to be held and protocols for collecting gametes may
have to follow several available methods required to induce ovulation (Doroshov et al.
1983; Parauka et al. 1991; Doroshov et al. 1997; Webb et al. 2002). All three of these
collection methods are labor intensive requiring both ﬁeld and hatchery components.
Greater numbers of naturally produced eggs could have been obtained if additional effort
was made to increase the number of spawning sites sampled. Collection numbers of

dispersing larvae could have been higher if additional nets and personnel were employed.

39

Furthermore, collecting dispersing larvae has been shown to be transferable and effective
in a number of rivers (Kynard et al. 1999; Auer and Baker 2002; Smith and King 2005).
Collecting eggs, whether from spawning adults or from the stream substrate, requires
effort during the incubation stage to reduce mortalities and improve hatching success.
Capturing dispersing larvae bypasses the incubation stage by collecting offspring that are
developed to the extent where they can begin feeding immediately upon entry into the
hatchery environment. One positive aspect of rearing progeny at a streamside hatchery is
that the facility can be designed based on the restoration needs of a speciﬁc system or

population.

CONCLUSIONS

To maximize genetic diversity in offspring used in supplementation, programs
should focus on developing collection methods that incorporate aspects of the species
reproductive ecology to ensure they adequately capture the genetic diversity of the adult
breeding population. Managers should acknowledge that populations are composed of a
mixture of individuals that reproduce at different times and in different locations. For
lake sturgeon, a species that has undergone signiﬁcant reductions in abundance,
conservation programs should be designed based on rigorous evaluations of the most
effective methods required to collect, rear, and stock progeny back into the natural
environment. We found that collecting dispersing larvae downstream of the spawning
grounds minimized estimates of coancestry while at the same time maximizing effective

population size. Difference between the hatchery rearing environment and the wild have

40

the ability to inﬂuence the genetic variation observed in the supplemental progeny.
Methods and results described in this study provide a framework for evaluating
alternative strategies for managers designing and implementing conservation programs

for lake sturgeon and other long-lived iteroparous species.

4l

Chapter 2
COMPLEXITY OF REARING ENVIRONMENT AF F ECTS SURVIVAL AND
GROWTH OF LAKE STURGEON EMBRYOS, LARVAE, AND JUVENILES

INTRODUCTION

Declines in numerical abundance have been described for the majority of ﬁsh
populations’ worldwide (Myers and Worm 2003). Natural populations are increasingly
augmented by releases of hatchery offspring. Hatchery supplementation has been used
as a management prescription to augment existing stocks (Levin et al. 2001), mitigation
for habitat destruction (W aples 1991 a), enhance ﬁshing opportunities through
introduction of desirable sport ﬁsh, and conservation (Waples 1991a; Ryman 1991). The
ability to raise large numbers of ﬁsh at all life history stages in hatchery environments
bypassing bottlenecks imposed by egg of size-speciﬁc mortality (Secor and Houde
1998).has fostered widely held beliefs by the public and managers of hatchery abilities to
compensate for losses due to both ﬁshing and natural mortality.

Due to the large numbers of ﬁsh produced in hatcheries and stocked into natural
environments, there is great concern about the potential impacts of hatchery ﬁsh on
native populations (W aples 1991a; Thomas and Mathisen 1993). Increasing evidence for
negative impacts of hatchery stocks on wild populations has prompted much discussion
and research regarding the efﬁciency of hatchery use for conservation and management
(W aples 1991; Ryman 1991; Hilbom 1992; Ryman et al. 1995; Waples 1999). Recent
evaluations of hatchery practices as they relate to management goals to maintain genetic

diversity have raised concerns regarding the negative effects on individual ﬁtness and

42

population sustainability if hatchery ﬁsh interact with wild counterparts (Araki et al.
2007a; Miller et al. 2004; Ford 2002; Araki et al. 2007b; Araki et al. 2008). Concerns
have prompted calls for research directed at evaluating the efficacy of different culture
techniques to produce ﬁsh that are ecologically and genetically compatible with intended
release environments.

Conservation hatchery programs have been developed for a few species that have
become listed as endangered, threatened, or have the potential to become threatened due
to habitat loss or alteration (Flagg and Nash 1999; Maynard et al. 2004). The role of
conservation hatcheries is to restore populations through supplemental stocking by
lessening the genetic and ecological impacts of hatchery released ﬁsh on wild
populations. Many of these programs have developed ﬁsh culture techniques that enable
hatcheries to produce ﬁsh with more wild-like characteristics that result in increased post-
release survival (Flag and Nash 1999; Maynard et al. 2004). Techniques include rearing
supplemental ﬁsh in semi-natural habitats. Semi-natural rearing incorporates components
of the natural environment which may include adding structure and cover into ﬁsh culture
tanks or exposing ﬁsh to a more natural ﬂuctuation in water temperature. Many different
semi-natural culture techniques have been suggested based on the assumption that rearing
ﬁsh hatchery environments that emulate natural environmental conditions will promote
expression of traits that improve probabilities of survival following release into the wild
(Wiley et al. 1993; Olla et al. 1994; Maynard et al. 1995; Olla et al. 1998; Brown and
Laland 2001). However, alternative culture techniques should be developed and

evaluated on a species by species speciﬁc basis. Research aimed at empirically evaluating

43

the effectiveness of natural culture techniques will be important for the restoration of
species of conservation concern.

The lake sturgeon (Acipenserﬁrlvescens) is a native ﬁsh in the Great Lakes and
has undergone signiﬁcant declines in numerical abundance over the past century due to
anthropogenic factors including overﬁshing, habitat destruction, and the presence of
dams on many rivers that historically supported spawning runs impacting access to
historical spawning grounds (Holey et al. 2000). Current lake sturgeon abundance in the
Great Lakes is believed to be less than 1% of historic levels (Hay-Chmielewski and
Whelan 1997). The lake sturgeons unique life history characteristics (late age at maturity,
infrequent spawning, low recruitment rate) combined with low abundance of
reproductive adults complicates recovery efforts. Management and conservation
strategies have been implemented throughout the Great Lakes that focus on restoring
remnant populations through supplementation of hatchery production and release of
juveniles (Holey et al. 2000).

Considerable uncertainty exists among lake sturgeon managers and researchers
about the impacts of current lake sturgeon culture techniques and stocking efforts on
remnant populations. A variety of techniques are currently implemented throughout the
region that differ in progeny collection methods, (larval collection, direct gamete takes)
and rearing environments (streamside versus traditional hatcheries). The variety of
methods currently used by managers and the lack of data collected on the success of each
strategy reﬂects of the lack of consensus of the most effective methods. Currently,
management prescriptions for lake sturgeon hatchery production are largely adopted

from other well-studied species (e.g., salmonids). However, the reproductive ecology

44

and early life history of lake sturgeon are very different from salmonids. Therefore,
rehabilitation strategies should be designed, experimentally tested, and developed for
lake sturgeon rather than adopted from a species with dissimilar life histories.
Increasing the effectiveness‘of culture techniques for lake sturgeon will help
guide current and future restoration programs. There is a critical need to determine if
rearing lake sturgeon supplemental progeny in different hatchery environments has an
effect on stage speciﬁc survival. We raised lake sturgeon in two different hatchery
environments to quantify differences in stage based survival and growth between the two
hatchery environments and three methods of gamete/larval collection. Our objectives
were to determine the effects of rearing environment and complexity and methods of
gamete/larval acquisition on survival and growth at different life stages. Speciﬁc
objectives were to, 1) quantify and compare survival of eggs, larvae, and juveniles
between hatchery rearing environments and methods of gamete/larval collection, 2)
quantify and compare growth across the three collection methods and hatchery
environments, 3) determine the effects of refuge (habitat complexity) and resource
availability on growth and yolk-sac utilization in larval lake sturgeon, and 4) determine

the effects of refuge availability on juvenile lake sturgeon growth.

METHODS
Study Population
Work was conducted during each of two years (2005 and 2006) using gametes
and larvae from the lake sturgeon population residing in the Black Lake Michigan. This

system offers a unique ability to sample large numbers gametes and individuals across all

45

lake sturgeon life history stages, and is founded by previous research on egg (Forsythe,
unpublished data), larval (Smith and King 2005a), juvenile (Smith and King 2005b,
Crossman 2008 chapters 3 and 4), and adult stages (Baker and Borgeson 1999; Smith and

Baker 2005; Forsythe, unpublished. data).

Hatchery Rearing Environments

Lake sturgeon eggs, larvae, and juveniles were reared in two different
environments, a streamside hatchery on the natal Upper Black River and at a traditional
hatchery. Streamside hatcheries have been shown to provide an environment for
successful growth and survival for lake sturgeon and offer adaptability in facility design
that isn’t available in more traditional hatchery facilities (Holtgren et al. 2007). More
traditionally, gametes have been incubated, hatched, and larvae reared at hatcheries
geographically removed from natal streams and returned when stocked. Lake sturgeon
are highly philopatric and rearing that occurs at non-natal environments has led to
concerns regarding disruption of natal imprinting and straying among populations
following release into the natal environment. Straying may artiﬁcially elevate levels of
gene ﬂow and affect the genetic conservation programs for lake sturgeon, as remnant
populations of lake sturgeon within the Great Lakes basin exhibit genetic structuring
(DeHaan et al. 2006).

The streamside hatchery used in this study was constructed on the Upper Black
River in the spring of 2005 using a steel building (US Buildings; Dimensions:
7.3/12.2/4.3m). Water for the streamside hatchery was pumped directly from the Upper

Black River. Water passed through a primary settling tank, sand ﬁltration, a second

46

settling tank, and a degassing column before entering egg incubation and juvenile rearing
tanks via gravity feed. Flow rate through the system measured approximately 350L/min
allowing for two total water turnovers per hour. Aeration was provided to the entire
facility ﬁom a single unit (Sweetwater Regenerative Blower, Aquatic Eco-Systems) to
maintain levels of dissolved oxygen during periods of warm summer temperatures and
elevated ﬁsh respiration. System speciﬁcs for rearing different lake sturgeon life stages at
the streamside hatchery are described below.

Lake sturgeon reared at the Michigan Department of Natural Resources, Wolf
Lake State Hatchery, representing a traditional hatchery environment. Lake sturgeon have
been reared in this multi-species facility since the late 1970’s following consistent rearing
protocols. Water for the hatchery is provided from both well and natural spring sources in

a ﬂow through design. Water temperature in the egg incubation trays remains relatively

constant year round (11.10C) and the hatchery has the capacity to warm a small

percentage of the water to rear cool-water species such as lake sturgeon. Heated water for

juvenile lake sturgeon rearing ranges from 11.1 to 220°C during the summer months

with dissolved oxygen levels maintained at greater than 9mg/L.

Progeny Collection

We collected gametes and larval lake sturgeon using three different methods.
Methods were 1) direct gamete takes (DGT) from adults spawning in the Upper Black
River, Michigan, 2) collection of naturally produced eggs (NPE) from the stream
substrate, and 3) collection of larval lake sturgeon passively dispersing downstream (DL)

from the spawning grounds. For the DGT collection method, adults were captured while

47

spawning using large hand held dip nets. Eggs were collected by hand stripping females
that were caught in the act of spawning and kept in sealed plastic bags at ambient river
temperatures. Milt was collected ﬁom as many males as possible using a syringe and kept
on ice. Fertilizations were conducted at the streamside hatchery the same day as the eggs
were collected. Eggs from an individual female were divided into two lots equally based
on volume and fertilized separately with milt from two randomly chosen males. Milt was
diluted at a ratio of 1:200mls and immediately poured over the eggs. The milt and eggs
were mixed for 1 minute at which time a mixture of water and diatomaceous earth was
added and mixed with the eggs by hand to remove the adhesiveness of the eggs. This de-
adhesion process was conducted for a period of one hour at which time the eggs were
rinsed with clean water. Immediately following fertilization and de-adhesion, half of the
eggs from each paternal half-sib family were transported to the traditional hatchery
environment. Fertilized embryos were transported from the streamside hatchery to the
traditional hatchery in plastic bags kept at ambient river temperatures. Eggs were placed
in incubating trays immediately following fertilization at the streamside hatchery and
immediately following arrival at the traditional hatchery (approximately 5hr post
fertilization).

Systematic kick-net sampling was conducted below observed spawning
aggregations to collect eggs that were naturally fertilized and deposited on stream
substrate. Transects were conducted across the stream at 1 meter intervals. At each
interval we conducted kick-net sampling for 10 seconds. Transects continued
downstream in intervals of 5 meters until no eggs were collected in consecutive transects.

The number of sampled versus non-sampled locations varied across years. Eggs were

48

counted at the streamside hatchery. Eggs from this collection method were incubated and
hatched at the streamside rearing environment because managers at the traditional
hatchery environment had concern regarding potential pathogens that eggs taken from
stream could contain. Larvae were divided equally between the two hatchery rearing
environments within one week post hatch.

Sampling for larval lake sturgeon larvae passively dispersing downstream has
been conducted in several studies to quantify recruitment and chronology and duration of
dispersal (Auer and Baker 2002; Smith and King 2005) and represents a viable collection
strategy. Systematic larval drift sampling was conducted during a 5-hour period,
beginning at dusk (21 :00) and ending in the early morning (03:00). The sampling location
was located approximately 2 km downstream from all spawning areas on the Upper
Black River (Fig. 1) allowing for collection of offspring ﬁ'om all spawning adults.
Sampling began seven days after the ﬁrst observation of reproduction. We deployed ﬁve
D—framed drift nets across the stream channel in equal intervals to capture dispersing
larval sturgeon. Nets were checked hourly and the larvae within each net were
enumerated. This design was replicated nightly each year for the entire drifting period.
We sampled the entire distribution of dispersing larvae due non-randomness described in
other work (DeHaan 2003). Sampling was terminated after adult spawning ceased and
zero larval captures were observed for consecutive nights. Larval lake sturgeon captured
each evening were reared at the streamside hatchery until the end of the drifting period.

The larvae were then divided equally between the two hatchery rearing environments.

49

Experimental Design

Mortality and growth rates of ﬁsh during egg, larval, and juvenile stages are both
high and variable (Houde 1989; Pepin 1991). It is difficult to obtain precise estimates of
mortality for many teleost ﬁshes and such estimates are critical for evaluating the extent
and causes of stage-speciﬁc mortality (Houde 1997). For species with complicated life
histories, such as lake sturgeon, events occurring during one stage can have important
implications for the performance and survival of individuals in subsequent stages.
Experimental and analytical protocols conducted in this study are described on a life

stage by life stage basis.

Egg Stage: The eﬂects of hatchery rearing environment on egg survival

Rationale: The environmental conditions (i.e. temperature) in which a sturgeon
embryo is reared, have been shown to affect development and survival rates (Wang et al.
1985; Gershanovich and Tauﬁk 1992). Lake sturgeon reproduce over a broad range of
time that can last for up to 6 weeks (F orsythe, unpublished data). Eggs incubating in
natural river conditions may experience environmental ﬂuctuation in water temperature
and chemistry. Furthermore, eggs that are deposited early in the spawning season likely
experience environment conditions (temperature and ﬂow) that are different from eggs
deposited late in the season. Hatchery incubation of lake sturgeon eggs has traditionally
been conducted at facilities that use a water source that is different from that of the natal
river. Typically, these hatcheries use a well water source which remains at a constant
temperature during the incubation period allowing for accurate predictions on the time to

hatch which improves hatchery management efﬁciency. The effects of ﬂuctuating and

50

O

constant environments on egg survival during incubation have not been compared for
lake sturgeon. Streamside hatcheries have been recently developed for this species in an
attempt to minimize straying following release of juveniles back into the wild and offer a
unique approach to examine differences in egg survival relative to more traditional
hatchery environments. Streamside hatcheries mimic the natural temperature ﬂuctuations
of the natal river. We examined overall and daily egg survival in two different hatchery
environments.

Methods: Fertilizations of embryos followed protocols discussed for the DGT
collection method above. Eggs were incubated separately by family group using heath
tray stacks at both hatchery rearing environments. Family groups were randomly assigned
to one of 8 trays within a heath tray stack. Egg volumes larger than 200mls were divided
between multiple trays. Eggs were treated with a formalin drip (17ppm) for 15 minutes
daily to minimize pathogen and fungal infection (Bouchard and Aloisi 2002). Daily

measurements recorded included the number of dead eggs and the number of dead eggs

with microbial infection. We also recorded Temperature (0C, Data loggers every hour)

and dissolved oxygen (mg/L, YSI) within the heath trays.

Analysis: Analyses were conducted using program R (Version 2.1.1). We used a
two-way analysis of variance (ANOVA) to test for differences in overall egg survival
after all eggs had hatched as a ﬁrnction of hatchery environment and year. Separate
analyses were performed for each year because we used different subsets of females in
each year. We used a two-way ANOVA to test for overall differences in egg survival
between hatchery environments and years. We were not able to collect daily survival

information at the traditional hatchery in 2005. We calculated daily survival at both

51

hatchery environments in 2006. We used a general linear model (GLM) to test for
differences in daily survival based on both ﬁxed (hatchery, day, heath tray location) and
random (female) effects. Daily measurements of temperature and dissolved oxygen were
used as covariates. We also examined the proportion of dead eggs that were infected by
microbes between the two hatcheries over time using a GLM. Factors included hatchery
and day. Temperature and dissolved oxygen were included as covariates. Finally, since
we had daily survival at the streamside hatchery for two years we tested for the effect of
year and day on daily survival. Environmental covariates were incorporated. We also
examined the proportion of dead eggs infected by microbes as a function of year and day
for the streamside hatchery. Data were tested for normality and homogeneity of variances
by comparing residuals versus ﬁtted values and using a Shapiro-Wilk test in program R.
Analysis were conducted using arcsine-square-root transformed data in cases of non-

norrnality or heterogeneity of variance.

Larval Stage: The eﬂects of hatchery rearing environment on larval survival

Rationale: Under artiﬁcial conditions, once larval sturgeon hatch from the egg
and deplete their yolk sac reserves, their survival depends on multiple factors including
the tank design (i.e. water ﬂow, access to cover) and hatchery management (i.e. feed,
feeding rates, cleaning; (Conte et al. 1988). Streamside hatcheries have been
implemented as restoration tool lake sturgeon and represent more natural ﬂuctuation in
incubation and rearing environments. However, there is no empirical evidence to suggest
that streamside hatcheries result in elevated numbers at hatch or higher survival during

the critical larval stage. We were interested in examining the effects of hatchery

52

environments on larval hatch and subsequent survival by comparing a streamside
hatchery to a traditional hatchery environment. Results ﬁ'om this work will be important
for the development of future streamside hatchery programs and will help to improve
culture techniques for larval lake sturgeon.

Methods: Larvae were reared in circular tanks (diameter = 1.2m) that were each
divided to include 2 replicates of each of the three collection methods described above.
Tanks were replicated ten times at the streamside hatchery and six times at the traditional
hatchery. Larvae were fed live brine shrimp nauplii (Artemia spp) beginning ﬁve days
following hatch. Brine shrimp were hatched in multiple cone shaped tanks (19 liters) over
a 24 hour period set to be harvested at different times. Brine shrimp were harvested by
stopping the aeration to the hatching cone and affixing a light source to the bottom of the
cone. Brine shrimp are positively phototactic and nrigrate towards the light source while
empty casings and un-hatched eggs ﬂoated to the surface. Brine shrimp were then
harvested through a valve at the bottom of the tank. We concentrated the brine shrimp
using a plankton screen bag (100 um) and transferred them into an aerated bucket of fresh
water. Brine shrimp were fed in 200ml aliquots once every two hours from dawn (0700)
until dusk (2100). Tank sections were cleaned daily by purging all the water in the tank to
remove excess food and waste and the tank sides were wiped with providone-iodine
(Betadine) for disinfection.

Analysis: All analyses were conducted using program R (Version 2.1.1).
Differences in the number of larvae that hatched between the two rearing environments in
each of 2005 and 2006 were examined using a 2-way ANOVA. Furthermore, we used a

2-way ANOVA within each year to examine if the number of hatched larvae differed

53

between individual females reared in equal initial numbers between the two hatchery
environments. We calculated daily larval survival for a period of 14 days at which time
the yolk-sac had been absorbed and the ﬁsh were actively feeding. We used a GLM to
test for differences in daily survival between hatcheries with day and hatchery location as
ﬁxed effects. Daily measures of temperature were used as a covariate. The effects of
female could not be incorporated into the analysis of daily larval survival because
families had to be mixed at the traditional hatcheries due to a smaller number of rearing
tanks. Data were tested for normality and homogeneity of variances by comparing
residuals versus ﬁtted values and using a Shapiro-Wilk test in program R. Analysis were
run on arcsine-square-root transformed data in cases of non-normality or heterogeneity of

variance.

Larval Stage: The eﬂects of within hatchery habitat complexity on larval growth and
yolk-sac utilization.

Rationale: Starvation is viewed as a signiﬁcant source of mortality following
absorption of endogenous reserves during the yolk-sac stage of development in larval
ﬁshes (Folkvord and Hunter 1986). Mortality has been attributed to patchy distributions
of food resources available during period when larvae switch ﬁom endogenous reserves
to exogenous feeding (Letcher et al. 1996). Resource limited larvae can either decrease
metabolic expenditures (reduced activity) (W ieser et al. 1992) or increase activity until
suitable resources are located (Mehner and Wieser 1994). It is unknown how larval lake
sturgeon utilize yolk-sac reserves when they are in the presence of reﬁrge or resources. It

has been noted that larval sturgeon will utilize refuge when presented in an aquaculture

54

setting (Aloisi et al. 2006). The refuge is subsequently abandoned when endogenous
reserves become low to search for resources. In a typical hatchery setting larval sturgeon
are not presented with available refuge and food is administered to the tank
approximately two weeks following hatch. Information on differential growth and
utilization of yolk-sac reserves as a function of refuge and resource availability is lacking
for larval lake sturgeon. We examined four treatment groups including; 1) tanks with no
refuge, 2) tanks with simulated refuge (shag carpet), 3) tanks with no refuge but presence
of a food resource, and 4) tanks with simulated refuge and presence of a food resource.
Results from this work address concerns of how artiﬁcial rearing conditions within a
hatchery inﬂuence variation in individual size and will help to improve future lake
sturgeon culture techniques in conservation hatcheries.

Methods: Each of the four treatment groups was replicated across paternal half-
sib families (n = in 2006; n = 2007). Larvae in two treatments including the presence of
resources were fed live brine shrimp every two hours from 0700 until 2100. Larvae that

hatched from family groups were measured (n=10) for total length (TL; mm), yolk-sac
length (Y SL; mm), yolk-sac height (Y SH; mm), and body area (BA; m2) and then

allocated to experimental tanks. Treatment groups were stocked with 100 larvae at the
start of the experiment. We then removed 10 random individuals every 2 days for
measurements. Individuals were only used' once and were not returned to the tank. Tanks
were cleaned each day to remove mortalities and uneaten food. We used Image J analysis
software (Version 1.34, freeware) to conduct all digital measurements. Information
recorded ﬁ'om each larva included: YSL, YSH, YSA, BA, and TL. Total length was

measured from the anterior most part of the developing rostrum to the posterior most part

55

of the notochord. We calculated elliptical yolk-sac volume (Y SV in mm3) following the

formula of Blaxter and Hempel (1963),

st = (n/6)*L*H2

where L = yolk-sac length and H = yolk-sac height. Prior to image analysis larvae were
anesthetized with 25mg/L M8222.

Analysis: All analyses were conducted using program R (Version 2.1.1). We used
a one-way analysis of variance to test for differences in both total length and yolk-sac
area between treatment groups on the day of hatch. Furthermore, we also used a one-way
AN OVA to test for differences in total length and yolk-sac area for larvae produced ﬁom
different females at hatch. We used a mixed effects model to examine differences in the
response variables (growth and yolk-sac utilization) between the four treatment groups
over time. Fixed effects of the model included treatment, day since hatch, and year.
Female was included as a random effect to account for variation between females and
because we used a limited number of females representing a large population of breeding
adults. Female was also nested within year as different females were used in 2006 and
2007 so comparisons between years as a function of female could not be conducted. Data
were tested for normality and homogeneity of variances by comparing residuals versus
ﬁtted values and using a Shapiro-Wilk test in program R. Analysis were run on log10

transformed data in cases of non-normality or heterogeneity of variance.

Juvenile Stage: The effects of rearing environment on juvenile growth and survival
Rationale: The environments in which a ﬁsh is reared and into which it will be

released are very strong determinants of a restoration programs success (Travis et al.

56

1998). Ideally, we want to minimize the possibility that, between hatching and ﬁnal
release, the hatchery is selecting for distinct alleles or allelic combinations that would be
disadvantageous in the natural environment. Phenotypic plasticity in response to
variations in environmental conditions occurs frequently in a natural setting but may be
more critical in an aquaculture setting (Brokordt et al. 2006). This may be particularly
evident at early life history stages and can affect abilities to respond to environmental
cues following release into the natural environment (Olla et al. 1998; Travis et al. 1998).
Streamside hatcheries offer a more natural rearing environment. However, they have
gone into operation without research evaluating their effects on juvenile growth and
survival relative to more traditional production hatcheries. We evaluated growth and
survival of juvenile lake sturgeon collected using three gamete/larval collection methods
in two different hatchery environments in each of two years. Results ﬁ'om this work will
help to reﬁne culture methods for this species and provide guidelines for the use of
streamside hatcheries as a management option for the restoration of lake sturgeon
populations.

Methods: The experiment was initiated after all gamete/larval collection methods
had been distributed evenly between the two hatchery environments. Fish in both
hatchery environments were reared in circular tanks (diameter=1.2m) divided into six to
include two replicates of each of the three gamete/larval collection methods. Each
collection method was randomly assigned to two of the six slots located within each tank.
Tanks were replicated ten times at the streamside hatchery and six times at the traditional

hatchery. Water volume within each section was 90 liters. Maximum rearing densities

were held to a maximum of 400 ﬁsh/m2 of tank space. Faj fer et al. (1999) found no

57

signiﬁcant difference in growth between three different juvenile lake sturgeon rearing

densities and concluded that densities of 150—450 ﬁsh/m2 are equally acceptable for

rearing lake sturgeon. Juvenile lake sturgeon were transitioned from brine shrimp onto
live cultured blackworrns (tubiﬁcid annelids; J .F . Enterprises, CA, USA) starting at day
21 following hatch. Blackworrns were fed for approximately one week prior to the start
of feeding frozen bloodworms (chironomid rnidge larvae). A 1kg sheet of frozen blood
worms was fed per 1000 juvenile sturgeon per day. The worms were thawed and divided
evenly (~1 5 grams per section) among all tank sections. The amount of bloodworms fed
increased approximately .5kg every 2 weeks for the duration of rearing. Tanks were
cleaned once daily in both environments to remove waste and excess food. Protocols in
both hatcheries involved purging the water volume in each tank through the removal of a
central stand pipe. The tank walls were disinfected using betadine and the stand pipe
replaced and the tank reﬁlled. At the streamside hatchery all ﬁsh were given a salt bath
once per week as a preventative measure because the streamside hatchery water source
was untreated. Salt was administered at Sppt for 15 minutes to the entire system
simultaneously by adding salt to the ﬁnal settling tank. Tank sections with ﬁsh exhibiting
signs of stress (i.e. reduced feeding and movement or increased respiration) were treated
with additional salt baths. Mortalities were recorded and removed daily at both hatchery
environments. A subset of ﬁsh (n=25) from each tank section and corresponding

collection method were measured for total length (mm) weekly throughout the rearing

period. Daily measurements of temperature (0C) and dissolved oxygen (mg/L) were

recorded using data loggers and a handheld YSI instrument, respectively.

58

Analysis: We examined two response variables, survival and growth, as a function
of hatchery rearing environment and different collection methods for gametes/larvae. All
analyses were conducted using program R (Version 2.1.1). We used a GLM to exarrrine
daily survival of juvenile lake sturgeon as a function of both ﬁxed (hatchery, day, year,
treatment) and random effects (tank section). Temperature was included as a covariate.
Insigniﬁcant factors were eliminated and the model was rerun. We used the same model
to examine differences in weekly growth (total length). Data were tested for normality
and homogeneity of variances. In cases of non-normality or heterogeneity of variances,
analyses were run using arcsine-square-root-transformed survival data and log

transformed growth data.

Juvenile Stage: The eﬂects of within hatchery habitat complexity on juvenile survival,
growth, and body coloration.

Rationale: Cultured and wild ﬁsh grow in different environments and different
experiences are likely to produce behavioral and physiological differences at all life
history stages (Huntingford 2004). Providing ﬁsh in hatcheries with simulated habitat
within rearing tanks has been shown to increase post-release survival (N aslund 1992;
Tipping 1998; Berejikian et al. 2000; Tipping 2001; Maynard et al. 2004) and inﬂuence
morphological expression by producing ﬁsh with the appropriate camouﬂage colorations
for the habitat backgrounds found in natural habitats. We were interested in examining
the effects of three levels of habitat complexity on juvenile lake sturgeon growth. These
three treatment groups included 1) tanks with no cover, 2) tanks with 50% cover

(Simulated with large rocks), and 3) tanks with 100% cover.

59

Methods: This experiment was conducted entirely at the streamside facility.
Juvenile lake sturgeon were reared in circular tanks (diarneter=1.2m) divided into six to
include two replicates of each of the three treatment groups we evaluated. We replicated
each treatment group six times. A total of 25 ﬁsh were stocked into the experiment and
were reared for a total of six weeks. We recorded weekly measurements of total length
for all individuals within each tank section. Tanks were cleaned and mortalities removed
and recorded daily. Juvenile lake sturgeon were fed bloodworms as previously described.

Analysis: We examined growth and survival of juvenile lake sturgeon reared in
the three different treatment groups. All analyses were conducted using program R
(Version 2.1.1). Survival was examined using a GLM with day tank, treatment, and the
interactions as ﬁxed factors. We examined weekly differences in growth using a GLM
with week, tank, treatment, and the interactions ﬁxed factors. Data were tested for
normality and homogeneity of variances. In cases of non-normality or heterogeneity of
variances, analyses were conducted using arcsine-square-root-transformed survival data

and log transformed growth data.

RESULTS
Egg Stage
There was no signiﬁcant difference in the number of eggs collected ﬁom females
for incubation (mean :b ISD = 11,157 7883) between 2005 and 2006 (df = 16, t = 2.12, P

= 0.99). Overall egg survival to hatch was low in 2005 and did not differ signiﬁcantly

(F 1,20 = 0.12, P = 0.73) between the streamside hatchery (mean i SE: 0.17 d: 0.02) or the

traditional hatchery (0.15 i 0.02). Egg survival was signiﬁcantly different among females

60

in 2005 (F930 = 3.52, P = 0.009; Fig.6), though the interaction between female and

hatchery environment was not signiﬁcant (F130 = 0.69, P = 0.71). In 2006, eg survival

was signiﬁcantly higher at the streamside hatchery compared to the traditional hatchery

(F130 = 11.37, P = 0.003). Egg survival was twice as high at the streamside hatchery

(0.35 d: 0.05) compared to the traditional hatchery (0.17 d: 0.04). As in 2005, eg survival

differed signiﬁcantly among females (F730 = 2.83, P = 0.03; Fig. 6) but was not a result

of the interaction with hatchery environment (F130 = 0.47, P = 0.85). The number of egg

incubation days varied considerably at the streamside hatchery over the two years (7-14

days) and was a constant 19 days at the traditional hatchery in both years. Temperature

was a signiﬁcant predictor (F134 = 334.2, P < 0.001, r2 = 0.843) for egg incubation time

(Fig. 7) and may explain inter-female differences.
We compared the daily survival of eggs incubated between hatcheries in 2006.

Daily survival rates of eggs reared at the streamside hatchery in 2006 were signiﬁcantly

higher than eggs reared at the traditional hatchery (F 1,354 = 6.52, P = 0.011) in the same

year. We found a signiﬁcant difference between groups of eggs within different heath

trays (F225 54 = 2.63, P < 0.001) but the signiﬁcance was not attributed to individual
females (F 1,3 54 = 0.42, P = 0.52). Furthermore, the interaction between egg group and

hatchery environment was not signiﬁcant (F173 54 = 0.42, P = 0.98). Microbial infection

resulted in a large portion of daily egg mortality in both hatchery environments though

the prevalence of infection was signiﬁcantly lower at the traditional hatchery (F1321 =

6l

18.24, P < 0.001; Fig. 8). Several factors inﬂuenced the occurrence of microbial infection

in both hatchery environments. As observed with daily survival, groups of eggs between

heath trays differed signiﬁcantly in infection rates (F22,321 = 3.39, P < 0.001) which
again was not driven by either individual females (F1321 = 1.45, P = 0.23) or the

interaction between egg group and hatchery rearing environment (F 16,321 = 2.82, P <
0.001). Eggs at the streamside hatchery that experienced higher mean temperatures
during incubation had signiﬁcantly higher rates of microbial infection (F1321 = 27.06, P

< 0.001).

We tested daily survival and the proportion of dead eggs with microbial infection

between years at the streamside hatchery. Year was not signiﬁcant (F 1,5 14 = 0.01, P =

0.95) and was dropped from the model. Daily survival was signiﬁcantly different

(F15,514 = 10.89, P < 0.01) among different incubation times (Fig. 9). For example,

signiﬁcant peaks in mortality were observed at day 3 for eggs incubating for 10 days, and

at day 4 for eggs incubating for 11 and 12 days. There was no difference in mean daily

temperatures over the entire period of incubation between years (F1514 = 0.77, P = 0.40).

Microbial infection was not different between years at the streamside hatchery (F1514 =

0.68, P = 0.41) and similar to above, was found to be positively inﬂuenced by

temperature (F1514 = 32.17, P <0.001).

62

Larval Stage

Hatchery Environments: The number of larvae at hatch was signiﬁcantly higher

for the streamside hatchery compared to the traditional hatchery in 2006 (F130 = 26.76, P
< 0.001), but was not signiﬁcantly different in 2005 (F130 = 1.73, P = 0.20). Larval hatch
was signiﬁcantly different between females (Fig. 10) in both 2005 (F930 = 4.28, P =

0.003) and 2006 (F730 = 11.99, P < 0.001). Furthermore, in 2006 the interaction between

female and hatchery environment was signiﬁcant (F7920 = 3.52, P = 0.01) with certain

females having higher numbers of larvae at the streamside hatchery compared to the
traditional hatchery (Fig. 10).
Larval survival was followed over a period of 14 days at which time the yolk-sac

had been absorbed and feeding commenced. Larval survival in 2005 did not differ
signiﬁcantly between hatchery environments (F1516 = 0.68, P = 0.41) with daily survival
rates being 0.957 i 0.002 at the streamside hatchery and 0.970 :b 0.010 at the traditional
hatchery. Survival differed signiﬁcantly by day (F 19,516 = 4.31, P < 0.01) with mortality

occurring in the ﬁrst three days following hatch. In 2006, larval survival was high at the

traditional hatchery (0.966 i 0.001), however larval survival was signiﬁcantly higher

(F 1,918 = 30.32, P < 0.01) at the streamside hatchery (0.982 :1: 0.001). The interaction

between hatchery location and day was marginally signiﬁcant (F13,913 = 1.52, P = 0.073)

with survival at the traditional environment lower in the ﬁrst three days following hatch.

63

Hatchery Complexity: A total of 3088 and 3887 individual larvae were measured

for total length and yolk—sac area in 2006 and 2007, respectively. We found a signiﬁcant

difference in total length between females at hatch in both 2006 (F5378 = 21.4, P <

0.001) and 2007 (F7350 = 106.92, P < 0.001). Since all families were represented in all

treatment groups there was no difference in total length between treatments group at the
start of the experiment. In both years the experiment was terminated after 15 days due to
complete absorption of the yolk-sac. Larval lake sturgeon grew rapidly in total length in
both 2006 and 2007. However, differences were found between the four treatment

groups. There was a signiﬁcant interaction between the treatments over time between the

two years (F21,6545 = 9.72, P < 0.001) with larvae in the open treatment group being

smaller in total length at the end of the experiment compared to the treatment group with
available refuge. In 2006, both the open treatment and the open treatment with resources
were signiﬁcantly smaller compared to the treatment with refuge (Fig. 11). In 2007, the
difference was only observed between the treatment with refuge and the open treatment
without food. In 2007 larvae in the treatment with refuge grew 9.5mm in total length over
the course of the experiment while larvae in the treatment with no refuge grew 8.6mm
over the same period. Differences between the two years were attributed to differences
between treatment groups on certain days but did not differ in their ﬁnal result. For

example, in 2006 the treatment group with refuge was signiﬁcantly different than the

treatment group without refuge at day 9 (F31’6545 = 2.02, P = 0.04) while the same result

was not found until day 11 in 2007 (F31,6545 = 2.32, P = 0.03). We found a signiﬁcant

64

difference between females in total length over time in both years (F366, 6545 = 13.67, P
< 0.001; Fig. 12).
We found a signiﬁcant difference in yolk-sac volume between females at hatch in

both 2006 (125,478 = 10.93, P < 0.001) and 2007 (F7350 = 48.8, P < 0.001) with earlier

spawning females having larger yolk-sac reserves at hatch. Since treatment groups were
replicated across families there was no difference between treatments at the outset of the

experiment. There was a signiﬁcant effect of treatment over time between years on larval
yolk-sac volume (F31,6545 = 6.89, P < 0.001). As with total length, differences were
attributed within different days but yolk-sac area was not signiﬁcantly different between
the treatment groups at the termination of the experiment (F31,6545 = 0.331, P = 0.74). In
2006, both the open treatment and the open with food treatment had signiﬁcantly smaller
yolk-sac volumes on days 5 (F31, 6545 = 2.58, P = 0.01; Fig. 7) and 7 (F31,6545 = 3.052, P

= 0.002; Fig 11) implying greater expenditures (allocation) of resources related to

metabolic activity. We found a signiﬁcant difference between females in yolk-sac volume

over time in both years (F 366, 6545 = 7.53, P < 0.001; Fig. 12).

Juvenile Stage
Hatchery Environments: We tested for differences in daily survival of juvenile

lake sturgeon from three different gamete/larval collection method reared in two different

environments in each of two years. Both tank within location (F59,725 5 = 0.79, P = 0.43)

and the interaction of collection methods over time across years (F 638,725 5 = 0.01, P =

65

0.90) were not signiﬁcant and were removed ﬁom analyses. Subsequently, we found a

signiﬁcant difference in survival between collection methods over time at the two

different hatchery environments (F284,7671 = 1.72, P < 0.001). In both cases, offspring

from the dispersing larvae collection method had the lowest mean daily survival (Fig.

13). Daily survival was signiﬁcantly lower at the traditional hatchery for the ﬁrst week of

juvenile rearing (F234,7671 = 2.992, P = 0.003) and then mortalities decreased (Fig. 13).

Mortalities approached zero rapidly at the traditional hatchery following day 50 post
hatch while we still observed minor ﬂuctuations in survival at the streamside hatchery in

both years (Fig. 13). Mean daily temperature did not signiﬁcantly inﬂuence survival as a

function of hatchery environment (F1=0.1840, P = 0.67). Mean daily temperature was

22.01 :t 0.3 and 21.57 i 0.23 at the streamside hatchery in 2005 and 2006 respectively.
Mean daily temperature was 21.4 i: 0.1 and 21.1 i 0.03 at the traditional hatchery in 2005
and 2006 respectively.

We evaluated juvenile growth for each of three different methods of gamete/larval

collection reared in two different hatchery environments in each of two years. The effects

of tank and tank sections within tank were dropped from the ﬁnal model due to

insigniﬁcance. Average rearing densities per tank slot were 390.6 i 4.2 ﬁsh/m2 at the

streamside hatchery and 362.3 :t 5.4 ﬁsh/m2 at the traditional hatchery. Juvenile lake

sturgeon reared at the traditional hatchery in 2006 grew signiﬁcantly faster among all

three treatment groups compared to individuals reared at the streamside hatchery in the

same year and all individuals from both hatchery environments in 2005 (F98, 29494 =

66

85.7, P < 0.001). For example, in 2006 juvenile sturgeon produced as a product of direct
gamete takes grew 7 .29cm per week at the streamside hatchery and 11.96cm per week at

the traditional hatchery. There was no difference in growth between treatment groups at

the traditional hatchery in 2006 (F95, 29493 = 9.62, P = 0.31). However, juveniles in the

NPE treatment group were signiﬁcantly larger than DL juveniles at the end of the

experiment at the streamside hatchery (F 95, 29494 = 27.2, P = 0.008). There was no

difference between the growth of juvenile lake sturgeon between the three treatment

groups at either hatchery environment in 2005 (F 95,29493 = 2.89, P = 0.18). Weekly

growth of juvenile sturgeon in 2005 was 7.21cm at the streamside hatchery and 7.30crn at
the traditional hatchery. Weekly growth varied signiﬁcantly between treatments over
time in both hatchery environments in both years (F95,29493 = 67.4, P <0.001). Variation
in total length within a treatment increased over time.

Hatchery Complexity: We evaluated the effects of three different levels of tank
complexity (open, half cover, total cover) on daily survival and weekly growth of

juvenile lake sturgeon at the streamside hatchery in 2005. There was no signiﬁcant

difference in daily survival among the three treatment groups over the course of the

experiment (F32,504 = 0.96, P = 0.57). Survival was high among all treatments with total

survival at the end of the experiment being 0.952, 0.968, and 0.968 for open, 50%
covered, and 100% covered treatment groups respectively. There was no signiﬁcant

difference in the total length among the three treatment groups at the outset of the

experiment (F2322 = 1.89, P = 0.15). We found a signiﬁcant effect of treatment over time

67

(F10,1253 = 1.86, P = 0.04) with juvenile lake sturgeon in the open treatment group

reaching a larger size by the end of the experiment (Fig. 14).

DISCUSSION

For recruitment-limited and numerically depressed populations such as the lake
sturgeon, conservation hatcheries are an important technique in the return of populations
to sustainable levels. Furthermore, conservation hatcheries for sturgeon species will need
to employ management techniques that maximize the probability of imprinting to natal
streams following release. We evaluated the effects of two different hatchery rearing

environments on growth and survival. Results are discussed on a stage by stage basis.

Egg Stage

Egg survival was relatively low in both years in both hatchery environments with
the highest egg to hatch survival being 0.35 at the streamside hatchery in 2006. We
documented signiﬁcantly higher survival of lake sturgeon eggs reared at the streamside
hatchery in 2006 compared to the traditional hatchery in the same year. This also
translated to higher daily survival at the streamside hatchery compared to the traditional
hatchery. Stream environmental conditions including temperature, stream physical and
biotic conditions vary temporally within a season and between years and are expected to
affect embryonic development and survival in ﬁshes (Houde 1987). Higher egg survival
at the streamside hatchery is likely attributed to differences in incubation temperatures as

all other rearing methods and females used were consistent between the two hatcheries.

68

Eggs incubating at the two different hatchery environments incubated for different

lengths of time. Eggs incubating in 11°C water at the traditional hatchery did not hatch

until day 19 while a much broader range was seen at the streamside hatchery where the
embryos were subject to more natural temperature regimes. Within the streamside
hatchery we documented minimum and maximum incubation times of 7 and 14 days
respectively. Temperature proﬁles did not differ between years within each hatchery
environment. Furthermore, daily egg survival was different across different incubation
lengths. Experiencing natural variations in water temperatures, within tolerable limits,
may have both beneﬁcial and deleterious effects on survival. Previous research on
sturgeon species has shown that rates of development at the embryonic stage are
temperature dependant (Wang et al. 1985). Eggs incubating at lower temperatures may
beneﬁt by larvae being bigger at hatch but suffer a tradeoff by being exposed to higher
levels of predation at the egg stage. Eggs incubating at higher temperatures will hatch
rapidly avoiding additional risks of predation at the egg stage but will produce larvae of
smaller size. For conservation purposes it is critical that hatcheries capture this diversity
during the incubation stage. If spawning time is repeatable for adult female lake sturgeon
(Forsythe, unpublished data) then this would suggest that females choose when to deposit
eggs and collecting eggs ﬁom only one of multiple spawning groups would limit overall
diversity captured. Furthermore, the impacts of collecting eggs from females spawning in
late periods with high water temperatures and rearing them in conditions reﬂective of
earlier spawning females is still unknown in lake sturgeon culture. Rearing lake sturgeon
eggs under constant and cool water conditions eliminates natural variability in incubation

observed-in both the river and the streamside hatchery environment. Constant hatchery

69

environments create a logistically efﬁcient hatchery environment where eggs hatch more
synchronously and on a predicted date. However, if timing of spawning is further
inﬂuenced by the conditions the egg is reared in then preference for spawning time could
be shifted in one direction with eggs incubated in constant traditional hatchery
conditions. This has yet to be evaluated in lake sturgeon.

We also documented differences in egg survival between different females but the
interaction between hatchery environments was not signiﬁcant. Maternal effects have
become increasingly recognized (Mousseau and Fox 1998) as having signiﬁcant effect on
survival during early life stages including the egg stage (Heath et al. 1999). These effects
are still relatively understudied in sturgeon. Differences in egg survival could have been
due to differences in egg quality between females spawning at different times. Egg
quality is an important component for fertilization and egg survival (Bromage et al. 1994;
Brooks et al. 1997). Differences between female lake sturgeon eggs have been
documented in the levels of total lipids and thiamine (K. Debrowski, personal
communication), and yolk-sac reserves (this study).

Microbial infection was found to be signiﬁcantly higher at the streamside
hatchery compared to the traditional hatchery in 2006. Furthermore, incidences of
microbial infection increased with water temperature. Microbial infection, primarily the
fungal order saprolegniales and other aquatic fungi, are widespread in water supplies of
ﬁsh hatcheries and often result in elevated mortality during the egg stage (Marking et al.
1994). Microbial infection has been noted in studies on sturgeon (Smith et al. 1980;
Dettlaff and Goncharov 2002) and research has been conducted on developing efﬁcient

methods to treat eggs prior to and during incubation (Rach et al. 1997; Rach et al. 1998;

70

Bouchard and Aloisi 2002). We used heath trays as incubation vessels in both hatcheries.
However, MacDonald jars are typically used for sturgeon egg incubation as the eggs
remain suspended and rolling which helps to reduce fungal infection. We used heath trays
because they provided an efﬁcient way to quantify egg mortality and microbial infection
on a daily basis. Furthermore, we were using fairly small numbers of eggs which were
reared separately by family. Small volumes of eggs typically do not remain suspended as
easily in MacDonald jars as large batches of eggs. It was not surprising that microbial
infection was higher at the streamside hatchery. Direct river water was minimally treated
before entering the incubation system and rrright expect to have higher levels of
prokaryotic and eukaryotic egg pathogens relative to cooler ground water. Finally,
different batches of eggs composed of eggs taken from females spawning at different
time periods differed in the time it took for microbial infection to occur (Fig. 8). Later
batches of eggs had earlier incidences of microbial infection. Differences might be
expected at the streamside hatchery where different batches of eggs throughout the
season will be subject to different daily mean temperatures. However, even within a
constant temperature proﬁle at the traditional hatchery we saw microbial infection arrive
earlier in one batch then the other. One previously mentioned cause may be related to egg
quality between females spawning at different times (Bromage et al. 1994; Brooks et al.
1997). Future research designed to identify egg quality between females as a ﬁrnction of

reproductive timing is warranted.

Larval Stage

7l

Larval hatch was higher at the streamside hatchery in 2006 compared to both the
traditional hatchery in the same year and both hatcheries in 2005. This corresponds to the
increase in egg survival observed over the same time period. Daily larval survival was
high at both hatcheries in 2006 but was signiﬁcantly higher at the streamside hatchery. In
both years we saw the bulk of the larval mortalities in the three days immediately
following hatch at both hatchery environments. Larval survival was then very high for the
remainder of the period in which they had endogenous reserves. We did not observe a
signiﬁcant difference in survival at the end of the larval stage when the yolk-sac was
completely absorbed. This is not surprising as other sturgeon species have been shown to
resist starvation for 18 days (Hardy 2000).

Larvae grown in different levels of rearing habitat complexity responded
differently in growth and utilization of yolk-sac reserves. We found that larvae in tanks
with refuge reached larger total lengths by the end of the experiment compared to those
that were reared in open tank environments. In natural conditions, larvae hatch and
immediately burrow into the stream substrate. They utilize their endogenous reserves
during this period. When endogenous reserves get low larvae then emerge ﬁ'om the
substrate at night and drift passively downstream into feeding areas. This hiding phase is
likely a critical stage in the development of a larval lake sturgeon. We showed a
difference in size of larvae that were provided with available refuge. Larvae reared in
open treatments remained active during the daylight hours, apparently searching for
reﬁrge. The presence of resources (food) in the tanks only differed between as a function
of available refuge. Our initial hypothesis was that if a larval sturgeon could perceive a

resource rich environment then they may utilize yolk reserves at a faster rate. The timing

72

of when a larval sturgeon can perceive this type of variable in the environment it still
relatively unknown. We also found a difference in the size at hatch and growth rates
between families, irrespective of treatment groups. Females that had larvae which
hatched at smaller sizes remainedsmaller throughout the course of the experiment (Fig.
12). The effects of refuge on phenotypic expression should be considered by managers
conducting conservation hatchery programs for lake sturgeon.

Larval lake sturgeon utilized endogenous yolk-sac reserves over a period of 15
days. Yolk-sac reserves between treatments did not differ in the initial or ﬁnal days of the
experiment. We did ﬁnd however that yolk-sac reserves were used at a faster rate during
days 5 and 7 in the open treatments. The efﬁciency of yolk utilization depends on the
relationship of the rate at which yolk reserves are used in metabolic processes with the
rate at which they are incorporated into body tissue (Blaxter and Hempel 1966). Larval
lake sturgeon reared in tanks without cover spent a large majority of their time during the
day actively swimming which likely resulted in higher amounts of yolk reserves put
towards metabolic processes. This was observed up until day 7 which at that time the
majority of the larvae in the open tanks remained on the bottom and clustered in the
corners of the tanks. As previously mentioned, other species of sturgeon have been
shown to be able to resist starvation for long periods. Therefore, this delay in growth,
caused by increased use of yolk reserved in metabolic processes, may result in a

reduction in survival in later stages (Bailey and Houde 1989).

Juvenile Stage

73

We examined differences in daily survival of juvenile lake sturgeon as a function
of year, method of gamete/larval collection, and hatchery environments. One important
result was that survival of offspring collected as dispersing larvae was signiﬁcantly lower
during the ﬁrst week of rearing in both hatchery environments and years compared to the
other two collection methods (Fig. 13). Dispersing larvae were captured when yolk-sac
reserves were diminished and they were passively dispersing downstream. These
individuals could have been feeding on natural food sources and the transition onto
hatchery feed (brine shrimp) was not as easy compared to individuals obtained using the
other collection methods that were exposed to only brine shrimp since hatch. DiLauro et
al. (1998) found that once larval lake sturgeon had imprinted on brine shrimp they would
not consume a different and formulated diet. Future programs interested in utilizing this
collection method may need to explore different mixtures of natural and artiﬁcial food
sources in order to wean a greater proportion of dispersing larvae onto suitable hatchery
diets. Anderson (1984) found high survival using a larval lake sturgeon diet consisting of
natural sources (mixed zooplankton species and aquatic annelid worms). Differences in
daily survival between hatchery environments occurred in the ﬁrst week with streamside
hatchery ﬁsh having higher daily survival. We were limited to feeding during the daylight
hours at both hatchery environments when small hourly feedings over a 24 hour period
using automatic feeders would likely have been favorable (Cui et al. 1997). Water at the
streamside hatchery may have provided an additional and natural source of food.
However we did not quantify natural sources of food. One observation to support this
hypothesis was four larval lake sturgeon that were found in a heath tray several weeks

after hatch that had never been offered brine shrimp but were equal in size to individuals

74

that had been on food for weeks. An analysis of the potential larval food sources
(zooplankton) contained in ﬁltered river water supplying streamside hatcheries would be
an important future research.

Once all juvenile lake sturgeon transitioned to blood worms (~day 50) survival
was very high and close to 1 for the remaining rearing period. We observed minor peaks
of mortality at the streamside hatchery in both years following this period. These
mortalities occurred in seemingly healthy ﬁsh which may indicate the potential for either
parasitic or disease issue at the streamside hatchery. If these mortalities were indeed a
ﬁmction of natural parasites, then weekly salt bath treatments may have helped reduce the
occurrence of parasite-mediated mortality to a minimum. At the end of the rearing period
and prior to release into the wild, a subset of juvenile lake sturgeon ﬁ'om both hatchery
environments were required by the Michigan Department of Natural Resources to
undergo routine disease testing. Results of these tests were inconclusive. Programs
utilizing streamside hatcheries for lake sturgeon culture may need to have more frequent
testing done to ensure ﬁsh quality and reduce mortalities.

Juvenile lake sturgeon grew rapidly in all years at both hatchery environments.
Growth in both hatcheries was a minimum of 7.2mm per week which is similar to
estimates from other studies on lake sturgeon (Fajfer et al. 1999). Holtgren et al. (2007)
also reared juvenile lake sturgeon in a streamside hatchery and experienced higher
growth compared to the streamside hatchery used in this study with a limited number of
ﬁsh being fed continuously with automated feeders. Growth was signiﬁcantly higher at
our traditional hatchery in 2006 compared to the streamside hatchery in the same year

and both hatcheries in 2005. Fish ﬁ'om all collection methods grew approximately the

75

same within a hatchery environment. Inter-individual variation in growth increased
throughout the rearing period in both years. Differences may be attributed to increased
competition within tank sections as certain individuals grow faster than others. In most
production hatchery settings these ﬁsh would be graded in order to keep growth rates of
all ﬁsh as close to maximum as possible. Our goal was not to maximize growth but to
compare growth and survival across two hatchery environments and three different
collection methods.

Juvenile lake sturgeon grown under different levels of habitat complexity
exhibited growth trends that were opposite from those observed during the larval stage.
Fish in the open treatments grew signiﬁcantly more over a six week period than those
reared in tanks with available cover. Observation of juvenile feeding behavior from this
experiment indicated that ﬁsh actively searched for food for extended periods of time
between feedings. Fish reared in the open treatments were able to ﬁnd and consume
100% of the provided food while lake sturgeon grown in treatments with cover may not
have been successful in ﬁnding food. Other studies have found no difference in growth
between traditional and semi-natural rearing conditions (Berejikian et al. 2001; Maynard
et al. 2004). However, Berejikian et al. (2000) found that steelhead fry reared in an
enriched hatchery environment grew faster than those from a conventional environment
when the two treatments were stocked together but not when they were stocked
separately. Results from this and other studies suggest that individuals reared in semi-
natural hatchery environments may not exhibit a growth advantage until stocked into the
wild. Unfortunately we could not evaluate juvenile lake sturgeon used in this study

beyond the hatchery environment. Studies that have used a more semi-natural rearing

76

approach suggest that factors other than growth may be important for supplemental ﬁsh
following release. For example, available structure and habitat in rearing tanks has been
shown to inﬂuence coloration patterns in ﬁsh (Ellis et al. 1997). Lake sturgeon have
evolved patterns of body coloration that reﬂect the bentlric (sand) environment that they
live in. If hatchery produced juveniles exhibit different coloration patterns than this may
increase probabilities of survival. This research question is one that will be important for

conservation hatchery programs to evaluate.

CONCLUSIONS

We documented differences in growth and survival of lake sturgeon between two
different hatchery environments. Survival ﬁ'om the egg stage until the end of the rearing
process was highest at the streamside hatchery (7.7%) in 2006 and lowest at the
traditional hatchery in 2005 (4.8%). The majority of mortality was incurred during the
egg stage at both hatchery environments. The main difference between the two hatchery
environments was that offspring reared at the streamside hatchery experienced a natural
temperature regime. We found that this natural temperature regime resulted in
maintaining natural variation in egg incubation time and larval size at hatch. The
traditional hatchery environment minimized this variation. Increasing temperature had
negative effects on egg incubation and larval size at hatch. Rearing eggs across a more
natural temperature regime is important for hatchery programs wanting to maximize

survival and natural variation in offspring phenotype. Repeatability for spawning time in

77

adult lake sturgeon may be inﬂuenced if eggs that are deposited in late conditions are
reared only in early environmental conditions. Stocked offspring may all return at the
same spawning time and diminish any adaptations that had evolved due to reproductive
isolation between spawning groups. Production of lake sturgeon in hatcheries has been
limited in certain years due to poor success obtaining eggs from the wild breeding
population. We demonstrated that rearing offspring ﬁ'om three collection methods results
in similar levels of survival and growth. However, natural food sources need to be
evaluated for offspring collected as dispersing larvae. Conventional hatchery practices
purposely reduce individual size variability in juveniles and we found that by rearing
different groups of offspring produced ﬁom different reproductive periods we were able
to maintain variation in ﬁsh size which increased over time. The natural variability in
development and timing characteristic of wild ﬁsh may be inherent factors which enable
them to adapt to changing conditions.

Results from this study provide the ﬁrst comparison between streamside and
traditional hatchery rearing environments. Streamside rearing and the addition of habitat
complexity within rearing tanks produce lake sturgeon offspring that are morphologically
similar to their wild counterparts. Future research directed at identifying the effects of
different rearing environments on behavior following release as well as the mechanisms
and timing of imprinting would be important next steps in evaluating streamside rearing

for lake sturgeon.

78

Chapter 3
HATCHERY REARIN G ENVIRONMENT AND AGE AFFECT SURVIVAL AND
MOVEMENTS OF STOCKED JUVENILE LAKE STURGEON.

INTRODUCTION

Hatchery programs have been widely used as both an enhancement and
management tool to address declines in wild ﬁsh populations (W aples 1999). More
recently, hatchery pro grams are receiving increased scrutiny because. of the potential
negative impacts of captive reared individuals to natural populations (Ford 2002; Araki
et al. 2007a; Miller et al. 2004). Criticisms have been raised based on empirical
evaluations of the relative effects of different management options including; hatchery
rearing environments (Berejikian et al. 2000), age of release into the natural environment
(Paragarnian and Kingery 1992), and gamete, juvenile, or broodstock collection and
maintenance protocols (Flag and Nash 1999).

Stocking protocols developed from hatchery programs are used to increase wild
population abundance and probabilities of survival during critical life history stages
(Alverson 2002; Brown and Day 2002). Furthermore, they have the potential to
supplement recruitment when unfavorable environmental conditions result in high
embryo and larval mortality (Secor and Houde 1998). Although stocking programs have
been used to establish and enhance many ﬁsheries, programs have failed to increase the
numerical abundance of others (Secor et al. 2000; Brown and Day 2002; Myers et al.
2004). General conclusions concerning the effectiveness of stocking programs have not

been reached and challenges remain to reconcile beneﬁts and potential costs to

79

population dynamics, genetic integrity of resident populations, and to ecosystem
processes and resource economics (Travis et al. 1998; Utter 1998). Supplementation
programs are becoming increasingly important for conservation efforts of threatened and
endangered species (Brown and Day2002), including regionally imperiled lake sturgeon
(A cipenserﬁtlvescens).

The lake sturgeon is considered as threatened throughout much of its range. This
species has experienced dramatic declines in both numerical abundance and in
distributional range, and numbers have been projected to be less than 1% of historic
levels (Hay-Chmielewski and Whelan 1997). Current impediments to lake sturgeon
restoration include the sensitivity of adults to anthropogenic factors such as over-
harvesting, degradation in water quality and spawning habitat and loss of connectivity
due to impoundment (Holey et al. 2000). In recent years, lake sturgeon rehabilitation
through stocking has become a high priority throughout the Great Lakes (Peterson et al.
2007). Despite this, considerable uncertainty remains regarding the efﬁcacy of different
egg and larval collection methods, and the appropriate age of ﬁsh to stock. Success of
supplementation programs is typically based on contributions of stocked ﬁsh to older age
classes (Li et al. 1996) or return to post stocking assessments (Walters et al. 1997).
However, these methods are not amenable to lake sturgeon as they are not widely subject
to harvest and younger age classes are not efﬁciently captured in surveys using standard
sampling methods (Benson et al. 2005). Late age at maturity, infrequent spawning, low
recruitment rates (N ilo et al. 1997), and occupancy of geographically separated breeding
and nonbreeding areas also confound interpretations, dictating that other methods be

employed to evaluate supplementation programs. Stocking prescriptions have typically

80

been based on a large salmonid literature though species such as lake sturgeon have very
different ecologies, mandating that research investigating the success of alternative
methods of culture and release be conducted on this species explicitly.

A number of factors contribute to the success of hatchery based stocking
programs. Stocking of lake sturgeon throughout the Great Lakes region has employed
numerous techniques for collecting progeny (e.g., larval collection (Auer and Baker
2002) and direct gamete takes), and juveniles have been stocked at different ages (ﬁ'y,
fall ﬁngerling, and yearlings; Schramm et al. 1999). Stocking studies that have attempted
to identify optimal ages for release into natural environments have reported mixed results
with respect to survival based on age at stocking (Amtstaetter and Willox 2004; Elrod et
al. 1988; Margenau 1992; Secor and Houde 1998). Studies have indicated that rearing
hatchery ﬁsh to larger sizes leads to increased survival upon release (Gunn 1987;
McKeown et al. 1999; Yule et al. 2000). Recently streamside hatcheries have been
advocated as a restoration tool for lake sturgeon throughout the Great Lakes (Holtgren et
al. 2007). Streamside hatcheries use water from streams targeted for release to maximize
probabilities of imprinting and to help reduce domestication selection. Conservation
hatcheries have been used for trout and salmon and have attributed positive results to
more natural rearing conditions (Maynard et al. 1996; Berejikian et al. 2000). In this
study, we examined the effects of different hatchery environments by rearing lake
sturgeon in two different hatchery settings, a streamside hatchery and a traditional
hatchery, in order to quantify differences in post stocking success.

We tested the null hypothesis that hatchery rearing environment and age at release

are not signiﬁcant predictors of post-stocking survival for juvenile lake sturgeon. Studies

8l

identifying the success of stocking strategies typically deﬁne success by the proportion of
ﬁsh recaptured in subsequent assessments (Leber et al. 2005). We adopted this criteria
and directed research objectives to: 1) determine contributions of ﬁsh stocked ﬁ'om two
different rearing environments and three different collection methods on the rate of
recapture using in-stream assessment, 2) determine which stocking age results in higher
rates of capture (a surrogate measure of survival), and 3) quantify rates of downstream
dispersal as a function of age and size. Results of this study demonstrate how lake
sturgeon rearing conditions, collection methods, and size and age at stocking contribute
to the variance in probability of survival, and movements after release into the natural

environment.

METHODS

Study Site

Research was conducted on the Black Lake system, Michigan (Fig 15, see details
in Smith and Baker 2005; Smith and King 2005). Fish were stocked into the Upper Black
River, a fourth order stream located in the northeastern corner of the state of Michigan
that is highly inﬂuenced by dams. The hydrology of the Upper Black River provides a
unique opportunity to enumerate a large proportion of the annual adult spawning run as
well as collect gametes and juveniles. Physical variables (water depth and ﬂow; Table 7)
provide wadeable conditions for several kilometers downstream of the ﬁrst impoundment

allowing for instream experimental work.

82

Rearing environment

Juvenile lake sturgeon were reared in two hatchery environments. Half of all ﬁsh
were reared at a streamside hatchery using water from the natal stream on the Upper
Black River, Michigan. The remaining ﬁsh were reared using ground water at a state
hatchery in southwestern Michigan, representing a traditional hatchery environment. The
streamside hatchery used a ﬂow-through design where water was pumped directly from
the Upper Black River. Water was mechanically ﬁltered to remove sediments. F low rates
were kept constant and at a level that resulted in approximately two complete water
turnovers every hour. Cleaning, feeding, lighting, and water ﬂow protocols were
consistent between the two hatchery environments following common lake sturgeon
rearing methods (Ceskleba et al. 1985; Fajfer et al. 1999). Fish in both hatcheries were
kept in 1.22m diameter tanks (0.5m in depth) divided to include 2 replicates of three

different gamete/juvenile collection methods (discussed below).

Collection of progeny

We used three methods to collect lake sturgeon progeny. The ﬁrst two methods
are commonly used for many sturgeon species and included directed egg takes from
spawning females (Birstein 1993), and collection of recently hatched larvae dispersing
downstream of the spawning grounds (Auer and Baker 2002). The third method involved
collecting eggs naturally fertilized and deposited on the stream substrate by spawning
adults.

1) Direct take spawning: Lake sturgeon were captured on the spawning grounds

using large dip nets. Eggs were removed by hand stripping females captured in the act of

83

spawning. Eggs were placed in zip lock bags and transferred to a cooler to maintain
ambient river temperatures. Milt was collected using a 30ml syringe and was
immediately placed on ice. Fertilizations were conducted within 12 hours of egg
collection. Egg volumes ﬁ'om each female were measured, divided into two equal lots,
and fertilized separately with milt from two randomly selected males to create paternal
half-sib family groups (n=26 in year 1, n=25 in year 2). Half of the eggs ﬁom each half-
sib family were transported to the traditional hatchery and half were maintained at the
streamside hatchery. Eggs were incubated and hatched out in heath trays at both
hatcheries.

2) Drift Larvae: Sampling for larval lake sturgeon passively dispersing
downstream has been shown to be a viable collection strategy (Auer and Baker 2002,
Smith and King 2005). Larval sampling was conducted during a 5-hour period, beginning
at dusk (21:00) and ending in the early morning (03:00) at a sampling location
approximately 2 km downstream (Figs 1,15) ﬁ'om the primary spawning areas on the
Upper Black River. Five D-framed drift nets were set across the stream channel to
capture dispersing larval sturgeon. This design was replicated nightly each year for the
entire drifting period. Nets were checked hourly and the larvae within each net were
enumerated. Larval lake sturgeon captured each evening were reared at the streamside
hatchery until the end of the drifting period. The larvae were then divided between the
two hatchery environments. Minimum number of adults contributing to larval drift based
on capture data was 154 and 234 for year one and two respectively.

3) Naturally fertilized and deposited eggs: Systematic kick-net sampling was

conducted below observed spawning locations to collect eggs that were naturally

84

fertilized and deposited in the stream substrate. Transects were run across the stream at 1
meter intervals. At each interval we conducted kick-net sampling for 10 seconds.
Transects continued downstream in intervals of 5 meters until no eggs were collected in
consecutive transects. Eggs were incubated and hatched at the streamside rearing
environment because of pathogen concerns of the traditional hatchery. Larvae were then

divided between the two rearing environments.

Fish stocking and assessment

We used the Upper Black River as an experimental release site to test the
hypothesis that hatchery rearing environment and age were not signiﬁcant predictors of
post stocking survival. The release and assessment area encompassed a stream region of
approximately 4 km (Fig. 15). Three age classes were stocked into the Upper Black
River. Fish at 8 and 13-weeks of age were released during 2005 and l7-week old ﬁsh
were released in 2006. Prior to release, all individuals were measured (total length (cm)),
weighed (g) and tagged with a colored implant elastomer dye (Northwest Marine
Technology, WA, USA) unique to their rearing environment and gamete/juvenile
collection method. Elastomer was injected on the ventral side of the rostrum where the
colors were most easily distinguished. Juvenile lake sturgeon were transported from the
traditional hatchery to the streamside hatchery one day prior to each release. All age
classes were acclimated to ambient stream conditions within the streamside hatchery for
>12hours prior to being transported to the release site. All ﬁsh were released

simultaneously at the same location on the Upper Black River.

85

We chose a region of the river because of increased catchability and ease of
replication of methods across the different stocking ages. The release location did not
represent a stream region where juvenile sturgeon would naturally be present at the three
ages of release. Prior to release, fourcapture locations were established downstream of
the release site (Fig. 15). Within each site four D-framed drift nets were deployed across
the stream at equal intervals. Net position was recorded using a GPS unit. Water velocity
(m/s) and depth were recorded at the mouth and comers of each net. Nets were checked
and emptied at hourly intervals following release. Sampling persisted for approximately
24hours following release. Sampling at the most upstream assessment site was
discontinued for all release ages after no juvenile sturgeon were captured in consecutive
hours. Each captured individual was examined for elastomer marks indicating collection
method, rearing location, and total length (cm) was recorded. Fish were then released
immediately downstream of the drift nets.

Block nets were deployed downstream of the third site assessment site for both

the 13 and 17-week release ages. During the 17-week release a second block net was also

deployed below the fourth downstream site. The block nets consisted of 0.32cm2 mesh

leaders and large mesh (0.32cm2) fyke nets. Leaders were used to guide dispersing lake

sturgeon into fyke nets. Block nets were deployed to estimate drift net capture efﬁciency,
and to use as a second gear type to assess downstream dispersal. Block nets were not
installed during the 8-week release. Sampling protocols for the block nets were consistent
with those for the drift nets.

A DC tow-barge shocker was used to survey the stream predator community to

assess sources of mortality for released juvenile lake sturgeon. Shocking was conducted 1

86

hour following release. We conducted a single pass starting at the release location and
proceeded downstream past all assessment sites. Predators were captured and held in an
aerated container until the end of the pass. We used gastric lavage to expel the stomach
contents of all captured ﬁsh. All predators containing juvenile sturgeon were measured

for total length and gape width and height.

Statistical Analysis

All statistical analyses were performed using program R (R Development Core
Team 2007, http://www.rproj ect.org). Data were tested for normality and homogeneity of
variances using a Shapiro-Wilk test and by examining residuals versus ﬁtted values in R.
In cases of non-normality or heterogeneity of variances, analyses were run using arcsine-
square-root-transformed survival data and log transformed length data. For analyses
within and among release ages we only included data ﬁ'om the drift net captures. Two
dependant variables, the proportion of ﬁsh recaptured and the total length of recaptured
ﬁsh were examined. The proportion of ﬁsh ﬁom each collection method and rearing
environment captured was calculated for each hourly sampling interval at each
assessment site for all release ages. All recaptured proportion values were weighted by
the total number released ﬁom each hatchery environment and the different collection
methods within each release age. We used a general linear nrixed effects model to
examine the effects and interactions of six independent variables on the proportion of ﬁsh
recaptured. Fixed effects included assessment site of capture, gamete/juvenile collection
method, rearing environment, net location, physical stream variables (depth and ﬂow

measured at'the net), and the age of release. Time of capture was included as a random

87

variable in the model. We also estimated interactions among-age, rearing environment,
time of capture, and assessment site. Data collected ﬁ'om each block net used in the 13
and 17 week releases were analyzed separately due to lack of comparability to the drift
nets and because we were unable to replicate the sampling strategy across all release
ages. The same model was used for the block net as the drift nets except for the
assessment site variable. Efﬁciency estimates for the drift nets were calculated by
dividing the total amount of ﬁsh recaptured in the drift nets by the total amount of ﬁsh
recaptured within the same time period (both gear types).

For all stocking events we analyzed the response variable total length of
recaptured juvenile lake sturgeon using a general linear mixed effects model. In this
model the response was the observation of length for a speciﬁc ﬁsh at a distinct sampling
time and at a speciﬁc sampling site. Fixed effects included capture site, collection
method, and rearing environment. Time of capture was included as a random effect in the
model. We also examined the interaction effects between all independent variables. The
independent variable age was not examined in our analysis of total length because there
were signiﬁcant differences between ages in terms of total length at the time of release.
The block net data was analyzed using the same models.

We tested for mean differences in the total length of recaptured ﬁsh in relation to
the total lengths at the time of release for each treatment group at each release age using a
2 factor ANOVA. The response variable, length, was predicted by period (prior to release
or recapture), time (recapture hour), physical stream variables, net location, and we
examined the interaction between the independent variables. Analyses were conducted

separately for ﬁsh captured in the block nets and drift nets for each release age. To test

88

for biases in sized based catchability between the two gear types we used an unpaired t-
test to test for mean differences in the total lengths of recaptured ﬁsh from each gear

type. This was conducted separately for data from the 13 and 17 week releases.
RESULTS

Results are presented for dependant variables, recapture rates and size, as they
related to age, hatchery rearing environment, and gamete/juvenile collection method.
Stream depth and ﬂow as well as net position were removed from the ﬁnal model due to
insigniﬁcance. Though sampling was conducted for 24 hours, no juvenile lake sturgeon
were captured beyond 15 hours following release. Physical conditions, including water
depth and velocity were not statistical predictors of capture rates but varied between the

release ages (Table 7).

Variation in Recapture Rates
Based on 11,721 juvenile lake released across each collection method, hatchery

rearing environment, and age class (Table 8) we documented signiﬁcantly higher

recapture rates for ﬁsh released at 17 weeks of age (F 21,148 = 6.45, P < 0.01) compared

to ﬁsh released at 8 and 13 weeks of age (Fig 16). During the 8 and l3-week releases we

observed a signiﬁcant main effect of hatchery rearing environment on the proportion of

ﬁsh recaptured. Signiﬁcantly fewer ﬁsh (F 21,148 = 6.45, P = 0.03) recaptured were

reared in the traditional hatchery environment relative to the streamside hatchery (Fig.

16). Signiﬁcantly fewer ﬁsh were captured at increasing downstream distances from the

89

release site (F 21,148 = 6.45, P = 0.03) relative to the most upstream assessment sites (Fig.

16). We recaptured signiﬁcantly more ﬁsh at downstream assessments sites in the 13-
week release when compared to the 8-week release and in the l7-week release when

compared to both the 8 and 13 week releases (Fig. 16). There was a signiﬁcant effect of

time following release on the proportion of ﬁsh recaptured (1721,1453 = 6.45, P < 0.01). At

8 and 13 weeks of age we captured ﬁsh moving downstream rapidly after release during
the daylight hours (Fig. 17). At l7-weeks of age we captured very few ﬁsh during the
daylight hours immediately following the release (Fig. 17). A large proportion of total
captures occurred during evening hours >8 hours following the release (Fig. 17). There
was no main effect of treatment group on recapture success among release ages. Distinct
differences among treatment groups (Table 8) occurred within sites at distinct time
periods but no overall trends emerged from the analysis.

The block net deployed during the 13-week and 17-week release provided
estimates of the efﬁciency of the drift nets as well as a second gear type for assessing
downstream dispersal. Within the 13-week stocking event the block net below the third
downstream site recaptured 10.3% of the total number of ﬁsh released. There were

signiﬁcantly more ﬁsh moving downstream of this assessment site that were from the
streamside hatchery environment (F137 = 3.04, P < 0.01). Efficiency of the drift nets
varied hourly with a mean (:t 1.SE) of 12.6 d: 2.8% during the 13-week release. There
were no distinct trends in treatment differences among sampling times. Within the 17-

week stocking event the block net below the third downstream sampling location resulted

in recapturing 15.01% of the total number of ﬁsh released. Drift net efﬁciency was 14.3 i

90

5.2%. There was no effect of rearing environment on the proportion of ﬁsh moving

through this site at any time during the sampling period. As with the drift nets for this age

class, signiﬁcantly more ﬁsh were captured at later sampling hours (F 7,11 = 38.52, P <

0.01). The block net below the fourth downstream site produced recaptures totaling
19.26% of the ﬁsh released. Drift net efﬁciency was 6.2 :I: 2.1%. There was no effect of
rearing environment on the proportion of ﬁsh recaptured. Again, we caught signiﬁcantly

more ﬁsh at later sampling hours compared to hours immediately following release at this

downstream site (F 6,8 = 11.82, P < 0.01). Even though the drift net efﬁciency varied

hourly, results from the block nets are consistent with those from the drift nets.

Variation in Size

Documenting differences in length within and among release ages is important for
determining size based downstream dispersal and as a factor affecting differential
survival. We observed no signiﬁcant differences between rearing environments in the
total length of ﬁsh prior to release for the 8 and 13-week release ages (Table 8). At 17

weeks of age ﬁsh ﬁom the traditional rearing environment were signiﬁcantly larger

(F1509 = 11.52, P < 0.01) than those reared at the streamside hatchery at the time of

release (Table 8). There was no main effect of rearing environment or treatment type on
the size of recaptured ﬁsh in both the 8 and 13-week releases. There was a signiﬁcant
effect of rearing environment in the l7-week release with recaptured ﬁsh from the

traditional hatchery being signiﬁcantly larger than those from the streamside hatchery

(F 26,583 = 22.32, P < 0.01). In all release ages there was a signiﬁcant positive effect of

9l

time following release on the size of recaptured ﬁsh (Fig. 18). Fish captured in later post-

release were signiﬁcantly larger than ﬁsh captured soon after release in the 8-week
(Fug-02 = 4.06, P < 0.01), 13-week (F 18,302 = 7.46, P < 0.01), and l7-week (F26,583 =
22.32, P < 0.01) release (Fig. 18).

Sizes of ﬁsh captured in the block nets at both 13 and l7-weeks revealed trends
consistent with those from the drift nets. There was a signiﬁcant increase in the length of

ﬁsh captured in successive sampling periods post-release in the block during the 13-week

release. The same results were found for both block nets during the l7-week release
(F7,621 = 23.97, P < 0.001).
During the release at 8 weeks of age the mean total length of recaptured ﬁsh was

signiﬁcantly smaller (F 5,589 = 11.50, P < 0.01) than the mean length prior to release with

the exception of the ﬁnal recapture period. This was consistent in the 13 week release

with ﬁsh lengths prior to release being signiﬁcantly larger than recaptured individuals at
all hours except the ﬁnal net check (F 6,1 596 = 65.82, P < 0.01). The mean total length of
ﬁsh captured in the block net at 13 weeks of age was signiﬁcantly smaller compared to

mean total length prior to release for all hourly checks (F 7,1597 = 56.91, P < 0.01). At 17

weeks of age, juvenile lake sturgeon recaptured in the drift nets were signiﬁcantly smaller

on average than prior to release for all hourly checks except the ﬁnal recapture period (F

4,1191 = 7.07, P < 0.01). Fish recaptured in the block nets at 17 weeks of age were

signiﬁcantly smaller for all recaptured periods except the ﬁnal two net checks for both

the third (F6316 = 16.53, P < 0.01) and fourth most (F6,2859 = 17.59, P < 0.01)

92

downstream sites. There was no signiﬁcant difference in the size of juvenile sturgeon
recaptured in the block net compared to the drift nets (df= 392, t = 1.293, P = 0.197)
during the 13 week release. During the 17 week release there was no signiﬁcant
difference in sizes between the gear types at the third most downstream site (df= 691 , t =
0.092, P = 0.93) but a signiﬁcant difference was found at the fourth most downstream
assessment site (df= 848, t = 3.51 , P < 0.01) with ﬁsh in the block and drift nets

averaging 153.47 i 0.97mm and 138.35 d: 3.80mm respectively.

Evidence of mortality

We were unable to quantify sources or rates of mortality by ﬁsh predators
following release using the tow barge shocker. During the release at 8-weeks of age two
rock bass were identiﬁed as having juvenile lake sturgeon in their stomachs. Further
attempts to quantify predation at later release ages proved unsuccessful. At both 8 and
l3-weeks of age we visually observed crayﬁsh preying upon juvenile lake sturgeon after

release.

DISCUSSION

Sampling protocols providing quantiﬁable and comparable results are currently
lacking for age-0 lake sturgeon (Holey et al. 2000), despite recent advances in visual
surveys (Benson et al. 2005). Using this natural stream as an experimental tool was
positive for attempting to provide quantiﬁed estimates of survival and downstream

movement for a species that is otherwise difﬁcult to capture and enumerate. Assessments

93

of juvenile lake sturgeon released at 8, 13, and 17 weeks of age revealed that there were
low rates of recovery and inferentially high rates of mortality. Rates of recovery
increased with increasing age. We demonstrated an effect of hatchery rearing
environment on juvenile lake sturgeon survival and movements when juveniles were
stocked at 8 and 13 weeks of age. Our ability to detect this difference is an indication that
even moderate differences in hatchery rearing environments potentially related to
domestication selection (e.g. Lynch and O’Hely 2001) can play an important role in post
release movements and survival. The effects of rearing environment immediately
following release may be reduced by releasing ﬁsh at older ages as evident by the lack of
signiﬁcant differences attributed to hatchery environment for 17 week old ﬁsh. However,
theoretical (Lynch and O’Hely 2001) and empirical (Araki et al. 2007a) data on the
effects of hatchery rearing environment on reproductive success of hatchery reared ﬁsh
dictates that further research be conducted for sturgeon.

Signiﬁcant differences in recapture rates between different hatchery environments
indicates the need to tailor culture methods that maximize probabilities of survival
through important life-history transitions. This is especially important when dealing with
threatened or endangered species such as lake sturgeon. Modifying juvenile rearing
environments to approximate natural conditions is increasingly used to minimize the
degree of domestication for captive animals (Flag and Nash 1999). Studies of ﬁsh reared
in different environments have documented differences in behavior (Olla et al 1998;
Berejikian et al. 1999; Flagg and Nash 1999), post stocking survival (Wiley et al. 1993;
Maynard et al. 1996), growth rates (Mesick 1988), social rankings (Berejikian et al.

2000), and development of appropriate body camouﬂage coloration (Maynard et al.

94

1996). Olla et al. (1998) suggested that ﬁsh reared in a sensory deprived hatchery
environment have poor feeding efﬁciency and are less capable of avoiding predators.
Differences between the streamside hatchery and the traditional hatchery used in this
study were likely sensory because physical rearing conditions such as tank size, shape,
ﬂow rate, and feeding regimes were consistent between the two environments. Juvenile
lake sturgeon reared in the streamside hatchery may have beneﬁted from ﬂuctuations in
temperature mimicking natural environmental conditions. Mechanical ﬁltration within
the streamside hatchery, though effective at removing larger particulates, likely allowed
exposure to stream organic material and biological organisms. Furthermore, the water at
i the streamside hatchery potentially contained natural stream chemical cues ﬁ'om either
conspeciﬁcs, predators, or other species.

We observed a signiﬁcant effect of time on ﬁsh capture. During releases
conductal at 8 and 13 weeks of age, we found that ﬁsh were captured at signiﬁcantly
earlier times following release relative to recoveries at 17 weeks of age (Fig. 17). Fish
released at 17 weeks dispersed downstream after sunset. This negative photo-tactic or
nocturnal behavior has been noted with lake sturgeon larvae dispersing downstream of
spawning grounds (Auer and Baker 2002) and similar behavioral trends have been noted
in ﬁeld (Chiasson et al. 1997; Benson et a1. 2005) and hatchery studies on juveniles
(Peake 1997). This is the ﬁrst documentation of nocturnal behavior patterns in hatchery
reared juvenile lake sturgeon immediately following release into the natural environment.
It has been shown that nocturnal behavior of other sturgeon species (Kynard et al. 2005;
Kynard and Parker 2005) as well as other ﬁsh species (Bradford and Taylor 1997; Crisp

1991) is a dominant feature of migration and foraging in the ﬁrst year of life. Variability

95

in the proportion of juveniles recaptured was high during the early hours following
release at the 8 and 13 week age classes. This is likely due to the fact that most
individuals moving passively in the current were unable to control their rate of
downstream dispersal. We saw a noted decrease in the variability in sizes of ﬁsh
dispersing downstream later in the day at 8 and 13 weeks which can be attributed to the
movement of primarily larger ﬁsh (Fig. 18). There was a signiﬁcant effect of ﬁsh size on
dispersal at 17 weeks of age as well. However the variation surrounding the estimates
was noted in later hours, opposite to that of the earlier age classes (Fig. 18). This increase
in variation at later time periods following release indicates that the majority of juveniles
had an ability to choose dispersal time. Size-based downstream dispersal has been noted
with other species of ﬁsh. Bradford and Taylor (1997) found that larger Chinook salmon
fry had a higher probability of dispersal during nighttime hours than smaller individuals
of the same cohort. Age dependant diﬁerences in physical abilities, as suggested by
differences in timing and size distributions of dispersing juveniles should be used to
develop release strategies for juvenile lake sturgeon.

Our estimates of recapture success and stream movements among ﬁsh from
different collection method within each release age are robust because of the large
number of ﬁsh released (Table 8). However, comparisons of results across age classes
should be interpreted in the context of limitations imposed by our sampling design.
Catchability has been shown to increase with increases in body size of ﬁsh (Borgstroem
and Skaala 1993) and models assuming equal catchability typically underestimate the
actual number of ﬁsh in the population (Mantyniemi et al. 2005). Juvenile lake sturgeon

were stocked over two years due to a combination of rearing limitations and the need for

96

large sample sizes for release. The use of the block nets as a means of assessing the drift
net efﬁciency and as second gear type was important. Consistency of results between the
two gear types was encouraging because drift net efﬁciency was generally low.
Furthermore, we only documented differences in size of ﬁsh between the two gear types
at one assessment site. This was at the most downstream assessment site and slower
water velocities (Table 8) likely allowed the juvenile lake sturgeon to elude this passive
sampling gear. Finally, transport of juveniles reared at the traditional hatchery to the
streamside hatchery may have posed increased levels of stress. However, ﬁsh were
acclimated at the streamside hatchery overnight (>12hrs) prior to release and we don’t
believe transport affected our results. ,

River conditions varied between release events with the lowest water velocities
occurring during the 17 week release. Lower drift net efﬁciency at this site can likely be
attributed to the decrease in water velocities. Studies have indicated that higher water
velocities can strongly inﬂuence downstream displacement of young ﬁsh (Dauﬁ'esne et
a1. 2005). This may have been the case for juvenile lake sturgeon at 8 and 13 weeks of
age however, we documented a subset of larger ﬁsh being able to control their rate of
dispersal in each release.

It is imperative that well deﬁned species speciﬁc protocols be developed for
stocking programs (Cowx 1999). Supplementation protocols should be tailored to species
speciﬁc ecologies and behaviors, are assessments should be made to quantify estimates of
survival following release. Protocols have been developed for many recreationally and
commercially important species that incorporate ecologically important information.

Largemouth bass (Micropterus salmoides) have been shown to have low dispersal

97

following release indicating that localized supplemental stocking in distinct areas of
complex reservoirs may increase the beneﬁts on the population level (Copeland and
Noble 1994). Ellison and Franzin (1992) indicate that matching stocking times and
locations to appropriate food resources are important in walleye (Stizostedion vitreum)
introductions. Our results indicate that supplementation strategies should focus on night
releases and in areas of moderate ﬂow. These parameters ensure that rates of dispersal are
high immediately following release and will likely have an indirect effect on levels of
predation by reducing mortality incurred by visual predators. It is critical that release sites
are evaluated on a system by system basis with emphasis placed on determining areas
that maximize suitable juvenile nursery habitat. For lake sturgeon, natural juvenile
rearing habitat has been deﬁned as predominantly shallow riverine habitat consisting
mostly of sand (Benson et al. 2005). Our research addresses the importance of tailoring
stocking programs to individual species and systems rather than focusing on numbers and
sizes of ﬁsh.
CONCLUSIONS

In conclusion, our study addresses the need for further empirical and
experimentally rigorous evaluations of hatchery rearing and stocking programs for little
studied species such as lake sturgeon. Streamside rearing may be advantageous to
small/young lake sturgeon by exposing them to an enriched environment prior to release.
However there may be an age/ size threshold where the effects of rearing environment on
movements and survival immediately following release are diminished. Further work that

characterizes mechanisms inﬂuencing differences among hatchery rearing environments

98

for lake sturgeon is important to the development of system speciﬁc management

prescriptions.

99

Chapter 4
OVERWINTER SURVIVAL OF STOCKED AGE-0 LAKE STURGEON REARED IN
NATAL AND NON-NATAL ENVIRONMENTS

INTRODUCTION

Survival through the ﬁrst winter of life is commonly viewed as a critical
determinant of year class strength (Oliver et al. 1979). Despite this, quantiﬁed estimates
of overwinter mortality are lacking for many species. Survival throughout early life
history stages, including the ﬁrst winter of life, is mediated by a number of selective
processes that operate as ﬁsh experience elevated rates of size dependant mortality due to
both predation and energy depletion (Miller et al. 1988). Physiological changes (e.g.,
metabolic rates, hematological parameters) during the winter months result in decreasing
condition and a depletion of energy reserves, the effects of which differ among
individuals ﬁ'om the same cohort (Cunjak, 1988). Overwinter energy reserves are highly
correlated with the length of the growing season, which is reduced for temperate species
(Post and Parkinson 2001). Furthermore, ﬁsh produced in later reproductive events have
even greater constraints with the shortened growing season acting as a strong selective
pressure against overwinter survival (Biro et al. 2003). Knowledge of age-speciﬁc
survivorship through this critical stage is particularly important as a measure of the
success of hatchery programs, which are increasingly used as a part of recovery programs

for imperiled species.

lOO

Populations of lake sturgeon (Acipenser fulvencens), a native ﬁsh of the Great
Lakes, have been numerically depressed throughout their range due to anthropogenic
factors and are believed to be at less than 1% of historical numbers (Hay-Chmielewski
and Whelan, 1997). Restoration activities have been initiated, largely in the absence of
data regarding their effectiveness, focusing on hatchery supplementation (Holey et al.
2000). Recently, there has been an interest in the use of streamside hatcheries (Holtgren
et al. 2007) as a restoration tool. Streamside hatcheries use water from streams targeted
for release, potentially reducing the degree of domestication and increasing the likelihood
of imprinting. Assessing the overall impact of stocking juveniles reared in streamside
versus traditional hatchery environments is important to project the effects of
supplementation on long-term population abundance. This assessment is complicated by
this species’ unique life history characteristics (e.g., delayed sexual maturity), and
inefﬁcient collection methods for young age classes (Benson et al. 2005a). Low sampling
efﬁciency affects accuracies of survival estimates which impedes evaluations of post-
release survival of juveniles reared in different hatchery environments. Most lake
sturgeon culture and stocking efforts in the Great Lakes have reared ﬁsh to fall ﬁngerling
size (~4 months) under the assumption that larger ﬁsh survive better following release
into the natural environment (Schram et a1. 1999). No quantitative data are available on
age-speciﬁc lake sturgeon survival, including survival through the ﬁrst winter.

Ultrasonic telemetry is a valuable ﬁsheries technique used to estimate parameters
such as mortality, distance traveled, and range size (Flavelle et al. 2002; Tavemy et al.
2002; Zamora and Moreno-Amich 2002). Despite some limitations (e.g. sample size,

battery life,'and handling-induced mortality), telemetry allows the collection of multiple

lOl

observations from a single individual, which is an advantage over passive capture
techniques such as gillnetting (Zamora and Moreno-Amich 2002). Recent advancements
in telemetry technology, particularly decreasing transmitter size, have increased
opportunities to study species over _a greater range of ages. This is especially important
for studies interested in assessing the survival of ﬁsh released during their ﬁrst year of
life. Telemetry studies on sturgeon have generally been conducted in the spring and
summer months (Hay-Chmielewski 1987; Holtgren and Auer 2004; Smith and King
2005). Research through the winter months has described overwinter movement patterns
and identiﬁed habitats that serve as both important nursery grounds for subadults and
staging areas for adults (Quist 1999; Sulak and Clugston 1999; Li et al. 2007).

There is a critical need for determining over-winter survival of age-0 sturgeon.
Our overall goal was to provide estimates of overwinter survival, however our speciﬁc
objectives were to, 1) estimate overwinter survival of hatchery reared juvenile lake
sturgeon released at a traditional (fall ﬁngerling) age and 2) determine differences in
survival between juvenile lake sturgeon reared in two different hatchery environments, a
streamside hatchery on the natal river and a traditional non-natal hatchery environment.
This study provides both the ﬁrst quantitative estimate of overwinter survival and the ﬁrst
attempt at assessing the effects of hatchery rearing environment on overwinter survival

for juvenile lake sturgeon.

l02

METHODS
Study Site
Juvenile lake sturgeon were collected ﬁ'om a population located in Black Lake,
Michigan (See Smith and King 2005 for site description). Black Lake encompasses
approximately 4,000 hectares. Black River, the primary habitable tributary to Black Lake
provides spawning habitat for adults and nursery habitat for juveniles. The population of
lake sturgeon is closed to immigration by dams on both the Lower Black River and the

Upper Black River.

Rearing Environment

Twenty juvenile lake sturgeon were reared ﬁom eggs to six months of age in each
of two environments, a streamside hatchery using water from the natal Upper Black
River, and a traditional hatchery (The Michigan Department of Natural Resources
(MDNR) Wolf Lake hatchery) using a non-natal (ground water) source. Rearing
conditions (circular ﬁberglass tanks 1.22m in diameter) and feeding regimes were
consistent between the two hatcheries. Fish were reared separately for three months at
which time all individuals from the streamside facility were transported to the traditional
hatchery and retained for an additional three months. Fish were relocated from the
streamside hatchery to the traditional hatchery for several reasons, the ﬁrst being that the
streamside hatchery could not be operated during the fall. Secondly, we needed to
increase growth to a size (~approximately 25 cm total length) that would allow insertion

of the ultrasonic transmitters.

103

Transmitter Implantation and Monitoring

Juvenile lake sturgeon were surgically implanted with small coded ultrasonic
transmitters; 1.7 i 0.01 cm (mean :t 1 SE) in length and 0.7 :t 0.001 cm in diameter with
a mass of 1.95 :l: 0.01 g in water (Vernco, model V7, Nova Scotia, Canada). Prior to
surgery, streamside hatchery ﬁsh had a mean :1: SE total length (TL) of 31.05 :I: 0.34cm

and a weight (W) of 106.39 d: 2.59g. Traditional hatchery ﬁsh were 31.45 :t 0.37cm in TL

and 106.44 3: 3.25g in W. All surgeries were conducted between December 8th and

December 12th, 2005. The weight of the transmitter did not exceed 2% of the ﬁsh weight,

which fell well within the generally accepted 2% rule (Winter 1996). Fish were
anesethized using tricaine methane sulfonate (MS222; 125 mg/l; Summerfelt and Smith
1990) in an aerated container prior to surgery. Sedation was maintained with a constant
dose of MS222 (50 mg/l) recirculating through a portable surgical table (LaVigne 2002).
Transmitters were anchored to the wall of the peritoneal cavity using non-absorbable
sutures to reduce movements of the transmitter in the body cavity. Incisions were closed
using 3-0 gauge non-absorbable monoﬁlament nylon sutures (Ethicon) in an interrupted
pattern. Ovadine antiseptic was applied to the wound to promote rapid healing. Following
surgery, ﬁsh were monitored for a 3-day period for complications from the surgical
procedure. Juvenile lake sturgeon were transported to the release site in an aerated

stocking trailer. Fish were acclimated into the wild using a net pen for a 3 hour period

prior to release into Black Lake on December 16th 2005. Sonic transmitters were set at an

operating frequency of 69 kHz and programmed to emit a coded pulse randomly between

403 and 1203.

l04

Manual tracking excursions followed transects that spanned the length of the lake.
Manual tracking was conducted on successive days during which time we attempted to
cover the total area of Black Lake. Attempts were also made to track upstream and
downstream in the Black River. When possible, tracking was conducted on a weekly
schedule. Movements of sonically tagged ﬁsh were monitored ﬁom a boat using a Vernco
receiver (Model VR100) equipped with both omni-directional (VH65) and directional
(V10) hydrophones. The directional hydrophone had a horizontal beam width of 22
degrees and a vertical beam width of 150 degrees. Transmitters had a potential detection
range of 1 km. Hydrophones were deployed every 500 m on calm days and every 250 m
or less during days with moderate waves. Recorded positions of detected ﬁsh were made
using a handheld WAAS enabled GPS receiver and bearings were taken using a magnetic
compass adjusted to obtain true magnetic north. Fish locations were triangulated with a
minimum of three geographic points. Three automated hydrophone receivers (Vemco,
Model VR2) were strategically placed within Black Lake to detect movement between
distinct zones. Collected positioning data were managed and analyzed using geographic
information system (GIS) software (ESRI, ArcMap v. 8.3). GIS data were managed in the
form of Universal Transverse Mercator (UT M) units. Temperature was monitored near
the release site for the duration of the observational period at l-hour intervals using an
Onset HOBO Pro underwater temperature logger (Onset Computer Corp., Bourne,

Massachusetts).

l05

Data Analysis

Juvenile lake sturgeon locations were calculated using the bearing angle and the
GPS coordinates from that observation using the telemetry software Locate II (N arns
2001), which uses the method of least squares to estimate the error associated with the
predicted location. Juvenile lake sturgeon exhibiting active movement patterns after ice-
out were assumed to have survived. Average inter-observational distance traveled
between tracking events was calculated for each individual as further evidence of
survival. This was calculated by taking straight line measurements between successive
detection points. We also calculated the inter-observational distance of each individual
from the release site by taking straight line measurements to each detection point from
the release location.

We used a chi-square test to determine signiﬁcance of differences in survival
attributed to rearing location. We used a t-test to examine differences in the size at
stocking for ﬁsh between the two hatchery enviromnents, to test for the sizes of surviving
ﬁsh relative to undetected ﬁsh, and to test for differences between hatchery environments

in distance traveled.
RESULTS

There were no signiﬁcant differences in length or weight of ﬁsh prior to release

between hatchery rearing environment (TL: df = 19, t = 0.42, P = 0.68; Weight: df = 19, t
= -0.29, P = 0.78). Tracking was conducted from April 15th until June lSt 2006. Sixteen

juvenile lake sturgeon were detected and determined to be actively moving during the

l06 »

period of observation (Table 9). This corresponds to a minimum survival of 40%. Seven
of these individuals were reared at the streamside hatchery environment and nine
individuals were ﬁom the traditional hatchery environment. A chi-square test revealed no

signiﬁcant difference in the number of ﬁsh surviving between the two rearing

environments (df = 1,12 = 0.25, P = 0.62). Of the remaining 24 ﬁsh that were released,

only two other ﬁsh were located but were assumed to be dead after several consecutive
tracking excursions identiﬁed them to be in the same location. This location was close to
the release site and a stationary automated receiver recorded these individuals over an
extended time period. All detected and surviving individuals remained within 1 km of the
release location (Table 9) until the end of May 2006. Manual tracking and automated
receivers documented a number of ﬁsh moving extensively within this zone. There was
no signiﬁcant difference between ﬁsh from the different hatchery environments in inter-
observational distance traveled (df = 14, t = -0.24, P= 0.81) or in inter-observational
distance from the release site (df = 14, t = -0.96, P = 0.351). Average distance traveled

between detection events was 0.19 i 0.14 km (Table 9). Water temperatures ranged

between 1.4 and 223°C (mean i SE, 14.4 i 0.50C) during the observational period.

The size distribution at the time of release of ﬁsh identiﬁed as surviving was not
signiﬁcantly different ﬁ'om ﬁsh that were not detected for TL (df = 38, t = -1.63, P =
0.11) and W (df = 38, t = -1.29, P = 0.20). A t-test revealed no signiﬁcant difference
attributed to hatchery environment in TL (df = 14, t = -0.88, P = 0.39) or W (df = 14, t =

0.86, P = 0.40) of surviving ﬁsh.

l07

DISCUSSION

This quantitative estimate of overwinter survival is the ﬁrst recorded for hatchery
reared juvenile lake sturgeon. Our study was not exhaustive and the estimate presented
(40%) should be considered as a minimum value of survival. This result is encouraging
considering the number of deleterious factors associated with long-term hatchery rearing,
including elevated levels of domestication (Huntingford 2004). Our survival estimate for
age-0 lake sturgeon is comparable to estimates produced for other non-Acipenseriform
species. For example, Biro et al. (2004) estimated overwinter survival for age-0 rainbow
trout (Oncorhynchus mykiss) stocked into two different lakes to be 27% and 33%
respectively. Overwinter survival of wild juvenile coho (Oncorhynchus kisutch) salmon
has been documented as high as 46% (Quinn and Peterson 1996) while monthly survival
estimates for juvenile largemoutlr bass (Micropterus salmoides) have been found to be as
low as 54% during the winter months (Miranda and Hubbard 1994). Finally, Jonas et al.
(1996) found that survival of fall stocked muskellunge (Esox masquinongy) was low
(2.3%) and attributed results to a signiﬁcant loss in energy reserves through the winter.

Greater stored energy reserves of larger age—0 individuals of some ﬁsh species
increase probabilities of over-winter survival relative to smaller individuals (Shuter et al.
1980; Miranda and Muncy 1987). We did not document a signiﬁcant difference in the
sizes of ﬁsh detected versus not detected in our study however it is important that we
compare the size of ﬁsh released in this study to wild ﬁsh of similar age. Often, estimates
of survival of hatchery stocked ﬁsh might be inﬂated because hatchery ﬁsh are generally

larger and in better condition than their wild counterparts. This is attributed to a longer

l08

hatchery growing season. Hatchery reared juvenile lake sturgeon in our study were
approximately 31.23 cm in total length and 106.41 g in wet weight at the time of release
in early December. Wild juvenile lake sturgeon captured in the fall months as late as
November in the Peshtigo River, Wisconsin, were as large as 31.6 cm in total length and
weighed as much as 134 g (Benson et al. 2005b) indicating that they were in comparable,
if not better, condition relative to the hatchery produced ﬁsh. Even though hatcheries
provide progeny for stocking that are of good condition at the time of release, tradeoffs in
elevated levels of domestication and reduced foraging ability might reduce probabilities
of survival if monitored over a longer period.

One observation from detected individuals was that juvenile lake sturgeon
maintained activity throughout the observational period. However individuals remained
within a very small area close to the release location. Smith and King (2005) found that
yearling sturgeon in Black Lake had individual areas of activity that were likely
associated with preferred habitats. Restricted movements have also been documented by
Fox et al. (2002), who found that Gulf sturgeon (Acipenser oxyrinchus desotoi) remained
in localized areas for extended time periods before rapidly dispersing due to competition,
food, or physical parameters. Tavemy et al. (2002) also documented similar restricted
movement behaviors when tracking juvenile European sturgeon (Acipenser sturio).
Juvenile lake sturgeon in our study could have preferred habitats adjacent to the river
mouth because of potentially higher productivity (e.g. ice free earlier). Furthermore,
dispersal following release in December might not have been very high as colder water

temperatures reduced activity.

l09

We failed to detect 55% of released ﬁsh. The non-detection of the remaining ﬁsh
could be attributed to several factors. First, equipment failure associated with the
ultrasonic transmitters could explain our inability to locate ﬁsh. Secondly, ﬁsh could have
passed the hydroelectric dam on the outlet from Black Lake in the Lower Black River.
Survival rates of ﬁsh attempting this is unknown but if there is a natural propensity to
disperse downstream to more productive feeding areas then this could have occurred. We
did not survey downstream of the dam on the Lower Black River or in adjoining Lake
Huron. Mortality due to predation is likely and has been noted with other hatchery
stocked ﬁsh (Olla et al. 1998). However, transmitters should have been identiﬁed either
within the predator itself or on the substrate after expulsion. Furthermore, if ﬁsh died near
the release location at the mouth of the river the transmitter could become buried in
sediment exiting the Upper Black River rendering the signal undetectable. This was
unlikely because the signal strength should penetrate through moderate amounts of
bottom substrate (V emco, personal communication). Finally, we did not document a
signiﬁcant difference in survival due to hatchery rearing environment though there are
several explanations for the lack of differences. First, streamside reared ﬁsh were
transported and kept at the traditional hatchery for approximately three months. This
movement back to traditional rearing conditions may have diminished any potential
ﬁtness advantage incurred at the streamside hatchery, like exposure to ﬂuctuating
temperature regimes and natural river odors. Secondly, data collected by Crossman
(chapter 3) suggests that the effects of different hatchery rearing environments have a
size/age threshold beyond which the effects of hatchery environment may not be

important. Our results support and extend this ﬁnding.

llO

CONCLUSIONS

In conclusion, our minimum estimate of over-winter survival (40%) is important
for the designs of restoration programs since it has been identiﬁed that survival of
juveniles past the young of the year stage is high and similar to adult stages (Gross et al.
2002). It is generally accepted that mortality rates of young the year sturgeon are high
(Nilo et al. 1997) with the young of the year age class having the strongest effect on
overall population growth (Gross et al. 2002). Hatchery restoration programs have the
ability to target these early age classes in an attempt to increase survival and augment
population abundance. Hatchery rearing environment did not affect survival probabilities
of age-0 lake sturgeon. However, ﬁrrther research is needed identifying the contribution
of hatchery reared sturgeon to population growth. Finally, ﬁrture studies aimed at
identifying overwinter survival of wild or hatchery reared juvenile lake sturgeon should
incorporate ﬁner scale monitoring immediately following release and assessments should

continue for as long as possible thereafter.

lll

Chapter 5
DIRECT AND INDIRECT EFFECTS OF PREDATION ON SURVIVAL AND
HABITAT USE OF JUVENILE LAKE STURGEON

INTRODUCTION

The ecological importance of predator-induced indirect effects is widespread and
represents a signiﬁcant component of a predator's impact on prey communities (Pecor
and Werner 2004; Werner and Anholt 1996), thus inﬂuencing community and ecosystem
level processes (Walsh and Rezrrick, 2008). Predators can inﬂuence the behavior of prey
species which in turn modiﬁes effects incurred ﬁom other predators (Miller and Kerfoot
1987). These indirect/non-lethal effects of predators may be as important as the direct
mortality of prey in understanding the outcome of species interactions (Mittelbach 1986;
Werner and Anholt 1996; Werner and Pecor 2003). Predation risk and effects on foraging
have been shown to inﬂuence habitat choice across a number of species (Lima 1998;
Belk et al. 2001; Bystrdm et al. 2003; Kneib 1987; Winkelman and Aho 1993).
Furthermore, alteration of prey behavior can affect probabilities of detection or capture
(Lima 1998). Estimating behavioral responses related to ontogenetic changes in body size
and morphology during early life history stages is particularly relevant to understanding
the ecology of predator-prey interactions (Werner and Gilliam 1984; Werner and Anholt
1993).

Documenting sources of mortality, as well as behavioral or phenotypic traits
associated with mortality, is complex in natural settings (Houde 1987) due to several

interacting factors of both the prey and predators (e. g. morphology, size, behavior, and

H2

distribution; Schlosser 1987). Laboratory mesocosm studies have provided a successful
arena to study direct and indirect effects on prey as a function of habitat types and
predator species (Bemot and Turner 2001; Stunz and Minello 2001; Werner and Hall
1988; Werner et al. 1983). However, the complexity of experimental conditions likely
plays a large role in the outcomes of predator-prey interactions. This is particularly
relevant for hatchery-raised ﬁsh, which are commonly used as prey in experimental
predation studies. The types or complexity of habitats simulated in mesocosm studies can
affect the outcomes of predator-prey interactions. Increasing the complexity of rearing
environments, including the addition of habitat, has been shown to improve survival
(Maynard et al. 1996) and can inﬂuence morphology and color patterns (Fuji 1993)
compared to ﬁsh grown in tanks devoid of cover. Furthermore, ﬁsh subjected to
chemical and extracted predator odorants prior to release exhibit antipredator responses
(Berejikian et al. 1999; Gazdewich and Chivers 2002), which subsequently increase
survival in the wild. Understanding how rearing environment and age interact to
inﬂuence both predation risk and anti-predator behavior is important for describing
phenomenon in natural systems and for improving restoration efforts of numerically
depressed species.

Estimating the direct effects of different predators on juvenile lake sturgeon
survival and the indirect effects on behavior are critical for the regionally imperiled lake
sturgeon (Acipenser ﬁtlvescens) as they are the center of extensive restoration work
throughout the Great Lakes. Sources of natural mortality, particularly mortality attributed
to predation, for juvenile lake sturgeon are currently unknown, despite recent work

conducted On the vulnerability of sturgeon eggs and larvae (Miller and Beckrnan 1996;

HS

Gadomski and Parsley 2005a), and juvenile sturgeon of other species (Gadomski and
Parsley 2005b; Gadomski and Parsley 2005c). Accordingly, we designed experiments to
test several hypotheses to determine, 1) how habitat use by juvenile lake sturgeon of
different age classes and body sizes varied with and without the presence of predators, 2)
the effects of age and body size on levels of predation by different predator species, 3)
the effects of hatchery rearing environment on predation vulnerability, and 4) rates of
predation on juvenile lake sturgeon in the presence of an alternate prey species. Results
deﬁne important mechanisms associated with predation risk in juvenile lake sturgeon

during the ﬁrst three months of life.

METHODS

Juvenile lake sturgeon were from hatched from paternal half-sib family groups
that were collected and produced from a population in Black Lake, Michigan. Families
were reared separately in two hatchery environments. Half of all ﬁsh used in predation
trials were reared at a streamside hatchery in water ﬁ'om the natal Upper Black River.
The remaining ﬁsh were reared using ground water at a state hatchery in southwestern
Michigan, representing a traditional hatchery environment. The streamside rearing
facility used a ﬂow-through design where water was pumped directly from the river.
Water was mechanically ﬁltered to remove sediments, and returned directly to the river
after passing through the juvenile rearing tanks. Flow rates were kept constant and at a

level that resulted in approximately two complete water turnovers every hour for all

ll4

rearing tanks (1.22m diameter, 0.5m deep). Cleaning, feeding, and lighting protocols
were consistent between the two hatchery environments.

Adult rock bass (Ambloplites rupestris), smallrnouth bass (Micropterus
dolomieue), northern pike (Esox lucius), and rusty crayﬁsh (Orconectes rusticus) were
used as predators in experimental trials. Predators were collected from the Upper Black
River using fyke nets, backpack electro-ﬁshing, and hook and line ﬁshing. Species were
held separately at the streamside hatchery in indoor, circular tanks ﬁberglass tanks
(0.61m diameter 0.7m deep). All indoor tanks were supplied with ﬂow-though river
water. Tanks had overhead lighting that mimicked a natural photoperiod. Predators were

collected _<4 days prior to the start of each experimental trial and were not fed during

captivity. Total length (TL; cm; mean i 1 SE) and gape width (CW) and height (GH) for
ﬁsh predators used in trials were as follows: Rock Bass: TL: 17.96 i 1.92, Gw: 2.75 i
0.24, GH: 2.62 i 0.22. Smallrnouth Bass: TL: 28.04 i 0.98, Gw: 3.4 i 0.07, CH: 3.45 i
012. Northern Pike: TL: 51.0 i 1.29, Gw: 3.9 i 0.09, GH: 4.1 i 0.14. Total carapace
length (CL; mm; mean i 1 SE) and pincher width (PW) for crayﬁsh predators was: CL:

28.75 i 2.11, PW: 13.44 i 1.0.

Experimental design
All experiments were conducted at the streamside hatchery. Experiments were
conducted in a circular (2.44m diameter, 0.6m deep) ﬁberglass tank that was encircled

with a black tarp to reduce human induced behavioral effects. The tank was divided into

ll5

two equally sized sections so 2 replicates could be run simultaneously. Each section
included three substrate types (sand, small gavel, and large rocks) divided into equal
contiguous segnents. Substrate types were removed and randomly redistributed prior to
the beginning of each trial. Trials were conducted using 30 juvenile lake sturgeon for a
duration of 24 hours beginning at 0730. We conducted trials for juvenile lake sturgeon
that were 8-9, 11-13, and 15—16 weeks of age. Age classes were incorporated because we
could only run two trials simultaneously and several weeks were needed to complete all
combinations of juvenile lake sturgeon and the different predators. Fish reared at the
traditional hatchery environment were transported to the streamside rearing facility a few
days prior to the start of the trials. Fish from the different hatchery environments, and
families were marked uniquely with a visible implant elastomer dye (Northwest Marine
Technology, WA, USA). Elastomer was injected on the ventral side of the rostrum where
the colors were most easily distinguished. All ﬁsh were measured for total length before
introduction into the tank and surviving ﬁsh were re-measured at the termination of the

trial.

Experiment I .' Tests of indirect eﬂects of predation risk on substrate use

This experiment was conducted to test two hypotheses: l) The presence of
predators did not signiﬁcantly affect the distribution of juvenile lake sturgeon among
habitats, and 2) There was no affect of rearing environment on the distribution of juvenile
lake sturgeon among habitats with and without the presence of predators. Fish from
traditional and streamside hatcheries were used in separate trials because we were unable

to follow individual ﬁsh during the trial. Within each trial we used juveniles (n=5) ﬁ'om 6

H6

families marked correspondingly. Juvenile lake sturgeon were introduced into tanks
without a predator. Observations of ﬁsh locations, either on one of three substrate types
or in the water column, were taken hourly until nightfall and then again in the morning.
Sturgeon were removed and the predator was introduced. The predator was allowed to
acclimate to the tank conditions for a period of 24 hours and a new set of sturgeon were
introduced into the tank with the predator. The same juvenile lake sturgeon were never
used twice. The same sets of observations were taken including predator location. The
predator was removed immediately following the termination of the trial.

We used a multifactor analysis of variance (ANOVA) to analyze the data. The
response variable was the proportion of juvenile sturgeon observed using each substrate
type. The response variable was transformed using a square root transformation to
achieve normality. Factors in this analysis included the presence or absence of a predator,
the age class, rearing environment, and the trial number. Trial was not a sigriﬁcant factor
within either the presence or absence of a predator so all the trial data were pooled. Data
were reanalyzed using repeated measures ANOVA to incorporate hour of observation as
an additional factor. We conducted separate analyses for the visual water column

predators (ﬁsh) and the benthic predators (crayﬁsh).

Experiment 2: Direct eﬂects of predation on age-speciﬁc survival

This experiment was conducted to test the hypothesis that juvenile lake sturgeon
reared in different hatchery environments and at different ages were not differentially
vulnerable to predation. Experimental design was consistent with experiment 1 with the

exception that we could follow individual survival across hatchery environments and

H7

families. Within each trial we used juveniles (n=5) ﬁom 6 families uniquely marked. All
ﬁsh were measured for total length prior to the trial. Surviving ﬁsh were re-measured at
the termination of each trial.

We used a general linear model to predict survival using several independent
variables including: Age, predator species, rearing environment, and family. Survival was
square root transformed prior to analysis to achieve normality. To determine if there was

size selective predation within each age class we used a paired t-test to test for mean

differences in the total lengths of ﬁsh that were killed and those that survived.

Experiment 3: Alternate species trials

This experiment was conducted to test the hypothesis that predation rates on
juvenile lake sturgeon were not sigriﬁcantly different with and without the presence of
an alternate prey species. An equal mixture of juvenile lake sturgeon (n=15) and emerald
shiners (Notropis atherinoides; n=15) were introduced into a tank containing an
acclimated predator (smallrnouth bass (n=1 per trial) and crayﬁsh (n=4 per trial)). Only
age classes 11-13 and 15-16 were represented in these trials. Family and rearing
environment treatments were not included in alternate species trials. Total length of
sturgeon and minnows was measured for prior to the trial. A t-test was conducted to test
for differences in the size distributions between sturgeon and minnows for each age class.
Substrate use and survival were recorded hourly until nightfall and then again the
following morning. A three-factor ANOVA was used to analyze variation in survival as

a function of age, alternate prey species presence/ absence, and predator species.

l18

RESULTS

Experiment 1: Tests of indirect eﬂects of predation risk on substrate use
Juvenile lake sturgeon strongly prefer sand when provided with a choice of

substrate types (Peake 1999). Without the presence of a predator, sand was signiﬁcantly

preferred (F3,1276 = 4670, P < 0.01). Proportion of substrate used by juvenile lake

sturgeon changed in the presence predators. Juvenile lake sturgeon exposed to three

different ﬁsh predators used sand substrate at a signiﬁcantly higher proportion (F 59,1220

= 172.6, P = 0.005) when compared to gravel, larger rocks, and the water column (Fig

19). Furthermore, juvenile lake sturgeon preference for sand remained signiﬁcant

(F 59,120, P < 0.001) following increased time periods after being introduced into the

tank while preference for the water column signiﬁcantly decreased over the same time
period (F59,1220 = 172.6, P = 0.016; Fig. 19). There were no signiﬁcant differences
between the three predatory ﬁsh species or the main effect of hatchery rearing
environment in our analysis. Juvenile lake sturgeon at 8-9 weeks of age spent
signiﬁcantly less time (F 59,1220 = 172.6, P = 0.006) in the water column compared to
ﬁsh at both 11-13, and 15-16 weeks of age. In general, substrate preference was divided
between sand and the water column with gravel and large rocks combining for <15%
total preference across all trials containing the three ﬁsh predators.

Proportional substrate use of juvenile lake sturgeon was signiﬁcantly different

when crayﬁsh were used as the predator species compared to all three ﬁsh predators

(F 1 1,755 = 373.6, P <0.01). Use of the water column increased signiﬁcantly (F 55,584 =

H9

65.39, P < 0.001; Fig. 19) over time following introduction, while use of sand decreased

(F 55,534 = 65.39, P < 0.001; Fig. 19). In the absence of crayﬁsh the preference for sand

was high and increased signiﬁcantly over time following being introduced (F 55,584 =

65.39, P = 0.004; Fig. 19). There was no effect of hatchery environment or age on ﬁsh

distributions with and without crayﬁsh predators.

Experiment 2: Size-dependant predation rates
Survival rates across all age classes were high when exposed to visual ﬁsh
predators. Survival rates ranged ﬁ'om 97.4 — 100% at 8-9 weeks of age, 98.5 — 100% at

11-13 weeks of age, and 100% for 15-16 weeks of age. Survival was signiﬁcantly less in

trials conducted with 8-9 week old juvenile lake sturgeon (F15,224 = 3.48, P < 0.001).

Rock bass had signiﬁcantly higher rates of predation among the three ﬁsh predator types

(F 2,237 = 12.35, P < 0.001). There were no signiﬁcant effects of predator size, rearing

environment, or time on the survival of juvenile lake sturgeon exposed to the ﬁsh
predators. Family was not signiﬁcantly related to mortality, potentially due to overall low
predation rates. However in three trials where only 1 juvenile lake sturgeon was
consumed all mortalities were from the same family. The probability of this occurring by
random chance is 3.7x10'5. The total length individuals that died were not signiﬁcantly
different ﬁ'om the surviving ﬁsh across all ages (df = 162, t = 0.468, P = 0.64).

Survival rates were signiﬁcantly less for juvenile lake sturgeon exposed to

crayﬁsh (171,286 = 1393.6, P < 0.001) compared to the visual ﬁsh predators. There was a

signiﬁcant effect of time on juvenile sturgeon survival with a higher proportion of

120

 

mortality occurring at hours immediately following release into the tank (F 1 1,36 = 3.94, P

< 0.001). There was no effect of age on survival with rates being 0.66 :i: 0.01 and 0.65 i
0.09 for age 11-13 and 15-16 weeks respectively. There was no effect of hatchery rearing
environment on survival and we could not test for a relationship with crayﬁsh
morphometric data as multiple crayﬁsh were in each trial simultaneously. We did not
document any differences attributed to family or size across different ages or trials. The
total length individuals that were killed in the crayﬁsh trials were not signiﬁcantly

different from the surviving ﬁsh across all ages (df = 100, t = 1.187, P = 0.238).

Experiment 3: Alternate species trials

Age was not a signiﬁcant factor (F1,138 = 1.41, P = 0.238) in the analysis so the
data were pooled. Predators differed signiﬁcantly in prey choice (F 1,140 = 313.94, P <

0.01). We documented a signiﬁcant difference (F 1,92 = 94.52, P < 0.01) in smallmouth

bass predation rates on sturgeon and minnows with larger proportions of minnows being
consumed over the 24hour period (Fig. 20).All juvenile lake sturgeon survived in all

trials and average survival of minnows was 0.69 i 0.21. We found the opposite result

with crayﬁsh. Signiﬁcantly more juvenile lake sturgeon (F1 ’44 = 457.8, P = <0.01) were

consumed during the 24hour period with survival rates being 0.67 i 0.08 and 1.0 for
sturgeon and minnows respectively (Fig. 20). There was no signiﬁcant difference in the
total length of sturgeon compared to minnows at the 11-13 age class (df = 118, t = 0.092,
p = 0.927) however sturgeon were signiﬁcantly larger during the 15-16 age class trials (df
= 118, t = 34.04, p < 0.001).

121

DISCUSSION

We documented a shift in juvenile lake sturgeon habitat use as a result of the
. indirect effects of two predator types. Indirect effects of predation have been documented
to inﬂuence habitat shifts in numerous studies (Lima 1998). The shift in habitat
preference was most pronounced in the presence of crayﬁsh (Fig 19). Studies have
indicated that ﬁsh learn anti-predator behaviors rapidly after witnessing only a few
attacks on conspeciﬁcs (Berejikian et al. 1999). Predation rates for juvenile lake sturgeon
exposed to crayﬁsh were much higher relative to the visual ﬁsh predators. This result
likely explains the dramatic distribution shift in the crayﬁsh trials. Even with the higher
survival rates, juvenile lake sturgeon still chose sand at a signiﬁcantly greater proportion
when in the presence of the ﬁsh predators (Fig 19). These indirect effects may play a
large role in survival in the wild as one predator type might force juvenile lake sturgeon
to choose habitat that has a higher risk of exposure to predation by other predator types.
Rahel and Stein (1988) showed that a prey darter species (Etheostoma nigrum) avoided
predatory bass by hiding under provided cover, however became vulnerable when ﬂushed
ﬁ'om the cover by a second species, a benthic crayﬁsh. In such instances, selection might
be expected to favor alternative avoidance responses such as shifts to other reﬁrge, which
would be an important extension of this work. Previously, it has been hypothesized that
changing habitat to avoid predation risk is unlikely once juvenile sturgeon obtain a size
too large for most predators (Kynard et al. 2000). We observed habitat shifts in each

juvenile lake sturgeon age class evaluated.

l22

 

Documenting the rates of predation by ﬁsh predators is important for
understanding lake sturgeon ecology and for deﬁning barriers to recruitment. Our results
represent the ﬁrst replicated data on predation as it relates to age and avoidance for
juvenile lake sturgeon provided with habitat choices. Efforts have been made to
document rates of predation on embryos (Kynard and Horgan 2002; Miller and Beckman
1996; Nichols et al. 2003) and larvae (Gadomski and Parsley 2005a; Kynard and Horgan
2002) but limited work has been conducted on juveniles using age as an experimental
factor. Gadomski and Parsley (2005b) found that predation is a likely cause of mortality
in age-0 white sturgeon (Acipenser transmontanus) and that predation rates decreased
with age when exposed to four predator species. They documented higher rates of
predation for younger ages however direct comparisons with the results ﬁ'om our study is
difficult due to differences in experimental design. Gadomski and Parsley (2005b) used
smaller, larval, sized individuals initially, used large prey size ranges within each trail
and predators in their study were predominantly benthic feeding ﬁsh. We extended their
study by providing a heterogeneous tank environment and following antipredator
behavior, which are important ecological factors.

There is a lack of knowledge regarding predation mortality for different ages of
juvenile lake sturgeon but it is generally accepted that mortality rates of young the year
sturgeon are high (Nilo et a1. 1997). Rates of predation on juvenile lake sturgeon by the
visual ﬁsh predators were low. In all trials, none of the predators used were gape limited
for the size of juvenile lake sturgeon used. During the 8-9 week age class rates of
predation were signiﬁcantly higher than later ages. Juvenile lake sturgeon responded

behaviorally to the presence of predators by reducing activity levels, a behavior that has

123

been documented in other predator-prey studies (Skelly and Werner 1990; Anholt et al.
2000). Observations taken during trials with the visual ﬁsh predators revealed
unsuccessful attacks on juvenile lake sturgeon at all age classes by all predator types. We
included alternate species trials to help determine if the hatchery environment, or tank,
had any effect on the feeding behavior of the predator species. Signiﬁcantly higher
predation rates on the alternate prey species conﬁrmed the ﬁsh were actively feeding.
Another positive aspect of our experiment was that trials were conducted within a large
tank space including only one predator which was never reused, eliminating learned
behavior. This reduced issues surrounding small-scale experiments by creating conditions
where the rates of interaction between predators and prey were not elevated in
conﬁnement, which is a common problem in predation trials. Furthermore, crayﬁsh
densities were at a level that reduced aggressive interaction among individuals which is
common with higher crayﬁsh densities.

We found signiﬁcantly different results between predator types in the alternate
species trials (Fig. 20). Signiﬁcantly more minnows were consumed by the visual ﬁsh
predator compared to the same initial amount of juvenile lake sturgeon. The opposite
trend was documented in the crayﬁsh trials and was not a surprising result. Crayﬁsh were
very efﬁcient at capturing juvenile lake sturgeon at both 11-13 and 15-16 weeks of age,
despite large sturgeon body sizes. Gadomski and Parsley (20050) found a similar
difference in prey choice between two different predators, one preferring juvenile white
sturgeon over the alternative prey choice and the opposite with the other predator.
Predation rates would have been much higher if more than four crayﬁsh had been used.

Predation by crayﬁsh was immediate (Fig 19) and observations were that each crayﬁsh

I24

 

 

immediately captured a sturgeon. However the entire trial period was needed for a single
crayﬁsh to consume a single juvenile lake sturgeon. The high levels of vigilance,
schooling and avoidance behaviors of the minnows precluded their capture by crayﬁsh.
Results showing high levels of predation by crayﬁsh regardless of juvenile age suggest
that lake sturgeon remain vulnerable to crayﬁsh predation over prolonged periods. The
rusty crayﬁsh used in our trials are an invasive species (Olden et al. 2006) that are in high
abundance in the Upper Black River system. Crayﬁsh are active predators at night and
this could be detrimental for lake sturgeon due to their nocturnal behavior at juvenile ages
(Crossman chapter 3). Release locations for lake sturgeon restoration programs should be
assessed for crayﬁsh densities prior to stocking to reduce post release mortality.

There were no signiﬁcant results attributed to the different hatchery rearing
environments. This is likely due to the fact that the experiments were conducted at the
streamside hatchery. Fish from the traditional hatchery were transported to the streamside
facility a few days prior to the start of the predation trials allowing for acclimation to the
environment. Furthermore, if incoming water contains chemical odorants or cues from
predator species (as suggested by Berejikian et al. 1999), or injured conspeciﬁcs, then the
ﬁsh could have already been slightly conditioned prior to introduction into the trial. This
would be another beneﬁt for streamside rearing of lake sturgeon. Because experience
inﬂuences the behavioral responses to different stimuli and can shape the future behavior
of captive animals, pre-release environmental enrichment may be successful in
facilitating the expression of adaptive behavioral responses following release (Watters
and Meehan 2007). Further research that examines juvenile lake sturgeon mortality

within each of the different hatchery environments would be desirable. The signiﬁcant

I25

 

shift in habitat use with and without the presence of a predator suggests that even after
lengthy periods of captivity, predator avoidance was still observed which is important for

lake sturgeon restoration.

CONCLUSIONS

In conclusion, our results revealed a strong anti—predator response in juvenile lake
sturgeon when exposed to predators in a heterogeneous environment. Even though
overall predation was low for the visual ﬁsh predators, this source of mortality cannot be
discounted as a signiﬁcant barrier to recruitment. Furthermore, since many large ﬁsh
predators (e.g. smallmouth bass) prey upon crayﬁsh, examining rates of juvenile lake
sturgeon predation with two predator types simultaneously would be important, even
though predation rates on rusty crayﬁsh by ﬁsh are lower than native crayﬁsh species
(Kuhlmann et al. 2008). Finally, indirect effects are likely to predispose juvenile lake

sturgeon to higher rates of predation by alternate species.

126

 

 

FINAL CONCLUSIONS

Results presented in my dissertation address the need for further empirical and
experimentally rigorous evaluations of hatchery rearing and stocking programs for little
studied species such as lake sturgeon. It is evident based on currently available scientiﬁc
data that restoration programs could be signiﬁcantly improved if methods by which lake
sturgeon supplemental progeny are acquired, reared, and released are critically evaluated.
In order to collect supplemental progeny for hatchery programs I recommend that
programs maximize genetic diversity by focusing collection efforts on dispersing larvae
because of the collection methods evaluated, they represent the most unbiased sample of
progeny from natural adult reproduction. Collection of dispersing larvae may not be
possible across all systems given logistics and unknown population parameters such as
spawning locations or timing. If programs directly take gametes from adults or collect
naturally produced eggs from the streams substrate focus should be directed at
incorporating aspects of the lake sturgeon reproductive ecology including sampling
across spawning times and locations. Managers should also recognize the variation that
exists in growth and survival as a function of family (i.e., variation in important traits is
heritable). Quantifying additive genetic variation for adaptive phenotypic traits at the
time the progeny enters the hatchery and to the extent possible when they are released
into the wild will be important future predictors of future population viability.
Furthermore, differences between hatchery rearing environments and relative to natural
stream conditions will inﬂuence probabilities of survival and thus levels of genetic

variation in supplemental progeny.

127

 

Streamside rearing may be advantageous to small/young lake sturgeon by
exposing them to an enriched environment prior to release. However, data suggest there
may be an age/size threshold where the effects of rearing environment on movements and
survival are diminished at the time of stocking (chapter 3) and in terms of overwinter
survival (chapter 4). I recommend the use of streamside hatcheries for rearing lake
sturgeon supplemental progeny up to 13 weeks of age. The beneﬁt of rearing ﬁsh to older
ages (>13 weeks) and subsequently larger sizes at a traditional hatchery is outweighed by
the evidence indicating that lake sturgeon progeny survive better and exhibit natural
variation in incubation times and size at hatch when reared at a streamside hatchery.
Furthermore, researchers have suggested that streamside rearing may minimize straying
following stocking by maximizing probabilities of imprinting on the natal river, as has
been observed with other species. Future work that characterizes factors inﬂuencing
differences among hatchery rearing environments for lake sturgeon are needed to develop
system-speciﬁc management prescriptions.

My research emphasizes the importance of tailoring stocking programs to
individual species and systems rather than focusing on management prescriptions solely
based on numbers and body sizes of supplemental ﬁsh released. Results indicate that
supplementation strategies for juvenile lake sturgeon should focus on night releases in
areas of sand and moderate ﬂow. Releases if conducted under these conditions will
ensure that rates of dispersal are high immediately following release and will likely have
an indirect effect on levels of predation by reducing mortality incurred by visual
predators. Results of my predation work (chapter 5) revealed a strong anti-predator

response in‘ juvenile lake sturgeon when exposed to predators in a heterogeneous

I28

environment. Because crayﬁsh predation was documented as a signiﬁcant source of
mortality, releases should be conducted in areas of low crayﬁsh densities. It is critical that
release sites are evaluated on a system by system basis with emphasis placed on
determining areas that maximize suitable juvenile nursery habitat.

Hatchery programs have received increased scrutiny because of the potential
negative impacts of captive reared individuals to natural populations. Improving lake
sturgeon culture methods should be considered a priority for future management and
conservation programs. In long-lived iteroparous organisms such as the lake sturgeon,
longevity is required because per capita recruitment is expected to be low during a single
reproductive period and rates of mortality after early juvenile stages are expected to be
low. Therefore, small improvements in probabilities of individual survival will result in
substantial gains in probabilities of population sustainability. Managers must also
maintain diversity across reproductive time and locations to not diminish adaptations that
have evolved due to reproductive isolation that has likely evolved in response to temporal
and spatial environmental heterogeneity. Methods and results described in my
dissertation provide a framework for evaluating alternative strategies for managers
designing and implementing conservation programs for lake sturgeon. Accordingly,
results from this research may be applied to other systems and can be used to develop

predictions that can be tested for other species.

l29

APPENDIX I

TABLES AND FIGURES

I30

Table 1. Comparisons of sample size (N) and summary measures of genetic diversity
including coancestry (0) and effective population size for individuals sampled at different
developmental stages using each of three methods of gamete/larval collection over each
of three years. Effective population size estimates include calculations using the number

of males and females (Ne), male and female reproductive variance (Nev), and coancestry

(Neg). Individuals were raised in one of two hatchery environments (streamside or
traditional).

 

Hatchery Environment

 

 

 

and Developmental Stage Year N 0 Ne Nev Neg
2005
Streamside Hatchery
Egg Stagea 94,856 0.020 38.1 38.1 24.9
Larval Stagea 15,383 0.029 29.3 29.4 17.5
56 Days 1,500 0.034 29.3 29.4 14.7
90 Days 1,500 0.040 29.3 29.4 12.5
Traditional Hatchery
Egg Stage 79,983 0.026 29.3 29.3 19.5
Larval Stage 10,371 0.033 26.7 26.7 15.0
2006
Streamside Hatchery
Egg Stage 77,731 0.024 31.1 31.7 20.9
Larval Stage 19,129 0.030 26.2 26.2 16.8
40 days 1,000 0.032 26.2 26.2 15.9
140 Days 1,000 0.041 26.2 26.3 12.3
Traditional Hatchery
Egg Stage 47,702 0.037 21.3 21.3 13.6
Larval Stage 9,894 0.043 15.0 15.0 11.6
2007
Streamside Hatchery
Egg Stage 50,170 0.017 38.6 38.6 29.3
Larval Stage 13,243 0.019 38.6 38.6 26.3

 

a: Egg and larval stage refer to times of fertilization and hatch, respectively.

131

Table 2. Comparisons of sample size (N) and summary measures of genetic diversity
including coancestry (0) and effective population size for offspring collected as naturally

produced eggs in each of two years. Effective population size estimates include

calculations using the number of males and females (Ne), male and female reproductive

variance (Nev), and coancestry (Neg). Individuals were raised in one of two hatchery

environments (streamside or traditional).

 

 

 

 

Hatchery Environment Year N 0 Ne Nev Neg
2005 1,100

Streamside Hatchery 700 — —— — —

Traditional Hatchery 400 — — — —
2006 1,138 0.008 51.3 125.1 63.4

Streamside Hatchery 772 0.004 42.8 93.7 126.0

Traditional Hatchery 366 0.014 13.7 33.2 36.0
2007

Streamside Hatchery 275 0.007 14.9 59.0 72.0

 

132

Table 3. Comparisons of sample size (N) and summary measures of genetic diversity

including coancestry (0) and effective population size for offspring collecting as

dispersing larvae in each of three years. Effective population size estimates include
calculations using the number of males and females (Ne), male and female reproductive

variance (Nev), and coancestry (Neg). Individuals were raised in one of two hatchery

environments (streamside or traditional).

 

 

 

 

Hatchery Environment Year N 0 Ne Nev Neg
2005 7,800

Streamside Hatchery 3,000 — — — —

Traditional Hatchery 4,800 — — — —
2006 5,500 0.004 122.1 214.10 141.3

Streamside Hatchery 3,500 0.004 70.0 190.58 116.7

Traditional Hatchery 2,000 0.004 87.9 188.22 122.6
2007

Streamside Hatchery 1,400 0.005 80.1 181.3 105.9

 

133

Table 4. Reproductive success of males and females based on known pedigree
information for offspring produced from directed gamete takes and reared in two
hatchery environments (Streamside and Traditional) during each of 3 years. Variables

represent the mean i ISD.

 

 

 

Hatchery Variable 2005 2006 2007
Eggs/Female 6323.7 i 5056.2 6477.6 d: 4857.2 3345.0 i 2030.4
Larvae/Family 669.2 d: 666.2 863.4 :2 820.1 343.8 i 214.8
Streamside NmiNfa 26:15 22:12 27:15
Mates/Female 1.7 i 0.5 2.1 i 0.7 1.8 :t 0.4
Mates/Male 1 1.1 i 0.3 1
Eggs/Female 7271.1 i 5415.2 5962.8 :1: 4812.4 —
Larvae/Family 518.6 :1: 475.6 549.7 i 420.1 —
Traditional Nmsz“ 22:11 16:8 —
Mates/Female 2 2 —
Mates/Male 1 1 —

 

a: Ratio of the number of males (Nm) to females (Nf).

134

Table 5. Reproductive success of males and females based on genetically determined
pedigree information for offspring collected using two collection methods and reared in
two hatchery environments (Streamside and Traditional) during each of 3 years.
Variables represent the mean 2!: ISO

 

 

 

 

 

Collection Method and
Hatchery Environment Variable 2005 2006 2007
Naturally Produced Eggs
Offspring/Female — 1.9 i 1.1 1.1 d: 0.4
Offspring/Male — 1.4 :1: 0.7 1.1 :t 0.4
Streamside waf“ — 23:20 8:7
Mates/Female — 1.4 :t 0.8 1.1 i 0.4
Mates/Male — 1.2 i 0.5 1.1 :t 0.4
Offspring/Female — 1.50 :t 0.8 —
Offspring/Male — 1.13 i 0.4 —
Traditional Nm ija _ 8:6 _
Mates/Female — 1.50 i 0.8 —
Mates/Male — 1.13 i 0.4 —
. Dispersing Larvae
Offspring/Female — 1.67 :1: 1.1 2.2 i 1.3
Offspring/Male — 1.19 :1: 0.4 1.5 i 0.8
Streamside waf“ — 42:30 51:33
Mates/Female — 1.63 :1: 1.0 2.2 i 1.3
Mates/Male — 1.17 :1: 0.4 1.5 i 0.8
Offspring/Female —— 2.1 :t 1.4 —
Offspring/Male — 1.3 :1: 0.5 —
Traditional Nmrvf" —— 59:35 —
Mates/Female --— 2.0 d: 1.2 ——-
Mates/Male — 1.3 :t 0.4 —

 

a: Ratio of the number of males (Nm) to females (M).

135

Table 6. Estimates of mean relatedness (rxy) of parent pairs for offspring collected using
three different gamete/larval collection methods and the adult spawning population (SP)
during 2005-2007. Collection methods included direct gamete takes (DGT), naturally
produced eggs (NPE) and dispersing larvae (DL). Mean rxy (i 18B) is presented along
the diagonal. Uncalculated values are represented with —.

 

 

Collection Method 2005 2006 2007
DGT 0.014 :t 0.033 -0.084 :1: 0.06 0.034 :1: 0.04
NPE — 0.013 i 0.04 -0.013 :t 0.08
DL — 0.021 i 0.02 0.007 d: 0.03
SP 0.001 0.004 -0.004

 

136

Table 7. Physical conditions of the Upper Black River during stocking events of three
different age classes of juvenile lake sturgeon. Physical variables including water depth
(m) and velocity (111/ sec) are included for each downstream assessment site during each

release age.

 

 

 

 

Age Site Water Depth Water Velocity
1 0.62 i 0.05 0.79 i 0.04
8 2 0.52 d: 0.06 0.73 i 0.03
3 0.61 i 0.06 0.62 :t 0.04
4 0.74 :1: 0.01 0.32 d: 0.02
1 0.62 :1: 0.13 0.74 :t 0.03
13 2 0.68 :t 0.05 0.71 d: 0.10
3 0.77 i 0.09 0.59 :t 0.03
4 0.99 i 0.07 0.40 :t 0.01
1 0.65 :t 0.06 0.48 :1: 0.16
17 2 0.66 i 0.03 0.45 i 0.18
3 0.68 i 0.03 0.46 i 0.18
4 0.93 i 0.05 0.37 :t 0.24

 

137

 

Table 8. Numbers released and total length (cm, mean :1: ISE) corresponding to age,
rearing environment, and gamete/ juvenile collection method for three experimental
releases of juvenile lake sturgeon into the Upper Black River, Michigan. Letters (A, B, C)
correspond to statistically greater proportions of ﬁsh recaptured between the different
collection methods within each release age.

 

Age Hatchery Collection Total Proportion Total
(weeks) Environment Method Released Recaptured Length

 

 

 

 

8 Streamside Artiﬁcial*A 1257 0.09 7.87 r 1.06
DriﬁlB 485 0.05 7.51 r 0.82
NaturalIB 220 0.07 7.68 a 1.11
Traditional ArtiﬁcialB 1380 0.03 7.38 r 0.99
DriftB 725 0.04 7.42 :1: 0.73
NaturalC 240 0.02 7.77 :h 1.22
Total 4307 0.05
13 Streamside ArtiﬁcialA 999 0.16 10.84 :1: 0.08
DriftB 494 0.06 11.70 :1: 0.20
NaturalA 181 0.13 11.981017
Traditional ArtiﬁcialA 623 0.10 11.20 a 0.20
DriftB 692 0.04 11.70 a 0.14
NaturalA 237 0.10 12.90 a: 0.21
Total 3226 0.10
17 Streamside ArtiﬁcialA 1000 0.16 13.85 a 0.15
DriftB 1088 0.08 14.81 a 0.20
NaturalC 399 0.26 14.49 a 0.15
Traditional ArtiﬁcialA 1000 0.17 16.51 a: 0.19
DriftB 493 0.05 17.19 i 0.30
NaturalC 208 0.27 15.80 3 0.34
Total 4188 0.15
Grand Total 11721 0.10

*Artiﬁcial: Paternal half-sib families, each created with one female and two males
(11 = 26 and 25 for two years).

TDriﬁ: Larvae captured dispersing downstream from spawning areas.

INatural: Naturally fertilized and deposited eggs that were collected from the
stream.

138

Table 9. The number of observations (Obs), average inter-observational distance traveled

(Mean i 1 SD), and average inter-observational distance traveled (Mean i 1 SD) from
the release site for juvenile lake sturgeon surgically implanted with ultrasonic
transmitters and released into Black Lake, Michigan. Other characteristics of the
experimental ﬁsh represented include: individual ﬁsh ID number, hatchery rearing

location (Loc: S: streamside; T: traditional), total length (TL) and weight (W).

 

 

 

Distance Distance From

ID Lo TL (cm) W (g) Obs Traveled (Km) Release Site (Km)
36 S 28.58 102.10 4 0.46 i 0.03 0.12 :1: 0.01
42 S 29.21 93.06 5 0.11 :1: 0.01 0.03

44 S 29.21 92.90 6 0.08 i 0.01 0.04 i 0.01
49 S 31.75 101.80 3 0.14 :1: 0.10 0.08 i 0.02
50 S 33.66 118.60 4 0.21 i 0.05 0.03 :t 0.01
55 S 28.80 90.00 2 0.16 :1: 0.02 0.06 :t 0.03
56 S 31.75 112.60 7 010220.10 O.l3:1:0.01
Mean :1: SE 30.42 :1: 0.74 101.58 i 4.06 4.43 0.18 i 0.05 0.07 i 0.02
22 T 31.43 110.50 8 0.04 :1: 0.02 0.1 i 0.02
24 T 30.32 101.60 4 0.55 i 0.15 0.05 i 0.01
27 T 32.39 116.90 3 0.14 :1: 0.04 0.06 :t 0.01
28 T 30.80 99.10 5 0.09 d: 0.03 0.09 i 0.01
29 T 29.21 94.60 4 0.25 :t 0.11 0.11 i 0.02
30 T 30.48 97.70 2 0.17 d: 0.08 0.07 i 0.04
31 T 30.62 103.00 5 0.10 :1: 0.05 0.07 i 0.02
51 T 31.75 105.00 6 0.32 :1: 0.10 0.09 i 0.03
52 T 33.02 126.60 3 0.12 :t 0.04 0.11 d: 0.01
Mean :t SE 31.11 A: 0.39 106.11 1 3.41 4.44 0.20 i 0.05 0.08 :1: 0.01

 

139

Figure l. The Black Lake study site in Michigan, showing the adult spawning grounds,
the collection sites for naturally produced eggs (*), and the collection site for dispersing
larvae.

 

 

 

 

 

          

Dispersing Larvae
/ Collection Site

Adult Spawning
Grounds

0 0.5Km

_=

 

 

 

 

 

 

 

 

 

 

140

 

Figure 2. Total number of adult lake sturgeon captured by day on the spawning grounds
of the Upper Black River ﬁom 2005 to 2007.

 

 

 

 

 

 

50 J 2005
30 -
10 -
g .l - 2
a 50 -1 006
g .
E 10 -
° __ 2007
I- 50 -
30 -
10 - I III I I'll '-

20-Apr 27-Apr 4—May 1 1-May 1 8-May 25-May 1 -Jun
Date

141

Figure 3. Frequency distributions of inter-individual relatedness values (rxy) values for
adults producing offspring from A) each of three different gamete/larval collection
methods (Naturally Produced Eggs (NPE), Directed Gamete Takes (DGT), Dispersing
Larvae (DL)) and B) the total adult spawning p0pulation (SP) over three years (2005-

2007)

 

Frequency

 

 

-0.5 -O.4 -O.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

3000‘ .2005

132006
6000. 32007
I 1
2000- I I I |
I i 1 I I

0 ‘ _
-O.5 -O.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

 

 

 

 

Frequency

 

142

Figure 4. Estimates of inter-individual relatedness (rxy) values (Mean :1: ISE) between
adult female lake sturgeon concurrently breeding within distinct spawning zones in the
Upper Black River during each of three years (2005-2007). Numbers refer to the total

number of pairwise comparisons used to calculate the mean

 

 

 

 

 

 

 

 

 

0.2 r I 2005
21 . am
new
6
0.15 - "
g 0.1
a:
c
l
0
z
005 -
0
0.05 J
1 2 3 4 5 6

Spawning Zone

143

Figure 5. Estimates of inter-individual relatedness (rxy) (mean i lSE) between adult
female lake sturgeon concurrently breeding within distinct spawning runs in the Upper
Black River during each of three years (2005-2007). Numbers refer to the total number of
pairwise comparisons used to calculate the mean

 

 

 

 

 

 

 

 

 

 

 

0.15 ~ I m
D 2008
19 l 2007
10
36
L__1
-0.05 -
-0.1 4
1 2 3 4

Spawning Run

144

Figure 6. Overall egg survival (mean 5: lSE) to hatch for females incubated in two
different hatchery environments across two different years. Different sets of females were
used between years

 

 

 

 

 

 

 

 

 

 

Isuearrslde

0.8'DTradtloral

0.6-

0.4-
3
H 11 J
3 0. [L-
(D
30.81
I.“

0.61

0.4:

02‘ Ft. C. r:- 1“

11 22 33 44 55 66 77 88 99 1010

Female

145

Figure 7. The effect of temperature on egg incubation time for eggs reared at a streamside

hatchery in 2005 and 2006

 

Incwaﬁon The (Days)

 

v =0.1582x2 - 6.2949x +66.702

 

14‘ O 2
.. O R =0.8935

12:
O.

10- o

8.1

6: o

4-

2-

O I I 1 I I I I I

12 13 14 15 16 17 18 19 20
MearDﬂyTamaatrePC)

146

Figure 8. Daily proportions of dead eggs infected by microbes (mean 3: lSE) for 2
different lots of eggs (Early (E) and Late (L)) reared in streamside and traditional
hatcheries

 

0.9 -
0.8 i
0.7 '1
0.6 -

- - art:

1

0.4
0.3
0.2
0.1

l

-D-Traditional
+Streamside

Proportion of Dead Eggs with
Microbial Infection
6
U!

1

 

 

 

o q I I I I I I I I 1 m I T I I I I I I

1 2 3 4 5 6 7 8 9 1011121314151617181920
Incubation Day

147

Figure 9. Daily survival rates for lake sturgeon eggs incubating for a different number of

days until hatch

P
g
l

93‘” Survival
'm

9
UI

 

(L4

 

3 4 5 6 7 ,8 9
IncubationDay

148

I I 1 I

10 11 12 13 14

Figure 10. Total numbers of larvae hatched (mean :1: ISD) from females in two different

hatchery environments in A) 2005, and B) 2006. Different sets of females were used
between years

: lsueansicb

    

§

Mean Hatch
3

$3359

 

 

 

11 22 33 44 55 66 77 88 99 1010
Female

149

Figure 11. The effects of four treatments and time since hatch (days) on the change in
larval lake sturgeon total length (A) and yolk-sac utilization (B). Treatments correspond
to tanks that include refuge (cover), refuge and resources (cover-food), totally open tank
(open), and an open tank with resources (open-food)

 

24 - -a-Cover A

+Cover-Food
-C1—Open
+Open-Food

   
  

Total Length (mm)
3 3 13

d
a)
L

a
h

 

 

12 I I 1 I I

Yolk-Sac Area (m m3)
h

 

 

 

1 3 S 7 9 11 13 15
Days Since Hatch

150

Figure 12. Differences in total length (A) and yolk-sac utilization (B) over time among
larval lake sturgeon from six females reared at a streamside hatchery in 2006

 

Total Length (mm)

Yolk-Sac Area (mm3)
N w #5 Ol 09 N a) to

24

22

20

18

16

14

12

1O

.3
0

DA

 

 

 

 

 

 

Day Since Hatch

151

Figure 13. Daily survival of juvenile lake sturgeon collected using three gamete/larval
collection methods (dispersing larvae (DL), naturally produced eggs (NPE), and direct
gamete takes (DGT)) reared in two different hatcheries (streamside and traditional) in

2005

 

Mean Daily Survival

Mean Daily Survival

 

 

 

 

28 34 40 46

 

Streamside

I I I I T I I T

28 34 40 46 52 58 64 70 76 82 88 94

Age (Days)

 

Traditional

I I I I I I I I

52 58 64 70 76 82 88 94
Age (Days)

152

Figure 14. Growth of juvenile lake sturgeon reared in three different levels of habitat
complexity. Habitat complexity consisted of completely open, ﬁlled half way with large
rocks to simulate cover, and ﬁnally tanks that were completely covered

 

.1
O)
l

--I--Cover

.1
.5
1

Total Length (cm)

 

 

 

25 32 39 46 53 60
Age (Days)

153

Figure 15. The Black Lake study site in Michigan, showing the release location and
sampling sites (1—4) for release assessments in the Upper Black River. Site 4 also
corresponds to the larval sampling site

 

 

  
 

Upper Black River

—River Km 5

—River Km 10

\
Site 3
Site 2\ ’
Release Site 0 0.5 1
—= Km

 

 

 

154

Figure 16. The proportion of juvenile lake sturgeon recaptured at downstream assessment
sites (Sites 1-4; Fig 1) that were reared in different hatchery environments. Ages 8, 13,
and 17 weeks are represented within a downstream interval respectively reading left to
right. Asterisks represent signiﬁcantly greater proportions of recaptured ﬁsh that were
reared at the streamside hatchery versus the traditional hatchery. Letters (A, B, C)
correspond to statistically greater proportions of ﬁsh recaptured at the different release
ages

0.15 - I Streamside Hatchery
0,12 _ A 1:1 Traditional Hatchery

  

0.09 -

 

 

 

0.06 -
0.03

Proportion Recaptured

0.5 1 2 4
Distance Downstream (Km)

155

Figure 17. The proportion of juvenile lake sturgeon captured at consecutive hours
following release into the Upper Black River at 8, 13, and 17 weeks of age

 

Proportion Recaptured

 

l3

 

17

 

0.5 1.5 2.5 3.5 4.5 5.5 6.5 8.0 10.5 12.5 15.5
Hours Following Release

156

Figure 18. Total length (means i 1 SE) of recaptured juvenile lake sturgeon over time
following release at 8, 13, and 17 weeks of age into the Upper Black River

 

 

 

 

 

181 I
8. . 1
A 6- IIIII‘
54 8
:14- -
512- II I
5133111111‘ B
- 6
£13" 1 11 l
14-I IIIIIII‘l l l I
13' I’ n
0 2 4 6 8 10 12 14 16

Hours Following Release

157

Figure 19. The effect of two predator classes (Fish, Crayﬁsh) on the proportion of
juvenile lake sturgeon observed on sandy substrate or that were swimming in the water
column. Gravel and large rock substrates were also available choices in this experiment
but are not included due to insigniﬁcant choice for these two habitat types. Predators
included rock bass (Ambloplites rupestris), smallmouth bass (Micropterus dolomieue),
northern pike (Esox lucius) and the rusty crayﬁsh (Orconectes rusticus). Absent and
present refer to trials conducted with and without predators

14

0.8 -

o 6 —a-—Sand: Absent
. . - -u- - - Water. Absent

0.4 _ +Sand: Present

02- ‘

 

- - 1- -- Waten Present

.0

 

0.8 -
0.6 -
0.4 -
0.2 -

Proportion
O

 

 

 

0.5‘1 1.512 3 5 7 91123
Hours Following Release

* First observation of predation by crayﬁsh predators.
1' First observation of predation by ﬁsh predators.

158

 

Figure 20. Survival (Mean :1: 1 SE) of juvenile lake sturgeon and an alternate prey species

exposed to two predator species at two ages (smallmouth bass, Micropterus dolomieue,
and crayﬁsh, Orconectes rusticus)

0.8 -
0.6 i
0.4 J
0.2 '1

Survival

 

1:1 Sturgeon
I Alternate

 

 

 

 

 

 

 

   

 

 

Smallmouth Crayﬁsh Smallmouth Crayﬁsh
Bass Bass

1 1-13 Weeks 15-16 Weeks

159

 

LITERATURE CITED

Allendorf, F. W. 1993. Delay of adaptation to captive breeding by equalizing family size.
Conservation Biology 7:416-419.

Allendorf, F. W. and S. R. Phelps. 1980. Loss of genetic variation in a hatchery stock of
cutthroat trout. Transactions of the American Fisheries Society 109:537—543.

Aloisi, D., R. R. Gordon Jr., N. J. Starzl, J. L. Walker, and T. R. Brady. 2006. Genoa
national ﬁsh hatchery lake sturgeon culture standard operating procedures. US.
Fish and Wildlife Service Region 3 Fisheries Data Series. FDS 2006-003.

Alverson, D. L. 2002. Factors inﬂuencing the scope and quality of science and
management decisions The good, the bad and the ugly. Fish and Fisheries 3:3-19.

Amtstaetter, F. and Willox, CC. 2004. Survival and growth of lake Whiteﬁsh from two
stocking strategies in Lake Simcoe, Ontario. N. Am. J. Fish. Manage. 24: 1214—
1 220.

Anderson, E. A. 1984. Artiﬁcial propagation of lake sturgeon Acipenserﬁtlvescens
(Raﬁnesque), Under Hatchery Conditions in Michigan. Michigan Department of
Natural Resources Fisheries Research Report No. 1898.

Anderson, E. R. 1984. Formulated propagation of lake sturgeon Acipenserfulvescens
(Raﬁnesque), under hatchery conditions in Michigan. Michigan Department of
Natural Resources, Fisheries Division, Fisheries Report 1898, Ann Arbor.

Anholt, B. R., E. E. Werner, and D. K. Skelly. 2000. Effect of food and predators on the
activity of four larval ranid ﬁogs. Ecology 81(12):3509-3521.

Araki, H., Cooper, B., Blouin, M.S., 2007a. Genetic effects of captive breeding cause a
rapid, cumulative ﬁtness decline in the wild. Science, 318, 100-103.

Araki, H., W. R. Ardren, E. Olsen, B. Cooper, and M. S. Blouin. 2007b. Reproductive
success of captive-bred steelhead trout in the wild: Evaluation of three hatchery
programs in the Hood River. Conservation Biology 21(1):]81-190.

Araki, H., B. A. Berejikian, M. J. Ford, and M. S. Blouin. 2008. Fitness of hatchery-
reared salmonids in the wild. Evolutionary Applications 1(2):342-355.

Auer, NA. 1996. Importance of habitat and migration to sturgeons with emphasis on

lake sturgeon. Canadian Journal of Fisheries and Aquatic Sciences 53(1):]52—
160.

160

 

Auer, N. A. and E. A. Baker. 2002. Duration and drift of larval lake sturgeon in the
Sturgeon River, Michigan. Journal of Applied. Ichthyology 18:557—564.

Avise, J. C. 2004. Molecular Markers, Natural History, and Evolution, 2nd ed. Sinauer
Associates, Sunderland, Massachusetts.

Avise, J. C., A. G. Jones, D. Walker, and J. A. DeWoody. 2002. Genetic mating systems
and reproductive natural hiStories of ﬁshes: Lessons for ecology and evolution.
Annual Review of Genetics 36:19-45.

Bailey, K. M. and E. D. Houde. 1989. Predation on eggs and larvae of marine ﬁshes and
the recruitment problem. Advances in Marine Biology 25: 1—83.

Baker, E. A. and D. J. Borgeson. 1999. Lake Sturgeon Abundance and Harvest in Black
Lake, Michigan, 1975—1999. North American Journal of Fisheries Management
19(4):]080-1088.

Ballou, J ., and R. C. Lacy. 1995. Identifying genetically important individuals for
management of genetic diversity in pedigreed populations. In J. Ballou, M.
Gilpin, and L. Foose, eds. Population Management for Survival and Recovery:
Analytical Methods and Strategies for small Population Conservation, pp. 76—
111. Columbia University Press, New York.

Bartron, M. L. 2003. Assessment of historical and contemporary genetic diversity of
steelhead (Oncorhynchus mykiss) in the Lake Michigan basin. Doctoral Thesis,
Michigan State University, East Lansing.

Bearnesderfer, R. C. P. and R. A. Farr. 1997. Alternatives for the protection and
restoration of sturgeons and their habitat. Environmental Biology of Fishes
48:407—417.

Belk, M. C., M. J. Whitney, and G. B. Schaalje. 2001. Complex effects of predators:
determining vulnerability of the endangered June sucker to an introduced
predator. Animal Conservation 4(3):251—256.

Benson, A. C., T. M. Sutton, R. F. Elliott, and T. G. Meronek. 2005a: Evaluation of
sampling techniques for age-0 lake sturgeon in a Lake Michigan tributary. North
American Journal of Fisheries Management 25(4):1378-1385.

Benson, A. C., T. M. Sutton, R. F. Elliott, and T. G. Meronek. 2005b: Seasonal
movement patterns and habitat preferences of age-0 lake sturgeon in the Lower
Peshtigo River, Wisconsin. Transactions of the American Fisheries Society
134(5):1400-1409.

161

 

Benson, R. L., S. Turo, and B. W. McCovey Jr. 2007. Migration and movement patterns
of green sturgeon (Acipenser medirostris ) in the Klamath and Trinity rivers,
California, USA. Environmental Biology of Fishes 79(3-4):269-279.

Berejikian, B. A., E. P. Tezak, T. A. Flagg, A. L. LaRae, E. Kummerow, and C. V. W.
Mahnken. 2000. Social dominance, growth, and habitat use of age-0 steelhead
Oncorhynchus mykiss grown in enriched and conventional hatchery
environments. Canadian J0urnal of Fisheries and Aquatic Sciences 57:628-636.

Berejikian, B. A., R. J. F. Smith, E. P. Tezak, S. L. Schroder, and C. M. Knudsen. 1999.
Chemical alarm signals and complex hatchery rearing habitats affect antipredator
behavior and survival of chinook salmon (Oncorhynchus tshawytscha) juveniles.
Canadian Journal of Fisheries and Aquatic Sciences 56(5):830—838.

Bemot, R. J. and A. M. Turner. 2001. Predator identity and trait-mediated indirect effects
in a littoral food web. Oecologia 129(1):]39-146.

Biro, P. A., A. E. Morton, J. R. Post, and E. A. Parkinson. 2004. Over-winter lipid
depletion and mortality of age-0 rainbow trout (Oncorhynchus mykiss). Canadian
Journal of Fisheries and Aquatic Sciences 61 :1513—1519.

Biro, P. A., J. R. Post, and E. A. Parkinson. 2003. From individuals to populations: prey
ﬁsh risk-taking mediates mortality in whole-system experiments. Ecology
84(9):24l 9-243 1 .

Birstein, V. J. 1993. Sturgeons and Paddleﬁshes: Threatened ﬁshes in need of
conservation. Conservation Biology 7(4):773—787.

Blaxter, J. H. S. and G. Hempel. 1963. The inﬂuence of egg-size on herring larvae
(Clupea harengus L.). Journal du Conseil, Conseil International pour
l’Exploration de la Mer 28:211—240.

Blaxter, J. H. S. and G. Hempel. 1966. Utilization of yolk by herring larvae. Journal of
the Marine Biological Association U.K. 46:219-234.

Borgstroem, R., and O. Skaala. 1993. Size-dependent catchability of brown trout and
Atlantic salmon parr by electroﬁshing in a low conductivity stream. Nordic
J oumal of Freshwater Research 68:14-20.

Bott, J. K. 2006. Genetic analyses of dispersal, harvest mortality, and recruitment for
remnant populations of Lake Sturgeon, Acipenser fulvescens, in open-water and
riverine habitats of Lake Michigan. Masters Thesis, Michigan State University,
East Lansing.

162

 

 

Bouchard, H. J. and D. B. Aloisi. 2002. Investigations in Concurrent Disinfection and
De-adhesion of Lake Sturgeon Eggs. North American Journal of Aquaculture
64(3):212-216.

Bradford, M. J. and G. C. Taylor. 1997. Individual variation in dispersal behaviour of
newly emerged chinook salmon (Oncorhynchus tshawytscha) ﬁ'om the Upper
Fraser River, British Columbia. Canadian Journal of Fisheries and Aquatic
Sciences 54:1585—1592. ‘

Brokordt, K. B., M. Fernandez, and C. F. Gaymer. 2006. Domestication reduces the
capacity to escape ﬁom predators. Journal of Experimental Marine Biology and
Ecology 329:11—19.

Bromage, N., M. Bruce, N. Basavaraja, K. Rana, R. Shields, C. Young, J. Dye, P. Smith,
M. Gillespie and J. Gamble. 1994. Egg quality determinants in ﬁnﬁsh the role of
overripening with special reference to the timing of stripping in the Atlantic
halibut Hippoglossus hippoglossus. Journal of the World Aquaculture Society
25(1):]3-21.

Brooks, 8., C. R. Tyler and J. P. Sumpter. 1997. Egg quality in ﬁsh: what makes a good
egg? Reviews in Fish Biology and Fisheries 7(4):387-416.

Brown, C. and K. Laland. 2001. Social learning and life skills training for hatchery-
reared ﬁsh. Journal of Fish Biology 59:471-493.

Brown, C. and R. L. Day. 2002. The ﬁrture of stock enhancements: lessons for hatchery
practice ﬁ'om conservation biology. Fish and Fisheries 3:79-94.

Bruch, R. M. and F. P. Binkowski. 2002. Spawning behavior of lake sturgeon (Acipenser
ﬁzlvescens). Journal of Applied Ichthyology 18(4-6):570-579.

Busack, C. and C. M. Knudsen. 2007. Using factorial mating designs to increase the
effective number of breeders in ﬁsh hatcheries. Aquaculture 273:24—32

Bystriim, P., L. Persson, E. Wahlstriim, and E. Westman. 2003. Size- and density-
dependent habitat use in predators: Consequences for habitat shifts in young ﬁsh.
The Journal ofAnimal Ecology 72 (1):156-168.

Campton, D. E. 2004. Sperm Competition in salmon hatcheries: The need to
institutionalize genetically benign spawning protocols. Transactions of the
American Fisheries Society 133(5):1277—1289.

Caswell, N. M. D. L. Peterson, B. A. Manny, and G. W. Kennedy. 2004. Spawning by

lake sturgeon (Acipenserﬁtlvescens) in the Detroit River. Journal of Applied
Ichthyology 20(1):1—6.

163

 

 

Ceskleba, D.G., S. AveLallemant, and T. F. Thuemler. 1985. Artiﬁcial spawning and
rearing of lake sturgeon, Acipenserﬁtlvescens, in Wild Rose State Fish Hatchery,
Wisconsin, 1982-1983. Environmental Biology of Fishes 14(1):79-85.

Chesser, R. K. 1991. Inﬂuence of gene ﬂow and breeding tactics on gene diversity within
populations. Genetics 129:573-583

Chesser, R. K., O. E. Rhodes-Jr., D. W. Sugg and A. Schnabel. 1993. Effective Sizes for
Subdivided Populations. Genetics 1 3 5: 1221-1232.

Chiasson, W. B., D. L. G. Noakes, and F. W. H. Beamish. 1997. Habitat, benthic prey,
and distribution of juvenile lake sturgeon (Acipenserﬁtlvescens) in northern
Ontario rivers. Canadian Journal of Fisheries and Aquatic Sciences 54:2866—
2871.

Conte, F. S., S. I. Doroshov, P. B. Lutes, and E. M. Strange. 1988. Hatchery manual for
the white sturgeon Acipenser transmontanus Richardson with application to other
North American Acipenseridae. Division of Agriculture and Natural Resources,
University of California, Oakland, Calif.

Copeland, J. R. and R. L. Noble. 1994. Movements by young-of—year and yearling
largemouth bass and their implications for supplemental stocking. North
American Journal of Fisheries Management. 14(1):]19-124.

Cowx, I. G. 1994. Stocking strategies. Fisheries Management and Ecology. 1(1):15-30.

Cowx, I. G. 1999. An appraisal of stocking strategies in the light of developing country
constraints. Fisheries Management and Ecology 6(1):21—34.

Crisp, D. T. 1991. Stream channel experiments on downstream movement of recently
emerged trout, Salmo trutta L., and salmon, S. salar L. Effects of developmental
stage and day and night upon dispersal. Journal of Fish Biology 39(3):371—381.

Cui, Y., S. S. O. Hung, D. F. Deng, and Y. Yang. 1997. Growth performance of juvenile
white sturgeon as affected by feeding regimen. Progressive Fish Culturist 59:31—-
35.

Cunjak, R. A., 1988: Physiological consequences of overwintering in streams: The cost
of acclimatization? Canadian Journal of Fisheries and Aquatic Sciences
45(3):443-452.

Daufresne, M., H. Capra, and P. Gaudin. 2005. Downstream displacement of post-

ernergent brown trout: effects of development stage and water velocity. Journal of
Fish Biology 67:599-614.

164

 

DeHaan, P. W. 2003. Demographic and life history characteristics of remnant lake
sturgeon populations in the upper Great Lakes basin: inferences based on genetic
analysis. / by Patrick William DeHaan. Masters thesis, Michigan State University,
East Lansing Michigan.

DeHaan, P. W., S. V. Libants, R. F. Elliott, and K. T. Scribner. 2006. Genetic p0pulation
structure of remnant lake sturgeon populations in the Upper Great Lakes Basin.
Transactions of the AmeriCan Fisheries Society 135:1478—1492.

Dettlaff, T. A. and B. F. Goncharov. 2002 Contribution of developmental biology to
artiﬁcial propagation of sturgeon in Russia. Journal of Applied Ichthyology
18:266-270.

DeWoody , J. A., and J. C. Avise. 2001. Genetic perspectives on the natural history of
ﬁsh mating systems. The Journal of Heredity 92:167-172.

Dilaaro, M. N., W. F. Krise, and K. Fynn-Aikins. 1998. Growth and survival of lake
sturgeon larvae fed formulated diets. The Progressive Fish-Culturist 60:293—296.

Doroshov, S. I., W. H. Clark-Jr, P. B. Lutes, R. L. Swallow, K. E. Beer, A. B. McGuire,
and M. D. Cochran. 1983. Artiﬁcial propagation of the white sturgeon Acipenser
transmontanus Richardson. Aquaculture 32:93-104.

Duchesne, P., T. Castric, and L. Bematchez, L., 2005. PASOS (parental allocation of
singles in open systems): a computer program for individual parental allocation
with missing parents. Molecular Ecology Notes 5:701 -704.

Ellison, D. G. and W. G. Franzin. 1992. Overview of the symposium on walleye stocks
and stocking. North American Journal of Fisheries Management 12(2):271-275.

Elrod, J. H., D. E. Ostergaard, and C. P. Schneider. 1988. Comparisons of hatchery-
reared lake trout stocked as fall ﬁngerlings and as spring yearlings in Lake
Ontario. North American Journal of Fisheries Management 8(4):455-462.

Fajfer, S., L. Meyers, G. Willman, T. Carpenter, and M. J. Hansen. 1999. Growth of
juvenile lake sturgeon reared in tanks at three densities. North American Journal
oquuaculture 61 :331—335.

Fiumera, A. C., B. A. Porter, G. Looney, M. A. Asmussen, and J. C. Avise. 2004.
Maximizing offspring production while maintaining genetic diversity in
supplemental breeding programs of highly fecund managed species. Conservation
Biology 18:94—101.

F lagg, T. A. and C. E. Nash. 1999. A conceptual framework for conservation hatchery

strategies for paciﬁc salmonids. US. Department of Commerce, NOAA
Technical Memorandum NMFS-NWFSC-3 8, Seattle, WA.

165

Flavelle, L. S., M. S. Ridgway, T. A. Middel, and R. S. McKinley. 2002: Integration of
acoustic telemetry and GIS to identify potential spawning areas for lake trout
(Salvelinus namaycush). Hydrobiologia, 483:137-146.

Folkvord, A. and J. R. Hunter. 1986. Size-speciﬁc vulnerability of northern anchovy,
Engraulis mordax, larvae to predation by ﬁshes. Fishery Bulletin 84(4):859-869.

Folz, D. J ., D. G. Czeskleba, and T. F. Thuemler.]983. Artiﬁcial spawning of lake
sturgeon in Wisconsin. The Progressive Fish-Culturist 45(4):231-233.

Ford, M. J. 2002. Selection in captivity during supportive breeding may reduce ﬁtness in
the wild. Conservation Biology 16(3):815-825.

Fox, D. A., J. E. Hightower, and F. M. Parauka. 2002: Estuarine and nearshore marine
habitat used by Gulf sturgeon from the Choctawhatchee River System, Florida.
American Fisheries Society Symposium 28:111-126.

Fraser, D. J. 2008. How well can captive breeding programs conserve biodiversity? A
review of salmonids. Evolutionary Applications.

Fuji, R. 1993. Coloration and chromatophores. In The physiology of ﬁshes. Edited by
DH. Evans. Academic Press, New York. pp. 535—562.

Gadomski, D. M. and M. J. Parsley. 20058. Effects of Turbidity, Light Level, and Cover
on Predation of White Sturgeon Larvae by Prickly Sculpins. Transactions of the
American Fisheries Society 134(2):369-374.

Gadomski, D. M. and M. J. Parsley. 2005b. Laboratory studies on the vulnerability of
young white sturgeon to predation. North American Journal of Fisheries
Management 25:667—674.

Gadomski, D. M. and M. J. Parsley. 2005c. Vulnerability of young white sturgeon,
Acipenser transmontanus, to predation in the presence of alternative prey.
Environmental Biology of Fishes 74:3 89—396.

Gazdewich, K. J. and D. P. Chivers. 2002. Acquired predator recognition by fathead
minnows: Inﬂuence of habitat characteristics on survival. Journal of Chemical
Ecology 28(2):439-445.

Gershanovich, A. D. and L. R. Tauﬁk. 1992. Feeding dynamics of sturgeon ﬁngerlings
(Acipenseridae) depending on food concentration and stocking density. Journal of
Fish Biology 41(3):425-434.

Goodnight, KR 2000. Kinship (version 1.3.1) http://gsoft.smu.edu/GSoft.html

166

 

Goudet, J. 2001. FSTAT, a Program to Estimate and Test Gene Diversities and Fixation
lndices, Version 2.9.3. Available from:
http://www.unil.ch/izea/softwares/fstathtml.

Gross, M. R., J. Repka, C. T. Robertson, D. H. Secor, and W. V. Winkle. 2002. Sturgeon
conservation: Insights from elasticity analysis. American Fisheries Society
Symposium 28:13—30.

Gunn, J. M., M. J. McMurtry, J. N. Bowlby, J. M. Casselrnan, and V. A. Liimatainen.
1987. Survival and grth of stocked lake trout in relation to body size, stocking
season, lake acidity, and biomass of competitors. Transactions of the American
Fisheries Society 116(4):6l 8—627.

Hanson, K. C., S. J. Cooke, C. D. Suski, and D. P. Philipp. 2007. Effects of different
angling practices on post-release behavior of nest-guarding male black bass,
Micropterus spp. Fisheries Management and Ecology 14(2): 141-148.

Hardy, R. S. 2000. Factors inﬂuencing survival, development, and behavior of Shortnose
(Acipenser brevirostrum) and Atlantic sturgeon (A. oxyrhynchus) larvae. Master’s
thesis. University of New Brunswick, NB Canada.

Hay-Chmielewski, E. M. 1987. Habitat preferences and movement patterns of the lake
sturgeon (Acipenserﬁlvescens) in Black Lake, Michigan. Michigan Department
of Natural Resources, Fisheries Research Report 1949, Ann Arbor. -

Hay-Chmielewski, EM. and G. Whelan. 1997. Lake sturgeon rehabilitation strategy.
Michigan Department of Natural Resources, Fisheries Division, Special Report
18, Lansing, MI.

Heath, D. D., C. W. Fox, and J. W. Heath. 1999. Maternal Effects on Offspring Size:
Variation Through Early Development of Chinook Salmon. Evolution
53(5):]605-1611.

Heidinger, R. C. 1999. Stocking for sport ﬁsheries enhancement. In C.C. Kohler & W.A.
Hubert eds. Inland Fisheries Management in North America. p. 375-401. 2nd
Edition. American Fisheries Society, Bethesda.

Hendry, A. P. and T. Day. 2005. Population structure attributable to reproductive time:
isolation by time and adaptation by time. Molecular Ecology 14:901—91 6

Herbinger, C. M., R. W. Doyle, E. R. Pitrnan, D. Paquet, K. A. Mesa, D. B. Morris, J. M.
Wright, and D. Cook. 1995. DNA ﬁngerprint based analysis of paternal and
maternal effects on offspring growth and survival in communally reared rainbow
trout. Aquaculture 137:245-256.

167

Hilbom, R. 1992. Hatcheries and the Future of Salmon in the Northwest. Fisheries
1 7(1):5-8.

Holey, M.E., E. A. Baker, T. F. Thuemler, and R. F. Elliott. 2000. Research and
assessment needs to restore lake sturgeon in the Great Lakes. Great Lakes Fishery
Trust, Workshop Results, Muskegon, Michigan.

Holtgren, J. M., and N. A. Auer. 2004. Movement and habitat of juvenile lake sturgeon
(Acipenserﬁtlvescens) in the Sturgeon River/Portage Lake system, Michigan.
Journal of Freshwater Ecology 19:419-432.

Holtgren, J. M., S. A. Ogren, A. J. Paquet, and S. Fajfer. 2007. Design of a portable
streamside rearing facility for lake sturgeon. North American Journal of
Aquaculture 69:317—323.

Houde, E. D. 1987a. Fish early life dynamics and recruitment variability. American
Fisheries Society Symposium. 2:17-29.

Houde, E. D. 1987b. Patterns and trends in larval-stage growth and mortality of teleost
ﬁsh. Journal of Fish Biology 51:52-83.

Houde, E. D. 1989. Subtleties and episodes in the early life of ﬁshes. Journal of Fish
Biology 35:29—38.

Huntingford, F. A. 2004. Implications of domestication and rearing conditions for the
behaviour of cultivated ﬁshes. Journal of Fish Biology 65: 122-142.

Hutchings, J .A., and D. J. Fraser. 2008. The nature of ﬁsheries- and farming-induced
evolution. Molecular Ecology 17, 294-313.

Ireland, S. C., R. C. P. Bearnesderfer, V. L. Paragamian, V. D. Wakkinen, and J. T. Siple.
2002. Success of hatchery-reared juvenile white sturgeon (Acipenser
transmontanus) following release in the Kootenai River, Idaho, USA. Journal of
Applied Ichthyology 18:642-650.

Jonas, J. L., C. E. Kraft, and T. L. Margenau. 1996. Assessment of seasonal changes in
energy density and condition in age-0 and age-1 muskellunge. Transactions of the
American Fisheries Society. 125(2):203-210.

Kalinowski, S. T., M. L. Taper, and T. C. Marshall. 2007. Revising how the computer
program CERVUS accommodates genotyping error increases success in paternity
assignment. Molecular Ecology 16:1099-1006.

Kempinger, J. J. 1988. Spawning and early life history of lake sturgeon in the Lake

Winnebago system, Wisconsin. American Fisheries Society Symposium
5:110.122.

168

 

King, T. L., B. A. Lubinski, and A. P. Spidle. 2001. Microsatellite DNA variation in
Atlantic sturgeon (Acipenser oxyrinchus oxyrinchus) and cross-species
ampliﬁcation in the Acipenseridae. Conservation Genetics 2(2): 103-1 19.

Kneib, R. T. 1987. Predation risk and use of intertidal habitats by young ﬁshes and
shrimp. Ecology 68(2): 379- 386.

Kuhlmann, M. L., S. M. Badylak, and E. L. Carvin. 2008. Testing the differential
predation hypothesis for the invasion of rusty crayﬁsh in a stream community:
laboratory and ﬁeld experiments. Freshwater Biology 53(1):113—128.

Kynard, B. and M. Horgan. 2002. Ontogenetic behavior and migration of Atlantic
sturgeon, Acipenser oxyrinchus oxyrinchus, and shortnose sturgeon, A.
brevirostrum, with notes on social behavior. Environmental Biology of Fishes
63: 1 3 7—1 50

Kynard, B. and Parker, E. 2005. Ontogenetic behavior and dispersal of Sacramento River
White Sturgeon, Acipenser T ransmontanus, with a note on body color.
Environmental Biology of Fishes 74: 19-30.

Kynard, B., E. Parker, and T. Parker. 2005. Behavior of early life intervals of Klamath
River green sturgeon, Acipensermedirostris, with a note on body color.
Environmental Biology of Fishes 72(1):85-97.

Kynard, B., M. Horgan, M. Kieffer, and D. Seibel. 2000. Habitats used by shortnose
sturgeon in two Massachusetts livers, with notes on estuarine Atlantic sturgeon: A
hierarchical approach. Transactions of the American Fisheries Society 129:487—
503.

Kynard, B., M. Kieffer, M. Horgan, M. Burlingame, and P. Vinogradov. 1999. Studies on
Connecticut River shortnose sturgeon. Final Report, NE Utility Service Co., p.84.

Kynard, B., R. Suciu, and M. Horgan. 2002a. Migration and habitats of diadromous
Danube River sturgeons in Romania: 1998—2000. Journal of Applied Ichthyology
18:529-535.

Lande, R. and G. F. Barrowclough. 1987. Effective population size, genetic variation,
and their use in population management. In Soule, M.E., ed. 1987. Viable
Populations

LaVigne, H. R., 2002. An Improved Portable Surgical Table for the Field and
Laboratory. North American Journal of Fisheries Management. 22(2):571-572.

Leber, K. M., R. N. Cantrell, and P. S. Leung. 2005. Optimizing cost-effectiveness of

size at release in stock enhancement programs. North American Journal of
Fisheries Management 25(4): 1596—1608

169

 

 

Lehmann, N., P. Clarkson, L. D. Mech, T. J. Meier, and R. K. Wayne. 1992. A study of
the genetic relationships within and among wolf packs using DNA ﬁngerprinting
and mitochondrial DNA. Behavioral Ecology and Sociobiology 3:83—94.

Letcher, B. H., J. A. Rice, L. B. Crowder, and F . P. Binkowski. 1996. Size-Dependent
Effects of Continuous and Intermittent Feeding on Starvation Time and Mass
Loss in Starving Yellow Perch Larvae and Juveniles. Transactions of the
American Fisheries Society 125:14-26.

Levin, P. S., R. W. label, and J. G. Williams, J .G. 2001. The road to extinction is paved
with good intentions: negative association of ﬁsh hatcheries with threatened
salmon. Proceedings of the Royal Society 268:1153-1158.

Li, J ., Y. Cohen, D. H. Schupp, and I. R. Adelman. 1996. Effects of walleye stocking on
year-class strength. North American Journal of Fisheries Management 16(4):840-
850.

Li, X., M. K. Litvak, and J. E. H. Clarke. 2007. Overwintering habitat use of shortnose
sturgeon (Acipenser brevirostrum): deﬁning critical habitat using a novel
underwater video survey and modeling approach. Canadian Journal of Fisheries
and Aquatic Sciences 64(9):]248-1257.

Lima, S. L. 1998. Nonlethal effects in the ecology of predator-prey interactions.
BioScience 48(1):25-34.

Lynch, M. and M. O’Hely. 2001a. Captive breeding and the genetic ﬁtness of natural
populations. Conservation Genetics 2(4):363-378

Lynch, M. and M. O’Hely. 2001b. Supplementation and the genetic ﬁtness of natural
populations. Conservation Genetics 2:363-378.

Mantyniemi, S., A. Romakkaniemi, and E. Arjas. 2005. Bayesian removal estimation of a
population size under unequal catchability. Canadian Journal of Fisheries and
Aquatic Sciences 62(2):29l-300.

Margenau, T. L. 1992. Survival and cost-effectiveness of stocked fall ﬁngerling and
spring yearling muskellunge in Wisconsin. North American Journal of Fisheries
Management 12:484-493.

Marking, L. L., J. J. Rach, and T. M. Schreier. 1994. Evaluation of antifungal agents for
ﬁsh culture. The Progressive Fish-Culturist 56:225-231.

May, B., C. C. Krueger, H. L. Kinkaid. 1997. Genetic variation at microsatellite loci in

sturgeon: Primer sequence homology in Acipenser and Scaphirhynchus. Canadian
Journal of Fisheries and Aquatic Sciences 54:1542-1547.

170

 

 

Maynard, D. J ., S. C. Riley, and T. A. Flag. 2004. Development of a natural rearing
system to improve supplemental ﬁsh quality. Final Report. US. Department of
Energy, Bonneville Power Administration, Project Ntunber 91-055.

Maynard, D. J ., T. A. Flagg, and C. V. W. Mahnken. 1995. A review of seminatural
culture strategies for enhancing the post-release survival of anadromous
salmonids. American Fisheries Society Symposium 15:307-314.

Maynard, D. J., T. A. Flagg, C. V. W. Mahnken, and S. L. Schroder. 1996. Natural
rearing technologies for increasing post release survival of hatchery-reared
salmon. Bulletin of National Research Institute of Aquaculture 2:71-7 7.

McCabe, G. T., Jr. and L. G. Beckrnan. 1990. Use of an artiﬁcial substrate to collect
white sturgeon eggs. California Fish and Game 76:248-250.

_ McKeown, P. E., J. L. Fomey, and S. R. Mooradian. 1999. Effects of stocking size and
rearing method on muskellunge survival in Chautauqua Lake, New York. North
American Journal of Fisheries Management 19(1):249-257.

McQuown, E. C., B. L. Sloss, R. J. Sheehan, J. Rodzen, G. J. Tranah, and B. May. 2000.
Microsatellite analysis of genetic variation in sturgeon: New primer sequences for
Scaphirhynchus and Acipenser. Transactions of the American Fisheries Society
129:1380-1388.

McQuown, E., G. A. E. Graham, B. May. 2002. Characterization and inheritance of
six microsatellite loci in lake sturgeon. Transactions of the American Fisheries
Society 13 1 :299-307.

Mehner, T. and W. Wieser. 1994. Energetics and metabolic correlates of starvation in
juvenile perch (Percaﬂuviatilis). Journal of Fish Biology 45(2):325-333.

Mesick, C. E. 1988. Effects of food and cover on numbers of Apache and brown trout
establishing residence in artiﬁcial stream channels. Transactions of the American
Fisheries Society 117:421-431.

Miller, T. E., and W. C. Kerfoot. 1987 . Redeﬁning indirect effects. Pages 33-37 in W. C.
Kerfoot and A. Sih, editors. Predation: direct and indirect impacts on aquatic
communities. University Press of New England, Hanover, New Hamsphire, USA.

Miller, T. J ., L. B. Crowder, J. A. Rice, and E. A. Marschall. 1988. Larval size and

recruitment mechanisms in ﬁshes: towards a conceptual ﬁamework. Canadian
Journal of Fisheries Aquatic Sciences 45:1657-1670.

171

 

Miller, A. I., and L. G. Beckman. 1996. First record of predation on white sturgeon eggs
by sympatric ﬁshes. Transactions of the American Fisheries Society 125(2):338-
340.

Miller, L. M., A. R. Kapuscinski. 2003. Genetic guidelines for hatchery supplementation
programs. Pages 329—355 in E. Hallerrnan, editor. Population genetics: principles
and practices for ﬁsheries scientists. American Fisheries Society, Bethesda.

Miller, L. M., T. Close, and A. R. Kapuscinski. 2004. Lower ﬁtness of hatchery and
hybrid rainbow trout compared to naturalized populations in Lake Superior
tributaries. Molecular Ecology 13(11):3379-3388.

Miranda, L. E., R.J. Muncy. 1987. Recruitment of young-of-year largemouth bass in
relation to size structure of parental stock. North American Journal of Fisheries
Management 72131—137.

Miranda, L. E., W. D. Hubbard. 1994. Winter survival of age-0 largemouth bass relative
to size, predators, and shelter. North American Journal of Fisheries Management
14(4):790-796.

Mittelbach, G. 1986. Predator-mediated habitat use: some consequences for species
interactions. Environmental Biology of Fishes. 16: 159-1 69.

Moser, M. L., S. W. Ross. 1995. Habitat use and movements of shortnose and Atlantic
sturgeons in the Lower Cape Fear River, North Carolina. Transactions of the
American Fisheries Society 124(2):225-234.

Mousseau, TA. and C. W. Fox. 1998. The adaptive signiﬁcance of maternal effects.
Trends in Ecology and Evolution 13(10):403-407.

Myers, R. A. and B. Worm. 2003. Rapid worldwide depletion of predatory ﬁsh
communities. Nature 423:280.

Myers, R. A., S. A. Levin, R. Lande, F. C. James, W. W. Murdoch, and R. T. Paine.
2004. Hatcheries and endangered salmon. Science 303:1980.

Nams, V. O. 2001. Locate 11 users guide. Pacer Computer Software, Truro, Nova Scotia,
Canada.

Nasulund, I. 1992. Survival and distribution of pond- and hatchery-reared 0+ brown
trout, Salmo trutta L., released in a Swedish stream. Aquaculture and Fisheries
Management 23:477-488.

Nichols, S. J ., G. Kennedy, E. Crawford, J. Allen, J 111 French, G. Black, M. Blouin, J.
Hickey, S. Chemyak, R. Haas, and M. Thomas. 2003. Assessment of lake
sturgeon (Acipenserfulvescens) spawning efforts in the Lower St. Clair River,
Michigan. Journal of Great Lakes Research 29(3):383-391.

172

Nickelson, T. E., M. F. Solazzi, S. L. Johnson. 1986. Use of hatchery coho salmon
(Oncorhynchus kisutch) presmolts to rebuild wild populations in Oregon coastal
streams. Canadian Journal of Fisheries and Aquatic Sciences 43(12):2443-2449.

Nilo, P., P. Dumont, and R. Fortin. 1997. Climatic and hydrological determinants of year-
class strength of St. Lawrence River lake sturgeon. Canadian Journal Fisheries
and Aquatic Sciences 54:774-780.

Oliver, J. D., G. F. Holeton, K. E. Chua. 1979. Overwinter mortality of ﬁngerling
smallmouth bass in relation to size, relative energy stores and environmental
temperature. Transactions of the American Fisheries Society 108:130—136.

Olden, J. D., J. M. McCarthy, J. T. Maxted, W. W. Fetzer, and M. J. Vander Zanden.
2006. The rapid spread of rusty crayﬁsh (Orconectes rusticus) with observations
on native crayﬁsh declines in Wisconsin (U.S.A.) over the past 130 years.
Biological Invasions 8:1621—1628.

Olla, B. L., M. W. Davis, and C. H. Ryer. 1994. Behavioural deﬁcits in hatchery-reared
ﬁsh: potential effects on survival following release. Aquaculture and Fisheries
Management 25:19-34.

Olla, B. L., M. W. Davis, and C. H. Ryer. 1998. Understanding how the hatchery
environment represses or promotes the development of behavioral skills. Bulletin
of Marine Science 62:531-550.

Page, K. S. 2001. Genetic diversity and interrelationships of wild and hatchery lake trout
in the upper Great Lakes: inferences for broodstock management and
development of restoration strategies. Master’s thesis. Michigan State University,
East Lansing.

Page, K. S., K. T. Scribner, D. Bast, M. E. Holey, and M. K. Burnham—Curtis. 2005.
Genetic evaluation of a Great Lakes lake trout hatchery program. Transactions of
the American Fisheries Society 134:872—891.

Paragamian, V. L., and R. Kingery. 1992. A comparison of walleye fry and ﬁngerling
stockings in three rivers in Iowa. North American Journal of Fisheries
Management 12(2):313-320.

Paragamian, V. L., G. Kruse, and V. Wakkinen. 2001. Spawning habitat of Kootenai
River white sturgeon, post-libby darn. North American Journal Fisheries
Managrnent 21(1):22-33.

Parauka, F. M., W. M. J. Troxel, F. A. Chapman, and L. G. Mcbay. 1991. Horrnone-

induced ovulation and artiﬁcial spawning of Gulf of Mexico sturgeon (Acipenser
oxyrhynchus desotoi). The Progressive Fish-Culturist 53(2):]13-117.

173

 

Peacor, S. D., and E. E. Werner. 2004. How dependent are species-pair interaction
strengths on other species in the food web? Ecology 85(10):2754-2763.

Peake, S. 1999. Substrate preferences of juvenile hatchery-reared lake sturgeon,
Acipenserﬁtlvescens. Environmental Biology of Fishes 56(4): 1 573-5133.

Pepin, P. 1991. Effect of temperature and size on development, mortality, and survival
rates of the pelagic early life history stages of marine ﬁsh. Canadian Journal of
Fisheries and Aquatic Sciences 48(3):503-518.

Peck, J. W., T. S. Jones, W. R. MacCallum, and S. T. Schram. 1999. Contribution of
hatchery-reared to chinook salmon populations and sport ﬁsheries in Lake
Superior. North American Journal of Fisheries Management 19:155—164.

Post, J. R., E. A. Parkinson. 2001: Energy allocation strategy in young ﬁsh: allometry and
survival. Ecology 82(4): 1040-1051.

Pyatskowit, J. D., C. C. Krueger, H. L. Kincaid, and B. May. 2001. Inheritance of
microsatellite loci in the polyploidy lake sturgeon (Acipenserﬁtlvescens).
Genome 44: 1 85—191.

Queller, D. C., K. F. Goodnight. 1989. Estimating relatedness using genetic markers.
Evolution 43(2):258-275.

Quinn, T. P., and N. P. Peterson. 1996. The inﬂuence of habitat complexity and ﬁsh size
on over-winter survival and growth of individually marked juvenile coho salmon
(Oncorhynchus kisutch) in Big Beef Creek, Washington. Canadian Journal of
Fisheries and Aquatic Sciences 53:1555-1564.

Quist, M. C., J. S. Tillrna, M. N. Burlingame, and C. S. Guy. 1999: Overwinter habitat
use of shovelnose sturgeon in the Kansas River. Transactions of the American
Fisheries Society 128:522—527.

Rach, J. J ., M. P. Gaikowski, G. E. Howe, T. M. Schreier. 1998. Evaluation of the
toxicity and efﬁcacy of hydrogen peroxide treatments on eggs of wann- and
coolwater ﬁshes. Aquaculture 165:1 1-25.

Rach, J. J ., G. E. Howe, and T. M. Schreier. 1997. Safety of formalin treatments on
warm- and coolwater ﬁsh eggs. Aquaculture 149:183-191.

Rahel, F. J ., and R. A. Stein. 1988. Complex predator-prey interactions and predator

intimidation among crayﬁsh, piscivorous ﬁsh, and small benthic ﬁsh. Oecologia
75:94-98.

174

Ratnayeke, S., G. A. Tuskan, and M. R. Pelton. 2002. Genetic relatedness and female
spatial organization in a solitary carnivore, the raccoon, Proqyon lotor. Molecular
Ecology 11:1115—1124.

Ryman, N. 1991. Conservation genetics considerations in ﬁshery management. Journal of
Fish Biology 39(Supp. A):211-224.

Ryman, N., and L. Laikre. 1991. Effects of supportive breeding on the genetically
effective population size. Conservation Biology 5(3):325-329.

Ryman, N., P. E. Jorde, and L. Laikre. 1995. Supportive breeding and variance effective
population size. Conservation Biology 9(6): 1 61 9-1628.

Schaffter, R. G. 1997. White sturgeon spawning migrations and location of spawning
habitat in the Sacramento River, California. California Fish Game 83:1—20.

Secor, D. H., and E. D. Houde. 1998. Use of larval stocking in restoration of Chesapeake
Bay striped bass. ICES Journal of Marine Science 55(2):228-239 .

Secor, D. H., P. J. Anders, W. Van Winkle, and D. A. Dixon. 2002. Can we study
sturgeons to extinction? What we do and don't know about the conservation of
North American sturgeons. American Fisheries Society Symposium 28:3-10.

Secor, D. H., V. Arevyev, A. Nikolaev, and A. Sharov. 2000. Restoration of sturgeons:
Lessons from the Caspian Sea sturgeon ranching program. Fish and Fisheries
1 :21 5-230.

Serge I., S. I. Doroshov, G. P. Moberg, and J. P. Van Eenennaarn. 1997. Observations on
the reproductive cycle of cultured white sturgeon, Acipenser transmontanus.
Environmental Biology of Fishes 48:265-278.

Schlosser, I. J. 1987. The role of predation in age- and size-related habitat use by stream
ﬁshes. Ecology 68(3):651-659.

Schram, S. T., J. Lindgren, and L. M. Evrard. 1999. Reintroduction of lake sturgeon in
the St. Louis River, Western Lake Superior. North American Journal of Fisheries
Management 19:815—823.

Shuter, B. J ., J. A. MacLean, F. E. J. Fry, and H. A. Regier. 1980: Stochastic simulation
of temperature effects on ﬁrst year survival of smallmouth bass. Transactions of
the American Fisheries Society 109:1—34.

Siepker, M. J ., K. G. Ostrand, S. J. Cooke, D. P. Philipp and D. H. Wahl. 2007. A review
of the effects of catch-and-release angling on black bass, Micropterus spp.:
implications for conservation and management of populations. Fisheries
Management and Ecology 14(2):91—101.

175

Skelly, D. K., and E. E. Werner. 1990. Behavioral and life-historical responses of larval
American toads to an odonate predator. Ecology 71 (6):2313-2322.

Smith, K. M., and E. A. Baker. 2005. Characteristics of spawning lake sturgeon in the
Upper Black River, Michigan. North American Journal of Fisheries Management
25:301—307.

Smith, K. M., and D. K. King. 2005. Dynamics and extent of larval lake sturgeon
Acipenserfulvescens drift in the Upper Black River, Michigan. Journal of Applied
Ichthyology 21 :161—168.

Smith, K. M. and D. K. King. 2005. Movement and habitat use of yearling and juvenile
lake sturgeon in Black, Michigan. Transactions of the American Fisheries Society
134:1159—1172.

Smith, T. I. J ., E. K. Dingley, and D. E. Marchette. 1980. Induced spawning and culture
of Atlantic sturgeon. The Progressive Fish-Culturist 42(3): 147-151.

Stunz, G. W., and T. J. Minello. 2001. Habitat-related predation on juvenile wild-caught
and hatchery-reared red drum Sciaenops ocellatus (Linnaeus). Journal of ,
Experimental Marine Biology and Ecology 260:13—25.

Sulak, K. J ., and J. P. Clugston. 1999. Recent advances in life history of Gulf of Mexico
sturgeon, Acipenser oxyrinchus desotoi, in the Suwannee river, Florida, USA: a
synopsis. Journal of Applied Ichthyology 15:116—128.

Summerfelt, R. C., and L. S. Smith. 1990: Anesthesia, surgery, and related techniques.
In: Methods for ﬁsh biology. Eds: C. B. Schreck; P. B. Moyle, American
Fisheries Society, Bethesda, Maryland. pp. 213-272.

Suski, C. D., J. H. Svec., J. B. Ludden, F. J. S. Phelan, and D. P. Philipp. 2003. The effect
of catch-and-release angling on the parental care behavior of male smallmouth
bass. Transactions of the American Fisheries Society 132:210—218.

Tavemy, C., M. Lepage, S. Piefort, P. Dumont, and E. Rochard. 2002. Habitat selection
by juvenile European sturgeon Acipenser sturio in the Gironde estuary (France)
Journal of Applied Ichthyology 18:536—541.

Thomas, G. L. and O. A. Mathisen. 1993. Biological interactions of natural and enhanced
stocks of salmon. Fisheries Research 18.

Tipping, J. M. 1998. Return rates of hatchery-produced sea-run cutthroat trout reared in a

pond versus a standard or baffled raceway. The Progressive Fish Culturist 60: 109-
1 1 3.

176

Tipping, J. M. 2001. Adult returns of hatchery sea-run cutthroat trout reared in a
seminatural pond for differing periods prior to release. North American Journal of
Aquaculture 63:131-133.

Travis, J., F. C. Coleman, C. B. Grimes, D. Conover, T. M. Bert, and M. Tringali. 1998.
Critically assessing stock enhancement: An introduction to the mote symposium.
Bulletin of Marine Science 62(2):305-311.

Tringali, M. D., and T. M. Belt. 1998. Risk to genetic effective population size should be
an important consideration in ﬁsh stock-enhancement programs. Bulletin of
marine science 62(2):641-659.

Unwin, M. J ., M. T. Kinnison, N. C. Boustead,and T. P. Quinn. 2003. Genetic control
over survival in Paciﬁc salmon (Oncorhynchus spp.): experimental evidence
between and within populations of New Zealand chinook salmon (0.
tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences 60:1—1 1.

Utter, F. 1998. Genetic problems of hatchery-reared progeny released into the wild, and
how to deal with them. Bulletin of Marine Science 62:623-640.

Vandeputte, M., M. Kokour, S. Mauger, M. Dupont-Nive, D. De Guerry, M. Rodina, D.
Gela, D. Vallod, B. Chevassus, and O Linhart. 2004. Heritability estimates for
growth-related traits using microsatellite parentage assignment in juvenile
common carp (Cyprinus carpio L.). Aquaculture 235:223—236.

Walsh, M. R., and D. N. Rezrrick. 2008. Interactions between the direct and indirect
effects of predators determine life history evolution in a killiﬁsh. Proceedings of
the National Academy of Sciences 105(2):594—599.

Walters, J. P., T. D. Fresques, and S. D. Bryan. 1997. Comparison of creel returns ﬁom
rainbow trout stocked at two sizes. North American Journal of Fisheries
Management 17(2):474-476.

Wang, Y. L., F. P. Binkowski, and S. I. Doroshov. 1985. Effects of temperature on early
development of white and lake sturgeon, Acipenser transmontanus and A.

fulvescens. Environmental Biology of Fishes 14:43-50.

Wang, J. 2004. Monitoring and managing genetic variation in group breeding populations
without individual pedigrees. Conservation Genetics 5:813—825.

Waples, R. S. 1991. Genetic interactions between wild and hatchery salmonids: lessons
ﬁom the Paciﬁc Northwest. Canadian Journal of Fisheries and Aquatic Sciences
48(Supplement 1):124-133.

Waples, R. S. 1999. Dispelling some myths about hatcheries. Fisheries 24(2):2-21.

177

Watters, J. V., and C. L.Meehan. 2007. Different strokes: Can managing behavioral types
increase post-release success? Applied Animal Behaviour Science 102:364—3 79.

Wayne, R. K., N. Lehman, D. Girman, P. J. P. Gogan, D. A. Gilbert, K. Hansen, R. 0.
Peterson, U. S. Seal, A. Eisenhawer, L. D. Mech, and R. J. Krumenaker. 1991.

Conservation genetics of the endangered Isle Royale gray wolf. Conservation
Biology 5(1):41—51.

Webb, M. A. H., G. W. Feist, J. M. Trant, J. P. Van Eenennaam, M. S. Fitzpatrick, C. B.
Schreck, S. I. Doroshov. 2002. Ovarian steroidogenesis in white sturgeon
(Acipenser transmontanus) during oocyte maturation and induced ovulation.
General and Comparative Endocrinology 129:27-38.

Welsh, A. B., M. Blumberg, and B. May. 2003. Identiﬁcation of microsatellite loci in
lake sturgeon, Acipenserfulvescens, and their variability in green sturgeon, A.
medirostris. Molecular Ecology Notes 3:47-55.

Werner, E. E., J. F. Gilliam, D. J. Hall, and G. G. Mittelbach. 1983. An experimental test
of the effects of predation risk on habitat use in ﬁsh. Ecology 64(6):]540-1548.

Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in
size-structured populations. Annual Review of Ecological Systems 15:393-125.

Werner, E. E., and D. J. Hall. 1988. Ontogenetic habitat shifts in bluegill: The foraging
rate-predation risk trade-off. Ecology 69(5): 1 352-1366.

Werner, E. E., and B. R. Anholt. 1993. Ecological consequences of the trade-off between
growth and mortality rates mediated by foraging activity. The American
Naturalist l42(2):242-272.

Werner, E. E., and B. R. Anholt. 1996. Predator-Induced Behavioral Indirect Effects:
Consequences to Competitive Interactions in Anuran Larvae. Ecology 77(1): 1 57-
169.

Werner, E. E. and S. D. Peacor. 2003. A review of trait-mediated indirect interactions in
ecological communities. Ecology 84(5): 1 083-1 100.

Whitlock, M. C. and D. E. McCauley. 1999. Indirect measures of gene ﬂow and
migration: FST¢1/(4Nm+1). Heredity 82:117-125.

Wieser, W., G. Krurnschnabel, and J. P. ijang—Okwor. 1992. The energetics of

starvation and growth after refeeding in juveniles of three cyprinid species.
Environmental Biology of Fishes 33:63-71.

178

 

Wiley, R. W., R. A. Whaley, J. B. Satake, and M. Fowden. 1993. An evaluation of the
potential for training trout in hatcheries to increase poststocking survival in
streams. North American Journal of Fisheries Management 13:171-177.

Winkelman, D. L. and J. M. Aho. 1993. Direct and indirect effects of predation on
mosquitoﬁsh behavior and survival. Oecologia 96(3):300-303.

Winter, J. D. 1996. Advances inunderwater biotelemetry. In: Fisheries techniques. 2"‘1
Edition. Eds: B. R. Murphy; D. W. Willis, American Fisheries Society, Bethesda,

Maryland. pp. 555-590.

Yule, D. L., R. A. Whaley, P. H. Mavrakis, D. D. Miller, and S. A. Flickinger. 2000. Use
of strain, season of stocking, and size at stocking to improve ﬁsheries for rainbow
trout in reservoirs with walleyes. North American Journal of Fisheries
Management 20( 1):10-1 8.

Zamora, L. and R. Moreno-Amich. 2002. Quantifying the activity and movement of

perch in a temperate lake by integrating acoustic telemetry and a geographic
information system. Hydrobiologia 483:209-218.

179