Jfixfif:

-.,~.-x:z. my.

ӣ1.15:

.H ,_ .3] 1“ ; I;-‘-.> arr-4
-. .fs‘ia‘a‘Jq'fiR: .x-é'wu'fii’l' refiffl‘ . ‘

5»
“A

p =
5.“.

. '_::~\
J: 33$”
rfftt‘
11.

' 143,.
“5'32:

u‘.

.1 .

; ' !

3%- H
4.1:“:

{£5712
. u v-

~‘IL

u

s“ ‘- Y
L. n .
. Sac-«w
\ . 4‘5.

‘3‘:
.leRM

3455
A?

,_ .fi,
‘«

- n
-

v.31?» ‘
.. .45. .
A {Azalczw

‘3

4"

A ii?)
‘ “i
=w;

1

J

rjf
:Jfii’é" ’ ‘

2??

§

5" i
a “h f“l§afl
”r a

i!

z; .1 I. ~_ lift,

awfifi‘fig‘ ‘
at " 1.13 '3 1‘ ‘

‘ v‘ "m‘fiw}3$%$$'

V I”!
,«4 ”55‘

Wfigfif

.\-
_ is.

“if-“{—
~33

r"

: (mi, \7

2-!“

 

 

g

1 “ I
a 35%;.
n a, fi};
3?~!“> . at}.
3'

13'
, ‘ ‘an " ”J“ 1
I. ~ g .3 3;: ”‘13“! I,
Q ‘ ., _ 9 v
w_ r1. " «Ugg‘a'F-Jm‘sfi '

3

«
1.1
t,‘
\2’

M5

3 33,51 r
'6‘. 9». w
a l

3556???
L

.-
c
‘1

SITY UBRARIE 8

mess ‘ A lllllllllllllllllllllllllllllllllllllllllll

I 3 1293 01409

(‘3“§:>

 

 

 

This is to certify that the

thesis entitled

Contribution of Hatchery and Natural Chinook Salmon
to the Eastern Lake Michigan Fishery, 1992-1993

presented by

Jay A. Hesse

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Fish. & Wildl.

 

 

Major professor

Date November 29, 1994

 

0.7539 MS U is an Affirmative Action/Equal Opportunity Institution

 

.LIBRARY
M'Chlgan state
Unlverslty

 

 

 

MACE ll RETURN BOX to remove thin checkout {tom your record.
To AVOID FINES Mum on or baton duo duo.

DATE DUE DATE DUE . DATE DUE

_'

     
   
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CONTRIBUTION OF HATCHERY AND NATURAL CHINOOK SALMON TO
THE EASTERN LAKE MICHIGAN FISHERY, 1992-1993

By

Jay A. Hesse

A THESIS

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

MASTER OF SCIENCE

Department of Fisheries and Wildlife
1994

ABSTRACT

CONTRIBUTION OF HATCHERY AND NATURAL CHINOOK SALMON TO THE
EASTERN LAKE MICHIGAN FISHERY, 1992-1993

By

Jay A. Hesse

All hatchery produced chinook salmon (Oncorhynchus tshawytscha) stocked into the
Lake Michigan Basin from 1990 to 1993 were marked with various fin clips and/or an
internal mark with the antibiotic oxytetracycline (OTC). As a result, individual chinook
salmon could be identified as being of either hatchery or natural origin.

The chinook salmon sport harvest was sampled at three locations to determine: 1) the
validity of ageing chinook salmon using vertebrae, 2) the effectiveness of OTC marking, 3)
the contribution of naturally reproduced chinook salmon to the sport harvest, and 4) the
regional variation of that contribution. Additional gill net samples were used to compare
various sampling techniques and to examine the occurrence of bacterial kidney disease (BKD)
in hatchery and natural chinook salmon.

A total of 703 and 1,374 chinook salmon were sampled from the sport fishery during
1992 and 1993, respectively. Annular bands in vertebrae were used to age samples with 97
percent accuracy. Using chinook salmon marked with fin clips and OTC, an OTC mark
failure of 5 percent was established. Regional differences existed in the percentage of
naturally produced age 3 chinook salmon that contributed to the sport fishery. N0 regional
difference existed at ages 1 and 2, with about 30 percent of the sport harvest consisting of

naturally produced chinook salmon. Both gill net samples and sport fishery samples exhibited
the same percentage of hatchery and natural chinook salmon. BKD affected hatchery and

natural chinook salmon equally.

ACKNOWLEDGEMENTS

I would like to thank my major professor, Dr. Niles Kevem, for his support
and guidance throughout my college education and on this graduate research project.
I would also like to thank my committee members, Dr. Thomas Coon and Dr.
Richard Hill, for their advice and educational example. Their never ending
enthusiasm for teaching will always be an inspiration to me.

Funding sources for this research project were provided by: Michigan Sea
Grant College Program, Michigan Department of Natural Resources Fisheries
Division, Michigan Agricultural Experiment Station, West-Metro Chapter of
Michigan Steelheaders, and the Michigan Polar-Equator Club.

A very special thanks is given to the fisherman who supplied samples and
especially to the charterboat captains for their cooperation and hospitality. I thank
Meryl Keller and the other MDNR personnel for their support and assistance. Robert
Elliott and Jay Wesley helped immensely in the development and execution of this
research project, thank you. I will never forget the friendships and support of my lab
mates: Rob Elliott, Jay Wesley, Bruce Peffers, and Nan Kelly. I will always try to
create my future professional relationships after this lab. I thank my entire family for

their examples and support.

iii

I have an inexpressible gratitude for my wife Linda not only for her emotional
support and friendship during my graduate education, but also for her hours upon
hours of free labor as a field assistant on the weekends. This research would not

have been possible without her help.

iv

TABLE OF CONTENTS

List of Tables ......................................... viii

List of Figures .......................................... x

Chapter One, Introduction .................................. 1
Chapter Two, An Alternative Age Determination for Lake Michigan Chinook

Salmon Using Vertebrae ............................... 8

Introduction ...................................... 8

Methods ....................................... 10

Sample Source ............................... 10

Laboratory Preparation .......................... 10

Vertebrae Description ........................... 11

Test for Accuracy of Vertebrae Age Estimates ............ 13

Test for Precision of Vertebrae Age Estimates ............ 13

Results ........................................ 14

Accuracy ................................... 14

Precision ................................... 16

Discussion ...................................... 16

Chapter Three, Oxytetracycline Detection and Quality of Marks in Lake
Michigan Chinook Salmon ............................ 22

Introduction ..................................... 22

Methods ....................................... 24
Sample Source ............................... 24
Laboratory Preparation .......................... 24
Mark Permanency ............................. 25
Mark Retention ............................... 26
Repeatability ................................ 26

Results ........................................ 27
Mark Permanency ............................. 27
Mark Retention .................... . ........... 27
Repeatability ................................ 30

Discussion ...................................... 30

Chapter Four, Percentage of Hatchery and Natural Chinook Salmon in the

Lake Michigan Sport Fishery ........................... 36
Introduction ..................................... 36
Methods ....................................... 37
Sampling Ports Description ........................ 37
Field Methods ............................... 39
Laboratory Analysis ............................ 41
Statistical Analysis ............................. 41
Results ........................................ 41
Discussion ...................................... 47
Seasonal and Port Trends .......................... 47

vi

Cohort Trends ............................... 48
Relative Survival .............................. 49
Natural Contribution ............................ 51

Chapter Five, Occurrence of Bacterial Kidney Disease and Comparison of
Sampling Techniques Using Hatchery and Natural Chinook Salmon . . . . 54

Introduction ..................................... 54
Methods ....................................... 55
Sample Source ............................... 55

Field Methods ............................... 5 6

laboratory Methods ............................ 5 6

Results ........................................ 5 6
Discussion ...................................... 58
Chapter Six, Summary ................................... 60
Appendix A .......................................... 62
Appendix B .......................................... 63
Appendix C ......................................... 64
Appendix D ......................................... 71
References ......................................... 77

vii

LIST OF TABLES

Table 1. Coded wire tag and vertebral age determination results for

chinook salmon sampled from Lake Michigan and harvest weirs . 15
Table 2. Chi-square values for comparison of vertebral age determination

accuracy. .................................. 15
Table 3. Difference in estimated ages between fish aged from vertebrae

and from Coded-Wire Tag(CWT). ................... 15
Table 4. List of estimated vertebral ages and associated APE, V, and D

from 3 readers for 100 chinook salmon. ................ 17
Table 5. Weighted average of the quality of OTC marks for all hatchery

chinook salmon in the 1990-1992 cohorts over time. ........ 27
Table 6. Correlation between fin clips and OTC marks. ............ 29
Table 7. Percentage of double marked chinook salmon not exhibiting an

OTC mark. ................................. 29
Table 8. Oxytetracycline mark quality ratings for samples which multiple

readers assigned different OTC mark qualities. ............ 31
Table 9. Primary cleaning stations targeted for sample collection. ...... 40

Table 10. Summary of the percentage of hatchery and natural chinook
salmon at Grand Haven, Ludington, and Leland during 1992 and
1993. ..................................... 45

Table 11. Percentages of hatchery and natural chinook salmon from the
1992 and 1993 Ludington area sport harvest, Michigan
Department of Natural Resources chinook salmon diet study gill
nets, and clinically sick chinook salmon sampled in diet study
gill nets. ................................... 57

viii

Table 12.

Table 13.

Table 14a.

Table 14b.

Summary of Lake Michigan hatchery chinook salmon marking,
1990-1993 stockings, numbers of fish in each category. ......

Summary of fin clips used by state and year. .............
Summary of the 1992 age 1 and age 2 daily samples for each

port. Percentages have not been corrected for 5 % occurrence of
unmarked hatchery chinook salmon. ..................

Summary of the 1993 age 1 and age 2 daily samples for each

port. Percentages have not been corrected for 5 % occurrence of
unmarked hatchery chinook salmon. ..................

ix

63

LIST OF FIGURES

Figure 1. Catch rates of chinook salmon in Michigan waters from 1985-

1993. ...................................... 4
Figure 2. Diagram of translucent and opaque annular bands in vertebrae of

Lake Michigan chinook salmon. ..................... 12
Figure 3. Timing and pattern of annuli formation in vertebrae of Lake

Michigan chinook salmon. ........................ 12

Figure 4. Range in quality and weighted average (wa) of OTC marks over
a two year period for the 1990 and 1991 cohorts (O=no mark,
1-3=poor, 4-6=good, 7=great). .................... 28

Figure 5. Lake Michigan and location of sampled ports. ............ 38

Figure 6a. Daily percentage of chinook salmon from natural reproduction

with respect to sample size for all ports combined at age 1 and
age 2 during 1992. ............................. 43

Figure 6b. Daily percentage of chinook salmon from natural reproduction
with respect to sample size for all ports combined at age 1 and
age 2 during 1993. ............................. 44

Figure 7. Percentage of naturally reproduced chinook salmon in the 1992
and 1993 sport harvest at Grand Haven, Ludington, and Leland.
Error bars represent +/- standard error. ................ 46

Figure 8. Yearly percentages of naturally reproduced chinook salmon in
the 1990 and 1991 cohorts at Grand Haven and Ludington.
Bars with dashed lines represent data from Elliott (1994). ..... 50

Figure 9a. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1992 sport harvest at Grand
Haven. Dots represent daily samples, bars represent monthly
averages, and dashed lines represents yearly average. Only
months in which sampling occurred are shown. ........... 71

Figure 9b. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1992 sport harvest at Ludington.
Dots represent daily samples, bars represent monthly averages,
and dashed lines represents yearly average. Only months in
which sampling occurred are shown. .................. 72

Figure 9c. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1992 sport harvest at Leland.
Dots represent daily samples, bars represent monthly averages,
and dashed lines represents yearly average. Only months in
which sampling occurred are shown. .................. 73

Figure 9d. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1993 sport harvest at Grand
Haven. Dots represent daily samples, bars represent monthly
averages, and dashed lines represents yearly average. Only
months in which sampling occurred are shown. ........... 74

Figure 9e. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1993 sport harvest at Ludington.
Dots represent daily samples, bars represent monthly averages,
and dashed lines represents yearly average. Only months in
which sampling occurred are shown. .................. 75

Figure 9f. Daily, monthly, and yearly percentages of age 1 and age 2
natural chinook salmon in the 1993 sport harvest at Leland.
Dots represent daily samples, bars represent monthly averages,
and dashed lines represents yearly average. Only months in
which sampling occurred are shown. .................. 76

xi

CHAPTERI

Introduction

Introduction of exotic salmonid species into the Great Lakes was attempted
numerous times between the 1870’s and 1950’s. Most of these introductions involved
one-time stockings to various habitats in an attempt to establish self sustaining
populations. Generally, these introductions were not successful due to competition
from native species, lack of a proper forage base, and degraded spawning habitat
(Parsons 1973). Historical reviews of the Great Lakes fish communities and
management can be found in Smith (1968), Parsons (1973), Christie (1974), Carl
(1980), and Keller et a1. (1990).

In the late 1960’s Great Lakes fishery managers started a stocking program of
Pacific salmon in an attempt to produce a put and take fishery. Coho salmon
(Oncorhynchus kisutch) and chinook salmon (Oncorhynchus tshauytscha), were
introduced in 1966 and 1967 respectively. These introductions were successful
because of the large forage base provided by the alewife (Alosa pseudoharengus)
which invaded the Great Lakes from the Atlantic Ocean. The presence of salmon in
Lake Michigan developed one of the most spectacular sport fisheries in the world and
contributed an estimated $1.4 billion annually to the Michigan economy (Robertson
1990).

Lake-wide annual stocking levels of chinook started around 700,000 during the

late 1960’s and increased to a peak of 7,859,000 in 1989 (Parsons 1973 and Holly

2
1994). Current 1993 stocking levels have decreased to 5,491,000, roughly equal to

1978 stocking levels (Holly 1994). The decision to reduce stocking levels was
triggered by a decease in the average size of sport harvested salmonids (Keller et a1.
1990). Around this same time (1988) disease problems in chinook salmon became
apparent and were then thought to be induced by a shortage in available forage.

In an attempt to maintain a balanced predator/prey ecosystem in Lake
Michigan a closer look at the interactions between forage fishes and salmonids was
examined by Stewart et a1. (1981), Eck and Brown (1985), Eek and Wells (1987),
and Brandt et a1. (1991). A computer program has been developed to simulate Lake
Ontario’s ecosystem dynamics (Jones et a1. 1993) and a similar model is currently
being developed for Lake Michigan. This model will help determine the optimal
chinook salmon population size. Stewart et a1. (1981) address the existence of some
natural reproduction of salmonids in the Great Lakes with estimates of 10-15% of the
total salmonid population coming from natural reproduction due to non-hatchery
recruitment. The contribution of non-hatchery fish is not currently incorporated into
Stewart et a1. (1981) or Jones et a1. (1993) population analysis, with the expectation
that modifications can be made later when good estimates of the contribution of
natural fish are available. Subsequent studies have focused mainly on estimating the
available forage and make the assumption that the actual population size closely
reflects stockings of salmonids (Eck and Wells 1987). However, if input of natural
reproduction of chinook salmon is not taken into account, the actual population size

will surpass the desired population size. Thus, estimates of lake-wide natural

3

reproduction input are needed to determine the actual population size and to adjust
stocking levels accordingly.

Bacterial Kidney Disease (BKD) was determined to be the cause of death for
high numbers of chinook salmon in Lake Michigan from 1988 to present (Nelson and
Hnath 1990). BKD occurs in natural systems in low incidence (Johnson and Hnath
1991), but it is believed that intensive culture of salmon in hatcheries has increased
the concentration of BKD in chinook salmon. As a result, BKD may affect hatchery
chinook salmon at higher concentrations than natural chinook salmon. If this is true,
offspring of natural chinook would be more desirable than prodigy from hatchery
propagated chinook due to increased survival.

Shown in Figure 1, catch per effort (fish per 100 angler hours) of chinook
salmon in Michigan waters of Lake Michigan has decreased since its peak in 1986 at
10.26 to a record low in 1993 of 1.80 (Rakoczy 1994). Knowledge about the proper
and efficient management of the Great Lakes fishery is critical for the future success
and existence of the chinook salmon fishery.

The management of the salmon fishery in the Great Lakes has been primarily
controlled by the stocking of hatchery propagated smolts. Prior to the introduction of
salmon into Lake Michigan, it was believed that the salmon would be unable to
naturally reproduce in the tributaries (Parsons 1973). Therefore, the population
would be entirely controlled and maintained by manipulating stocking levels, but as
early as 1970 natural reproduction of Pacific salmon was occurring in tributaries of
Lake Michigan (Rybicki 1973 and Taube 1974). Since that time studies by Carl

(1982a), Seelbach (1985), Zafft (1992), and Jennings (1992) have documented

 

12

10.26

3-.
O

A

(latch Rate (fish/100 angler hours)

 

N

 

 

0 1 1 1 1 1 1 1 1 1
1985 1986 1987 1988 1989 1990 1991 1992 1993

 

Figure 1. Catch rates of chinook salmon in Michigan waters from 1985 - 1993 (Rakoczy 1994).

5

chinook salmon natural reproduction in several Michigan rivers. These studies, plus
Carl (1984), also contain spawning and early (river) life history descriptions of Lake
Michigan chinook salmon.

Carl (1980) documented that the amount of natural reproduction is not uniform
but trends vary widely from river to river due to different habitat characteristics.
Zafft (1992), Jennings (1992), Seelbach (1985), and Carl (1984) also demonstrated
yearly changes of natural reproductive output to occur within rivers due to variable
environmental conditions. Results from these studies can be used to estimate the
contribution of natural reproduction to lake Michigan to range from 10 percent to 25
percent, but estimates are based only on initial numbers of juveniles entering the lake.

Not all tributaries and habitats have been directly evaluated for their
reproductive output and, as a result, the amount of natural reproduction occurring
may be underestimated. In-lake natural reproduction of chinook salmon was
documented in Lake Huron by Powell and Miller (1990). They observed and verified
chinook salmon spawning and presence of developing chinook salmon eggs on a shoal
traditionally used by lake trout (Salvelinus namaycush) for reproduction.

In 1990 all of the agencies responsible for stocking chinook salmon into Lake
Michigan, Michigan Department of Natural Resources (MDNR), Indiana Department
of Natural Resources (IDNR), Illinois Department of Conservation (IDOC), and
Wisconsin Department of Natural Resources (WDNR), began a 5 year marking
program in which all hatchery produced chinook salmon were marked with fin clips
and/or the antibiotic oxytetracycline (OTC). A total of 7 combinations of fin clips

was used in 1990 to externally mark 29% of the chinook salmon stocked into Lake

6
Michigan. Similarly, 11, 9, and 7 types of fin clips were used in 1991, 1992, and

1993 to externally mark 41%, 28%, and 26%, respectively. Fish with only the
adipose fin clipped also possessed a Coded-Wire Tag (CWT) in their nose (See
Appendix A for specific fin clip types and numbers stocked by each agency). For
these cohorts, all non-clipped and some fin clipped hatchery chinook salmon were
given an internal mark of OTC. This OTC mark appears as a fluorescent ring in the
vertebrae when viewed under ultraviolet light (Weber and Ridgeway 1962 and Weber
and Ridgeway 1967). This extensive marking program allows for a quick and precise
determination of the origin of individual chinook salmon.

With the ability to determine the number of hatchery and natural chinook
salmon in every sample, a much more detailed analysis of the contribution of natural
fish to the total p0pulation and to the fishery was examined in this study. Preliminary
results by Elliott (1994) indicate that 50 percent (North) and 30 percent (South) of age
0 chinook salmon from Eastern Lake Michigan are from natural reproduction. This
suggests a significant amount of natural reproduction. Each of the above mentioned
studies only accounts for naturally produced fish in early life stages and does not
answer the question of how extensively natural fish contribute to the fishery. If
naturally produced chinook salmon exhibit better survival than hatchery produced
chinook salmon, alternative management practices could be considered. Managing for
increased natural reproduction could increase the survival of chinook salmon by

reducing the effects of BKD and also reduce the hatchery costs.

Goal and Objectives

The goal of this project was to assess the contribution of hatchery and natural
chinook salmon to the Eastern Lake Michigan fishery during 1992 and 1993. This
could be accomplished by using the fin clips and OTC marks possessed by all
hatchery chinook salmon in the 1990 - 1993 cohorts. This goal involved four
objectives: 1) determining the lake-wide (Eastern Lake Michigan) contribution of
naturally reproduced chinook salmon to the Lake Michigan ecosystem, but focusing
on their contribution to the sport fishery, 2) determining the percentage of hatchery
versus natural chinook salmon in the sport harvest at three different locations in Lake
Michigan, 3) examining the relative survival rates for hatchery and natural chinook
salmon while in the lake, and 4) establishing the most feasible sampling methods and
procedures for future use of OTC marking in determining Great Lakes salmonid
population characteristics. The third objective was tested in two ways: first, by
looking at the percentage of hatchery and natural chinook salmon in the same cohort
at two time intervals and second, by comparing the percentage of both hatchery and
natural chinook salmon with clinical signs of BKD (high mortality factor).

Chapter 2 describes the age determination method used and its validity,
chapter 3 describes the OTC detection procedure used and its validity, chapter 4
addresses the contribution of natural chinook salmon, and chapter 5 compares gill net
and sport harvest sampling techniques along with the occurrence of BKD in hatchery
and natural chinook salmon. Objectives 1 and 2 are addressed in chapter 4, objective
3 is covered in chapters 4 and 5, and objective 4 is covered in chapters 2

through 4.

CHAPTER TWO

An Alternative Age Determination Method for Lake Michigan

Chinook Salmon Using Vertebrae

Introduction

In order to assess the population of hatchery and natural chinook salmon
(Onchorhynhus tshawytscha) in Lake Michigan it was necessary to compile age data.
Several age determination methods using different calcified structures including
otoliths, dorsal spines, pectoral fin rays, opercula, subopercula, cleithra, vertebrae,
and scales have been used to age fish with varying degrees of success (Menon 1950,
Chilton and Bilton 1986, Sharp and Bernard 1988, Baker and Timmons 1991, Hall
1991, and Rien and Beamesderfer 1994). Historically, scales have been used to age
chinook salmon from the Great Lakes. However, recent studies using known-age fish
from coded wire tags have raised validity concerns of scale-age estimates. This has
recently become a particular concern with regard to ageing mature chinook salmon in
the Great Lakes (Dan Anson, Personal Communication).

Regardless of the ageing method used, Beamish and McFarlane (1983 and
1987) stress the importance of validating age determination methods over all possible
ages. For an age estimation method to be useful in fisheries management it not only
has to be accurate but also precise. Beamish and McFarlane (1983) define precision
as the degree of agreement among readers, more simply referred to as "repeatability".

The existence of coded wire tagged known-age fish provided an opportunity to

9

examine the validity, both accuracy and precision, of age determination methods for
Lake Michigan chinook salmon. A valid age determination method using vertebrae
was desirable for this project since vertebrae were already being collected and
examined for the detection of oxytetracycline (OTC).

Vertebrae have been used for age determination in many species of fish.
Menon (1950) provides a summary of the pre-1950 literature on ageing using bones.
Fourteen of these studies, primarily on saltwater species, used vertebrae. More
recently, vertebrae of channel catfish (Ictalurus lacustris punctatus) were used by
Appelget and Smith ( 1951) and Sneed (1951) to determine age and growth rates.
These vertebrae studies all basically used the same methods for preparing and viewing
the vertebrae. The bones were boiled, soaked, and/ or digested to remove flesh, then
dried prior to observation through 6X magnification under reflected light. Prince et
a1. (1985) prepared thin cross-sections of vertebrae from Atlantic bluefin tuna
(Thunnus thynnus) to examine internal zones used for age determination.

Kusakari (1969) provides a description of the opaque and translucent zones
observed in flatfish (Kareius bicoloratus) vertebrae. This description is generally
acceptable with the previously mentioned studies as well. The laboratory methods
required to prepare samples in the manner previously published are very time
consuming and, as a result, these methods did not seem conducive for examining a
high number of samples. A similar pattern of opaque and translucent bands in Lake
Michigan chinook salmon vertebrae was observed by this author during examination

for OTC. The OTC/age detection method required very little sample preparation,

10

resulting in a much quicker age determination method than described in previous
papers.

The objective of this chapter is to present an age determination method for
Lake Michigan chinook salmon using vertebrae and test its validity by examination of

known-age coded wire tagged fish.

Methods

Sample Source

Adipose fin clipped/coded wire tagged chinook salmon were sampled from the
Lake Michigan sport fishery, from Michigan Department of Natural Resources
(MDNR) Survey Vessel (SV) Steelhead’s gill nets used for a chinook salmon-diet-
study, and from the Wisconsin Department of Natural Resources (WDNR) weir
harvest. Weight, length, and sex data, and scales, vertebrae, and the nose were
collected. Scale samples were taken just behind the dorsal fin and above the lateral
line. Scale samples were not collected from the WDNR weir fish. At least five
thoracic vertebrae were collected from below the adipose fin. Noses from the coded
wire tagged fish were cut off just behind the eyes. Vertebrae and noses were stored
frozen until lab examination.

W

The known-age from coded wire tags was provided by either the MDNR or
WDNR, and the estimated age from vertebrae samples was determined by myself
without prior knowledge of the coded wire tag results. The age estimate from

vertebrae was obtained by completely cleaning flesh and cartilage from one vertebra

1 1
with a scalpel (# 4 round tip scalpel blades work best). It is important to remove the

cartilage ring which separates vertebrae. Samples were viewed within 1 or 2 hours of
cleaning to prevent drying. The vertebra was placed on end, center hole vertical, on
a black background and covered with several drops of glycerin. The bone was then

viewed in a dark room through a dissecting microscope, with a magnification of 15 to

50 (using 10X occulars), under reflected ultraviolet light (365nm). The light used in
this study had an intensity of approximately 11,000 microwatts per square centimeter
at six inches (distance from bulb to bone). Magnification was adjusted to allow the
entire bone to be in the field of view.

Vegebrae descript'on

The observable annular patterns are similar to those described by Kusakari
(1969) in flatfish with an alternate pattern of translucent (dark, absorptive zones) and
opaque (white, reflective zones) bands (Figure 2). The focus of the centrum is dark
with the first white band being formed at age 1 during the second summer in the lake.
The white bands are the annuli being formed during the summer months of July and
August. This corresponds with formation of annuli in scales (Dan Anson, personal
communication). A second dark band starts to appear during the fall (September).
Each successive dark band becomes thinner. This pattern continues until the chinook
salmon reaches maturity (Figure 3). In weir harvest samples, the final band is white,
formed prior to return to river for spawning. The appearance of l to 3 "accessory
checks," thin white rings, within the first dark area is common in age 0 and in some
age 1 chinook salmon. These accessory checks are much thinner than the annular

bands and are not visible in age 2 and older chinook salmon.

12

 

Opaque Band (Age 1)

............ e 2

93

Figure 2. Diagram of translucent and opaque annular bands in vertebrae of lake Michigan

 

chinook salmon.
Accessory Checks
I But IWIIHO I D-k I Wit. I Dark I Wflta

for spawning

Hatdlory stodring and
Natural artmigratim
Return to river

_8
E;
i.

:l Annuli Formatim
Annuh Formatrm

IJPMAMJJAIONDIJFMAMJJABONDIJFMAMJJAIONDIJPMAMJJAIONDI

Age 0 Age 1 Age 2 Age 3

Figure 3. Timing and pattern of annuli formation in vertebrae of Lake Michigan chinook salmon-

13

On several trials multiple vertebrae from one fish were examined. All
vertebrae exhibited the same annuli patterns. The last 4-5 caudal vertebrae were not
readable due to fusing and the inability toseparate individual bones. The thoracic
vertebrae were the easiest to clean and read, while the pleural vertebrae located just
behind the head were the most difficult.

Tast far accuragy of yanebraa age esa'maga

A comparison of known-age to vertebral-age estimates was done on each age
class for both lake-caught and river—harvested mature samples. Bonferroni chi-square
contingency tables were used to test for differences in ageing accuracy between age
groups and ordinary chi-square tables were used to test for differences between
locations (Gill 1978). Both tests were done at a level of confidence of 95% (alpha of
0.05). Samples where estimated age deviated from the known-age were examined to
determine if any consistent bias of age estimates existed. Chi square statistics were
used to determine if misaged estimates differed from the non-biased 50% ratio.
Coded wire tag data were not known by the vertebrae reader at time of examination.

Tgst far prgiaign 9f vagabrag aga agg'mataa

A sample of 100 vertebrae was aged independently by 3 readers and the age
estimates compared to examine the repeatability among readers. Indices of average
percent error (APE), coefficient of variation (V), and index of precision (D) were
determined using methods described by Beamish and Foumier (1981) and Chang
(1982). The APE consists of obtaining an average age for each fish (Xj), either from

multiple readings by one reader or by readings from multiple readers (R). These

14

averages are used to establish an average error in aging each fish as follows:

.1.£____"‘s"9'.

R1-1 X]

An average of all error rates multiplied by 100 gives the index of APE. The V is the
standard deviation divided by the mean for each sample, then averaged for all samples
and expressed as a percentage. D is obtained by dividing V by the square root of the

number of times each sample is aged.

Results

A total of 197 chinook salmon possessing coded wire tags was sampled during
1992 and 1993 from Lake Michigan; 149 of these samples came from the sport
fishery and 48 from gill nets. The coded wire tag results indicated that 152, 37, and
8 of the samples were age 1, age 2, and age 3, respectively. An additional 168 (72
age 1, 74 age 2, and 22 age 3) coded wire tagged chinook were sampled from the
WDNR weir harvest (Table 1).

Aguragy

Of the lake-caught samples four age 1, four age 2, and one age 3 fish were
misaged, while one age 1, one age 2, zero age 3 weir harvested fish were misaged
(Table 1). Shown in Table 2, the accuracy of the vertebral ageing method did not
differ at the 0.05 level between any of the age classes and the method was equally
accurate for samples collected in the lake and at harvest weirs. A total of 365 coded
wire tagged chinook salmon were aged using the vertebral method with 11 (3%) being

incorrectly aged. Table 3 shows that three of the misaged samples were underaged

15

Table 1. Coded wire tag and vertebral age determination results for chinook salmon
sampled from Lake Michigan and harvest weirs.

 

 

Lake Caught Sample Size # Correct % Correct # Misaged % Error
(Sport Caught ’92-’93
and SV Steelhead ’93)
AGE 1 152 148 97.4 4 2.6
AGE 2 37 33 89.2 4 10.8
AGE 3 8 7 87.5 1 12.5
Weir Harvest
(Wisconsin ’93)
AGE 1 72 71 98.6 1 1.4
AGE 2 74 73 98.6 1 1.4
AGE 3 22 22 100.0 0 0.0
Combined Total 365 354 97.0 11 3.0

Table 2. Chi-square values for comparison ofvertebral age determination accuracy.

 

 

 

Lake Weir Between

Caught Harvested Locations
Age 1 vs. Age 2‘ 4.91 0.00
Age 2 vs. Age 3‘ 0.02 0.30
Age 1 vs. Age 3‘l 2.44 0.31

All Ages Combinedb 3.54

 

2 _
;X28,005,m=3,v=1 - 5731

Table 3. Difference in estimated ages between fish aged from vertebrae and

from Coded Wire Tag (CWT).

 

 

 

 

 

 

CWT Vertebrae Number of times age of Difference in age (yr)
aje age Vert<CWT I Vert>CWT of Vertebrae from CWT
l 2 5 +1
2 1 2 -1
2 3 3 +1
3 2 1 —1
Total 3 8

16

and eight of the samples were overaged. This distribution is not statistically different
(p<0.05) from that of an expected 50% ratio.

Praiaion

Of 100 fish aged independently by three individuals, only 8 samples showed
discrepancies in estimated age between readers (Table 4). The resulting index of

AFB was 2.17%, V was 2.81%, and D was 1.63%.

Discussion

Communication of the age of anadromous fishes is often accomplished with a
decimal system (Godfrey et a1. 1968, Seelbach and Beyerle 1984, Chilton and Bilton
1986), with the first digit being years spent in freshwater (stream) and the second
digit representing years in saltwater (ocean or lake). This system is necessary when
examining growth rates since most growth occurs during lake residency (Seelbach and
Beyerle 1984). Seelbach (1985) and Zafft (1992) documented age 0.0 and 1.0
naturally reproduced chinook smolts out-migrating from the Little Manistee and Pere
Marquette Rivers, respectively. The contribution of natural age 1.0 chinook salmon
tends to vary annually due to environmental conditions. Zafft (1992) estimated that
approximately 14% and 6% of the natural out-migrating chinook salmon were age 1.0
in 1989 and 1990, respectively.

For this paper I have not used this type of age expression for the following
reasons. First, only hatchery fish were used to validate the vertebral ageing method
and it is not known if hatchery fish express the same out-migration patterns as natural

chinook and a fairly low percentage of natural chinook salmon out-migrate at age 1.0.

17

Table 4. List of estimated vertebral ages and associated APE', V', and D' (Beamish
and Fournier 1981 and Chang 1982) from 3 readers for 100 chinook salmon

 

 

 

 

Reader

(n) 1 2 3 APE V D
(54) 1 1 1 0 0 0
(34) 2 2 2 0 0 0
(4) 3 3 3 0 0 0
(1) 1 1 2 0.3333 0.4330 0.25
(1) 2 2 1 0.2666 0.3464 0.20
(1) 3 2 3 0.1666 0.2165 0.13
(1) 2 1 2 0.2666 0.3464 0.20
(1) 2 1 2 0.2666 0.3464 0.20
(1) 2 2 1 0.2666 0.3464 0.20
(1) 2 1 2 0.2666 0. 3464 0.20
(1) 1 2 1 0.3333 0.4330 0.25

Average Percentage 2.17 2.81 1.63

' APE = Average Percent Error
' V = Coefficient of Variation
’ D = Index of Precision

18

Second, coded wire tags only express known-age in total years and do not convey any
information on amount of time spent in the river system. Third, vertebrae did not
possess obvious annular patterns that would allow for early life history determination.
Finally, only the total age of fish, which allowed for placement into cohorts, was
needed for the OTC marking project.

The importance of validating an age determination method for each species at
all possible ages is discussed by Beamish and McFarlane (1983). My research
demonstrates that age 1, 2, and 3 Lake Michigan chinook salmon can be accurately
aged with vertebrae. I have estimated 5 chinook salmon to be age 4 by their
vertebrae. Because no coded wire tagged age 4 chinook salmon existed in Lake
Michigan during the study period, I have not been able to prove the accuracy of
vertebral ageing on age 4 chinook salmon. However age 4 chinook salmon make up
a very small portion of the population. I encourage the continued collection of
vertebrae from coded wire tagged chinook salmon to check the accuracy of vertebral
age determination on age 4 chinook salmon.

The accuracy of vertebral age determination is not affected by the age of the
fish, as there were no statistical differences between any of the age classes. This is
not surprising since age determination accuracy problems are most often associated
with time of slow growth, such as the later years in long-lived fishes (Beamish and
McFarlane 1983). Slow growth is not a problem with chinook salmon which are fast
growing during their entire 34 year lake life cycle.

Problems associated with scale adsorption when ageing river harvested mature

chinook salmon by scales (Dan Anson, Personal Communication) were not apparent

19

when using vertebrae. The vertebral age estimates from river-harvested mature
chinook salmon were just as accurate as those made from lake-collected samples.
Since there were no accuracy differences between age classes or between lake-caught
and mature weir-harvested samples, all samples were combined resulting in one
estimate of vertebral age determination error to be 3 %.

Errors in estimated vertebral ages were made by both under- and over-
estimating true age. By examining the total number of under- and over-estimated
samples, no consistent bias was detectable. However, Richards et al. (1992) state
"we cannot assume that observed age is an unbiased estimate of true age because
observed ages must fall within the life span of the species. For example, due to
truncation in the age distribution, mean assigned age will overestimate true age for the
youngest fish and underestimate true age for the oldest fish. " In this study, both age
0 and age 4 fish do occur in the population and the possibilities did exist for age 1
fish to be underestimated as age 0 and likewise, age 3 fish could be overestimated as
age 4 fish. However, all errors of age 1 fish were overestimated and all errors of age
3 fish were underestimated, but the low number of misaged samples limited the
detection of any significant trends.

Richards et al. (1992) address how to incorporate ageing error estimates into
population analysis. I have not included this type of correction into subsequent
population analysis, as accuracy was good and age data were used in conjunction with
other data for individual samples. It will be assumed that the limited ageing errors

have random and unbiased effects on subsequent data analysis.

20

Testing precision of an age determination method does not have to be done on
known-age fish, as it is a measure of repeatability and does not convey anything about
accuracy. The 100 samples used to test vertebral ageing precision were randomly
selected and included age 1 through age 3 chinook salmon that were of both hatchery
and natural origin. This was done to mimic typical samples taken from Lake
Michigan waters. The resulting 2.17% APE and 2.81% V indicate that the vertebral
age determination method has a high degree of precision. As APE and V decrease,
the repeatability increases (Beamish and Foumier 1981 and Chang 1982). The D
expresses that an average of 1.63 % of the error is contributed by each sample. The
actual amount of error contributed by individual samples is expressed by D values for
each sample. These indices can be used to compare the precision of vertebral ageing
to other methods such as scale ageing.

Having determined that vertebral age determination of Lake Michigan chinook
salmon is a valid procedure does not imply that all studies of this population should
use the vertebral ageing method to estimate age. There are advantages and
disadvantages of verteme age determination when compared to the traditional scale
ageing method. The primary advantage of vertebral ageing arises when used in
conjunction with OTC studies. In this manner, both field and laboratory time is
saved by being able to use one structure and method to determine both the presence
or absence of OTC and estimate age. A second advantage deals with the ability to
estimate age on mature fish after they have returned to rivers for spawning. In this
situation, erosion of the annuli on scales occurs as the scales are reabsorbed making

ageing by scales difficult, while the structure and appearance of vertebrae remains

21

constant. Distinguishing vertebral annuli may also involve less guess work, as annuli
are broad bands, whereas identification of scale annuli involves recognition of slight
changes in circuli patterns. Both methods require some training; however, less
experience seems needed to become proficient at vertebral age determination.

Vertebral age determination also has disadvantages. 1) The use of vertebrae
necessitates sacrificing the fish: unlike scale ageing from which fish can be released
after sample collection. 2) Information on growth rates and early life history can only
be obtained from scales and not from vertebrae. This type of information may be
obtainable from vertebrae with further examination of internal zonation; however,
increased laboratory preparation would likely be required. 3) Vertebrae must be
stored frozen until time of lab examination and multiple readings require repetition of
the entire lab process. Scales on the other hand can be stored in thin envelopes and
multiple readings can be made from a single prepared slide.

As Chilton and Bilton (1986) discuss, the use of combinations of multiple
structures for ageing can reduce improper management of fish stocks. Whatever age
determination method or methods are used, periodic checks of the methods’ validity
should be applied (Beamish and McFarlane 1987). Depending on the desired data,

vertebrae represent a valid alternative for ageing Lake Michigan chinook salmon.

CHAPTER THREE

Oxytetracycline Detection and Quality of Marks in
Lake Michigan Chinook Salmon

Introduction

The ability to mass mark various species of hatchery fish with the antibiotic
oxytetracycline (OTC) has been demonstrated in several studies (Weber and Ridgeway
1962, Weber and Ridgeway 1967, Odense and Logan 1974, Koenings et al. 1983,
Bilton 1986, Lorson and Mudrak 1987, Bumguardner and Colura 1991). OTC can be
administered by injection, immersion, or through feeding. The most effective method
depends on the species of fish being marked. Likewise, the best structure for OTC
detection also varies by species.

Weber and Ridgeway (1963 and 1967), Odense and Logan (1974), and Bilton
(1986) examined marking salmonids (Oncorhynchus sp. and/or Salmo sp.) with OTC.
Weber and Ridgeway (1967) determined that an oral dosage of at least 250 mg of
OTC per kilogram of fwd at 2.0% of body weight for 4 or more consecutive days
produces a consistently distinguishable and permanent (at least 3.5 years) mark. OTC
is deposited in areas of osteogenesis and can be detected in scales, teeth, opercular
bones, ribs, otoliths, and vertebrae shortly after feeding (Weber and Ridgeway 1963).
Continued deposition of calcium on top of OTC marks and exposure to sunlight
reduces the ability for long-term detection of OTC in most of these structures.

Vertebrae are the best structure to use for long-term detection of OTC in chinook

22

23
salmon (Onchorhynhus tshawytscha) (Weber and Ridgeway 1967 and Odense and

Logan 1974).

Size of the fish at time of OTC administration can affect the quality of the
OTC marks. Poor marking success can occur if OTC is administered prior to the
development of calcified bone structure. Once the critical size where the onset of
calcification has been reached, mark quality increases as long as the dosage per body
weight remains constant (Odense and Logan 1974 and Bilton 1986). Depending on
the size of the fish at time of OTC administration and the total amount of OTC
consumed by each fish, the quality of OTC marks will be variable.

OTC and/or fin clips were used to mark all hatchery chinook salmon being
stocked into the Lake Michigan Basin, from 1990 to 1994, in an effort to distinguish
natural and hatchery fish. Appendix B contains a list of the marking patterns used by
each state agency. All OTC marking was done by feeding 350 mg OTC per kilogram
of feed at 2.0% body weight for 5 consecutive days approximately one month prior to
release (MDNR, interoffice communication). As with age determination (Beamish
and McFarlane 1983), the validity/quality of the marking methods being used must be
checked periodically on each population being studied. Likewise, to confidently use
data obtained for studies on the contribution of hatchery and natural chinook salmon
in Lake Michigan, the quality of the OTC marks must be established.

The objective of this chapter is to describe the methods used to detect the
presence or absence of the OTC marks and to determine the effectiveness/validity of
marking Lake Michigan hatchery chinook salmon with OTC. This objective included

examining the permanency of the OTC mark over time, establishing the percentage of

24

known hatchery fish not showing an OTC mark (marking error rate), and examining

the repeatability of OTC detection among readers.

Methods

5.9mm

Chinook salmon were sampled from the Lake Michigan sport fishery,
Michigan Department of Natural Resources (MDNR) Survey Vessel (SV) Steelhead
chinook salmon-diet-study gill nets, and from the Wisconsin Department of Natural
Resources (WDNR) weir harvest. Weight, length, sex and fin clip data, and scales,
vertebrae, and the noses from adipose fin clipped fish were collected. At least five
thoracic vertebrae were collected from below the adipose fin. Exposure of vertebrae
to direct sunlight was kept to a minimum and samples were stored frozen until lab
examination.

Win

The second vertebra from each sample was removed by cutting through the
cartilage rings. All flesh and cartilage were removed with a scalpel (11t 4 round tip
scalpel blades work best). It is important to remove the cartilage ring which separates
vertebrae. Prepared samples were placed in stacked tin weighing dishes and viewed
within 30 minutes of cleaning. This reduced exposure to light and prevented drying
of vertebrae. Each vertebra was placed on end, center hole vertical, on a black
background and covered with several drops of glycerin. The bone was then viewed in
a dark room through a dissecting microscope, magnification ranging from 15 to 50 X

(using 10 X oculars), under reflected ultraviolet light (365nm). The light used in this

25

study had an intensity of approximately 11,000 microwatts per square centimeter at a
distance of 15 cm (distance from bulb to bone). Except for the cleaning process,
these methods follow those described in Weber and Ridgeway (1963 and 1967).

The incidence and quality of OTC marks was recorded on an eight point scale.
This scale was broken down into four groups of marks: no mark (0), poor marks (1-
3), good marks (4—6), and a great mark (7). The samples were given a 0 rating if no
OTC mark was observed or if the reader harbored any doubt about the existence of an
OTC mark. Within the poor/ good categories mark quality was further broken down
into poor/good (minus), poor/good, and poor/good (plus) relating to the color
intensity of the OTC mark. The poor category describes marks that required some
degree of microscope focusing to detect mark. The poor minus rating was used when
the OTC mark was only visible in one focal position and may not have exhibited
yellow-green color. Good marks were detectable with the micros00pe slightly out of
focus, and appeared as a sharply defined yellow—green line. Great marks were
observable without the microscope and when viewed with the microscope appeared as
a wide fuzzy line. Some of the great marks also exhibited vertical rays extending
from the OTC ring towards the center hole. Multiple marks were recorded in order
of formation (inner ring first).

Classification of samples into respective cohorts was accomplished by vertebral
age determination, using the methods described in Chapter 2.

Mark Parmaaangy

OTC mark permanency was examined by observing the range in quality of the

OTC marks independently for the 1990 and 1991 cohorts over a two year period.

26
Only hatchery chinook salmon, those possessing an OTC mark and/or fin clip, were

used in the study. A weighted average of mark quality was used to compare quality
of marks in successive years.

Mark Ramntign

An OTC marking error rate was established as the percent of double marked
(both OTC and fin clipped) samples not showing an OTC mark. This was
accomplished by matching the OTC detection results with fin clip data. Since each
state used unique fin clips, a marking error rate was established for each state. A
lake-wide OTC marking error rate was determined by combining all state results into
one estimate. From this lake-wide estimate of OTC marking error, the percent of
non-marked hatchery chinook salmon was obtained by multiplying the lake-wide OTC
marking error rate by the number of chinook salmon marked only with OTC, then
dividing by the total number chinook salmon stocked. Fin clips were not known by
the OTC reader at the time of examination.

ili

A sample of 100 vertebrae was examined for OTC independently by 3 readers
and the OTC observations compared to examine the repeatability among readers.
Comparisons between readers were made with regard to 1) assigning mark quality
and 2) detection of OTC mark incidence. Results were expressed as percent

agreement among all three readers.

27

Results

Marki’mancncx
A total of 271 (146 at age 2 and 125 at age 3) known hatchery chinook salmon

from the 1990 cohort and a total of 462 (157 at age 1 and 305 at age 2) known
hatchery chinook from the 1991 cohort were examined for OTC mark quality

(Figure 4). The weighted average of mark quality, expressed on a scale of 0-7
defined earlier, for the 1990 cohort was 3.15 at age 2 and 3.0 at age 3. The
weighted average mark quality for the 1991 cohort was 3.69 at age 1 and 3.62 at

age 2. An additional 398 age 1 known hatchery chinook salmon from the 1992 cohort

were examined and had a weighted average mark quality of 4.14 (Table 5).

Table 5. Weighted average of the quality of OTC marks for
all hatchery chinook salmon in the 1990 - 1992
cohorts over time.

 

 

Age 1 Age 2 Age 3
1990 cohort -- 3.15 3.00
1991 cohort 3.69 3.62 --
1992 cohort 4.14 -- --
Mark geranrign

Table 6 shows the total number of chinook salmon sampled for each type of
fin clip and the number not showing an OTC mark. Fish from Indiana and samples
exhibiting non-possible fin clip patterns were excluded from subsequent analysis. A
total of 57 double marked chinook salmon were sampled from the 1990 cohort, three

(5.26%) of which did not exhibit a detectable OTC mark (Table 7). Likewise, 12

28

30

 

1990 Cohort
..... Age2 ..... w,a..§.,3.15. ...............

- Age3 wa = 3.00

 

 
    

 

 

10 ....................................................

 

_ 1991 Cohort
l Age 1 wa = 3.69
l Age 2 wa = 3.62

 

 

 

 

O 1 2 3 4 5 6 7

OTC Mark Quality Range

Figure 4. Range in quality and weighted average (wa) of OTC marks over a two year period for
the 1990 and 1991 Cohorts. (0 = no mark. 1-3 = poor. 4-6 = good. 7 = great).

29

Table 6. Correlation between fin clips and OTC marks.

 

 

 

 

1990 Cohort 1991 Cohort 1992 Cohort
I State Total # with I I State Total # with I I State Total # with I

Sample No OTC Sample N o OTC Sample N o OTC
Fin Clip Size Size Size
Used
Ad — CWT MI 21 1 MI 86 2 MI 122 5
Ad - CWT WI 22 0 WI 75 4 WI 73 3
LV 1 l IND‘ 15 10 6 4
RV MI 13 2 MI 41 6 6 0
LP 2 0 IND‘ 35 21
RP IND‘ 10 7 4 0 ILL 4 0
AdLV 1 1 IND' 6 6
AdRV [ND 6 4 3 2
AdLP IND‘ 4 3 IND‘ 5 5
AdRP IND‘ 1 0 1 0 IND‘ 1 1
BV IN D‘ 4 3 1 0 ILL 3 0
LPRV ILL 1 0
RPLV ILL 2 0
AdBV ILL 1 0
D ILL 1 0 [LL 5 2
DAd ILL 6 0
DRV ILL 2 0
DLP ILL l 0
DRP ILL
RMLV WI
LM

‘ Fish not marked with OTC

Table 7. Percentage of double marked chinook salmon not exhibiting an OTC mark.

 

 

 

 

1990 1991 1992
State Clip Cohort Cohort Cohort
MI Ad 4.76 2.33 4.10
RV 15.38 14.63 - -
WI Ad 0.00 5.33 4.11
ILL all clips 0.00 0.00 14.29
Cohort Total 5.26 5.66 4.76

TOTAL ACROSS COHORTS 5.22

30
out of 212 (5.66%) and 10 out of 210 (4.76%) double marked samples from the

1991 and 1992 cohorts, respectively, did not exhibit detectable OTC marks. Also
shown in Table 7 are the percent marking errors for each state. All three non-marked
samples from the 1990 cohort were Michigan stocked fish (3 of 34, 8.82%), while all
Wisconsin (22) and Illinois (1) double marked fish exhibit OTC marks. Of the 1991
and 1992 respective cohort samples, 127 and 122 were Michigan clips with eight
(6.29%) and five (4.10%) not showing an OTC mark, 75 and 73 were Wisconsin
clips with four (5.33%) and three (4.11%) not showing an OTC mark, and 10 and 14
were Illinois clips with zero (0%) and two (14.30%) not showing an OTC mark. A
total of 25 out of 479 (5.22%) double marked fish did not exhibit OTC marks.

R ili

Table 8 shows comparisons between samples for which discrepancies in OTC
mark quality between two or more readers occurred. Forty-two of the 100 (42%)
samples were assigned different OTC mark qualities by at least two readers, with four
(4%) of the discrepancies disagreeing about the presence of an OTC mark. Overall,

there was 96% agreement among readers in the detection of OTC marks.

Discussion

The mass marking of hatchery chinook salmon in Lake Michigan with OTC
provided an opportunity to independently examine hatchery and natural chinook
salmon populations. Ridgeway and Weber (1963 and 1967) demonstrated that Pacific
salmon can be effectively massed marked by feeding OTC. However, prior to using

this OTC data in Lake Michigan management decisions, the validity/effectiveness of

31

Table 8. Oxytetracycline mark quality ratings for samples which multiple readers assigned different

OTC mark qualities.

 

p—a

Reader

2

Us)

 

WWMNWQb’thNI—WO‘ch-FUIMNO‘MM

2.

3a
4a

##shHNahJiNNbbJQMbO‘bWhM-h

A¥MHNQMWWN$NQ¥MQM$MQ$

Difference

H

Reader

2

U)

 

5M$NMWQ§$MQW¥~WWA~HQN

38

4a

5‘

#M#wthM#O\&#MH#bMOOO\w

4a

5.

UGMW¥¢Q$¥me¥>O$MMOCQW

4a

5‘

Difference
Difference

Difference

’ Indicates the presence of a double mark (2 OTC rings). The first number is the inner ring.

32
this marking method for Lake Michigan chinook salmon must be established. Errors

in the OTC data could result from mark deterioration over time, failure to adequately
mark all hatchery chinook salmon, and poor repeatability among readers to detect
OTC marks.

Examination of OTC mark quality over time should be done within cohorts
and examined only for shifts in over-all mark quality (Figure 4). Due to the
subjectivity of assigning OTC mark quality, slight changes in individual quality
categories do not have any management implications. The use of a weighted average
of mark quality for each cohort reflects the over-all permanency of OTC marks over
time. The weighted average of OTC mark quality for the 1990 and 1991 cohorts
decreased as the fish aged 1 additional year by 0.15 and 0.07, respectively. This
Slight decrease in mark quality does not change the overall quality rating, indicating
that the OTC marks do not degrade significantly over time in terms of management
implications. This is supported by Weber and Ridgeway (1967) where OTC marks
were still detectable after 3.5 years. The 1993 results of OTC mark quality weighted
averages across cohorts, shown in Table 5, indicate that the 1992 cohort (4.14)
received higher quality OTC marks than did the 1991 (3.62) and the 1990 (3.00)
cohorts.

The percentage of hatchery chinook salmon not exhibiting an OTC mark
(shown with an OTC mark quality of 0 in Figure 4) does not represent a true OTC
marking error rate. Rather, this is representative of both the hatchery chinook salmon
that received only a fin clip and no OTC and those chinook salmon that were double

marked but failed to exhibit an OTC mark.

33

External marking with fin clips provides a means of identifying the state of
origin and establishing marking error rates. However, fin clips are not a fool-proof
method of identifying fish. Both the regeneration of clipped fins and the natural loss
of fins do occur, but the rates of regeneration and loss are unknown (Fry 1961). The
regeneration of all fin clips on a single fish does not adversely impact this OTC mark
evaluation study. However, the regeneration of a single fin on multiple fin clipped
fish or the natural loss of a non-clipped fin can result in the mis-identification of a
fish in terms of state and cohort of origin. The possibility of human marking error
also exists with the administration of incorrect clips. This mis-identification of fish
origin due to incorrect fin clips, regeneration, and natural fin loss could have effects
on this study. Samples were placed into cohorts based on vertebral age estimates and
it is possible that some samples were misaged.

The fact that errors do occur is evident in Table 6, as 25 samples were placed
into non-possible fin clip/cohort compartments. For example, 4 samples were
collected with a right pectoral (RP) clip and assigned to the 1991 cohort based on
vertebral age; this was not a fin clip used for that cohort. Several possible types of
error exist: 1) the samples could have been misaged as both the 1990 and 1992
cohorts used RP fin clips, 2) these samples could have been RP, left ventral (LV) fin
clip fish which regenerated LV fins, and 3) the RP fin could have been clipped by
mistake. These types of errors are easily detected when expressed as non-possible
combinations. These samples were not used in the marking error determination.

Errors could also occur and not easily be detected if the errors mimic used fin

clips. For example, Indiana marked chinook salmon with fin clips only and did not

34
use OTC, but 14 out of 35 left pectoral (LP) samples from the 1991 cohort exhibited

an OTC mark. This existence of OTC in samples that should not have OTC marks
can be explained by the regeneration of dorsal (D) fins from Illinois 1991 Stocking of
D, LP and OTC marked chinook salmon. An alternate explanation involves the
existence of false-positive OTC marks. However, the appearance and location of
OTC marks is very distinctive and uncontrolled exposure of chinook salmon to OTC
is not likely. This type of non-detectable error could be responsible for some of the
fin clipped samples not showing OTC if a non-hatchery chinook salmon naturally lost
fin(s).

The resulting estimates of each state’s OTC marking error (Fable 7) can be
used to monitor OTC marking effectiveness of individual states. Obtaining an
adequate sample size of each type of fin clip or each state’s fin clips is difficult due to
low numbers marked. For example, I only sampled 1, 10, and 14 Illinois fin clip
chinook salmon from the 1990, 1991, and 1992 cohorts, respectively. If an adequate
sample size of each state’s fin clips is obtained to provide an accurate estimate of
OTC marking error for each state (not the case with this study), then it would be
appropriate to adjust the lake-wide OTC marking estimate in respect to the numbers
of OTC-only marked chinook salmon stocked by each state. If the sample size of
each state’s fin clips is small, it is more appropriate to make the assumption that all
states obtained similar rates of OTC marking effectiveness and to lump all samples
into one estimate. It is not possible to determine and include samples that did not
take an OTC mark and regenerated fin clips with this type OTC marking evaluation.

The resulting lake-wide OTC marking error estimates were similar between cohorts

35
3.74%, 3.32%, and 3.52%. These data agree with marking error estimates by Elliott

(1994), where 2-5% of age 0 hatchery chinook salmon, sampled prior to and after
Stocking, did not exhibit OTC marks.

In accordance with Elliott’s (1994) findings, I used an estimate of 5% OTC
marking error for correcting hatchery and natural percentages. By using a slightly
inflated OTC marking error rate rather than that actually observed, some correction
for non-recognition of hatchery chinook salmon that regenerated fin clips may be
taken into account. This also produces a conservative estimate of the contribution of
naturally produced chinook salmon.

Correction of hatchery and natural estimates for 5 % non-marked hatchery
chinook salmon was accomplished by multiplying the number of marked (fin clip
and/or OTC mark) fishby 0.05. That product is then subtracted from the number of
unmarked fish and added to the number of marked fish. In this manner, the
correction method does not increase hatchery and decrease natural percentages
each by 5 %.

The relatively poor agreement (58%) between different readers assigning OTC
mark quality should not be a concern as the mark quality scale is very subjective.
The discrepancies in OTC mark quality ratings were small; with the exception of one
sample, ratings differed by only 1 or 2 quality categories. The important aspect of
repeatability among readers is the ability to distinguish OTC marks. The 4 samples
where the incidence of an OTC mark differed among readers were all rated as ”poor
minus" or "poor” quality marks. Experience and patience examining vertebrae will
help reduce non-detection of poor quality OTC marks. Samples not showing an OTC

mark could be saved and reexamined to reduce the non-detection of poor quality marks.

CHAPTER FOUR

Percentage of Hatchery and Natural Chinook Salmon in the

Lake Michigan Sport Fishery

Introduction

Since the introduction of chinook salmon (Oncorhynchus tshawytscha) into the
Great Lakes in 1967 several studies have documented the occurrence of successful
natural reproduction (Rybicki 1973, Taube 1974, Carl 1980 and 1982a, Seelbach
1985, Powell and Miller 1990, Zafft 1992, and Jennings 1992). The studies by Carl
( 1982a), Zafft (1992), and Seelbach (1985) estimated numbers of natural smolts out-
migrating from several streams in Michigan. There have been no lake-wide estimates
of the contribution or comparison between hatchery and natural chinook salmon in
Lake Michigan.

The identification and quantification of the hatchery and natural components of
the chinook salmon population is difficult after out-migration from rivers has
occurred. It is not economically feasible to externally mark the millions of the
hatchery chinook salmon being stocked each year. Hankin (1982) discussed
estimating escapement of hatchery and wild chinook salmon by marking a consistent
proportion of the hatchery fish being stocked, from which the relative contribution of
hatchery and wild fish can be estimated regardless of year and location of spawning
return. By marking only a portion of the hatchery fish stocked the non-marked

hatchery fish are still non-distinguishable from the natural fish, making comparisons

36

37
between hatchery and natural fish difficult. Hankin’s method also requires the

assumption of equal mortality/survial for hatchery and natural fish. Scale circuli
patterns of West Coast chinook salmon and Great lakes Steelhead (Oncorhynchus
mykiss) have been successfully used to distinguish hatchery and wild (natural)
individuals (Sneva and Knudsen 1989 and Seelbach and Whelan 1988). However Carl
(1982b) was not able to differentiate lake Michigan hatchery and natural chinook
salmon by scale patterns. In an attempt to distinguish hatchery and natural chinook
salmon, all hatchery chinook salmon stocked in the Lake Michigan basin from 1990
to 1994 were marked with the antibiotic oxytetracycline (OTC) and/or fin clips.

During 1992 and 1993 I sampled lake Michigan chinook salmon to determine
the occurrence of hatchery and natural chinook salmon in the sport fishery. The
objectives of this paper are to: 1) describe the contribution of hatchery and natural
chinook salmon to the Lake Michigan sport fishery, and 2) examine relative survival
rates for hatchery and natural chinook salmon during lake life stages.
Methods

Sampling Pam Dasgripg'gn

Three Michigan ports were used for sample collection: Grand Haven,
Ludington, and Leland (Figure 5). These ports incorporate areas that have different
amounts of natural reproductive input from nearby streams and also have sport
fisheries that catch high numbers of chinook salmon. Grand Haven, located in the
southern part of the lake and at the mouth of the Grand River, represents an area with
low natural reproductive input (Jennings 1992). Ludington, located in the central

portion of the lake at the mouth of the Pere Marquette River, is in an area with high

 

Wisconsin
Michigan
Grand Haven

{9
S

5 Ludington
2
i

Illinois ................

Indiana

 

 

 

Figure 5. lake Michigan and location of sampled ports.

39
natural reproductive input (Carl 1982a and Zafft 1992). Both Grand Haven and

Ludington have consistently been one of the top chinook salmon catch ports in
Michigan water’s (Rakoczy 1992). Leland, located in the northern part of the lake, is
in an area that has no major rivers and low natural reproduction potential, but has
traditionally had the highest chinook salmon catch of all lake Michigan Ports
(Rakoczy 1992).

W

Samples were collected from the sport fishery, primarily at the cleaning
stations where charterboat crews clean their fish (Table 9). With the permission of
the captains, all observed chinook salmon were sampled. Weight, length, and fin clip
data, and scales were collected prior to cleaning. Sex and maturity data, general
health condition, and a section of vertebrae were collected after cleaning. The name
of the captain or boat name was recorded. Weight was recorded to the nearest 1/4
pound on fish over 4 lbs and to the nearest 25 grams on fish under 4 lbs. Total
length (end of pinched caudal fin) was recorded in millimeters. The incidence,
including type, of fin clips was recorded. Noses from adipose fin clipped fish
(containing coded wire tags) of all species were collected and delivered to the
Michigan Department of Natural Resources. Scale samples were taken just posterior
of the dorsal fin just above the lateral line. In 1992 scales were collected from all
chinook salmon and in 1993 only from adipose fin clipped chinook salmon.
Observations of health condition followed a shortened procedure of Geode (1988) with
the condition of the kidney, liver, spleen, hindgut, and any gross abnormalities being

noted. A section of at least 5 thoracic vertebrae, from below the adipose fin, was

40
removed with shears. Vertebrae samples were placed in individually labeled bags and

frozen as soon as possible. Exposure of vertebrae to direct sunlight was minimized.

Table 9. Primary cleaning stations targeted for sample collection.

 

Grand Haven Chinook Pier Sportfishing. 301 N. Harbor Dr.
Bolhouse Charters. Holiday Inn M-21.
Ludington Abrahamsons Marina. S. Washington St.

Ray’s Auto Marina. 801 S. Washington St.
Leland Carlson’s Fish Market. Leland.

In an effort to collect additional samples, during 1993, "muck buckets"
(approximately 10 gallon garbage cans) were placed at the cleaning stations not being
personally attended. A sign was attached reading: ”Michigan State University
chinook salmon research. Please place all chinook salmon carcasses in green tub.
Tub is checked daily." These buckets were checked several times a day. Data
including total filleted length, sex and maturity, general heath condition, and fin clips
(if possible) were recorded, and a section of the vertebrae collected. It was observed
that not all chinook salmon carcasses were placed in these buckets; therefore all
samples collected in this fashion were distinguished on the data sheet as being not
representative of the harvest.

During 1992 sampling occurred mainly on weekends whereas sampling in 1993
occurred on all days. Ports were sampled proportionately to the success of chinook
salmon fishing. Grand Haven was sampled from May - September, Ludington June -

September, and Leland July - August. The number and pattern of sampling days at

41

each port were adjusted continually in attempts to keep total sample size
approximately equal between ports.

mum

Age determination and detection of OTC marks were accomplished using the
techniques described in chapters 2 and 3, respectively. It was determined that the
vertebral ageing method misages 3% of the samples and the ageing errors occur
randomly with no bias effects on subsequent data analysis. Data presented on the
percentage of hatchery and natural chinook salmon have been corrected for a 5%
occurrence of non-marked hatchery chinook salmon (see chapter 3).

tis ' An i

All tests were done individually for each cohort. Tests for seasonal variation
within locations and annual variation between locations were done with chi-square
contingency tables (p < 0.05) (Gill 1978). With the use of chi-square contingency
tables it was determined that a sample size of 400 fish per cohort per port per year
was needed to detect a 5 % change in the percentage of hatchery and natural chinook
salmon.
Results

Due to the low catch rates of chinook salmon during 1992 and 1993 the
desired sample size was not obtained. In 1992 and 1993, a total of 753 and 1,374
chinook salmon were sampled, (201 and 615 from Grand Haven, 249 and 697 from
Ludington, and 303 and 65 from Leland), respectively. Sampling effort was not
equal for the two years, with 41 sampling days in 1992 and 83 sampling days

in 1993.

42

Daily sample sizes ranged from 0 - 94 by cohort, with averages of 5 to 20 and
modes of 1 or 2 (Appendix C). Due to the frequency of small daily sample size no
statistically significant differences in the daily percentages of hatchery and natural
chinook salmon were detected (Figures 6a and b). Monthly totals were compared
within each location, samples were combined for each calendar month
sampled, and no statistically significant differences were detected. Samples were
further combined to produce one estimate of percent hatchery and natural at each port
and for each cohort. Shown in Appendix D are the daily, monthly, and yearly
percentages of natural chinook salmon for each cohort by port in 1992 and 1993.

Table 10 lists by port yearly total sample size, percentage of hatchery and
natural, and associated standard errors of hatchery and natural chinook salmon at:
ages 2 and 3 for the 1990 cohort, ages 1 and 2 for the 1991 cohort, and age 1 for the
1992 cohort. The percentages of naturally reproduced age 1 and age 2 chinook
salmon in 1992 and 1993 ranged from 20-39% and were not statistically different
within or between ports and cohorts (Figure 7). The 1990 cohort age 3 Ludington
sample did significantly differ (p<0.05) from both the 1990 cohort age 3 Grand
Haven sample (Figure 7) and the 1990 cohort age 2 Ludington sample. A limited
sample of six age 3 chinook salmon, from Leland in 1993, provided an uncertain
estimate of 67% from natural reproduction that was not significantly different from

other results.

43

10° 1992 Age 1

80—-

Daily Sample Size (#)
I

20+— I I.

 

20 40 60 80 100
Percent Natural

D.)
O

1992 Age 2

1’}.
1

8
1

Daily Sample Size (#)
I I

M

I—

 

 

I I
_ I I I
I I I
0 T 1 1 1 1 T
0 20 40 60 80 100
Percent Natural

Figure 6a. Daily percentage of chinook salmon from natural reproduction with respect to sample
size for all ports combined at age 1 and 2 during 1992.

U)
C

Daily Sample Size (#)
8 8

81288

Daily Sample Size (#)

u

 

44

 

g—s
O
I

 

 

1993 Age 1
l I
_ I
_ ' ......
_ ‘-
— -' I l~ i I I
I I I I
I I
I
I I
~ I . I l I I
I I I
n ' I
l l l l l
0 20 40 60 80 100
Percent Natural
1993 Age 2
" I
l I
_ I
P I
f ’. '
_ I .’
I I I
" I
_| ., H. M .
I I
_ . I
! 1 1 1 1 F
0 20 40 60 80 100
Percent Natural

Figure 6b. Daily percentage of chinook salmon from natural reproduction with respect to sample

size for all ports combined at age 1 and 2 during 1993.

45

Table 10. Summary ofthe percentage of hatchery and natural chinook salmon at Grand Haven,
Ludington, and Leland during 1992 and 1993.

 

 

 

 

 

 

 

 

 

 

 

1990 Cohort 1991 Cohort 1992 Cohort
Age 2 Age 3 Age 1 Age 2 Age 1
Grand Sample Size 80 90 97 165 339
Haven Percent Hatchery 67 61 79 73 67
Percent Natural I 33 39] I 21 27]
Standard Error 5.2 5.1 4.1 3.5 2.6
Ludington Sample Size 93 147 130 268 242
Percent Hatchery 63 46 69 71 68
Percent Natural I 37 54] r 31 29]
Standard Error 5.0 4.1 4.0 2.8 3.0
Leland Sample Size 64 6 280 22 37
Percent Hatchery 64 33 70 68 78
Percent Natural L 36 67] I 30 32J

 

 

Standard Error 6.0 19.2 2.7 9.9 6.8

46

50

 

- 1992

 

 

 

AGE 1 AGE 2
GRAND HAVEN LUDINGTON LELAND

 

80 —-

70 —

% NATURAL
' |

  

   

 

 

 

AGE 1 AGE 2 AGE 3
Figure 7. Percentage of naturally reproduced chinook salmon in the 1992 and 1993 sport

harvest at Grand Haven, Ludington, and Leland. Error bars represent
+/- 1 standard error.

47
Discussion

I have refered to Great Lakes naturally reproduced chinook salmon as
"natural" rather than "wild”. This was done for two reasons: 1) Great Lakes
chinook salmon are an introduced species which has become naturalized, and 2) the
methods used in this study cannot distinguish how many generations removed natural
fish are from hatchery produced parents. With only 68 generations of chinook
salmon having existed in the Great Lakes, combined with the likely continual cross
mating of hatchery and naturally produced fish, none of the natural fish are very far
removed from hatchery produced parents. Thus, the terminology of ”natural”
chinook salmon seems more appropriate than ”wild" when discussing Great Lakes
naturally reproduced chinook salmon.

Carl’s (1982b) inability to distinguish hatchery and natural Lake Michigan
chinook salmon using scales is not surprising due to the short period of time where
hatchery and natural chinook experience different environmental conditions.
However, Carl (1982b) did not have a sample of mature known natural chinook
salmon, but rather assumed that samples from the Pere Marquette River were natural
fish. Repetition of Carl’s study using the scales collected during my study where both
known hatchery and natural scale samples are available may prove beneficial.

WW

Catch per effort of chinook salmon at different locations in Lake Michigan
varies seasonally (Elliott 1994 and Rakoczy 1992). Chinook salmon were consistently
harvested at Grand Haven from May through August, likewise in Ludington from

June through August. Chinook salmon catches during June at both of these ports

48

were generally low. Chinook salmon have consistently contributed to the Leland
sport harvest from mid-July to Late August (Rakoczy 1992).

The ability to determine daily changes in the percentage of hatchery and
natural chinook salmon is limited by small daily sample sizes. As the daily sample
size increased above 10, most percentages of natural chinook salmon fell within the
95% confidence interval of the yearly mean (Figures 6a and b). However, a much
larger sample size was needed to produce confident estimates. The monthly
proportions of hatchery and natural chinook salmon did not change over the sampling
period. It was hypothesized that a change in the percentages of hatchery and natural
chinook salmon during the season might occur for age 3 fish as mature fish home in
on their native streams or stocking locations. This change was not apparent. The
lack of change over the season combined with increased precision as sample size
increases allowed for the combination of all samples within a port for each cohort.
Comparisons between ports based on these combined yearly samples revealed an
increased contribution of age 3 natural chinook salmon at Ludington over Grand
Haven.

Cohort Trends

Elliott (1994) used the OTC/fin clip marks to estimate that at age 0, 36%
(south) and 56% (north) of the 1990 cohort and 26% (south) and 44% (north) of the
1991 cohort were from natural reproduction, and at age 1, 18% (south) and 35%
(north) of the 1990 cohort were from natural reproduction. Elliott’s south samples

were collected from St. Joseph to Muskegon, Michigan, and the north samples were

49

collected from Pentwater to Leland, Michigan. These can be compared to my Grand
Haven and Ludington samples, respectively.

Combining Elliott’s age 0 and age 1 data with my age 2 and age 3 data for the
1990 cohort, the percentage of natural chinook salmon over an entire life cycle was
available (Figure 8). For both locations the age 0 percentages are similar to the age 3
percentages; however there is a significant drop to the age 1 and age 2 percentages.
This decrease in the percentage of natural fish at age 1 and 2 is probably due to
mixing of fish throughout the lake. With most natural reproduction occurring in
Michigan rivers, age 0 natural chinook salmon are concentrated on the Eastern side of
Lake Michigan. By the time the chinook salmon are age 1, they have time to move,
and the percentages of natural fish drop to 18% and 35% at age 1 and 33% and 37%
at age 2. The 18% percent natural fish for the Grand Haven age 1 sample is
probably artificially low due to the disproportional collection of adipose fin clipped
(hatchery) by anglers (Rob Elliott, personal communication). The age 3 chinook
salmon are sexually mature and are homing in on their native streams/ stocking
locations. As a result, the percentages of age 3 chinook salmon from natural
reproduction match those observed at age 0. This same trend, although not complete,
is also prevalent in the 1991 cohort at ages 0, 1, and 2 (Figure 8).

Relative Survival

If differential survival of hatchery and natural chinook salmon occurs during
the lake life stages corresponding shifts in the percentage of hatchery and natural
chinook salmon should be apparent. Due to the migration and mixing trends over the

life cycle of a cohort, annual differential survival could not be determined. However,

50

 

- 1990 COHORT

% NATURAL

       
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1990 (AGE 0) 1991 (AGE 1) 1992 (AGE 2) 1993 (AGE 3)
GRAND HAVEN LUDINGTON

1991 COHORT

8

% NATURAL
8

N
C

       

1991 (AGE 0) 1992 (AGE 1) 1993 (AGE 2)

Figure 8. Yearly percentages of naturally reproduced chinook salmon in the 1990 and 1991 cohorts
at Grand Haven and Ludington. Bars with dashed lines represent data from Elliott (1994).

51

differential survival over a cohort’s entire lake life cycle was examined by comparison
of age 0 and age 3 percentages of hatchery and natural chinook salmon. With the age
0 natural percentages (37% and 56 %) not being significantly different than the age 3
natural percentages (39% and 54%) for the 1990 cohort at both Grand Haven and
Ludington, it appears that no differential survival between hatchery and natural
chinook salmon occurs during the lake life stages.

Further examination for the possibility of differential survival was
accomplished by examination and determination of the origin of chinook salmon
showing clinical signs of bacterial kidney disease and will be discussed in chapter 5.

Ngeral Contribeg’en

By examining a cohort over an entire life cycle it became clear that sampling
at one point in time is not sufficient to answer all questions. If a single estimate of
the lake-wide contribution of naturally reproduced chinook salmon is desired it should
be obtained by sampling the age 1 and age 2 fish. The data collected in this project
were only collected from the Eastern side of Lake Michigan. As a result it requires
some extension of the data to estimate lake-wide contribution. It is acceptable to
estimate that 35% of the 1990 cohort, 29% of the 1991 cohort, and 32% of the 1992
cohort are contributed by natural reproduction to the Eastern Lake Michigan sport
fishery. These reduced estimates from the estimates at age 0 and age 3 indicate the
immigration of hatchery chinook salmon and/or the emigration of natural chinook
salmon at age 1 and age 2. Making the assumption that hatchery and natural chinook
exhibit the same in-lake behavior, it may be reasonable to extend the estimate of the

contribution of natural chinook salmon to Eastern Lake Michigan to the entire lake.

52

With significant differences occurring in the percentages of age 3 natural
chinook salmon between ports it is not appropriate to use age 3 fish for lake—wide
estimates. Rather these samples can be used to estimate the regional contribution of
natural chinook salmon. Either the age 0 or age 3 year classes can be used to
estimate regional contribution of natural chinook salmon. However, age 3 fish must
be used if samples are collected though the sport harvest. Only the 1990 cohort has
been sampled at age 3, with 54% of the Ludington area and 39% of the Grand Haven

area age 3 sport harvest of chinook salmon being contributed from natural

reproduction.

Using the premise that survival of hatchery and natural chinook is not different
throughout the entire life cycle and the premise that estimates of natural reproduction
at age 1 and age 2 from the Eastern portion of Lake Michigan are applicable to the
entire lake, estimates of the number of naturally produced chinook salmon entering
the lake can be made. Estimates are obtained by multiplying the number of hatchery
chinook stocked by the percentage of natural fish estimated in the lake population,
then dividing by the percentage of hatchery fish in the lake population. The resulting
estimate represents the number of stocked equivalent natural fish entering the Lake
Michigan ecosystem. Using this technique, approximately 3,838,500 and 2,549,800
natural chinook salmon entered Lake Michigan in 1990 and 1991 respectively.

These estimates of natural reproduction far exceed the estimates for the
streams in the lower peninsula of Michigan made by Car] (1982). This could
represent a significant contribution of natural chinook salmon from other streams.

However, it should be kept in mind that the current Lake Michigan chinook

53

population is very different than that it was even 5 years ago. As a result, my
estimates of the contribution of natural reproduction during 1990-1993 should not be
assumed to be constant or applicable for the entire time of existence of chinook
salmon in the Great Lakes. My estimates do represent a current significant
contribution of natural chinook salmon to the Lake Michigan ecosystem. Future
Great Lakes fisheries management strategies should account for and include the

contribution of natural chinook salmon in the decision making process.

CHAPTER FIVE

Occurrence of Bacterial Kidney Disease and Comparison of
Sampling Techniques Using Hatchery and Natural Chinook Salmon

Introduction

The examination of many population characteristics of Lake Michigan hatchery
and natural chinook salmon is possible with the existence of an extensive OTC/ fin clip
marking program from 1990 through 1994. In addition to samples collected for
examination of the percentage of natural reproduction discussed earlier in chapter
four, chinook salmon were being collected by the Michigan Department of Natural
Resources as part of a diet-study. This study used gill nets as the sampling method
and allowed for several additional aspects of Lake Michigan hatchery and natural
chinook salmon population characteristics to be examined without additional field
sampling. The gill net sampling permitted a comparison of sampling methods (gill
net and sport harvest) and determination of the origin of chinook salmon exhibiting
clinical signs of Bacterial Kidney Disease (BKD).

Gill nets have been extensively used as an efficient and relatively non-biased
method of sampling fishery resources (Nielsen and Johnson 1983). However,
depending on the type of data desired other sampling techniques maybe more
appropriate. Comparison of similar data sets collected with different types of
sampling techniques provides an opportunity to examine assumptions regarding the

representitivness of various sampling techniques.

54

55
Bacterial Kidney Disease (BKD) was determined to be the cause of death for

high numbers of chinook salmon in Lake Michigan from 1988 to present (Nelson and
Hnath 1990). BKD occurs in natural systems in low concentrations (Johnson and
Hnath 1991), but is believed that intensive culture of salmon in hatcheries has
increased the incidence of BKD in chinook salmon. As a result, BKD may affect
hatchery chinook salmon at higher rates than natural salmon.

The objectives of this paper are to briefly compare the percentage of hatchery
and natural chinook salmon: 1) estimated by gill nets and the sport fishery, and 2)

showing clinical signs of BKD.

Samples were collected from the sport fishery at Ludington and Pentwater and
from gill nets set off of Pentwater. The location of sport fishery fishing effort is
highly variable, but generally occurs between Little Point Sable, to the south, and Big
Point Sable, to the north, covering approximately 40 km of shore line. Gill net
sampling, conducted by the Michigan Department of Natural Resources Survey Vessel
(SV) Steelhead, occurred at various distances off shore in the Pentwater vicinity,
within the area targeted by the sport fishery. Due to the highly mobile and migratory
patterns of chinook salmon it was assumed that both collection methods had equal

probability of encountering the same groups of chinook salmon.

56
Field Methgs

Chinook salmon from the sport fishery and gill nets were sampled for weight,
length, sex and maturity, and general health condition; also fin clip data, scales, and
vertebrae were collected. At least five thoracic vertebrae were collected from below
the adipose fin. Observations of health condition followed a modified procedure of
Geode ( 1988) with the condition of the kidney, liver, spleen, hindgut, and any gross
abnormalities being noted.

2W

Procedures for age determination and the detection of OTC marks followed

those described in chapter 2 and chapter 3.

Results

A total of 117 clinically sick chinook salmon was collected in the MDNR gill
nets. No significant differences existed in the percentage of hatchery and natural fish
between the total gill net sample and the clinically sick gill net sample. Comparison
between the percentage of hatchery and natural chinook salmon with clinical signs of
BKD shows the only significant difference occurs in the 1991 cohort at age 1
(Table 11). The proportion of sick to heathy chinook salmon increased within each
cohort over time; the 1992 cohort (age 0 to age 1) increased from 0 to 12 percent, the
1991 cohort (age 1 to age 2) increased from 6 to 33 percent, and the 1990 cohort (age
2 to age 3) increased from 18 to 33 percent.

A total of 223 and 657 chinook salmon were sampled from the
Ludington/Pentwater sport harvest in 1992 and 1993 respectively. Similarly, 214 and

547 chinook salmon were collected in the Pentwater area by gill nets. Table 11 lists

57

 

489.com £32.sz MZQE .3 conga”... 29» moEEww .
.585. 38:30 .0053: Butane: Co 85:38 new a .8 3:88 9 nomads coon 25a mommaoeom g

 

:33... a 32 3 822.. .23: 82° 3

 

 

w e R c an a can 5:. .83.;— $8829.
$222. a. 13.3 3.5.. e22... a... .o E
a o R S a 2 SE 53 conga: omuaoeom
mm 3. 2. SN a: Locum gauge
2 am a 8 an 22:32 i
we 2. 82 9. Na bemoan &
G a 2. .w m . 2 25 038% new 28:50
3 3 3. .3 ad 3 Sam usage
an 2 mm 2 mm 3 .8932 a.
8 mm «A 8 S. 9. bones: a.
2:. 3. ME .42 S . S 3; 038$ .38.
.oz :5
3“ ma 3. S. 3 5cm magma
mm mu 3. 2 S .822 a.
me E a. 3. 8 55:5 e
as. I n MEN .5 2: mo SE 295$
33:: team
em? a a? N owe. _ a? m om< N a?

:25 Na:

tosoU 2mg

 

20:00 ama—

 

éo: Em 33¢. 36 E BEES

nos—8 goon—Eu no; 33233 new .83 E» 33... 36 .5823 goon—Eu AMZQZV moocsomom .9332 Ho Eva: uaoQ
cmeBZ 3322. to? no; “83895 mag new N93 2: So: cos—em xoano 22:3: one D0223 Lo momma—32$ .: 2an

58

for each sampling technique the number of samples in each cohort, the respective
percentages of hatchery and natural, and standard errors. Estimates of the

contribution of natural chinook salmon by the two sampling techniques are nearly
identical with no significant differences existing in the percentage of hatchery and
natural chinook salmon between the two sampling techniques, except for the 1990

cohort at age 2.

Discussion

Chinook salmon sampled from the sport fishery exhibited a low incidence of
clinical signs of BKD. This was probably due to the reduced feeding activity of fish
with advanced stages of BKD. Contrary to the sport fishery sample the ”non-biased"
gill sample of chinook salmon exhibited a high incidence of BKD. Only the samples
collected by gill nets were used to examine the occurrence of BKD in hatchery and
natural chinook salmon. The apparent difference in the rate of occurrence in BKD
between sport caught and gill net samples raises the question of the representativeness
of the sport fishery sample.

With the observed difference in the occurrence of BKD positive samples
collected in gill nets and in the sport fishery it is appropriate to first discuss the
effects of BKD on hatchery and natural chinook salmon prior to discussing the
comparison of sampling methods. The occurrence of BKD causing bacteria, Reni
salmonarium (R.S.), in chinook salmon is not equal to the rate occurrence of clinical
signs of BKD. Fish can harbor R.S. and not show clinical signs of BKD (Johnson
and Hnath 1991). However, it was assumed that the rate of clinical signs of BKD

reflects the rate of occurrence of R.S. for both hatchery and natural chinook salmon.

59

Hatchery chinook salmon may exhibit a slightly higher occurrence rate of
BKD at age 0 and 1 in comparison to naturally produced chinook salmon. However,
the difference is small and probably not significant in terms of management decisions.
At age 2 and 3 both hatchery and natural chinook salmon exhibit the same levels of
BKD occurrence. Although the occurrence of BKD is equal between hatchery and
natural chinook salmon, it appears that the rate of occurrence increased over the life
cycle of the fish with over 30 percent of the age 3 samples showing clinical signs of
BKD. With respect to objective 4, to determine if survival rates of hatchery and
natural chinook differ, it appears that the significant mortality factor in BKD does not
affect the survival of hatchery and natural chinook salmon differently.

Since there is not a difference in the occurrence of BKD between hatchery and
natural chinook salmon the selection of healthy fish by the sport fishery should not
affect the subsequent comparison of sampling techniques. All but one of the
cohort/age groups sampled by both sampling methods exhibited the same composition
of hatchery and natural chinook salmon. The age 2 - 1990 cohort showed the only
discrepancy with 37 percent of the sport harvest and 57 percent of the gill net samfie
being from natural reproduction. The OTC determination on this gill net sample was
done by MDNR personnel and could have included some age 3 - 1989 cohort samples
which were not marked with OTC, resulting in an inflated estimated of the
contribution of natural chinook salmon.

Depending on the type of samples and data needed for specific research
projects, it can be beneficial to utilize the sport harvest as a practical sample

collection method.

CHAPTER SIX

Summary

Natural reproduction of chinook salmon in the Great Lakes basin has been
documented by several studies since the introduction of chinook salmon in 1967.
These studies concentrated on estimating the production of natural chinook in
individual streams. Estimates of the lake-wide contribution of natural chinook salmon
to Lake Michigan was not possible due to the inability to distinguish natural chinook
salmon from unmarked hatchery chinook salmon after outmigration from streams.
Starting in 1990 and continuing through 1993 an extensive hatchery chinook salmon
marking program using external marks of fin clips and/or internal marks with
oxytetracycline (OTC) has provided the opportunity to examine the lake-wide
contribution of natural chinook salmon to the Lake Michigan ecosystem.

Successful completion of this study required: 1) all samples to be aged with a
high degree of accuracy, 2) a standard method for the detection of OTC marks be
established along with a detection failure rate of those OTC marks, and 3) an

adequate sample size of chinook salmon be obtained from Grand Haven, Ludington,

and Leland, Michigan.
Age was determined by counting annular bands appearing on the vertebrae.
The validity of this vertebral ageing method was determined by comparing known

ages from fish marked with coded wire tags to age estimates made from the

60

61

corresponding vertebrae. A sample of 197 tagged chinook salmon were age with
97% accuracy.

The existence of double marked (fin clipped and OTC fed) chinook salmon
enabled the calculation of an OTC marking/detection failure rate of 5 percent. By
rating the quality of the OTC marks observed and comparing the quality of those
marks over time in the same cohort, it was concluded that once an OTC mark is
administered the quality of that mark does not decrease over time.

The sampling of chinook salmon at three ports for two years allowed for 3 age
classes and 3 cohorts to be examined. It can generally be concluded that at ages 1
and 2 the fish are randomly mixed within the lake, with approximately 30 percent of
the chinook salmon sport catch being from natural reproduction at all three ports.
The percentage of naturally reproduced age 3 chinook salmon differed between ports,
from 39 percent in the Grand Haven area to 54 percent in the Ludington area. This
difference may represent the differences in contribution of natural fish between river

systems.

APPENDICES

APPENDIX A

Table 12. Summary of Lake Michigan hatchery chinook salmon marking,

1990 - 1993 stockings, numbers of fish in each category.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mark Type 1990 1991 1992 1993
[Michigan [RV/OTC 604,809 499,554 0 0

Ad/OTC 698,596 1,014,261 917,993 828,766

OTC 2,239,769 1,803,440 2,497,433 2,278,261

nm 96,582

Total 3,639,756 3,317,255 3,415,426 3,107,027
[Indiana [13v 153,590

LV 148,267

RP 288,234

LP 353,837

AdRV 95,181 83,770

AdLV 97,066 82,372

AdLP 93,688 60,888

AdRP 94,744 112,890

OTc 330,453 292,464

Total 630,256 694,351 504,231 458,606
[Illinois [1) 102,200 52,819

DLP/OTC 51,500

DRP/OTC 50,996

RP/OTC 51,221

DRV/OTC 100,499

BV/OTC 39,240

AdBV/OTC 51,000 49,873

RPLV/OTC 50,041 52,343

LPRV/OTC 75,817 50,517

AdD/OTC 76,500 49,044

OTC 377,200 191,480 57,894 162,420

Total 479,400 496,338 352,669 364,197
[Wisconsin [Ad/OTC 25,100 65,761 69,265 233,886

RMLV 166,989

OTC 2,354,211 1,668,857 1,286,885 1,327,234

Total 2,379,311 1,734,618 1,523,139 1,561,120
[GRAND TOTAL 7,128,723 6,242,562 5,795,465 5,490,950]
OTC Only 5,067,762 3,663,777 4,172,665 4,059,179
%clipped 28.9 41.3 28.0 26.1

62

APPENDIX B

Table 13. Summary Of fin clips used by state and year.

 

 

Fin Clip 1990 1991 1992 1993
Ad MI/WI MI/WI MI/WI MI
LV IND

RV MI MI

LP IND

RP IND ILL

AdLV IND IND
AdRV IND IND
AdLP IND IND

AdRP IND IND

BV IND ILL

LPRV ILL ILL
RPLV ILL ILL
AdBV ILL ILL
D ILL ILL

DAd ILL ILL
DRV ILL

DLP ILL

DRP ILL

RMLV WI

LM

63

APPENDIX C

Table 14a. Summary of the 1992 age 1 and age 2 daily samples for each port.

Percentages have not been corrected for 5% occurrence of
unmarked hatchery chinook salmon.

 

 

Age 1

Port Date Number # Natural # Hatchery % Natural % Hatchery

GH 1 05/09/92 1 1 0 100 0
GH 05/23/92 1 1 0 100 0
GH 05/24/92 2 l 1 50 50
GH 05/25/92 2 0 2 0 100
GH 05/28/92 5 2 3 40 60
GH 05/30/92 1 0 l 0 100
GB 05/31/92 2 0 2 0 100
GH 06/21/92 2 0 2 0 100
GB 07/17/92 14 4 10 29 71
GH 07/18/92 11 2 9 18 82
GH 07/25/92 5 1 4 20 80
GH 08/01/92 1 1 0 100 0
GH 08/15/92 11 2 9 18 82
GH 08/16/92 11 2 9 18 82
GH 08/22/92 27 7 20 26 74
GH 09/12/92 1 0 1 0 100
LU 2 06/20/92 2 1 1 50 50
LU 07/03/92 4 0 4 0 100
LU 07/04/92 5 0 5 0 100
LU 07/11/92 9 4 5 44 56
LU 07/12/92 21 11 10 52 48
LU 07/19/92 1 1 0 100 0
LU 07/23/92 2 1 1 50 50
LU 08/07/92 52 15 37 29 71
LU 08/14/92 3 0 3 0 100
LU 08/15/92 12 4 8 33 67
LU 08/17/92 6 3 3 50 50
LU 09/05/92 11 3 8 27 73
LU 09/12/92 2 l 1 50 50
LE 3 07/26/92 18 4 14 22 78
LE 08/02/92 43 11 32 26 74
LE 08/05/92 94 31 63 33 67
LE 08/06/92 87 28 59 32 68
LE 08/23/92 38 20 18 53 47

1 Grand Haven.
2 Ludington.
3 Leland.

Table 143. Continued

65

 

 

 

Age 2

Port Date Number # Natural # Hatchery % Natural % Hatchery

GH 05/09/92 8 4 4 50 50
GH 05/23/92 4 1 3 25 75
GH 05/24/92 4 2 2 50 50
GH 05/25/92 5 1 4 20 80
GH 05/28/92 8 4 4 50 50
GH 05/30/92 8 4 4 50 50
GH 05/31/92 3 0 3 0 100
GB 06/28/92 2 1 1 50 50
GH 07/05/92 2 0 2 0 100
GB 07/17/92 2 2 0 100 0
GH 07/18/92 1 0 l 0 100
GH 07/25/92 5 3 2 60 40
GH 08/01/92 2 1 1 50 50
GH 08/15/92 6 2 4 33 67
GH 08/16/92 9 2 7 22 78
GH 08/22/92 5 0 5 0 100
GH 09/05/92 3 0 3 0 100
GH 09/12/92 3 2 1 67 33
LU 06/06/92 1 0 1 0 100
LU 06/27/92 1 1 0 100 O
LU 07/03/92 1 1 0 100 0
LU 07/04/92 1 1 0 100 0
LU 07/11/92 11 4 7 36 64
LU 07/12/92 13 6 7 46 54
LU 07/19/92 3 0 3 0 100
LU 07/23/92 4 2 2 50 50
LU 08/07/92 26 11 15 42 58
LU 08/14/92 16 5 11 31 69
LU 08/15/92 8 3 5 38 63
LU 08/17/92 1 1 0 100 0
LU 09/05/92 3 l 2 33 67
LU 09/12/92 4 1 3 25 75
LE 07/26/92 17 9 8 53 47
LE 08/02/92 22 6 16 27 73
LE 08/05/92 12 4 8 33 67
LE 08/06/92 11 S 6 45 55
LE 08/23/92 2 1 1 50 50

66

Table 14b. Summary of the 1993 age 1, 2, and 3 daily samples for each port.
Percentages have not been corrected for 5% occurrence of
unmarked hatchery chinook salmon.

 

 

Age 1
Port Date Number # Natural # HatcherL% Natural % Hatchery
GH 1 05/08/93 2 o 2 o 100
GH 05/15/93 3 1 2 33 67
GH 05/16/93 5 2 3 40 60
GH 05/19/93 4 l 3 25 75
GH 05/20/93 2 1 1 50 50
GH 05/21/93 4 0 4 0 100
GB 05/22/93 20 6 14 30 70
GH 05/23/93 3 0 3 0 100
GH 05/26/93 3 1 2 33 67
GH 05/27/93 20 6 14 30 70
GH 05/28/93 2 0 2 0 100
GH 05/30/93 15 3 12 20 80
GH 06/02/93 8 2 6 25 75
GH 06/03/93 5 3 2 60 40
GH 06/04/93 4 2 2 50 50
GH 06/05/93 11 4 7 36 64
GH 06/20/93 1 0 1 0 100
GH 06/22/93 1 0 1 0 100
GH 06/27/93 2 l 1 50 50
GH 06/28/93 3 0 3 0 100
GH 07/07/93 1 0 1 0 100
GH 07/08/93 2 2 0 100 0
GH 07/09/93 1 0 1 0 100
GH 07/13/93 3 l 2 33 67
GH 07/14/93 2 0 2 0 100
GH 07/18/93 2 0 2 0 100
GH 07/19/93 1 0 1 0 100
GH 07/20/93 4 3 1 75 25
GH 07/23/93 20 10 10 50 50
GH 07/28/93 4 2 2 50 50
GH 08/01/93 5 3 2 60 40
GH 08/05/93 9 2 7 22 78
GH 08/06/93 3 1 2 33 67
GH 08/07/93 8 4 4 50 50
GH 08/13/93 8 2 6 25 75
GH 08/17/93 16 5 11 31 69
GH 08/18/93 13 4 9 31 69
GH 08/19/93 19 8 11 42 58
GH 08/20/93 7 0 7 0 100
GH 08/23/93 3 3 0 100 0
GH 08/26/93 26 16 10 62 38
GH 08/28/93 29 10 19 34 66
GH 08/29/93 35 14 21 40 60

1 Grand Haven

Table 146. Continued

67

 

Age 1 continued

 

Port Date Number # Natural # Hatchery % Natural % Hatchery
LU 2 05/29/93 1 0 r 0 100
LU 05/30/93 2 0 2 0 100
LU 06/03/93 5 0 5 0 100
LU 06/04/93 1 0 1 0 100
LU 06/19/93 17 10 7 59 41
LU 06/22/93 2 0 2 0 100
LU 06/23/93 1 0 1 0 100
LU 06/26/93 7 2 5 29 71
LU 07/03/93 6 2 4 33 67
LU 07/04/93 3 2 1 67 33
LU 07/10/93 6 3 3 50 50
LU 07/11/93 17 8 9 47 53
LU 07/17/93 21 5 16 24 76
LU 07/22/93 9 3 6 33 67
LU 07/28/93 5 4 l 80 20
LU 07/30/93 8 2 6 25 75
LU 07/31/93 13 7 6 54 46
LU 08/08/93 36 10 26 28 72
LU 08/12/93 19 9 10 47 53
LU 08/21/93 6 2 4 33 67
LU 08/22/93 20 7 13 35 65
LU 08/27/93 37 9 28 24 76
LE 3 07/21/93 10 o 10 0 100
LE 07/24/93 4 2 2 50 50
LE 07/25/93 8 2 6 25 75
LE 07/31/93 3 1 2 33 67
LE 08/02/93 2 1 1 50 50
LE 08/10/93 4 l 3 25 75
LE 08/14/93 4 2 2 50 50
LE 08/15/93 1 0 1 0 100
2 Ludington

3 Leland

Table 14b. Continued

68

 

 

Age 2
Port Date Number # Natural # Hatchery % Natural % Hatcheri
GH 05/12/93 1 l 0 100 0
GH 05/15/93 14 5 9 36 64
GH 05/16/93 10 3 7 30 70
GH 05/19/93 11 l 10 9 91
GH 05/20/93 2 0 2 0 100
GH 05/21/93 13 3 10 23 77
GH 05/22/93 10 3 7 30 70
GH 05/23/93 7 3 4 43 57
GH 05/26/93 5 1 4 20 80
GH 05/27/93 6 l 5 17 83
GH 05/28/93 1 1 0 100 0
GH 05/30/93 9 7 2 78 22
GH 06/02/93 3 0 3 0 100
GH 06/03/93 9 2 7 22 78
GH 06/04/93 5 3 2 60 40
GH 06/05/93 10 3 7 30 70
GH 06/22/93 1 0 1 0 100
GH 06/27/93 1 0 1 0 100
GB 07/13/93 1 0 1 0 100
GH 07/14/93 2 0 2 0 100
GH 07/18/93 1 0 l 0 100
GH 07/23/93 4 0 4 0 100
GH 08/05/93 2 2 0 100 0
GH 08/06/93 1 0 1 0 100
GH 08/07/93 6 l 5 17 83
GH 08/13/93 1 O 1 0 100
GH 08/18/93 7 3 4 43 57
GH 08/19/93 5 2 3 40 60
GH 08/20/93 3 0 3 0 100
GH 08/26/93 2 1 1 50 50
GH 08/28/93 4 2 2 50 50
GH 08/29/93 6 2 4 33 67
LU 05/29/93 5 0 5 0 100
LU 05/30/93 1 0 1 0 100
LU 06/03/93 3 0 3 0 100
LU 06/04/93 4 l 3 25 75
LU 06/06/93 4 1 3 25 75
LU 06/19/93 14 5 9 36 64
LU 06/22/93 4 1 3 25 75
LU 06/23/93 7 l 6 14 86
LU 07/03/93 6 0 6 0 100
LU 07/04/93 3 1 2 33 67
LU 07/10/93 10 6 4 60 40
LU 07/11/93 5 3 2 60 40

1 Grand Haven
2 Ludington

Table 14b. Continued

 

Age 2 continued

 

Port Date Number # Natural # Hatchery % Natural % Hatchery
LU 07/17/93 32 9 23 28 72
LU 07/22/93 1 l 3 8 27 73
LU 07/28/93 12 5 7 42 58
LU 07/30/93 19 8 11 42 58
LU 07/31/93 21 8 13 38 62
LU 08/08/93 30 1 1 19 37 63
LU 08/12/93 18 4 14 22 78
LU 08/21/93 4 1 3 25 75
LU 08/22/93 26 13 13 50 50
LU 08/27/93 28 7 21 25 75
LE 07/21/93 11 5 6 45 55
LE 07/24/93 2 l 1 50 50
LE 07/25/93 3 1 2 33 67
LE 07/31/93 1 1 0 100 0
LE 08/02/93 1 0 1 0 100
LE 08/10/93 1 0 1 0 100
LE 08/11/93 2 0 2 0 100
LE 08/14/93 1 0 1 0 100
LE 08/15/93 1 0 1 0 100
2 Ludington

3 Leland

, 70
Table 14b. Confirmed

 

 

Age 3
Port Date Number # Natural # Hatchery % Natural % Hatchery
GH 05/08/93 2 0 2 0 100
GH 05/12/93 1 0 1 O 100
CH 05/15/93 18 8 10 44 56
GH 05/16/93 11 3 8 27 73
GH 05/19/93 6 5 1 83 17
GH 05/20/93 2 1 1 50 50
GH 05/21/93 5 2 3 40 60
GH 05/22/93 5 3 2 6O 40
GH 05/23/93 1 1 0 100 0
GH 05/26/93 2 1 1 50 50
GH 05/27/93 2 0 2 O 100
CH 05/30/93 3 1 2 33 67
GH 06/02/93 1 0 1 0 100
CH 06/03/93 1 l O 100 0
GH 06/05/93 1 0 1 0 100
GH 07/08/93 1 0 1 0 100
CH 07/23/93 1 0 l 0 100
GH 07/27/93 1 0 1 0 100
GB 08/05/93 1 0 1 0 100
CH 08/06/93 3 2 1 67 33
GH 08/07/93 1 l 0 100 0
GH 08/13/93 1 0 1 O 100
GH 08/17/93 2 0 2 0 100
CH 08/18/93 1 1 0 100 0
GH 08/26/93 S 4 1 80 20
GH 08/28/93 5 2 3 40 60
GH 08/29/93 7 2 5 29 71
LU 05/29/93 1 0 1 0 100
LU 06/03/93 1 1 0 100 0
LU 06/04/93 1 1 0 100 0
LU 06/06/93 1 1 0 100 0
LU 06/19/93 2 0 2 0 100
LU 06/23/93 1 0 1 0 100
LU 06/24/93 2 2 0 100 0
LU 07/03/93 7 3 4 43 57
LU 07/10/93 5 2 3 40 60
LU 07/11/93 4 2 2 50 50
LU 07/17/93 16 10 6 63 38
LU 07/22/93 9 4 5 44 56
LU 07/28/93 3 2 1 67 33
LU 07/30/93 9 3 6 33 67
LU 07/31/93 16 12 4 75 25
LU 08/08/93 18 11 7 61 39
LU 08/12/93 12 7 5 58 42
LU 08/21/93 1 0 1 0 100
LU 08/22/93 19 15 4 79 21
LU 08/27/93 19 6 13 32 68
LE 07/21/93 4 2 2 50 50
LE 07/24/93 2 2 0 100 0

APPENDIX D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 e e 0
~ Grand Haven
80 _ Age 1 (1991 cohort)
’8 r
‘é
as 50 7
z 0
H
r:
8 4o — -
8
O-t o .
20 b O . O
0 l g l j 1 l l 1
May June July August September
Date
100 e
. Grand Haven
30 _ Age 2 (1990 cohort)
'8 t
H 0
g a _ .
z *- 0 O O. O O
H
r:
8 40
he -------------------------- l- ------------ l- --------------------------
0 e
n-i [
20 —- . '
0 l 1 L g l: l :1 Lf 1
May June July August September
Date

Figure 9a. Daily, monthly, and yearly percentages of age 1 and age 2 natural chinook salmon in
the 1992 sport harvest at Grand Haven. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

71

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 .
[-
80 e Ludington
.3 [ Age 1 (1991 cohort)
a .l
Z . g . . . .
‘5 .
8 4° 5
£ _ r . .
20 _
0 l l 1 F L l g l 1
May June July August September
Date
100 O Q Q
. Ludington
80 _ Age 2 (1990 cohort)
1;: .
60 .—
CB
2 '- O
H
t: . .
§ 40 ~ -------------- [----.--------.-----.‘ ....................
a: » . - .
20 r-
[_
0 1 l: l 1 g l l 1
May June July August September
Date

Figure 9b. Daily, monthly, and yearly percentages of age 1 and age 2 natural chinook salmon in
the 1992 sport harvest at Ludington. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

73

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
so _ Leland
.3 Age 1 (1991 cohort)
5 ,0 ,_
t6
2 .
‘a’
8 40 ~
8 ------------- 's --------
m _
20 — i
0 A l l l l l 1
May June July August September
Date
100
. Leland
30 _ Age 2 (1990 cohort)
3- .
g to __
Z _ a .
E .
8 40 ~ ...........................
8 .
9.. r ,
20 —
0 l l l l l l 1
May June July August September
Date

Figure 9c. Daily, monthly, and yearly percentages of age 1 and age 2 natural chinook salmon in
the 1992 sport harvest at Leland. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

100

8

8

Percent Natural

100

8

8

Percent Natural

100

S

8

Percent Natural

74

Grand Haven Age 1 (1992 cohort)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e
e
'- e . o
e e e e o e
- 0 ° c
"""" ..---..-:.----..--------------------...------------..-----:..----:.
e e e
: 1 e- e i: c :e' - M l I;
May June de Alum
Date
Grand Haven Age 2(1991 cohort)
e e e
'— e
- o
oo
o e
._ e
e
e
OOOOOOOO ‘..*....-.‘...........-..................‘................1r
0 e
e
L e e 1: 1 M e e 1 c c 1:
May Jlme lily Must
Date
Grand Haven Age3 (1990 cohort)
e e e e
e
,_. e
o
- e
e e
e
_----------: oooooooooooooooooooooooooooooooooooooo q ............... :0
e
. e
c i: : ‘fif 1 :1 c e I .g : :1
May J1me July Must

Date

Figure 9d. Daily, monthly, and yearly percentages of age 1, 2, and 3 natural chinook salmon in
the 1993 sport harvest at Grand Haven. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

100

 

75

Ludington Age 1(1992 cohort)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

lg 60 ‘- e
i [ . '
a e e
32 ‘° ”
a """""""""""""""" :"'f """"" '5'"""": """ I"---
20 ,_ O O
0 J May :: ;: line 1 c: 1 July 1 Must I
Date
Ludington Age 2 (1991 cohort)
100
80 t—
g m .— u .
g ‘0 _ . . .. 0
a: i_ --------------------------------- p -- ------ .- - -. ---------------------
20 __ O
0 gm 2 g In“ 1 .fi 1 My 1 m1 1
Date
Ludington Age 3 (1990 cohort)
100 00 e e
80 —
3 ° '
is 5° f .......................................... . ........... it __________
z 0
'5 , . .
g 40 ~ .
£ f o
20 '— o
0 I May Ar Jute :1 : 1 My 1 Amust I A
Date
Figure 9e. Daily, monthly, and yearly percentages of age 1, 2, and 3 natural chinook salmon in

the 1993 sport harvest at Ludington. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

 

76

Leland Age 1 (1992 cohort)
100

8
l

8
l

 

Percent Natural

 

 

 

 

 

 

May Jule July AEust

Date

Leland Age 2 (1991 cohort)
100 e

 

Percent Natural

20*-

 

 

 

 

 

May June July fiat

Date

Leland Age 3 (1990 cohort)
100 e

 

Percent Natural

20 —

 

 

 

 

May June J fly Aqust

Date

Figure 9f. Daily, monthly, and yearly percentages of age 1, 2, and 3 natural chinook salmon in
the 1993 sport harvest at Leland. Dots represent daily samples, bars represent

monthly averges, and dashed line represents yearly average. Only months in which
sampling occured are shown.

REFERENCES

LIST OF REFERENCES

Anson, Dan. Personal Communication. Michigan Department of Natural Resources,
Plainwell Office.

Appelget, J. and L.L. Smith, Jr. 1951. The determination of age and rate of growth
from vertebrae of the channel catfish, Ictalurus lacustris punctarus. Transactions
American Fisheries Society. Volume 80. pp. 119-139.

Baker, T.T. and S. Timmons. 1991. Precision of ages estimated from five bony
structures of artic char (Salvelinus alpinus) from the Wood River system, Alaska.
Canadian Journal of Fisheries and Aquatic Science. Volume 48. pp. 1007-1014.

Beamish, R.J, and DA. Foumier. 1981. A method for comparing the precision of a set
of age determinations. Canadian Journal of Fisheries and Aquatic Science.
Volume 38. pp. 982-983.

Beamish, R.J, and G.A. McFarlane. 1983. The forgotten requirement for age validation

in fisheries biology. Transactions American Fisheries Society. Volume 112. No.
6. pp. 735-743.

Beamish, R.J, and G.A. McFarlane. 1987. Current trends in age determination
methodology. Pages 15-42 In The Proceedings of an International Symposium on
the Ageing of Fish. Edited by Bagenal, T. B., Gresham Press, Old Working,
Surrey, England.

Bilton, HT. 1986. Marking Chum Salmon Fry Vertebrae with Oxytetracycline. North
American Journal of Fisheries Management. Volume 6(1): 126-128.

Brandt, S.B., D.M. Mason, E.V. Patrick, R.L. Argyle, L. Wells, P.A. Unger, and DJ.
Stewart. 1991. Acoustic measures of the aboundance and size of pelagic
planktivores in Lake Michigan. Canadian Journal of Fisheries and Aquatic
Science. Volume 48. pp. 894-908.

Bumguardner, B.W. and R.L. Colura. 1991. Chemical marking of fmgerling striped

bass otoliths. Texas Parks and Wildlife Department, Fisheries and Wildlife
Division Management Data Series No. 64. pp. 1-4.

77

78

Carl, L.M. 1980. Aspects of the population ecology of chinook salmon in Lke Michigan
tributaries. Ph.D. dissertation, University of Michigan. 174 pp.

Carl, L.M. 1982a. Natural reproduction of coho salmon and chinook salmon in some

Michigan streams. North American Journal of Fisheries Management. Volume 4.
pp. 375-380.

Carl, L.M. 1982b. Scale parameters of wild and hatchery chinook salmon (Oncorynchus
tshawytscha) from Lake Michigan. Canadian Technical Report of Fisheries and
Aquatic Sciences. No. 1105. pp. 19.

Carl, L.M. 1984. Chinook salmon (Oncorynchus tshauytscha) density, growth,
mortality, and movement in Two Lake Michigan tributaries. Canadian Journal of
Zoology. Volume 62. pp. 65-71.

Chang, W.Y. 1982. A statistical method for evaluating the reproducibility of age

determination. Canadian Journal of Fisheries and Aquatic Science. Volume 39:
pp. 1208-1210.

Chilton, DE. and HT. Bilton. 1986. New method for ageing chinook salmon
(Onchorhychus tshawytscha) using dorsal fin rays, and evidence of its validity.
Canadian Journal of Fisheries and Aquatic Science. Volume 43. pp. 1588-1594.

Christie, W.J . 1974. Changes in the fish species compositon of the Great Lakes. Journal
of the Fisheries Research Board of Canada. Volume 31. pp. 827-854.

Eck, G.W. and EH. Brown. 1985. Lake Michigan’s capacity to support lake trout
(Salvelinus namaycush) and other salmonines: an estimate based on the status of

prey populations in the 1970’s. Canadian Journal Fisheries and Aquatic Science.
Volume 42. pp. 449-454.

Eck, G.W., and L. Wells. 1987. Recent changes in Lake Michigan’s fish community
and their probable causes, with emphasis on the role of alewife (Alsoa
pseudoharengus). Canadian Journal of Fisheries and Aquatic Science. Volume 44
(Supplement 2). pp. 53-60.

Elliott, R. F. Personal Communication. US. Fish and Wildlife Service, Fishery
Resources Office 1015 Challenger Court. Green Bay, WI 54311.

Elliott. 1994. Early life history of chinook salmon in Lake Michigan. Michigan Sport
Fish Restoration Program, Project N0. F-53—R-10 Annual Report.

Fry, D.H. Jr. 1961. Some problems in the Marking of Salmonids. Pacific Marine
Fisheries Commission, Bulletin 5. pp. 77-83.

79

Gill, J. L. 1978. Design and analysis of experiments in the animal and medical sciences.
Iowa State University Press, Ames, Iowa.

Godfrey, H., D.D. Worlund, and HT. Bilton. 1968. Tests on the accuracy of ageing
chinook salmon (Oncorhynchus tshawytscha) from their scales. Journal of the
Fisheries Research Board of Canada. Volume 25(9). pp 1971-1982.

Hall, D.L. 1991. Age validation and aging methods for stunted brook trout.
Transactions of the American Fisheries Society. Volume 120. pp. 644-649.

Hankin, D.G. 1982. Estimating escapement of Pacific salmon: marking practices to
discriminate wild and hatchery fish. Transactions American Fisheries Society.
Volume 111. pp. 286-298.

Holly, M,E. 1990. Summary of trout and salmon stocking in Lake Michigan 1976-1993.
Great lakes Fishery Commission Lake Michigan Committee Meeting, March 23-
24, 1994. Ann Arbor.

Jennings, HE. 1992. Natural reproduction of Pacific salmon in a southern Michigan
stream. M.S. Thesis. Michigan State University, East Lansing, Michigan. 69pp.

Johnson, DC. and J .G. Hnath. 1991. Lake Michigan chinook salmon mortality - 1988.
Michigan Department of Natural Resources, Fisheries Technical Report 91-4,
Ann Arbor.

Jones, M.L., J.F. Koonce, and R. O’Gorman. 1993. Sustainability of hatchery-
dependent salmonine fisheries in Lake Ontario: the conflict between predator
demand and prey supply. Transactions of the American Fisheries Society.
Volume. 122(5). pp. 1002-1018.

Keller, M, K.D. Smith, and R.W. Rybicld. 1990. Review of salmon and trout
management in Lake Michigan. Michigan Department of Natural Resources,
Fisheries Special Report No. 14. 345 pp.

Koenings, J .P. and J. Lipton. 1983. The fluorometric determination of the uptake and
retention of the antibiotic oxtetracycline in sockeye salmonn (Oncorhynchus
nerka) fry: a quantitative approach to tetracycline marking. Alaska Department
of Fish and Game. Division of Fisheries Rehabilitation, Enhancement and
Development. P.O. Box 3—2000 Juneau, Alaska 99802.

Kusakari, M. 1969. Microscopical observation on the opaque and translucent zones of
the vertebral bone of the flatfish Kareius bicoloratus. Bulletin of the Japanese
Society Scientific Fisheries. Volume 35. NO. 1 pp. 7-17.

80

Lorson, RD. and V.A. Mudrak. 1987. Use of Tetracycline to Mark Otoliths of
American Shad Fry. North American Journal of Fisheries Management. Volume
7(3): 453-455.

Nelson, DD. and J .G. Hnath. 1990. Lake Michigan chinook salmon mortality - 1989.
Michigan Department of Natural Resources, Fisheries Technical Report 90-4,
Ann Arbor.

Nielsen, L.A. and D.L. Johnson. 1983. Fisheries Techniques. American Fisheries
Society, Bethesda, Maryland.

Menon, MD. 1950. The use of bones, other than otoliths, in determining the age and
growth-rate of fishes. Cons. Intemat. Explor. Mer. Jour. Du. Cons. Vol. 16 No.
3 pp. 311-335.

MDNR, Interoffice communitation. Fisheries Division.

Odense, RH. and V.H. Logan. 1974. Marking Atlantic salmon (Salmo salar) with
oxytetracycline. Journal of the Fisheries Research Board of Canada. Volume
31(3). pp. 348-350.

Parsons, J .W. 1973. History of salmon in the Great Lakes, 1850-1970. U.S. Fish and
Wildlife Service, Technical Paper 68, Washington, D.C., USA.

Powell, M.J. and M. Miller. 1990. Shoal spawning by chinook in Lake Huron. North
American Journal of Fisheries Management. Volume 10. pp. 242-244.

Prince, E.D., D.W. Lee., and J.C. Javech. 1985. Internal zonations in sections of
vertebrae from atlantic bluefin tuna, Thunnus thynnus, and their potential use in
age determination. Canadian Journal of Fisheries and Aquatic Science. Volume

42. pp. 938-946.

Rakoczy, GP. 1992. Charter boat catch and effort from the Michigan waters of the
Great Lakes, 1991. Michigan Department of Natural Resources, Fisheries
Technical Report 92-9, Ann Arbor.

Rakoczy. 1994. Sport fishing catch and effort. Michigan Department of Natural
Resources. Great Lakes Fishery Commission Lake Michigan Committee
Meeting, March 23-24, 1994. Ann Arbor.

Richards, L.J., J.T. Schnute, A.R. Kronlund, and R.J. Beamish. 1992. Statistical
models for the analysis Of ageing error. Canadian Journal of Fisheries and
Aquatic Science. Volume 49. pp. 1801-1815.

81

Rien, T.A. and RC. Beamederfer. 1994. Accuracy and precision of white sturgeon age
estimates from pectoral fin rays. Transactions American Fisheries Society.
Volume 123: pp. 255-265.

Robertson, J .M. 1990. Preface pages i-iii in 1989-1990 Michigan fisheries: a foundation
for the future. Michigan Department of Natural Resources. Special Report 13. pp
130.

Rybicki, R.W. 1973. A summary of the salomonid program (1969-1971). Pages 33—42
in Michigan’s Great Lakes trout and salmon fishery 1969-1972. Michigan
Department of Natural Resources, Fisheries Management Report 5, Lansing,
Michigan, USA.

Seelbach, P.W. and GB. Beyerle. 1984. Interpretation of the age and growth of
anadromous salmonids using scale analysis. Michigan Department of Natural
Resources, Fisheries Division, Technical Report 84-5. 9pp.

Seelbach, P.W. 1985. Smolt migration of wild and hatchery-raised coho and chinook
salmon in a tributary of northern Lake Michigan. Michigan Department of
Natural Resources, Fisheries Report 1935, Lansing, MI USA.

Seelback, P.W. and G. E. Whelan. 1988. Identification and contribution of wild and
hatchery steelhead sotcks in Lake Michigan Tributaries. Transactions American
Fisheries Society. Volume 117. pp. 444-451.

Sharp, D. and DR. Bernard. 1988. Precision of estimated ages of lake trout from five
calcified structures. North American Journal of Fisheries Managment. Volume
8. pp. 367-372.

Smith, S. H. 1968. Species succession and fishery exploitation in the Great Lakes.
Journal of the Fisheries Research Board of Canada. Volume 25(4). pp. 667-693.

Sneed, K.E. 1951. A method for calculating the growth of channel catfish, Ictalurus
lacustris punctatus. Transactions American Fisheries Society. Volume 80. pp.
174-179.

Sneva, J.G. and OM. Knudsen. 1989. Scale pattern analysis of 1987 and 1988 Idoaho
spring and summer chinook salmon. Washington Department of Fisheries. May
10, 1989.

Stewart, D.J., J .F. Kitchell, and LB. Crowder. 1981. Forage fish and their salmonid
predators in Lake Michigan. Transactions of the American Fisheries Society.
Volume 110. pp. 751-763.

82

Taube, GM. 1974. Transfer releases of coho salmon and trout into an upper part of
Platte River, and observations of salmonid spawning. Michigan Department of
Natural Resources, Fisheries Report 1815, Lansing, MI USA.

Weber, DD. and Ridgway, G.J. 1962. The deposition of tetracycline drugs in bones
and scales of fish and its possible use for marking. Progressive Fish-Culturist
(Bethesda). Volume 24. pp. 150-155.

Weber, DD. and Ridgway, G.J. 1967. Marking Pacific Salmon with Tetracycline
Antibiotics. Journal of the Fisheries Research Board of Canada. Volume 24(4).
pp. 849-856.

Zafft, DJ. 1992. Migration of wild chinook and coho salmon smolts from the Pere
Marquette River, Michigan. Masters Thesis. Michigan State University, East
Lansing, Michigan. 91pp.

"711111111[111111111