This is to certify that the
thesis entitled

PARENTAGE ASSIGNMENT OF BROWN TROUT (SALMO
TRUTTA L.) JUVENILES FOLLOWING STOCKING OF
MULTIPLE DONOR POPULATIONS

presented by
LAUREN STANCHEK

has been accepted towards fulfillment
of the requirements for the

MS. degree in Forensic Science

I..-“ m. .

<§<’

 

M“\ .

Major Professor’s Signature .___

// fizz/0 er
Date

MSU is an Affirmative Action/Equal Oppoztunity Institution

 

 

 

 

 

 

LIBRARY
Michigan State
University

PLACE IN RETURN BOX to remove this checkout from your record.
T 0 AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DaloDuopGS-p. 1 5

 

fl

PARENTAGE ASSIGNMENT OF BROWN TROUT (SALMO TRUTTA L.)
JUVENILES FOLLOWING STOCKING OF MULTIPLE DONOR
POPULATIONS
By

Lauren Stanchek

A THESIS

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

MASTER OF SCIENCE
Department of Forensic Science

2004

ABSTRACT

PARENTAGE ASSIGNMENT OF BROWN TROUT (SALMO TRUTTA L.)
J UVENILES FOLLOWING STOCKING OF MULTIPLE DONOR POPULATIONS

By
Lauren Stanchek

Forensic analyses have great utility beyond forensic science when applied to other
biological disciplines. For instance, parentage analysis is applied extensively in the field
of Molecular Ecology as a means of better understanding mating systems. In order to
adjust stocking strategy to more appropriately suit brown trout (Salmo trutta L.) mating
behavior, the current study sought to assign parentage to a juvenile cohort of brown trout
following stocking of adults from each of several donor populations. Parentage analysis
was based on adult and juvenile genotypes, which consisted of six microsatellite loci.
Juvenile genotypes were compared to those of putative parents using likelihood-based
statistical methods. Estimates of male and female reproductive success were calculated
based on parentage assignment. Several explanatory variables, including population
source and body size, were examined as potential sources of individual variation in male
and female reproductive success. All source populations contributed to the juvenile
generation, but twenty-one percent of the juveniles could not be associated with any
adults due to either low statistical power or failure to acquire samples from all adults. No
correlation between body size and reproductive success could be established. Mating
occurred randomly among males and females from all locations and the number of mates
per individual did not differ significantly between males and females. Several
suggestions are made regarding stocking strategy for the brown trout, as well as extensive

comparisons between human forensic and wildlife parentage analysis.

Copyright by
LAUREN STAN CHEK
2004

ACKNOWLEDGEMENTS

Many thanks to my committee, Dr. Kim Scribner, Dr. David Foran, and Dr.
Christina DeJong, whose time and expertise were essential to finishing this thesis.
Special thanks are extended to Dr. Scribner, who graciously Shared his lab’s resources
with me and allowed me to do this project. To Andy Nuhfer and the Hunt Creek
Research Station I also offer much gratitude for the sample collections and information
freely shared. I further extend many thanks to the members of the Scribner lab for their
support and suggestions, especially Scot Libants for answering many questions and Kristi
Filcek for acting as my sounding-board. I am also forever grateful to Hao Lu, my

parents, and my friends for their support and encouragement.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. v

LIST OF FIGURES ................................................................... ‘ ........... vii

INTRODUCTION ................................................................................. 1

DNA Analysis in Forensic Science—Overview ....................................... 1

DNA Analysis in Forensic Sub-disciplines ............................................ 2

Wildlife Forensics ................................................................ 2

Human Parentage Analysis ...................................................... 3

Non-human Parentage Analysis ................................................ 7

DNA Analysis in Molecular Ecology .................................................. 10
Parentage Analysis in Fish ...................................................... 10

Parentage Analysis in Brown Trout ........................................... 12

The Current Study ........................................................................ 14
MATERIALS AND METHODS ................................................................ 16
Study Area ................................................................................. 16
Stocking Timeline ......................................................................... 16

Data Collection ............................................................................ 18
Analysis ..................................................................................... 21
Cervus 2.0© Output ........................................................................ 23
Guidelines for Parentage Assignment ................................................... 23
RESULTS ............................................................................................ 25
Allele Frequency Results .................................................................. 25
Simulation Results ......................................................................... 28
Parentage Analysis Results ............................................................... 29
DISCUSSION ....................................................................................... 42
Allele Frequency Results .................................................................. 42
Simulation Results ......................................................................... 43
Ecological Applications of Parentage Analysis ........................................ 44
Stocking Strategy ........................................................................... 45
Expansion of the Study .................................................................... 46
Human Versus Non-human Parentage Analysis ....................................... 47
APPENDIX A ....................................................................................... 50
APPENDIX B ....................................................................................... 51
WORKSCITED ..................................................................................... 54

LIST OF TABLES

. Table 1. Primer characteristics and PCR conditions for each microsatellite locus

used ......................................................................................... 20

. Table 2. Estimates of population characteristics for each microsatellite locus (N:
sample size; No. of Alleles: the number of observed alleles for each locus;

Ho=observed heterozygosity; HE = expected heterozygosity; HW= tests for
deviations of genotype from Hardy-Weinberg equilibrium, where N S: not
significant; **= p<0.05) .................................................................. 26

. Tables 3 a—g. Summary measures of genetic diversity separated by adults and
progeny, as well as by donor source group (N: sample size; No. of Alleles: the

number of observed alleles for each locus; Ho=observed heterozygosity; HE =
expected heterozygosity; HW= tests for deviations of genotype from Hardy-
Weinberg equilibrium, where NS: not significant; **= p<0.05) ................... 26

. Table 4. Simulation summary measures of parentage assignment accuracy.

Levels of confidence are based on the distribution of simulated Delta scores
(based on known population allele frequencies). The strictest level was set at
95%, and the relaxed level was set at 80%. Any parentage assignment made
below the 80% confidence level was considered unresolved. The Delta Criterion
is the minimum Delta value defining the strict and relaxed confidence levels. The
Expected Success Rate refers to the expected rate of successful parentage
assignment at each confidence level ................................................... 29

. Table 5. Assignment summary from parentage analysis, including the number
and percentage of offspring that were unassigned or assigned to one or both
parents at 80% and 95% confidence levels ............................................ 3O

. Table 6. Estimates of parental reproductive success from each source location.
The number of offspring assigned refers to the number of offspring that can be
assigned to one parent within the respective donor group. The percent of
individuals with offspring represents the number of individuals within each donor
group that produced offspring. The percent of total assigned offspring reflects
contributions of the donor group to all of the 53 assigned offspring (both one and
two parents assigned) .................................................................... 34

. Table 7. Inferred sexes of parents based on pairing with individuals of known
sex .......................................................................................... 38

. Appendix Table A. 1. Example of output from Cervus 2.09. The table provides
the candidate parental pairs for each offspring individual, their LOD and Delta
scores, as well as the confidence level into which the respective Delta score falls.
O ID: offspring ID; KP ID: known parent ID; O—KP mismatching: number of
mismatched loci between offspring and known parent; Prob. of exclusion:

vi

10.

probability of exclusion; CP II): candidate parent ID; O—CP mismatching: number
of mismatched loci between offspring and candidate parent; LOD: logarithm of
odds score; Delta: Delta score, difference between LOD scores; Confidence: level
of confidence into which the Delta score falls, 80% (+) relaxed level, 95% (*)
strict level .................................................................................. 50

Appendix Table B.l. Assignment guidelines and the juveniles defined by them.
For each guideline the identification label of each juvenile individual assigned
under that guideline is listed. Identification labels are split into the sections of the
research area from which the juveniles were sampled ............................... 52

Appendix Table B.2. Juveniles defined by guideline 4. The candidate parental
pair ratio is the ratio of candidate parental pairs containing the assigned parent to
candidate parental pairs which do not contain the assigned parent. The assigned
parent is the most commonly found candidate parent amongst all of the putative
parental pairs listed for that particular juvenile ....................................... 53

vii

LIST OF FIGURES

. Hunt Creek research area, northern Lower Peninsula of Michigan. a. The Hunt
Creek Research Station’s location— Montgomery County. b. The research section
of station, split into segments A, B, and Z ............................................. 17

. Figure 2 a—e. Estimated reproductive success of each stocked adult. Bars
represent the number of progeny to which each adult contributed. a. Reproductive
success of transplanted adults from Hunt Creek, all of unknown gender. b-d.
Reproductive success of transplanted adults from Houghton Creek (b), East
Branch Au Sable (c), and Gilchrist Creek (d). The first seven individuals for each
source location are females and the second seven are males. e. Reproductive
success of Hunt Creek residents, all of unknown sex ................................. 31

. Figure 3 a—b. Estimates of reproductive success. a. Reproductive success for each
donor population as represented by the mean number of offspring per individual.
b. Reproductive success for donor populations of known sexes is represented by
the number of offspring to which females and males contributed .................. 35

. Figure 4 a—b. Number of mating partners. Only individuals of known sex, those
from Houghton Creek, East Branch Au Sable River, and Gilchrist Creek donor
populations, were considered. a. Minimum number of partners per known female.
b. Minimum number of partners per known male .................................... 36

. Figure 5. Minimum number of partners per parent for adults of unknown sex. . .38

. Figure 6. Source origins of mating partners. The percentage of individuals mating
with another individual of the same source (white column) is compared to the

percentage of individuals mating with an individual from a different donor source
(black column) ............................................................................. 39

. Figure 7 a-b. Body size (length in inches) and reproductive success. a.
Reproductive success in relation to body size for all stocked females (R2: 0.003,
p>0.05) and males (R2: 0.078, p>0.05). b. Reproductive success in relation to

body size for both females (R2: 0.0007, p>0.05) and males (R2: 0.1536, p>0.05),
when individuals who produced no offspring were removed ....................... 40

. Appendix Figure B.1. Parentage assignment guidelines. Direction of Assignment
refers to the fact that parentage analysis is achieved by using one sex as the
known parent and the other sex as the candidate, then running the analysis again
with the sexes switched .................................................................. 51

viii

INTRODUCTION

Forensic science, or scientific analyses as applied to circumstances concerning the
public, has evolved into a discipline of immense scope. This broad definition of forensic
science has encouraged the use of scientific analyses in situations ranging from murder
investigations, to identifications after mass disasters, computer crimes, detection of
deception, parentage analysis, and wildlife forensics. With such a broad spectrum of sub-
disciplines, forensic science can be applied to any judicial inquiry that requires the use of
scientific analyses.

DNA Analysis in Forensic Science—Overview

DNA analysis became firmly integrated into forensic science as a result of Alec
Jeffreys’ development of Multi-Locus Probe (MLP) DNA ‘fingerprinting’ in the mid-
19803 (Jeffreys, 1985; Butler, 2001; Lynch, 2003). Since its genesis in forensic science,
DNA analysis has evolved to the prevalent use of Short Tandem Repeat (STR) loci.
When multiple STR loci are used concurrently, statistical power in forensic analyses can
be extreme];] high (Balding, 1999).

STRs, also known as microsatellites, are tandem repeats of a short, simple DNA
sequence, usually from two to six base pairs in length. The advantages of microsatellites
include the fact that they are based on polymerase chain reaction (PCR) technology.
DNA samples of insufficient quantity or poor quality can be amplified until they are of
ample quantities for analysis. Also, in contrast to Jeffreys’ DNA MLP technique, which
targeted repeat motifs in multiple regions of the genome simultaneously, microsatellites
amplify very specific regions of the genome. Microsatellites greatest benefit, however, is

that they provide tremendous statistical power in individual identification, or when used

in parentage analysis. The statistical power inherent in microsatellites is a function of the
large number of microsatellites that are available for analysis, high levels of
heterozygosity, and large numbers of alleles at each locus. Loci can and should be
selected from different chromosomes. This independent assortment of loci means that
the product rule can be applied to the probabilities of genotype matches for each locus,
generating probabilities sufficient for exclusion and identification.

In 1997, a standardized core of 13 human STR loci were established for DNA
profiling and provided the foundation for the nation-wide CODIS (Combined DNA Index
System) system. CODIS now contains the DNA profiles of nearly 2 million convicted
offenders and has aided in over 3,000 cases (FBI, 2004). Furthermore, DNA analysis has
found remarkable use beyond the identification of offenders in a number of other forensic
sub-disciplines.

DNA Analysis in Forensic Sub-disciplines
Wildlife Forensics

Wildlife forensics constitutes one of the currently burgeoning sub-disciplines of
forensic science. Wildlife forensics originally developed from the need to analyze
evidence received in cases of illegal collection, possession, and sale or trade of pieces and
goods originating from organisms considered protected, threatened, or endangered
(Goddard and Espinoza, 2000). For example, traditional Chinese medicine frequently
uses parts from the tiger and rhinoceros, both of which currently hold endangered status
(Singh et al., 2004). This field has expanded to now include cases beyond just those
affecting species in peril, such as use of a prohibited weapon in making a kill, hunting

outside of an established season, poaching on private property, collecting too many

animals during a season, and killing an animal of the wrong sex (Goddard and Espinoza,
2000). The requirement for forensic analyses in many wildlife cases stems from the
condition of evidence, which has typically been altered to the extent that simple
morphological identification is impossible.

Like many other sub-disciplines of forensic science, wildlife forensics has
evolved in methodology to meet the challenges presented by the types and conditions of
evidence. Pathology, molecular biology, morphologic examinations, ballistics and
toolmarks, questioned documents, analytical chemistry, and fingerprints all play a part in
analyzing the evidence from wildlife cases (Goddard and Espinoza, 2000). DNA, one of
the newest and most commonly applied analyses in wildlife forensics, most commonly
functions in taxonomic identification, sex determination, and in some cases, identification
of individuals (Goddard and Espinoza, 2000). Cases requiring DNA analysis primarily
utilize PCR-based methods, such as mitochondrial DNA sequencing and STRs (Goddard
and Espinoza, 2000; Singh et al., 2004; Verma et al., 2003; Wan and Fang, 2002).
Human Parentage Analysis

Another sub-discipline within forensic science that is grounded in DNA
technology is parentage testing. Parentage testing most commonly involves paternity
analysis, where the genotype of an alleged father undergoes comparison to the genotypes
of a child and the child’s biological mother. However, the identification of a child’s
mother or both parents may also prove necessary in some cases (Schanfield, 2000). In
humans, parentage determination is typically employed in resolving child support
disputes. Parentage analysis is used not only in cases attempting to identify biological

parents and relatives, but also in situations where identification of a body cannot be

achieved simply by visual means, such as in mass disasters (Corvach et al., 1996; Leclair
et al., 2004; Marcotte et al., 1996; Martin et al., 1996; Primorac et al., 1996). Some cases
involving crime scenes with blood stain evidence, but the missing body of the victim,
may also require parentage analysis to identify the source of the blood stain (Mevag et
al., 1996; Schanfield, 2000).

Like wildlife forensics, the evolution of technological changes in parentage
testing has resulted in a shift from allozymes and blood groups to the prevailing use of
restriction fragment length polymorphisms (RFLPS) and PCR-based methods such as
VNTR loci, including long terminal repeats (LTRs) and STRs (Arroya et al., 1994;
Bjerre, 1997; Dobosz, 1990; Domenici et al., 1998; Geada et al., 2001; Hallenberg and
Morling, 2002; Lobbiani, 1991; Morling et al., 2002; Schanfield, 2000; Thomson, 1999).
Also like wildlife forensics, statistical probabilities provide greater certainty to
conclusions based on DNA analyses. If an alleged parent cannot possibly be a true
parent, that individual is eliminated from further consideration. If the alleged parent
cannot be excluded, a probability of parentage, or Parentage Index, is assigned in relation
to a reference population (Bias, 1983; Brooks, 1982; Bryant, 1980; Dawid, 2001; Dykes,
1982; Morris, 1982; Pohl, 1982; Schanfield, 2000; Thomson, 1999). Analysts typically
calculate probabilities using a computer program designed to account for such factors as
mutations and the unavailability of one parental genotype (Cowell, 2003).

Statistics ofParentage Analysis

The statistical probability of parentage, or likelihood of parentage, is determined

by means of LCD (Logarithm of Odds) scores. The calculation of a LOD score requires

the assessment of hypotheses in relation to a given data set (Balding and Nichols, 1997;

Evett and Weir, 1998; Marshall et al., 1998; National Research Council, 1996; Pohl,
1982). In parentage analysis, a likelihood ratio (Eq.l) is estimated by comparing the
hypothesis that the candidate parental pair is the true parental pair (Hl) to the hypothesis
that an arbitrary candidate parental pair from the population is the true parental pair (H2).
The likelihood of each hypothesis is determined from the probability of observing an
offspring genotype in the candidate parental pair’s genotypes (D).

L(H1,H2ID) = PIDIH] )
P(DIH2) (EQ- 1)

An overall likelihood ratio for each candidate parent consists of the product of likelihood
ratios for each locus. The LOD score itself is calculated by taking the natural log of the

overall likelihood ratio for each candidate parent.

A positive LOD score indicates that the candidate parent has a higher likelihood
of true parentage than an individual selected at random from the population. A zero LOD
score indicates that the likelihood of a candidate’s true parentage equals that of an
individual selected at random from the population, and a negative LOD score indicates
that the candidate parent has a lower likelihood of true parentage than a random
individual. Positive likelihood scores represent potential parents, even if a mismatch
exists between the genotypes of that potential parent and the progeny individual. In this
way, analytical programs account for the possibility of mutation, related individuals, as
well as the possibility of typing and data entry error. In addition, more than one potential
parental pair may have a positive likelihood score. In these situations, some programs
calculate additional statistics in order to quantify the magnitude of differences in

likelihood.

A number of additional estimates are used in order to generate LOD scores, as
well as to evaluate the quality and power of the data. These additional estimates include
the number of alleles per locus, expected and observed population heterozygosity, Hardy-
Weinberg expectations, and the probability of exclusion for each locus and across all loci.
The number of alleles per locus is a straightforward count of the number of different
alleles present at each locus taken directly from the population genotypes. The expected
heterozygosity (HE; Eq. 2) is a measure calculated by subtracting the sum of the
homozygote genotypic frequencies from one (Estoup et a1, 1998). This value is then
multiplied by the number of individuals in the population, divided by one less than the
number of individuals in the population. In this equation, n is the total number of

individuals in the population and pi2 is the frequency of homozygotes.

He=_n_2 (1 -2135) (Eq. 2)
n- 1
Observed heterozygosity (Ho; Eq. 3) is a direct measure, the actual count of heterozygote
genotypes in the population divided by the total number of individuals sampled. In this
equation, NAB is number of observed heterozygotes in population, where A and B are

alleles, and N is the total number of individuals in the population.

Ho = NAB/N (Eq. 3)

Hardy—Weinberg equilibrium (HW), is another indication of forces acting on a
single locus in a population, such as a founder effect or inbreeding, or may indicate the
presence of null alleles (Brooks, 1982). Hardy-Weinberg assumes that the organisms

under study are sexually-reproducing, diploid organisms who randomly mate and exhibit

discrete generations. This measure also assumes a sizable population with no mutation,
migration, or selection acting upon it. In brief, if a population is in Hardy-Weinberg
equilibrium, population allele frequencies are predictive of genotype frequencies in the
same population. Deviation from expectations may indicate a discord with the Hardy-
Weinberg assumptions. More specifically, violations of Hardy-Weinberg will appear as

an excess of homozygotes or heterozygotes.

Additional statistics, such as probabilities of exclusion, serve as indicators of
statistical power. They reflect probabilities of eliminating non-parents from
consideration. Exclusion is determined by comparison of the progeny individual’s
genotype to a known parent’s genotype and a candidate parent’s genotype (Dykes, 1982;
Marshall et al., 1998). Exclusion occurs when a candidate’s genotype includes a
mismatch with the offspring genotype. The actual probability of exclusion is calculated
based on the probability of finding a particular genotype in a population with specific
allele frequencies. Finding a specific genotype in the population hinges upon the number
of loci examined, the degree of polymorphism for each locus, the allele frequency
distribution, the number of potential parents, and the number of progeny (Weir, 1996).
Parentage Analysis in Non-humans

Non-human parentage analysis appears occasionally in forensic science, when
unique circumstances require the combination of wildlife forensics and DNA analysis.
Parentage analysis may determine whether an allegedly poached animal belongs to a
specific, protected group or whether trespassing occurred in order to obtain an animal
(Poetsch et al., 2001). It may .also determine ownership in disputes over livestock (Lirén

et al., 2004), as well as to verify cat, dog, and horse breeding lines.

In one case study, a dispute developed between two Holstein dairy farmers
because one farmer claimed that some calves of the neighboring farmer were sired by his
bulls (Lirén et al., 2004). Paternity analysis was applied using DNA collected from the
known mother of the calves, the calves themselves, and the potential fathers. The
parentage analyses performed in this case required the use of ten microsatellites to
develop genotypes for each animal. A computer program, Cervus© 2.0 (Marshall of al.,
1998), applied the genotypes obtained from the suite of microsatellites to calculate the
likelihood of each possible father for each calf. The analysis demonstrated a high
likelihood of paternity in the assignment of three of the disputed offspring (Lirdn et al.,
2004). The other two calves could not be attributed to any of the sampled bulls, implying
that another bull in the herd must have sired them.

Parentage Analysis Software in a Non-human Context

The Holstein case of Lir6n et a1. (2004), a non-human forensic case requiring
parentage analysis, utilized Cervus 2.0© (Marshall, 1998), a popular computer program
designed to assign potential parents to progeny individuals when only one parent or
neither parent is known. The program compares the genotypes of the putative parents to
the progeny and assigns parentage based on the likelihood of achieving a particular
progeny genotype in relation to the available parental genotypes. This likelihood is

directly based on the allele frequencies in the population.

Cervus 2.0© incorporates all of the statistical approaches commonly used for
human parentage analysis. However, the program includes one additional statistic, the
Delta score, to aid in assigning parentage. The development of this additional statistic is

due to the complex circumstances surrounding many cases of non-human parentage

analysis. The Delta score is a value used to assist in assigning parentage by
distinguishing the candidate parent of highest relative likelihood. The Delta statistic is
simply the difference between the candidate with the largest LOD score and the candidate
with the next largest LOD score. Zero and negative LOD scores are not included in
calculating Delta. Putative parental pairs receive rank based first on the LOD score, and
secondly on the Delta score. The pair with the highest LOD and Delta represents the

most likely pair.

As with many statistical analyses, the significance of results must be evaluated by
levels of confidence. Since Delta values do not exhibit a standard distribution, the
Simulation module of Cervus 2.0© uses Delta scores from many replicated simulations of
parentage tests to create a distribution. Delta scores generated during the real parentage
analysis are then compared to the distribution created by the program. A relaxed
confidence level of 80% means that four out of five Deltas fall within the distribution of
Deltas which belong to candidates of true parentage. Likewise, a strict confidence level
of 95% means that 19 out of 20 Deltas fall within the distribution of Deltas which belong
to candidates of true parentage. The Delta Criterion is the value at which actual Delta
scores become significant at a specific level of confidence. A Delta criterion is also
calculated for both levels of confidence for cases in which one parent is known, as well

as cases where neither parent is known.

DNA Analysis in Molecular Ecology

Ecologists embraced DNA analysis in the late 19805, not long after forensic
science adopted the techniques (Loeschcke et al., 1994). To date, DNA analysis has been
utilized extensively by molecular ecologists to examine relatedness, genetic diversity, and
evolution. Molecular techniques have been applied to species ranging from plants (e. g.
pine, Pinus sylvestris), to fishes (e. g. topminnows, Poeciliopsis spp.), and mammals (e. g.
Speke’s gazelle, Gazella spekei) (Loeschcke et al., 1994). Microsatellites have been used
in studies of species hybridization, population history, and phylogeography (Beaumont
and Bruford, 2000). These types of studies can reveal evidence of bottlenecks,
inbreeding, social structures, dispersion, and population genetic structure. More
specifically, molecular ecology studies have proven incredibly beneficial in evaluating
the effects of reproductive behaviors (Couvet and Ronfort, 1994; DeWoody and Avise,
2001).

Parentage Analysis in Fish

Molecular Ecologists, particularly those interested in fishes, have employed
parentage analysis extensively to estimate reproductive success and to quantify aspects of
the environment or phenotype that explain variance in individual reproductive success
(Bekkevold et al., 2002; Bentzen, 2001; Blanchfield et al., 2003; DeWoody and Avise,
2001; Fiumera, 2001; Fiumera et al., 2002; Garcia-Vasquez et al., 2001; Garrant et al.,
2001; Neff, 2001; Planes and Lenfant, 2002; Taborsky, 2001). Other fish parentage
studies have focused on estimating the mean and variance in male and female
reproductive success (Ostergaard et al., 2003; S'aisa et al., 2003). Characteristics of

interest to ecologists regarding reproduction include the number of partners per

10

individual, as well as selective mating based on phenotypic characteristics such as body
size (Bentzen et al., 2001; Blanchfield et al., 2003; DeWoody and Avise, 2001; Fiumera
et al., 2002). An understanding of fish mating systems is essential to ensuring the
preservation of diversity and determining the best methods for stocking (Cowx, 1994;
Gross et al., 2002).

Fisheries ecologists have widely embraced molecular methods due to the
difficulty in making accurate behavioral observations for aquatic organisms, as well as
the complex nature of fish mating systems (Blanchfield et al., 2003; DeWoody and
Avise, 2001; Fiumera et al., 2001; Planes and Lenfant, 2002; Utter and Ryman, 1993).
The application of genetic markers to wildlife has evolved in synchrony with other fields
that exploit genetic markers. Microsatellites (STRs) deserve the attention they currently
inspire, as they provide the best available resolution in statistical analyses (DeWoody and
Avise, 2000; Estoup et al., 1998; O’Reilly et al. 1998). Fish parentage analyses, like
. human parentage analyses, seek to establish probabilities of a parent-offspring match
based on available genotypic data, taking into account similar confounding factors such
as mutations and close relatives. Also like human parentage analyses, computer
programs typically provide the means of applying definitive statistical tests needed in
generating probabilities of correct parental assignment (Bekkevold et al., 2002; Bentzen
et al., 2001; Fiumera et al., 2001; Garrant et al., 2001; Marshall et al., 1998; Planes and
Lenfant, 2002; salsa et al., 2003).

Scientific research, including non-human parentage analysis, typically focuses on
species that are easy to study or that hold some level of importance to society. Economic

interests are often rooted in raising large numbers for food or for tourism. However,

11

recent trends emphasize additional motivations, such as the preservation of genetic
diversity in order to maintain healthy and sustainable natural populations (Ryman, 1991;
Tanksley and McCouch, 1997). One of the most studied fish families is the salmonids
(family Salmonidae), due to their importance as a food source and as a sporting
commodity (Bagliniére and Maisse, 1991; Hardy, 1972; MacCrimmon and Marshall,
1968). Also, in accordance with current trends toward environmental sustainability,
many salmonid populations have drawn attention to the need for preserving population
numbers and habitats. Fisheries geneticists have also sought to more fully grasp the
potential effects of stocking fish into wild populations (Allendorf and Ryman, 1987;
Altukhov and Salmenkova, 1987; Crisp, 1989; Thibault, 1983; Stanfield and Jones,
2003).

Parentage Analysis in Brown Trout

Among salmon, the brown trout (Salmo trutta L.) has been purposely spread by
humans well beyond its original distribution, stirring increasing research interest in this
species (MacCrimmon and Marshall, 1968). The brown trout adapts well to a variety of
environments and demonstrates a high degree of both phenotypic and behavioral
variation (Allendorf et al., 1976; Bagliniére and Maisse 1991: Frost and Brown, 1967).
S. trutta displays a range of scale colors and patterns and thrives in either freshwater or
saltwater (Campbell, 1977; Elliot, 1994; Frost and Brown, 1967). Brown trout are
observed in lakes, rivers, tributaries, and the ocean. Due to these qualities, brown trout
have been introduced in places far beyond their original distribution of Europe, northern
Africa, and north-westem Asia. Introductions to North America began in 1883 when

eggs were dispatched from Germany to a hatchery in New York, followed shortly

12

thereafter by more eggs from England (Frost and Brown, 1967). The fish reared from
these eggs were released into Canadian waters and spread further into Canada and the
United States and represent the forerunners of modern North American brown trout

populations.

The brown trout is a species of commercial, recreational, and ecological
significance. Highly desired in sport fisheries, S. trutta bolsters the tourism of a number
of European countries (Laikre et al., 2000). In addition, commercial fisheries produce
brown trout for stocking, either to found new populations or boost the size of existing
natural populations. In terms of ecology, S. trutta fills a unique aquatic niche in its native
habitat. In Scotland, larvae of the freshwater pearl mussel, Margaritifera margaritifera,
live on the gills of the brown trout (Young and Williams 1984 a,b). This element of the
freshwater pearl mussel’s reproductive strategy is absolutely critical in order to produce
the next generation. In addition to its interaction with other species, the brown trout
embodies an exceptionally genetically sub-structured species (Allendorf et al., 1976;
Bouza et al., 1999; Crozier and Ferguson, 1986; Fahy, 1989; Ferguson, 1989; Ferguson
and Taggart, 1991; Hindar et al., 1991; Karakousis and Triantaphyllis, 1990; Osinov and
Bematchez, 1976; ProdOhl et al., 1992). For instance, Hindar et a1. (1991) found
significantly large genetic differences between landlocked populations of Norwegian
brown trout and their counterparts with access to the sea. The rare between-population
diversity has captured the attention of conservationists, as brown trout habitats face
depletion and influxes of hatchery fish into wild populations threaten to decrease genetic

diversity (Laikre et al., 2000; Meffe, 1986; Togan et al., 1995).

13

Declining brown trout abundance and distribution throughout its native range
exemplify recent focal points in fisheries conservation. Many brown trout populations
have been destroyed in the past century, decreasing the availability of resources for
angling and aquaculture (Ferguson et al., 1989). The most frequent causes of
environmental degradation include overexploitation, pollution, loss of genetic diversity,
and destruction of habitat (Allendorf, 1988; Frankel, 1974; Laikre et al., 2000). Diversity
may be lost when large numbers of hatchery fish of limited genetic variation are mixed
with individuals from wild populations (Allendorf, 1991; Garcia-Marin et al., 1991;
Hindar et al., 1991; Taggart and Ferguson, 1986). The influx of many genotypes can
cause genetic ‘swamping’ of wild populations, causing wild populations to become more

and more like the hatchery fish with each passing generation.

The Current Study

The current study utilizes the forensic methodology of STR-based parentage
analysis of a cohort of brown trout recruited from adults introduced into Hunt Creek,
Michigan, USA. The fish introduction represented a means of evaluating a stocking
‘prescription’ for brown trout, and ultimately as a response to local demand for brown
trout sport fishing. Hunt Creek is the inland fish ecology research area for the Michigan
Department of Natural Resources. Long-term research has been conducted on a native
salmonid species, the brook trout (Salvelinus fontinalus). Introductions of brown trout

were performed in order to shift the species composition within the research area.

Proper management practices include not only meeting the demands of the public,

but also establishing a self-sustaining, healthy population. In order to develop a

14

flourishing population, the founding group of fish should exhibit sufficient genetic
diversity (Cowx, 1994). Thus, the founding group was previously genotyped to ensure
genetic heterogeneity. Determining the most efficient means of achieving the
aforementioned goals of public satisfaction and population health and sustainability is
also an integral part of proper management practices. Efficiency of stocking may depend
on such factors as the source of the stocked individuals and their physical phenotypic
characteristics, such as body size (Stanfield and Jones, 2003). Mate selection by body
size may further contain a sex-related component, with one sex choosing a mating partner
because it is larger than other potential mating partners (Bekkevold et al., 2002;

Blanchfield et al., 2003).

In this study, parentage analysis was used to evaluate the outcome of stocking
strategies designed to maximize the genetic diversity of progeny produced by stocked
adults. Specifically, the objectives of this study were: 1. to apply the methods of forensic
parentage analysis to a non—human case study; 2. to maximize the genetic diversity of the
parental generation and hence, the offspring produced following stocking; 3. to determine
the reproductive success of stocked adults, as well as the overall reproductive success of
each source location; 4. to use reproductive success, number of mating partners, and
origin of mating partners to characterize the species’ mating behavior; 5. to use the
information gleaned from measures of reproductive success and mating behavior to offer
management suggestions to guide future stocking efforts; and 6. to evaluate whether
forensic parentage analysis can serve as a valuable tool in fisheries management. Genetic
diversity, reproductive success, and mating behavior are critical informational

requirements in designing stocking strategies.

15

MATERIALS AND METHODS

Study Area

The study area is located within Hunt Creek Research Station, Michigan and
consists of a 2.5-mile stretch of Hunt Creek (Figure 1). The research station strives to
preserve, maintain, or restore environmental quality and population characteristics,
observe changes in habitat or population numbers, and to communicate and establish
policy for Michigan fisheries. Located in the Lower Peninsula of Michigan, the station
consists of 3,000 acres and contains approximately 5 miles of freshwater streams and
lakes. The research section of Hunt Creek is divided into segments, three of which were
stocked with brown trout—sections A, B, and Z. These segments of the creek are not
physically separated from each other and fish can move freely from one section to

another.

Stocking Timeline

Forty-two reproductively mature brown trout were introduced into Hunt Creek in
September of 2001. Fish originated from Houghton Creek (‘B’, N =14), East Branch Au
Sable River (‘C’, N=14), and Gilchrest Creek (‘D’, N=14), all of which are drainages in
the northern lower peninsula of Michigan (Figure 1). Individuals received distinctive
location-specific fin clips. Throughout the study a number and letter code was used to
identify each fish. The first letter corresponds to the source location and the number
identifies each individual within that source group (e.g., B8 is the eighth individual from

Houghton Creek).

16

 

 

 

 

 

 

 

 

3'5:
“’5 ' ‘
O ‘
‘0
c
\ .
\I '
n. u
‘ . - .‘f I
0% _ . on", ’.
n ‘Q..."f‘ T
.' .' a, -‘.
.. ‘w. I a
'5: ‘
Z 'r- e. . -‘ ~
v . . .1 I"
W . . .‘ . I}; s
”‘5 n 'I u'
’ of. ..-r
.. ,A .n
- g - .; ‘ ""
- .~;'.. _ I
‘ ' “I .Is.‘ *
..I; .-
D l‘. \
a. -* ‘

 

   

 

Michigan Department of Conservation 3%

Fish Division
HUNT CREEK FISHERIES EXPERIMENTAL AREA
Montmorency County

 

 

 

- sin: CABIN ........
univnz cum: AP..n: »
.---— .

 

. E V '. .'
‘ 3' .
. u
on! My L/
LEGEND
. .. LAaquvgm..pEyoewge .-.u-= ~ "(er rnAIL
b t caste CENSUS swam m » ”mars LAND
. n < our auvrn
L S h

STREAM Mic HON MmDAiIE'S
, .z

s
n + e as" Inuuaguegu 59mm “. I m “J. I
E «Imam um. Ina.

Figure 1. Hunt Creek research area, northern Lower Peninsula of Michigan. a. The Hunt
Creek Research Station’s location— Montmorency County. b. The research section of
station, split into segments A, B, and Z.

 

 

In addition, 41 mature adults, identified by ‘A’ and ‘E’, were transplanted from a
different area of the Hunt Creek drainage into the research portion of the creek. All fish
introduced into the research area were stocked simultaneously and at the same location.
Hunt Creek already contained a small but unknown number of resident brown trout
(identification label ‘HC’) in the research area, some of which were collected and

sampled to be included in the parentage analysis.

Data Collection

In the summer of 2002, biologists from the Hunt Creek Research Station made
collections of young of the year (YOY) juveniles for genetic analysis. Hunt Creek
resident adult brown trout were collected at the same time as the YOY. However, not all
of the resident Hunt Creek adults were collected and sampled. Brown trout spawn for a
period of approximately two to three weeks in the fall and winter, and the progeny reach
the “fry” stage, or feeding stage, as early as mid-March of the following year. The
present study utilized the collections made in the summer of 2002, which would include
progeny of Hunt Creek resident or adult fish stocked in the fall of 2001. Progeny and
resident brown trout identification labels contain ‘HC’ and also indicate the letter of the
research area section from the individual was sampled (e. g., HCA- 43 is a Hunt Creek

resident or offspring individual sampled in section A).

Collections of juveniles (N=109) and residents (N=16) were performed using
electrofishing and fish traps (Dolan and Miranda, 2003; Miranda and Dolan, 2003;

Miranda and Dolan, 2004; Nielsen et al., 2003; Peterson et al., 2004). A fin clip, taken

from each fish collected, served as the sample for genetic analysis. Fin clips were taken

18

from stocked fish before introduction into Hunt Creek (right pectoral clips for fish from
East Branch Au Sable, left pelvic clips from Houghton Creek, and right pelvic clips from
Gilchrist Creek). Dried fin clips were stored in separate paper envelopes at room

temperature.

DNA was extracted from fin clips using a QIAGEN DNeasy Tissue Extraction
Kit®. The extraction protocol consisted of a piece of fin, equivalent to half of the total
fin size, placed in a tissue lysis buffer and Proteinase K, followed by an overnight
incubation at 55° C. After the incubation, the supernatant was submitted to a series of
ethanol and ethanol-buffer washes. Finally, the DNA was eluted in a Tris-buffer (pH 8.5)
to a volume of 75 pl. Quantification of the DNA required the use of a spectrophotometer.
Five microliters of the DNA was diluted in 495 pl of distilled water and the reading taken
at 260)». The DNA was then diluted to 20 ng/ul concentrations in distilled water and
stored at -20°C.

A suite of six microsatellite loci was selected for developing genotypes, including
Omy301 (Wenburg et al., 1996), Sfol (Angers et al., 1995), One9 (Scribner et al., 1996),
Ssa85 and Ssal97 (Hansen et al., 2000), and Og02 (Olsen et al., 1998) (Table 1). These
loci were initially cloned from Rainbow trout (Oncorhynchus mykiss), Brook charr
(Salvelinusfontinalis), Sockeye salmon (Oncorhynchus nerka), Atlantic salmon (Salmo
salar), and Pacific salmon (Oncorhynchus gorbuscha). Loci were selected because they

were found to be polymorphic in other brown trout from Michigan streams.

19

do? £62 25830.82: some .8.“ 2522.8 mom 98 855.835 SEE A 2an

 

U<<P<U<OE<OOO<UOOH ”m

 

 

 

 

 

 

 

o2 I om. cm <95 otooo<ooo<eo<otooo a 83mm
oE<E<oDoDoooo< ”m

c2 I o2 om .5 o<eoo<<oo8oeoooeoo< a 3.3m
E<o<oogoooeoooeo ”m

9: I m: on .5 0505883553 a 5m

Do<o<8oo<o<oeotoe<oo “a
com I m2 on <0 Eoe<<oooooeoooEoDD ”a $5
558085800338 “a
2 _ I ow cm .5 Eo<<oooD<o<<E< a 53:0
EN I can o<ofioeoDeooEoD<ooEEo a m
cum I com 8 <0 95355055823 a mo 0
60v
gum OhfluEQQn—HQH. EOE 9M Ia MV
Eon—mach «5.355 “some: oouosaom non—F— mucou—

 

 

 

 

 

 

20

Each 25 pl PCR reaction contained 10 pmol each of forward and reverse primers,
10X LGL buffer (lOmM Tris-HCL, pH 8.5, 1.5mM MgClz, SOmM KCl, lOug ml'l

nuclease—free BSA, and 0.0025% Tween-20), 100-200 uM dNTPs, 0.5 units of AmpliTaq
DNA polymerase, and 100 ng of DNA. Four of the six microsatellites required 100 uM
dNTPs, while One9 required 150 M and Og02 required 200 uM dNTPs. The
amplification program consisted of two minutes of denaturation at 94° C; 30 cycles of
one minute at 94° C, one minute at each respective annealing temperature (Table 1), and
one minute of extension at 72° C, and one final extension cycle of two minutes at 72° C.

Each reaction was amplified using a Stratagene Robocycler® Gradient 96.

Amplification products were separated on 6% acrylamide gels and visualized
using the Hitachi FMBio II® genotyping system (Resolution: 150 x 150 dots per inch;
Repeat scan: 150 times; Focusing point: -0.4mm). The primers were labeled with either
Hex or Flo, dyes that fluoresce at 585 nm and 505 nm, respectively. Genotypes were
scored manually in comparison with standards for each locus. Standards included
reference individuals of known genotype, selected to demonstrate alleles over a broad
range in size for each locus. Each gel consisted of four standards equally spaced

throughout 32 samples, as well as a standard size ladder placed in the center of the gel.
Analysis

Parentage analysis using the program Cervus 2.0© consisted of three stages. The
‘Allele Frequency’ portion of the program calculates the frequencies of each allele for
each locus and uses those allele frequencies to calculate expected heterozygosity, conduct

a test of genotype frequency deviations from Hardy-Weinberg expectations and estimate

21

null allele frequencies. This portion of the program required the input of files containing

all population genotypes.

Calculations made during the ‘Allele Frequency’ portion are then input into the
‘Simulation’ step of the program, which uses the allele frequencies to generate possible
genotypes, assuming no deviations from Hardy-Weinberg and no linkage between loci.
Simulated multi-locus genotypes are used to conduct 10,000 replicated parentage
analyses to generate expected rates of successful parentage assignment at strict (95%) and

relaxed (80%) confidence levels.

The final stage of Cervus 2.0© is ‘Parentage Analysis’. Input files are separated
into three genotype files, consisting of a parental female file, a parental male file, and a
progeny file. One sex of putative parental individuals is first used as the ‘Known Parent’
and the other as the ‘Candidate Parent’. Due to the nature of this study, where neither
parent is known, the data had to be re-analyzed by switching the sexes used as ‘Known
Parent’ and ‘Candidate Parent’ and examining whether parental pairs were assigned with
equal confidence in both parentage analyses. In addition, sexes were known for only the
stocked individuals from Houghton Creek, East Branch Au Sable River, and Gilchrist
Creek. So those individuals for whom sexes remained unknown were included in both

the parental female and parental male files.

In addition to Cervus 2.0©, chi-square (X2) tests were performed in order to detect
significant deviations of observed values from expected values. The chi-square test was
specifically applied in detecting differences between the reproductive success of males

and females and differences in reproductive success between population sources.

22

Cervus 2.0© Output

Parentage analysis produced output listing each progeny individual and all
possible parental pairs that could have yielded that individual. The only putative parental
pairs of interest, those demonstrating no mis-matches with the progeny genotype,
received rank and assignment based on their LOD score, Delta score, and statistical

significance.

Cervus 2.0© produces output in the form of a spreadsheet containing the
necessary information for making parentage assignments (Appendix Table A.1). The
spreadsheet provides the most likely parental pairs with any number of allele mis-
matches, from none to all loci mis-matching. For ease in assignment and due to sheer
volume of putative parental pairs for each individual, all parental pairs containing mis-
matches were eliminated from further consideration as parents. Putative parental pairs
with Delta scores of statistical significance were also isolated since they represent the

most likely parents.
Guidelines for Parentage Assignment

Guidelines for parentage assignment (Appendix Figure B.1) provided an objective
means of determining the most likely parental pair among a number of other parental
pairs with no parent-offspring genotype mis-matches. Individuals with the following
described characteristics received a specific assignment. ‘Direction of Assignment’
refers to the fact that parentage assignment was achieved by using one sex as the known

parent and the other sex as the candidate parent, then running the analysis again with the

23

sexes of known and candidate parents switched. Table 1 in Appendix B provides a

summary of the guideline under which each offspring individual was categorized.

A small number of juveniles received parental assignments based on guideline 4.
The assigned parent is the most commonly found candidate parent amongst all of the
putative parental pairs listed for that particular juvenile. If the ratio of the number of
putative parental pairs containing a common individual to the number of putative parental
pairs not containing that individual was large, the common candidate parent was
assigned. Assignments also considered the level of confidence of candidate parental
pairs containing the common adult. Table 2 in Appendix B contains a summary of the

offspring defined by guideline 4 and the reasons for the assignments made.

24

RESULTS

Allele Frequency Results

The allele frequency estimation, simulation, and parentage analysis components
of Cervus 2.0© yielded the critical values necessary for assigning parentage. Ninety-nine
percent of the 208 total individuals were typed. One putative parent fish could not be
typed at the Sfol locus. The mean number of alleles per locus was 10.33 and the mean
expected heterozygosity was 0.746. The total exclusionary power of the first parent was
over 94%, while the total exclusionary power of the addition of the second parent, given

the first parent was assigned, was over 99%.

For each locus, the total number of alleles, the observed and expected
heterozygosity, concordance with Hardy-Weinberg, and presence of null alleles are
summarized in Table 2. In comparing observed and expected heterozygosity, an overall
excess of heterozygotes was observed. Two loci, One9 and Og02, demonstrated a
significant deviation from Hardy-Weinberg equilibrium in relation to this excess of
heterozygotes. In addition, the estimated frequency of null alleles for each locus was
essentially zero (slightly negative), suggesting no evidence of PCR artifacts that could

affect parentage assignments.

In order to determine the possible source of excess heterozygosity, allele
frequencies were calculated for the all adults, adults by location, and juveniles (Tables 3

a—g). In a majority of comparisons of observed and expected heterozygosity for each

25

frequencies were essentially zero.

group, observed heterozygosities (H0) were higher than expected (HE) and null allele

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N no HE HW Null

Alleles Freq.
Omy301 15 208 0.846 0.822 NS -0.0140
Ssa85 8 208 0.702 0.694 NS -0.0124
Ssal97 8 208 0.712 0.651 NS -0.0474
Sfol 13 207 0.841 0.843 NS -0.0017
One9 9 208 0.740 0.667 ** —0.0694
Og02 9 208 0.865 0.796 ** -0.0500

 

 

Table 2. Estimates of population characteristics for each microsatellite locus (N: sample
size; No. of Alleles: the number of observed alleles for each locus; Ho=observed

heterozygosity; HE = expected heterozygosity; HW= tests for deviations of genotype
from Hardy-Weinberg equilibrium, where NS: not significant; **= p<0.05).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N H0 HE HW Null
Alleles freq.
Omy301 14 99 0.808 0.781 NS -0.0181
883197 8 99 0.747 0.691 NS -0.0475
88385 8 99 0.727 0.681 NS -0.0418
Sfol 12 98 0.827 0.849 NS +0.0099
One9 9 99 0.747 0.654 ** -0.0977
Og02 9 99 0.848 0.795 NS -0.0433
a. Summary measures of genetic diversity for all adults
Locus N o. of N HO HE HW Null
Alleles freq.
Omy301 11 109 0.881 0.846 NS -0.0230
Ssal97 4 109 0.679 0.606 N 8 -0.0558
Ssa85 6 109 0.679 0.708 NS +0.0140
Sfol 10 109 0.853 0.821 NS -0.0233
One9 9 109 0.734 0.669 NS -0.0523
Og02 8 109 0.881 0.789 ** -0.0620

 

 

26

b. Summary measures of genetic diversity for all progeny

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N no HE HW Null

Alleles freq.
Omy301 7 41 0.707 0.676 NS -0.0207
Ssal97 6 41 0.707 0.666 NS -0.0429
Ssa85 6 41 0.683 0.628 NS -0.0587
Sfol 8 40 0.825 0.789 NA -0.0298
One9 6 41 0.610 0.497 NA -0. 1553
Og02 6 41 0.854 0.731 NA -0.0946

 

c. Summary measures of genetic diversity for Hunt Creek transplant adults

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N H0 HE HW Null

Alleles freq.
Omy301 6 16 0.813 0.760 NA -0.0434
Ssal97 4 16 0.563 0.563 NA -0.0209
Ssa85 4 16 0.750 0.611 NA -0. 1532
Sfol 6 16 0.875 0.798 NA -0.0658
One9 5 16 0.688 0.587 NA -0.1 177
Og02 4 16 0.875 0.706 NA -0. 1285

 

d. Summary measures of genetic diversity for Hunt Creek resident adults

 

 

 

 

 

 

 

Locus No. of N HO HE HW Null

Alleles freq.
Omy301 8 14 0.929 0.839 NA -0.0700
853197 5 14 0.643 0.802 NA +0.0949
SsaSS 6 14 0.786 0.788 NA -0.0207
Sfol 8 14 0.857 0.820 NA -0.0369
One9 5 14 0.929 0.664 NA -0.2387
Og02 7 14 1.000 0.810 NA -0. 1344

 

 

 

 

 

 

 

 

 

6. Summary measures of genetic diversity for East Branch Au Sable River
transplant adults

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N HO HE HW Null

Alleles freq.
Omy301 8 14 0.857 0.807 NA -0.0667
Ssal97 6 14 1.000 0.741 NA -0.1844
88385 5 14 0.714 0.738 NA +0.0045
Sfol 7 14 0.857 0.788 NA -0.0721
One9 5 14 0.929 0.762 NA -0. 1273
Og02 6 14 0.643 0.709 NA -0.0023

 

f. Summary measures of genetic diversity for Gilchrist Creek transplant adults

27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Locus No. of N HO HF. HW Null

Alleles freq.
Omy301 1 l 14 0.929 0.844 NA -0.0722
Ssal97 6 14 0.929 0.693 NA -0. 1835
Ssa85 6 14 0.786 0.717 NA -0.0654
Sfol 11 14 0.714 0.897 NA +0.1016
One9 7 14 0.857 0.831 NA -0.0347
Og02 7 14 0.857 0.804 NA -0.0471

 

 

g. Summary measures of genetic diversity for Houghton Creek transplant adults

Tables 3 a—g. Summary measures of genetic diversity separated by adults and progeny, as
well as by donor source group (N: sample size; No. of Alleles: the number of observed

alleles for each locus; Ho=observed heterozygosity; HE = expected heterozygosity; HW=
tests for deviations of genotype from Hardy-Weinberg equilibrium, where NS: not
significant; **= p<0.05).

Simulation Results

The simulation estimated expected success rates for assigning parentage when
analyses were conducted with one parent known, as well as neither parent known. Each
level of confidence, the strictest confidence level at 95%, relaxed at 80%, and an
unresolved level at <80% were presented for the parent-offspring scenarios when either
one or neither parent is known (Table 4). For cases in which one parent is known, and
under the strictest level of confidence (95%) the program estimated a success of
assignment rate of 68%. At the relaxed level of confidence (80%) the program estimated
a success of assignment rate of 100%. For cases in which neither parent is known,
expectations were dramatically lower, with an estimate of only 9% success of assignment
at the strictest level of confidence and a success rate of 34% at the relaxed level of

confidence, leaving an estimated 66% of progeny with no parents assigned.

28

 

 

 

 

 

 

 

 

 

 

 

Simulation Summary Statistics—Asflnment Success Estimates
Critical values and expected success rates (one parent known)
Confidence Confidence Delta Expected
Level (%) Criterion Success Rate
Strict 95.00 0.86 68%

Relaxed 80.00 0.00 100%
Unresolved 0%

Critical values and expected success rates (neither parent known)
Confidence Confidence Delta Expected
Level (%) Criterion Success Rate
Strict 95.00 1.86 9%

Relaxed 80.00 0.90 34%
Unresolved 66%

 

 

 

 

 

 

Table 4. Simulation summary measures of parentage assignment accuracy.

Levels of confidence were based on the distribution of simulated Delta scores
(based on known population allele frequencies). The strictest level was set at
95%, and the relaxed level was set at 80%. Any parentage assignment made
below the 80% confidence level was considered unresolved. The Delta Criterion
is the minimum Delta value defining the strict and relaxed confidence levels. The
Expected Success Rate refers to the expected rate of successful parentage
assignment at each confidence level.

Parentage Analysis Results

Over 52% of the progeny could not be assigned to even one parent (Table 5).
These individuals could not be assigned because either no sampled parental pairs were
compatible (21.10%), many parental pairs were compatible with offspring genotypes but
none were of statistical significance (9.17%), or too many statistically significant pairs
were given and a most likely pair could not be distinguished (22.02%). Nearly 25% of
the progeny could be assigned to one parent because one putative parent individual was
common to all or the vast majority of the potential pairs listed (see ‘Guidelines for

Parentage Assignment’, Appendix Figure B.1). Finally, almost 23% of progeny

29

individuals could be assigned to a single parental pair at either the strict or relaxed

confidence levels.

 

 

 

 

 

Number
Number of Unassigned Juveniles of Number of Juveniles with Both
Juveniles Parents Assigned
with One
Parent
Assigned
N o No Too many One pair, but
possible statistically statistically not 80% 95%
parental significant significant statistically
pairs pairs pairs significant
23 10 24 27 8 4 13
21.10% 9.17% 22.02% 24.77% 7.34% 3.67% 11.93%

 

 

 

 

 

 

 

 

Table 5. Assignment summary from parentage analysis, including the number and

percentage of offspring that were unassigned or assigned to one or both parents at 80%
and 95% confidence levels.

Offspring assignments were utilized to estimate reproductive success of each
putative parent, as well as the overall reproductive success of all adults from each source
location. Considerable variation in reproductive success was observed on a individual
level. The parent with the highest reproductive success was a female, B6, originally
transplanted from Houghton Creek (Figure 2 a—e). Individual B6 could be assigned to
seven of the progeny (Figure 2b). Many adults, 60 in total, contributed no offspring to
the YOY. The mean number of offspring per female was 1.0 (o = 1.80) and the mean

number of offspring per male was 0.81 (o = 1.07).

30

Adults from each donor source contributed proportionally to the offspring
generation relative to the size of each donor source population (Xz=2.902, p>0.05, d.f.=
4) (Table 6 and Figure 3a). However, 50% of the juveniles could not be assigned to any
of the sampled parents. For donor source populations with known genders, reproductive
success was compared between males and females. Across all donor source groups with
known parental sexes, no significant difference existed between the reproductive success

of males and females (X2: 3.677, p>0.05, d.f.= 2) (Figure 3b).

 

Reporoductlve Success of Hunt Creek Transplants

 

“8
.57
3-6
55
.3
32
E
:1
z0
\hfib‘bb‘b‘b‘b‘b'x ’\\
vrvsssseeeee‘es

Parental ID

 

 

 

a. Estimated reproductive success of parental brown trout from Hunt Creek transplants,
N=41.

31

 

Reproductive Success of Houghton Creek Transplants

Number of Ottsprlng
N 00 b- 0! 0') \I O)

_A

 

O

BI B6 BB 39 B10 812 BIS B2 B3 B4 BS B7 B11 B14

‘ Females l l Males |
Parental ID

b. Estimated reproductive success of parental brown trout from Houghton Creek
transplants, N=14.

 

 

 

 

Reproductive Success of East Branch Au Sable
Transplants

Number of Offspring

 

C102 C4 06 C7 C8 09 CS CS C10 C11 012 C13 C14

I Females I l Males l
ParentlD

 

 

 

c. Estimated reproductive success of parental brown trout from East Branch Au Sable
River transplants, N=14.

32

 

Reproductive Success of Gilchrist Creek Transplants

Number of Offspring

 

D1 03 D4 D7 D8 D9 011 D2 D5 D6 DIO D12 D13 D14

I Females I Parent ID I Males I

d. Estimated reproductive success of parental brown trout from Gilchrist Creek
transplants, N=14.

 

 

 

 

 

Reproductive Success of Hunt Creek Residents

 

2'33

S6

3 5

- 4

3 3

3 2

E1

3

z o
N‘s-‘3@fiy%yge°e\>°-’e°e‘a‘ka‘5:9$v°fl>
evevr‘ewvoc'o'e'e'ttt
Qg) Q9 ‘29 go go go go go go go go go go .29 .29 .29

Parent ID

 

 

 

e. Estimated reproductive success of parental brown trout resident to the research area of
Hunt Creek, N=l6.

Figure 2 a—e. Estimated reproductive success of each stocked adult. Bars represent the
number of progeny to which each adult contributed. a. Reproductive success of
transplanted adults from Hunt Creek, all of unknown gender. b-d. Reproductive success
of transplanted adults from Houghton Creek (b), East Branch Au Sable (c), and Gilchrist
Creek (d). The first seven individuals for each source location are females and the second
seven are males. e. Reproductive success of Hunt Creek residents, all of unknown sex.

33

Goa—mama 355$ 92 EB 28 Sony wctmmto nocwumm mm 05 we =« 2 95%

8:8 05 mo gown—stucco 88GB metamto Boa—mama :32 .«o 2582. SE. .weramto Boosted 35 95%
cocoa some 555 22636.: no 598:: 05 $5858 wetmmto 53» $5636.: mo “coupon 05. 95%
coco—u 968%»: 05 E53» 388 one 9 noewmmme on 238 85 metamto we confine 05 8 388 wocwmmma
metamto .«o 528:: of. dogwoo— 858 some Soc 3083 0362558 3:23 we 8383mm .0 035.

 

 

 

 

 

 

 

 

 

 

 

 

3.523%
«US$06 650W?“ wmvd 5 ©— x080
=5: -0:
053.2 exeowdv amod 2 3 x080
EEO—5 ID
Swwb— Scwdv omod 2 3 03am =<
Seam seem -O
Swmd. axe—5mm >36 9 3 x85
=oszom -m
Sewage;
0500:» eSmodm mowd mm 3 x85
:5: -m as <
wetnmflc S mew—mag 8.52
23:35.80 as. wereeeo 3 Ce
finch. £55355 Eavgm—E—hfihamg 23355—30 Lea—=52 ”558?—
.«o Begum no Eugen me .328: Z 532 no 3:552 39—. eon—5m

 

 

34

 

Reproductive Success Across Donor Sources

Individual

 

Mean Number of Offspring Per

A&E- Hunt B- Houghton C- Au Sable D- Gilchrist HC- Hunt
Creek Creek River Creek Creek
Transplants Residents

Donor Source

 

 

 

a. Reproductive success for each donor source group.

 

Reproductive Success by Donor
Source and Sex

 

 

 

 

 

 

 

or 10

.E

is. 8

g

o 6

:3 4

E 2

3

Z O

B- Houghton C- Au Sable D- Gilchrist
Creek River Creek

Donor Source
I Number of Offspring by Females
Number of Offspring by Males

 

b. Reproductive success for each donor source for which sexes are known.

Figure 3 a—b. Estimates of reproductive success. a. Reproductive success for each
donor population as represented by the mean number of offspring per individual.
b. Reproductive success for donor populations of known sexes is represented by
the number of offspring to which females and males contributed.

35

Based on parentage assignments, the number of mates per male and female were
determined. Both males and females exhibited mating with more than one partner
(Figures 4 a—b). Males appear to be more likely to mate with more than one female
(polygamy) than females were to mate with more than one male (polyandry). However,

the number of partners per female and number of partners per male did not differ

significantly (X2: 1.845, p>0.05, 2 d.f.).

 

Number of Partners for Each Female I

 

 

 

Number of Partners
N

 

 

82 B5 B11 C7 CB D1 D4 D7 DB 09
Female 10

 

 

 

a. Number of mating partners for each known female.

36

 

Number of Partners for Each Male

 

 

 

 

Number of Partners

 

 

 

B1 36 CS C10 013 C14 DZ D6 D10
MaIeID

 

 

 

b. Number of partners for each known male.

Figure 4 a—b. Number of mating partners. Only individuals of known sex, those
from Houghton Creek, East Branch Au Sable River, and Gilchrist Creek donor
populations, were considered. a. Minimum number of partners per known female.
b. Minimum number of partners per known male.

All but four individuals of unknown gender mated with only one partner, and
genders were inferred when a mating occurred with an individual of known gender
(Figure 5 and Table 7). Sex of individuals who mated with another individual of

unknown gender could not be inferred.

37

 

 

Number of Partners for Parents of Unknown Sex

 

 

 

Number of Partners

 

 

Parent ID

 

Figure 5.

Minimum number of partners per parent for adults of unknown sex.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unknown Gender Inferred Sex
Parent ID
A4 F
A7 Unkown
A8 F
A9 M
A 17 F
A20 Unknown
A23 M
A27 Unknown
A3 1 M
A33 F
A34 Unknown
A35 M
A37 Unknown
El Unknown
E2 F
HCA-47 Unknown
HCA-48 Unknown
HCA-Sl F
HCB-SO Unknown
HCB-52 Unknown

 

 

 

 

Table 7. Inferred sexes of parents based on pairing with
individuals of known sex.

38

 

 

Due to an interest in determining whether adults would mate randomly or
assortively with respect to source location, the number of mating events between
individuals from the same donor source was compared to the number of mating events by
individuals of different donor sources. For all donor sources, with the exception of Hunt
Creek transplants, matings among individuals of differing donor populations appear far
more common than matings among individuals from the same donor population

(X2=4.006, p<0.05, d.f.=4) (Figure 6). Results were consistent with random mating

given the predominance of potential inter-location possibilities for mating.

 

Donor Source of Mating Partners

Percentage of Individuals

 

Hunt Creek Houghton Au Sable River Gilchrist Creek Hunt Creek
Transplants Creek Residents

Donor Source

 

 

Figure 6. Source origins of mating partners. The percentage of individuals mating with
another individual of the same source (white column) is compared to the percentage of
individuals mating with an individual from a different donor source (black column).

Correlation analysis was conducted between adult body size (length in inches)

and reproductive success. No significant correlation was found for either females or

39

males [Rzz 0.003 (y = -0.0157x + 1.0661) for females and R2: 0.078 (y = -0.0985x +
2.7363) for males (Figure 7a)]. When adults with no assigned parentage were removed,
no significant correlation was observed [R2= 0.0007 (y = -0.0145x + 2.5569) for females

and R2: 0.1536 (y = -0.1295x + 3.924) for males (Figure 7b)].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reproductive Success In Relation
to Body Size and Sex
8
2’ 7 I
': 6
n.
g 5 e
'8 4 I
§ 3 e OReproductive
E 2 Success in
g RelatIon to Body
0 - . I Reproductive
10 15 20 25 30 Success by
Body Size
Length (In.) (Males)

 

 

 

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Body Size and Reproductive Success by Sex (Zero-
offspring Individuals Removed)
8
.E ‘
a 6
e 5 s
O
.5 4 I
. 3 . 0 Body Size and
.3 Reproductive
g 2 ‘ Success of
z 1 H I 4 Females
0 . . In Body Size and
10 15 20 25 Reproductive
Length (i n.) Success of Males

 

 

 

 

Figure 7 a—b. Body size (length in inches) and reproductive success. a. Reproductive
success in relation to body size for all stocked females (R2: 0.003, p>0.05) and males
(R2: 0.078, p>0.05). b. Reproductive success in relation to body size for both females

(R2: 0.0007, p>0.05) and males (R2: 0. 1536, p>0.05), when individuals who produced
no offspring were removed.

41

DISCUSSION

The present study used parentage analysis to detect factors influencing
reproductive success and to characterize the mating system of brown trout introduced into
Hunt Creek, Michigan. Information regarding the reproductive success and mating
system of these brown trout can be used to adjust future stocking strategies. Molecular
data were collected for a total of 208 total individuals, 99 putative parents and 109
juveniles (YOY). The YOY represented a large proportion of the first juvenile cohort
produced in Hunt Creek following stocking (Personal communication, A. N uhfer).
Information regarding the number of adults contributing to the offspring recruited
provides an estimation of levels of relatedness among offspring and probabilities of

inbreeding in future generations.
Allele Frequency Results

Evaluation of the molecular data revealed negative null allele frequencies for the
vast majority of loci across all adults and progeny. These negative null allele frequencies
resulted from excess heterozygosity in both the parental and offspring generations. Since
null alleles would result in an excess of homozygosity, the data provide little or no

evidence for the presence of null alleles.

Genotype frequencies at two loci, One9 and Og02, demonstrated a deviation from
Hardy-Weinberg expectations due to an excess of heterozygotes. This deviation is
presumably an artifact of stocking genetically differentiated adult populations.

Geographically separated populations may be expected to exhibit high frequencies of

42

fixed alleles, with the fixed alleles of interest differing between populations. Since
brown trout have been introduced to North America via a series of founding events,
excess homozygosity is particularly expected. In this case, however, all donor
populations displayed extremely high amounts of genetic heterzygosity. Brown trout, as
a species, exhibit high levels of genetic differentiation among geographically isolated
populations (e.g. Allendorf et al., 1976; Bouza et al., 1999; Crozier and Ferguson, 1986;
Fahy, 1989; Ferguson, 1989; Ferguson and Taggart, 1991; Hindar et al., 1991;
Karakousis and Triantaphyllis, 1990; Osinov and Bematchez, 1976; Prodohl et al., 1992).
Given that current North American brown trout populations were generated from
multiple European stocks, the interbreeding between differentiated source stocks may

have resulted in high heterozygosity in subsequent generations.

Simulation Results

Though body size was recorded for all adults in the parental generation,
incomplete records were available on the sex of some stocked adults and all resident
adults. Lack of information decreased the power of the assignment tests because adults
of unknown sex were treated both as males and females. Furthermore, individuals of
unknown sex could not be included in any data summarization requiring separation by

sex, leaving small sample sizes.

Cervus 2.0© requires sexes of potential parents be known and is better able to
assign parentage with high probability when the identity of one parent is known
(simulation results, Table 4). As a program designed to perform parentage analysis,

Cervus 2.0© exhibits some limitations based on the availability of information. In this

43

case, with the identity of neither parent known, the simulations conducted in Cervus 2.0©
estimated success rates for parental assignment at the strictest level of confidence to be
only 9% and at the relaxed level, 34%. Empirically, the actual success rate at each of
these levels was 11.93% and 15.60%, respectively. Empirical success rates differed from
those expected and represent a considerably lower rate of success than if all sexes of the
parental generation were known with certainty, and if one parent had been known for
each juvenile. Only slightly more than half (55.77%) of the assigned offspring had one
or both parents of known sex. Again, this difference between the expected and observed

success rates may have been caused by limitations in available information.

Ecological Applications of Parentage Analysis

Results from this study demonstrate trends in male and female reproductive
success and mating strategies. The most successful individual was B6, a female
originating from Houghton Creek, who contributed to at least seven offspring. However,
Hunt Creek residents appeared to be the most reproductively successful group when
source populations were considered collectively. More than 50% of juveniles could not
be assigned parents and were thus most likely produced by unsampled resident adults. A
portion of unassigned juveniles would likely be assigned to genotyped adults given
greater statistical power, for example, if additional loci were scored. However, over 21%
of the unassigned juveniles could not be associated with any genotyped adults, suggesting
that Hunt Creek residents exhibited a relatively higher reproductive success when
compared to each of the other source groups. This success is potentially a consequence
of familiarity with the Hunt Creek environment (Cowx, 1994; Hansen et al., 2000). This

expectation particularly applies to male residents, who should have already built nests

44

and established territories prior to the introduction of stocked adults. However, an
important finding was that all donor source stocks still contributed toYOY recruitment.
Further, all donor source stocks contributed approximately equally relative to the size of

the introduced adult population.

Across all source locations, reproductive success was not correlated with body
size for either males or females (Figures 7a and 7b). A number of studies have found,
contrary to previous ideas based on behavioral observations, that reproductive success
does not correlate to body size in brown trout and other related species (Blanchfield et al.,

2003; Bekkevold et al., 2002; Garcia-Vasquez et al., 2001; Garant et al., 2001).

Mating strategies in fish often include assortative or preferential mating based on
size, and also commonly includes variation in number of partners. In the present study,
males and females routinely mated with more than one partner. The difference in the
number of males that mated with multiple partners and the number of females that mated
with multiple partners was not statistically significant. Since male brown trout build the
nest, the data revealed that more than one female may lay her eggs in the same male’s

nest, and that a single female frequently allocates her eggs to more than one male.

Stocking Strategy

The parentage assignments and subsequent measures of reproductive success have
direct relevance to stocking strategy for systems similar to that of Hunt Creek.
Information pertaining to probabilities of reproduction given the number of source
populations and individuals stocked is extremely important. Adults from all source

populations contributed proportionally to the juveniles recruited during 2002. Therefore,

45

if maximizing genetic diversity is a desired stocking strategy, the use of multiple donor
source stocks is recommended. However, as Hunt Creek residents represented the likely
group of adults for demonstrating highest reproductive success, supplementation by
stocking may provide little contribution to future generations unless stocking is
comprised of large numbers of fish relative to the size of the wild population. Body size
in brown trout was not significantly correlated to male or female reproductive success,
and may potentially be disregarded in future stocking events. In further
recommendations on stocking strategy, a 1:1 sex ratio may prove ideal, as males and
females exhibited equal reproductive success. Campbell (1977) demonstrated that a 1:1
sex ratio of brown trout yielded the highest reproductive success in a series of stocking

events which varied the ratio of stocked females to males.

Expansion of the Study

Information obtained in the study could have been improved by obtaining sexes
for all adults and increasing the number of loci used, which would increase the statistical
power to exclude incorrect putative parental pairs and increase the success rate of
offspring parental assignments. Use of additional loci for both parents and progeny
would increase exclusionary power. Ideally, maternal and paternal assignments to all
offspring would be based either on total exclusions of all parental pairs or on likelihood
at the highest level of confidence. A large number of potential loci from related species
are available (e.g. Angers et al., 1995; Hansen et a1. 2000; Olsen et al., 1998; Scribner et

al., 1996; Smith et al., 1998; Wenburg et al., 1996).

46

Power of exclusion could also be bolstered if the sex of all adults in the parental
generation were known. Since sexes of Hunt Creek transplants and residents were not
recorded and the samples received were only fin clips, the only means of sexing already-
sampled individuals would be by molecular means. Sex-linked loci have revealed sexes
in a number of different species of fish, some of which are closely related to the brown
trout, such as chinook salmon (Oncorhynchus tshawytscha, Devlin et al., 1989), coho
salmon (Oncorhynchus kisutch, Forbes et al., 1994), and rainbow trout (Oncorhynchus
mykiss, Iturra et al., 1998). This could provide a means of determining adult sex of Hunt

Creek transplant and resident samples used here.

Human Versus Non-human Parentage Analysis

Information limitations encountered in this study are not common to human
forensic parentage analysis. However, human and non-human parentage analysis share
the same conceptual foundation. Assignment of parentage to wildlife species operates
under the same population genetic principles as those implemented in human parentage
analysis. Analysis of human and non-human parentage assumes Mendelian inheritance,
namely segregation and independent assortment, meaning that the product rule can be
applied in generating probabilities. Wildlife and human parentage analysis must account
for the possibility of mutation, relatedness of putative parents, and human error in
genotyping and data entry (Marshall et al., 1998; Schanfield, 2000). Furthermore,
parentage analysis is most profitably based on the use of molecular genetic technology,
usually in the form of STRs (Schanfield, 2000). Statistical analysis requires a reference
population, considering the fit to Hardy-Weinberg expectations, estimations of

heterozygosity, and the power of false parental exclusion.

47

The goals of human and non-human parentage analyses differ greatly. Human
parentage analysis seeks only to identify rightful parents. In contrast, non-human
parentage analysis often seeks to determine parents as a means of exploring additional
ecological and evolutionary questions. In other words, parentage assignment is only the

beginning of the hypothesis testing process.

Further differences between human and non-human parentage analyses lie in the
availability of information. Microsatellites used in human parentage analysis are
standardized and have been isolated directly from human DNA. In contrast, many
microsatellite loci used in non-human parentage studies are not standardized. Also in
non-human studies, microsatellites that have been cloned from one species may be used
in analyses regarding a different, related species. Additional differences appear in
relation to reference populations. Much is known of human populations regarding
geographic locations and racial distinctions, and information is easy to obtain by simple
questioning. On the contrary, little is typically known about non-human populations

unless the species has been studied extensively.

Another major difference lies in the statistical threshold mandated for assignment
of parentage. The American Association of Blood Banks (AABB) requires that human
parentage analysis must have an exclusionary power of 99%, and in some countries that
threshold is even higher (Schanfield, 2000). In the present study, parentage was assigned
in cases that fell within an 80% or 95% confidence level. The power of exclusion in this
study was estimated at only 9% for the strictest confidence level (95%), in the event that
neither parent is known. The assignments made here, while acceptable in molecular

ecology, would not serve as sufficient evidence of parentage in a court of law.

48

Most importantly, the present study differs from most cases of human parentage analysis
due to the unique circumstances under which individuals were obtained. A typical
human parentage dispute involves one child whose mother is known and father who is
uncertain (Schanfield, 2000). In this scenario, the only variable is the genotype of
potential fathers. Assignment of Hunt Creek brown trout parentage was seriously
complicated by not knowing either parent, by the large number of potential parents and
offspring (99 putative parents, 109 progeny), and by not knowing the sex of all putative
parents. Despite these complications, almost 50% of the sampled juveniles could still be
assigned one or both parents, testifying to the power of genetic data to resolve parentage,

even in the absence of critical information.

49

._o>o_ 85.: At oomo ._o>o_ woos—o: At oeow i=3 88m 330 2: £023 35 8:09.38 mo _o>o_ 6302.300

3203 D04 5950: 8:285: .088 SEQ ”SEQ 688 mg: mo EEEwE “DOA 3:83 8:228 :5 395.3 :3an
62 85363: Co 598:: ”wfinofifimmfi n5.0 we “:03: 836:8 ”A: no 603388 .8 baring: £06288

.8 :05 3:83 5505. :SW MET—amt: 5233 62 pea—89:3: no 52:5: ”wfizoafimmfi AUTO ”n: E33 5505— ”9 AM

Me wfiamto ”e 0 i=8 0:08 SEQ @3838: 05 2033 BE _o>o_ 8:02.28 2: w: :95 m: .888 BED :5 Q0..— :65
.wfiamto some :8 who: 13:88 03228 05 3:38: 2:8 23. .oo.~ £3.50 Soc 89:: no ofifiwxm .~.< 29¢. x6:o&<

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

< NuDZmEh<

+ _o-m3.w oo+m$d e co dammed o w< _ _-<U:
co+mood 844:6.— o nm< mompd o :0 ~ T<UI
oo+mood 8+mw _ .m o :< mompd c :0 _ _-<Um
oo+mood oo+mmv.m o wm< mommfim o v0 _ T<Um
Rims: oo+mow.~ o cm mommfim o v0 _ Tao:

... oo+mwo._ oo+mm9v o m< mommfim o VD _ T<Uz
8+mood 8+mm _ ._ o wm< vodcwd o 3100: aka:
8+mood 8+m _ m._ O St. com—own o Elton o-<UI
8+mood 8+mow._ o h _ < voimowfl o 3&0: o-<UE
8+mood oo+m~ _ ._ o wm< 3-Mcw.m o mm< 0&0:
8+mood 8+m _ m._ o nm< vomcwd o wm< o-<UI
oo+mood Rim—cw; o N. _ < 3%.ch o wm< o-<UI
8+mood .o-mno.m o om< 3-9.0.: o wm< a-<UI
oo+mood Sflmws 0 3.5V: ”combos o wm< 943:
8+mood 5&9: c m~< defines o wm< o-<0m
8+mood _o-mmm.m o E< vommtum o om< o-<U:

m-<U:
oo+mood _o-m _ Ow o _m< No-mmo.m o mm v-15:
853:5 as: :3 35385:: E .8 8:225 manages B a. E o
.8. .56 do: 52: moo. .26
5:5: 9..” 52:5

50

APPENDIX B

ASSIGNMENT GUIDELINES

 

H

 

ASSIGNMENT GUIDELINES

. No candidate parental pairs: ‘Unassignable’.

Numerous candidate parental pairs, none of statistical significance:
‘Unassignable’.
Numerous candidate parental pairs of statistical significance:
‘Unassignable’.
Numerous candidate parental pairs of statistical significance, MOST of
which contain one common individual: ‘One parent assigned’.
Numerous candidate parental pairs of statistical significance, ALL of
which contain one common individual: ‘One parent assigned’.
One candidate parental pair, but not of statistical significance: ‘Both
parents assigned by exclusion’.

a. Pair provided by one-directional assignment.

b. Pair provided by both directions of assignment.
One candidate parental pair, statistically significant with 80% confidence:
‘Both parents assigned at 80%’.

a. Pair provided by one-directional assignment.

b. Pair provided by both directions of assignment.
One candidate parental pair, statistically significant with 95% confidence:
‘Both parents assigned at 95%’.

a. Pair provided by one-directional assignment.

b. Pair provided by both directions of assignment.

 

Appendix Figure B.1. Parentage assignment guidelines. Direction of Assignment refers
to the fact that parentage analysis is achieved by using one sex as the known parent and
the other sex as the candidate, then running the analysis again with the sexes switched.

51

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Assignment Guidelines and the Juveniles Defined by Them
Guideline Number of Juvenile ID Juvenile ID
Individuals Group
Defined by (Stream
Guideline Section)

HCA 2,8,l4,16,21,26,29,30,35,38,39

l 23 HCB 10,19,24,26,27,33,37,39,4l
HCZ 2,6,13
HCA 5,10,19,27

2 10 HCB 29,30
HCZ 3,4,14,17
HCA 4,6,11,12,15,28,31,36,37

3 24 HCB 3,5,6,9,18,22,31,35,38,43
HCZ 8,9,10,15,16
HCA

4 7 HCB 4,11,12,21,36,46
HCZ 11
HCA 1,9,13,18,42

5 20 HCB 1,7,8,14,15,16,23,28,32,34,44,47
HCZ 1,5,18
HCA 3,7,33

6 8 HCB 2,40,45,48
HCZ 12
HCA 43

7 4 HCB 17,20
HCZ 7
HCA 17 ,20,22,23,24,25,32,34,40,4l

8 13 HCB 13,25,42
HCZ

 

 

Appendix Table B.l. Assignment guidelines and the juveniles defined by them. For each
guideline the identification label of each juvenile individual assigned under that guideline
is listed. Identification labels are split into the sections of the research area from which
the juveniles were sampled.

52

 

Juveniles Defined by Guideline 4

 

Juvenile ID

Candidate Parental
Pair Ratio

Assignment
Decision

Reasoning

 

HCB-4

16:2

One parent, A8

All at 80% confidence,
pairs not containing A8
did not have a common
individual

 

HCB-ll

7:2

One parent, E1

Mixed 80% and 95%
confidence levels, pairs
not containing E1 did not
have a common individual

 

HCB-12

11:2

One parent, HCB-
5 1

All at 80% confidence,
pairs not containing HCB-
51 contain HCB-53, ratio
was deciding factor

 

HCB-21

17:4

One parent, A7

All pairs containing A7
were at 95% confidence,

while all pairs with A23
were at 80%

 

HCB-36

5:2

One parent, A9

All pairs containing A9
were at 95% confidence,
while all pairs with HCB-
53 were at 80%

 

HCB -46

9:3

One parent, HCA-
47

Mixed 80% and 95%
confidence levels,
deciding factor was ratio

 

 

HCZ-l l

 

11:4:1

 

One parent, B6

 

All containing B6 were at
95% confidence, all
containing A5 are at 80%
confidence, one pair
containing neither B6, nor
A5

 

Appendix Table B.2. Juveniles defined by guideline 4. The candidate parental pair ratio
is the ratio of candidate parental pairs containing the assigned parent to candidate
parental pairs which do not contain the assigned parent. The assigned parent is the most
commonly found candidate parent amongst all of the putative parental pairs listed for that
particular juvenile.

53

 

WORKS CITED

Allendorf, F.W.; Ryman, N .; Stennek, A.; Stahl, G. Genetic variation in Scandinavian
brown trout (Salmo trutta L.): evidence of distinct sympatric populations. Hereditas
83:73-82, 1976.

Allendorf, F.W.; Ryman, N. Genetic management of hatchery stocks. In Ryman, N .;
Utter, F. (eds.). Population Genetics and Fishery Management. Washington Sea Grant
Publications/University of Washington Press, Seattle and London, pp.141-159, 1987.

Allendorf, F.W. Conservation biology of fishes. Conservation Biology 2: 145-148, 1988.

Allendorf, F.W. Ecological and genetic effects of fish introductions: Synthesis and
recommendations. Canadian Journal of Fisheries and Aquatic Sciences 48(Suppl.
1):l78-181, 1991.

Altukhov, Y.P.; Salmenkova, E.A. Stock transfer relative to natural organization,
management, and conservation of fish populations. In Ryman, N.; Utter, F. (eds.).
Population Genetics and Fishery Management. Washington Sea Grant
Publications/University of Washington Press, Seattle and London, pp. 333-343, 1987.

Angers, B.; Bematchez, L.; Angers, A.; Desgroseillers, L. Specific microsatellite loci for
brook charr reveal strong population subdivision on a microgeographic scale. Journal of
Fish Biology 47(Supplement A): 177-185, 1995.

Arroyo, E.; Garcia-Sfmchez, F.; Gomez-Reino, F.; Ruiz de la Cuestra, J .M.; Vicario, J .L.
Prenatal exclusion of paternity by PCR-FLP analysis of VNTR. Journal of Forensic
Sciences 39:566-572, 1994.

Balding, D. Forensic applications of microsatellite markers. In Microsatellites: Evolution
and Applications. Eds. Goldstein, DB. and Schlotterer, C. Oxford University Press, NY,
pp. 198-201, 1999.

Balding, D.J.; Nichols, R.A. Significant genetic correlations among Caucasians at
forensic DNA loci. Heredity 78:583-589, 1997.

Bagliniére, J—L.; Maisse, G. Introduction: The brown trout (Salmo trutta L)—its origin,

distribution and economic and scientific significance. Biology and Ecology of the Brown
and Sea Trout. INRA, Paris, 1991.

Beaumont, M.A.; Bruford, M.W. Microsatellites in conservation genetics. In
Microsatellites: Evolution and Applications. Eds. Goldstein, DB. and Schlotterer, C.
Oxford University Press, NY, pp.165-182, 1999.

Bekkevold, D.; Hansen, M.M.; Loeschcke, V. Male reproductive competition in
spawning aggregations of cod (Gadus morhua, L.). Molecular Ecology 11:91-102, 2002.

54

Bentzen, P.; Olsen, J.B.; McLean, J .E.; Seamons, T.R.; Quinn, T.P. Kinship analysis of
Pacific Salmon: insights into mating, homing, and timing of reproduction. The Journal of
Heredity 92: 127-136, 2001.

Bjerre, A.; Court, D.S.; Lincoln, P.; Morling, N. A report of the 1995 and 1996 Paternity
Testing Workshops of the English Speaking Working Group of the International Society
for Forensic Haemogenetics. Forensic Science International 90:41-55, 1997.

Blanchfield, P.J.; Ridgway, M.S.; Wilson, CC. Breeding success of male brook trout
(Salvinus fontinalis) in the wild. Molecular Ecology 12:2417-2428, 2003.

Bouza, C.; Arias, J.; Castro, J .; Sénchez, L.; Martinez, P. Genetic structure of brown
trout, Salmo trutta L., at the southern limit of the distribution range of the anadromous
form. Molecular Ecology 8: 1991-2001, 1999.

Brooks, M.A. Hardy-Weinberg and other laws of probability. In Silver, H. (ed.).
Probability of Inclusion in Paternity Testing. American Association of Blood Banks,
Arlington, Virginia, pp.1-13, 1982.

Bryant, N.J. Disputed Patemity: The value and application of blood tests. Thieme-
Stratton Inc., New York, 1980.

Butler, J.M. Forensic DNA Typing: Biology & Technology behind STR Markers.
Academic Press, San Diego, CA. 2001.

Campbell, 1.8. Spawning characteristics of brown trout and sea trout Salmo trutta L. in
Kirk Burn, River Tweed, Scotland. Journal of Fish Biology 11:217-229. 1977.

Corach, D.; Sala, A.; Penacino, G.; Sotelo, A. Combined STR VNTR and MVR typing
and mtDNA sequencing, led to the identification of human remains emergent from the
AMIA explosion in Buenos Aires. Advances in Forensic Haemogenetics 6:337-339,
1996.

Couvet, D.; Ronfort, J. Mutation load depending on variance in reproductive success and
mating system. In Loeschcke, V.; Tomiuk, J.; Jain, S.K. (eds.). Conservation Genetics.

Birkhauser Verlang Basel, pp. 55-68.

Cowell, R.G. FINEX: a Probabilistic Expert System for forensic identification. Forensic
Science International 134: 196-206, 2003.

Cowx, I.G. Stocking strategies. Fisheries Management and Ecology 1:15—30. 1994.

Crisp, D.T. Some impacts of human activities on trout, Salmo trutta, populations. F resh-
water Biology, 1989.

55

Crozier, W.W.; Ferguson, A. Electrophoretic examination of the population structure of
brown trout, Salmo trutta L., from the Lough Neagh catchment, Northern Ireland.
Journal of Fish Biology 282459-477, 1986.

Dawid, A.P.; Mortera, J.; Pascali, V.L. Non-fatherhood or mutation? A probabilistic
approach to parental exclusion in paternity testing. Forensic Science International
124:55-61, 2001.

Devlin, R.H.; McNeil, B.K.; Groves, T.D.D.; Donaldson, E.M. Isolation of a Y-
chromosomal DNA probe capable of determining genetic sex in Chinook salmon
(Oncorhynchus tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences
48:1606-1612, 1991.

DeWoody, J .A.; Avise, J .C. Genetic perspectives on the natural history of fish mating
systems. The Journal of Heredity 92:167-172, 2001.

Dobosz, M.; Pascali, V.L.; D’Aloja, E.; Pescarmona, M.; Fiori, A. Exclusion of an
individual charged with rape by allele-specific DNA profiling on fetal tissue. Advances in
Forensic Haemogenetics 3:117-118, 1990.

Dolan, C.R.; Miranda, L.E. Immobilization thresholds of electrofishing relative to fish
size. Transactions of the American Fisheries Society 132:969-976, 2003.

Dykes, D.D. The use of frequency tables in parentage testing. In Silver, H. (ed.).
Probability of Inclusion in Paternity Testing. American Association of Blood Banks,
Arlington, Virginia, pp.15-44, 1982.

Estoup, A.; Rousset, F.; Michalakis, Y.; Comuet, J.-M.; Adriamanga, M.; Guyomard, R.
Comparative analysis of microsatellite and allozyme markers: a case study investigating
microgeographic differentiation in brown trout (Salmo trutta). Molecular Ecology 7:339-
353, 1998.

Evett, I.W.; Weir, B.S. Interpreting DNA Evidence: Statistical Genetics for Forensic
Scientists. Sinauer Assoc., Inc. Sunderland, MA, 1998.

Fahy, E. Conservation and management of brown trout, Salmo trutta, in Ireland.
Freshwater Biology 21 :99-109, 1989.

Ferguson, A. Genetic differences among brown trout, Salmo trutta, stocks and their
importance for the conservation and management of the species. Freshwater Biology
21:35-46, 1989.

Ferguson, A.; Taggart, J.B. Genetic differentiation among the sympatric brown trout
(Salmo trutta) populations of Lough Melvin, Ireland. Biological Journal of the Linnean
Society 43:221-237, 1991.

56

Fiumera, A.C.; DeWoody, Y.D.; DeWoody, J .A.; Asmussen, M.A.; Avise, J .C. Accuracy
and precision of methods to estimate the number of parents contributing to a half-sib
progeny array. The Journal of Heredity 92: 120-126, 2001.

Fiumera, A.C.; Porter, B.A.; Grossman, G.D.; Avise, J.C. Intensive genetic assessment of
the mating system and reproductive success in a semi-closed population of the mottled
sculpin, Cottus bairdi. Molecular Ecology 11:2367-2377, 2002.

Forbes, S.H.; Knudsen, K.L.; North, T.W.; Allendorf, F.W. One of two growth hormone
genes in coho salmon is sex-linked. Proceedings of the National Academy of Sciences
(USA) 91:1628-1631, 1994.

Frankel, O.H. Genetic conservation: Our evolutionary responsibility. Genetics 78:53-65,
1974.

Frost, W.E.; Brown, ME. The Trout. Collins, St. James Place, London, 1967.

Garant, D.; Dodson, J .J .; Bematchez, L. A genetic evaluation of mating system and
determinants of individual reproductive success in Atlantic Salmon (Salmo salar, L.). The
Journal of Heredity 92:137-145, 2001.

Garcia-Marin, J .L.; Jorde, P.E.; Ryman, N.; Utter, F.; Pla, C. Management implications
of genetic differentiation between native and hatchery populations of brown trout (Salmo
trutta) in Spain. Aquaculture 95:235-249, 1991.

Garcia-Vasquez, E.; Moran, P.; Martinez, J .L.; Perez, J.; de Gaudemar, B.; Beall, E.
Alternative mating strategies in Atlantic Salmon and Brown Trout. The Journal of
Heredity 92: 146-149, 2001.

Geada, H.; Ribeiro, T.; Brito, R.M.; Espinheira, R.; Rolf, B.; Hohoff, C.; Brinkmann, B.
A STR mutation in a heteropatemal twin case. Forensic Science International 123:239-
242.

Goddard, K.; Espinoza, E. Wildlife. In Encyclopedia of Forensic Science. Ed. Jay Siegel.
Academic Press, 2000.

Gross, M.R.; Repka, J .; Robertson, C.T.; Secor, D.H.; Van Winkle, W. Sturgeon
Conservation: Insights from Elasticity Analysis. American Fisheries Society Symposium
00:000-000, 2002.

Hallenberg, C.; Morling, N. A report of the 2000 and 2001 Paternity Testing Workshops

of the English Speaking Working Group of the International Society for Forensic
Genetics. Forensic Science International 129:43-50, 2002.

57

Hansen, M.M.; Loeschcke, V. Effects of releasing hatchery-reared brown trout to wild
trout populations. In Loeschcke, V.; Tomiuk, J .; Jain, S.K. (eds.). Conservation Genetics.
Birkhauser Verlang Basel, pp. 273-289.

Hansen, M.M.; Ruzzante, D.E.; Nielsen, E.E.; Mensberg, K-L.D. Microsatellite and
mitochondrial DNA polymorphism reveals life-history dependent interbreeding between
hatchery and wild brown trout (Salmo trutta). Molecular Ecology 9:583-594, 2000.

Hardy, C.J. South Island Council of acclimatization societies. Proceedings Quinnat
Salmon fishery Symposium 2-3 October 1971—Ashburton. N.Z.M.A. Technology Report
8321-298, 1972.

Hindar, K.; Jonsson, B.; Ryman, N .; Stahl, G. a. Genetic relationships among landlocked,
resident, and anadromous Brown Trout, Salmo trutta L. Heredity 66:83-91, 1991.

Hindar, K.; Ryman, N.; Utter, F. b. Genetic effects of cultured fish on natural fish
populations. Canadian Journal of Fisheries and Aquatic Sciences 48:945—957, 1991.

Iturra, P.; Medrano, J .F.; Bagley, M.; Lam, N .; Vergara, N.; Marin, J .C. Identification of
sex chromosome molecular markers using RAPDs and fluorescent in situ hybridization in
rainbow trout. Genetica 101:209-213, 1998.

Jeffreys, A.J.; Wilson, V.; Thein, S.L. Hypervariable ‘minisatellite’ regions in human
DNA. Nature 314:67-73, 1985.

Karakousis, Y.; Triantaphyllidis, C. Genetic structure and differentiation among Greek
brown trout (Salmo trutta L.) populations. Heredity 64:297-304, 1990.

Laikre, L., Ed. Conservation Genetic Management of Brown Trout (Salmo trutta) in
Europe, 2000.

Largiadér, C.R.; Scholl, A. Effects of stocking on the genetic diversity of brown trout
populations of the Adriatic and Danubian drainages in Switzerland. Journal of Fish
Biology 47(Supplement A):209-225, 1995.

Leclaire, B.; Frégeau, C.J.; Bowen, K.L.; Foumey, M. Enhanced kinship analysis and
STR-based DNA typing for human identification in mass fatality incidents: The Swissair
Flight 111 disaster. Journal of Forensic Science 49: 1-15, 2004.

Lee, J .W.; Lee, H-S.; Park, M.; Hwang, J-J. Paternity determination when the alleged
father’s genotypes are unavailable. Forensic Science International 123:202-210, 2001.

Liron, J .P.; Ripoli, M.V.; Garcia, P.P.; Giovambattista, G. Assignment of paternity in a

judicial dispute between two neighbor Holstein dairy farmers. Journal of Forensic
Science 49(1): 1-3, 2004.

58

Lobbiani, A.; Nocco, A.; Vedrietti, P.; Brambati, B.; Colucci, G. Prenatal paternity
testing by DNA analysis. Prenatal Diagnosis 11:343-346 1991.

Loeschcke, V.; Tomiuk, J .; Jain, S.K. Introductory remarks: Genetics and conservation
biology. In Loeschcke, V.; Tomiuk, J .; Jain, S.K. (eds.). Conservation Genetics.
Birkh'auser Verlang Basel, pp. 3-8.

Lynch, M. God’s signature: DNA profiling, the new gold standard in forensic science.
Endeavor 27(2):93-97, 2003.

MacCrimmon, H.R.; Marshall, T.L. World distribution of Brown Trout, Salmo trutta.
Joumal of Fish. Res. Board Can. 27:811, 1968.

Marcotte, A.; Hoste, B.; Fays, M.; De Valck, E.; Leriche, A. Disaster victims
identification by using the DNA technology on dental pulp: preliminary results. Advances
in Forensic Haemogenetics 6:334-336, 1996.

Marshall, T.C.; Slate, J.; Kruuk, L.E.B.; Pemberton, J .M. Statistical confidence for
likelihood-based paternity inference in natural populations. Molecular Ecology 7:639-
655, 1998.

Martin, P.; Alonso, A.; Albarran, C.; Sancho, M. Identification of the skeletal remains of
two 12-year old bodies by nuclear DNA polymorphisms analysis. Advances in Forensic
Haemogenetics 6:301-303, 1996.

Meffe, G.K. Conservation genetics and the management of endangered fishes. Fisheries
11:14-23, 1991.

Mervfig, B.; J acobsen, S.; Olaisen, B. Three intriguing identification cases. Advances in
Forensic Haemogenetics 6:310-312, 1996.

Miranda, L.E.; Dolan, C.R. Test of a power transfer model for standardized
electrofishing. Transactions of the American Fisheries Society 132:1 179-1185, 2003.

Miranda, L.E.; Dolan, C.R. Electrofishing power requirements in relation to duty cycle.
North American Journal of Fisheries Management 24:55-62, 2004.

Morling, N.; Allen, R.W.; Carracedo, A.; Geada, H.; Guidet, F.; Hallenberg, C.; Martin,
W.; Mayr, W.R.; Olaisen, B.; Pascali, V.L.; Schneider, P.M. Paternity Testing
Commission of the International Society of Forensic Genetics: recommendations on
genetic investigations in paternity cases. Forensic Science International 129: 148-157,

2002.
Morris, D.B.; Richard, K.R.; Wright, J .M. Microsatellites from rainbow trout

(Oncorhynchus mykiss) and their use for genetic study of salmonids. Canadian Journal of
Fish Aquacultural Science 53: 120-126, 1996.

59

Morris, J .W. Probability of paternity: Logic 1. In Silver, H. (ed.). Probability of Inclusion
in Paternity Testing. American Association of Blood Banks, Arlington, Virginia, pp.45-
59, 1982.

National DNA Index System. Revised August 2004. Federal Bureau of Investigation.
Accessed September 28, 2004. http://www.fbi.gov/hq/lab/codis/index1.htm.

National Research Council. The Evaluation of Forensic DNA Evidence. National
Academy Press, Washington, DC, 1996.

Neff, B.D. Genetic paternity analysis and breeding success in Bluegill Sunfish (Lepomis
macrochirus). The Journal of Heredity 92:111-119, 2001.

Nielsen, C.; Aarestrup, K.; Norum, U.; Madsen, S.S. Pre-migratory differentiation of wild
brown trout into migrant and resident individuals. Journal of Fish Biology 63: 1 184-1 196,
2003.

Olsen, J .B.; Bentzen, P.; Seeb, J .E. Characterization of seven microsatellite loci derived
from pink salmon. Molecular Ecology 7: 1083-1090, 1998.

Osinov, A.G.; Bematchez, L. “Atlantic” and “Danubian” phylogenetic groupings of
brown trout Salmo trutta complex: genetic divergence, evolution, and conservation.
Journal of Ichthyology 36:723-746, 1996.

Gstergaard, S.; Hansen, M.M.; Loeschcke, V.; Nielsen, E.E. Long-term temporal changes
of genetic composition in brown trout (Salmo trutta L.) populations inhabiting an
unstable environment. Molecular Ecology 12:3123-3135, 2003.

Peterson, J .T.; Thurow, R.F.; Guzevich, J .W. An evaluation of multipass electrofishing
for estimating the abundance of stream-dwelling salmonids. Transactions of the
American Fisheries Society 133:462-475, 2004.

Planes, S.; Lenfant, P. Temporal change in the genetic structure between and within
cohorts of marine fish, Diplodus sargus, induced by a large variance in individual
reproductive success. Molecular Ecology 11:1515-1524, 2002.

Poetsch, M.; Seefeldt, S.; Maschke, M.; Lignitz, E. Analysis of microsatellite
polymorphism in red deer, roe deer, and fallow deer—possible employment in forensic
applications. Forensic Science International 116: 1-8, 2001.

Pohl, B. Probability of paternity calculations: Logic II. In Silver, H. (ed.). Probability of

Inclusion in Paternity Testing. American Association of Blood Banks, Arlington,
Virginia, pp.61-70, 1982.

60

Primorac, D.; Andelinovic, S.; Definis-Gojanovic, M.; Drmic, I.; Rezic, B.; Baden,
M.M.; Kennedy, M.A.; Schanfield, M.S.; Skakel, S.B.; Lee, H.C. Identification of war
victims from mass graves in Croatia, Bosnia, and Herzegovina by the use of standard
forensic methods and DNA typing. Journal of Forensic Sciences 41:891-894, 1996.

Riancho, J .A.; Zarrabeitia, M.T. A Windows-based software for common paternity and
sibling analyses. Forensic Science International 135:232-234, 2003.

Ryman, N. Conservation genetics considerations in fishery management. Journal of Fish
Biology 39(Supplement A):211-224, 1991.

S'ais’a, M.; Kolijonen, M-L.; Tahtinen, J. Genetic changes in Atlantic salmon stocks since
historical times and the effective population size of a long-term captive breeding
programme. Conservation Genetics 4:613-627, 2003.

Schanfield, M.S. Parentage Testing. In Encyclopedia of Forensic Science. Ed. Jay
Siegel. Academic Press, 2000.

Scribner, K.T.; Gust, J .R.; Fields, R.L. Isolation and characterization of novel salmon
microsatellite loci: cross-species amplification and population genetic applications.
Canadian Journal of Fish Aquacultural Science 53:833-841, 1996.

Singh, A.; Gaur, A.; Shailaja, K.; Bala, B.S.; Singh, L. A novel microsatellite (STR)
marker for forensic identification of big cats in India. Forensic Science International
141:143-147, 2004.

Smith, C.T.; Koop, B.F.; Nelson, R.J. Isolation and characterization of coho salmon
(Oncorhynchus kiskutch) microsatellites and their use in other salmonids. Molecular
Ecology 7:1613-1621.

Stanfield, L.; Jones, M.L. Factors influencing rearing success of Atlantic salmon stocked
as fry and parr in Lake Ontario tributaries. North American Journal of Fisheries
Management 23: 1 175-1 183, 2003.

Taborsky, M. The evolution of bourgeois, parasitic, and cooperative reproductive
behaviors in fishes. The Journal of Heredity 92:100-110. 2001.

Taggart, J .B.; Ferguson, A. Electrophoretic evalutation of a supplemental stocking
programme for brown trout, Salmo trutta L. Aquaculture and Fisheries Management
17:155-162, 1986.

Tanksley, S.D.; McCouch, S.R. Seed banks and molecular maps: unlocking genetic
potential from the wild. Science 277: 1063-1066, 1997.

Thangaraj, K.; Reddy, A.G.; Singh, L. Mutation in the STR locus D21811 of father
causing allele mismatch in the child*. Journal of Forensic Science 49(1):1-5, 2004.

61

Thomson, J .A.; Pilotti, V.; Stevens, R; Ayres, K.L.; Debenham, P.G. Validation of short
tandem repeat analysis for the investigation of cases of disputed paternity. Forensic
Science International 100:1-16, 1999.

Togan, I.; Fidan, A.Z.; Yain, E.; Ergiiven, A.; Emre, Y. Genetic structure of two Turkish
brown trout populations. Journal of Fish Biology 47(Supplement A): 164-169, 1995.

Turchi, C.; Pesaresi, M.; Alessandrini, F.; Onofri, V.; Arseni, A.; Tagliabracci, A.
Unusual association of three rare alleles and a mismatch in a case of paternity testing.
Journal of Forensic Science 49(2): 1-3, 2004.

Utter, F.; Ryman, N. Genetic markers and mixed stock fisheries. Fisheries 18: 1 1-21,
1993.

Verma, S.K.; Prasad, K.; Nagesh, N .; Sultana, M.; Singh, L. Was elusive carnivore a
panther? DNA typing of faeces reveals the mystery. Forensic Science International
137:16-20, 2003.

Vuorinen, J. Reduction of genetic variability in a hatchery stock of brown trout, Salmo
trutta L. Journal of Fish Biology 24:339-348, 1984.

Wan, Q.-H.; Fang, S.-G. Application of species-specific polymerase chain reaction in the
forensic identification of tiger species. Forensic Science International 131:75-78, 2003.

Weir, B.S. Genetic data analysis II. 2nd ed. Sinauer Associates, Inc., Sunderland, MA,
1996.

Wenburg, J .K.; Olsen, J .B.; Bentzen, P. Multiplexed systems of microsatellites for
genetic analysis in coastal cutthroat trout (Oncorhynchus clarki clarki) and steelhead
(Oncorhynchus mykiss). Molecular Marine Biology and Biotechnology 5(4):273-283,
1996.

Young, M.; Williams, J. a. The reproductive biology of the freshwater pearl mussel
Margeritifera margeritifera (Linn.) in Scotland: 1. Field studies. Archives of
Hydrobiology 99:405—422, 1984.

Young, M.; Williams, J. b. The reproductive biology of the freshwater pearl mussel
Margeritifera margeritifera (Linn.) in Scotland, H. Laboratory studies. Archives of
Hydrobiology 100:29-43, 1984.

62

 

llililliljlilljljiflillWilli!

 

fiZMT.‘-_—— ———4 A-