UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The origin and maintenance of genetic variation in small populations : coastal cutthroat trout (Oncorhynchus… Costello, Allan 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-267028.pdf [ 19.11MB ]
Metadata
JSON: 831-1.0100382.json
JSON-LD: 831-1.0100382-ld.json
RDF/XML (Pretty): 831-1.0100382-rdf.xml
RDF/JSON: 831-1.0100382-rdf.json
Turtle: 831-1.0100382-turtle.txt
N-Triples: 831-1.0100382-rdf-ntriples.txt
Original Record: 831-1.0100382-source.json
Full Text
831-1.0100382-fulltext.txt
Citation
831-1.0100382.ris

Full Text

T H E ORIGIN A N D MAINTENANCE OF G E N E T I C VARIATION IN S M A L L POPULATIONS: C O A S T A L C U T T H R O A T T R O U T (ONCORHYNCHUS CLARKII CLARKI) A S A M O D E L S Y S T E M  by ALLAN COSTELLO B.Sc. (Hons), Memorial University of Newfoundland, 1997  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE R E Q U I R E M E N T S F O R THE D E G R E E O F  D O C T O R OF P H I L O S O P H Y in THE F A C U L T Y OF G R A D U A T E STUDIES (ZOOLOGY)  THE UNIVERSITY OF BRITISH COLUMBIA December 2006  © Allan Costello, 2006  ABSTRACT  Genetic variation is widely recognized as a major component of species biodiversity, contributing greatly to the maintenance of adaptive potential in populations and their evolutionary response to change. Despite years of study and their obvious importance to evolution and conservation biology, however, the biological processes responsible for the maintenance of genetic variation in the wild are still poorly understood. Throughout this thesis, I have attempted to describe the historical and contemporary forces contributing to the origins and maintenance of genetic variation in coastal cutthroat trout (Oncorhynchus clarkii clarki), a salmonid characterized by small population sizes and one that is of growing conservation concern in western North America. I found evidence for the existence of three, and possibly four glacial refugia for coastal cutthroat trout and a complex pattern of secondary contact between refugial groups in the northern portion of their range. Over subregional scales, both long term and contemporary estimates of gene flow between adjacent populations appear on the order of one migrant per generation, the theoretically optimal level to reduce the loss of genetic diversity in small populations. I also found evidence that interspecific gene flow between coastal cutthroat and rainbow trout (O. mykiss) may be widespread in populations throughout the study area. Finally, through an intensive demographic and genetic study of a single representative population, I found that the breeding system of coastal cutthroat trout may itself compensate for the small, fluctuating number of spawners typical of the subspecies. Mating patterns were found to be quite complex ranging from monogamy to polygynandry, with mating occurring between different cohorts and life history types. My results go further, suggesting that genetic compensation, in the form of an increase in levels of polyandry and a reduction in the variance associated with female reproductive success, partially countered fluctuations in population size. I discuss the key findings of each chapter as they pertain to the maintenance of genetic variation in small populations, the future conservation of coastal cutthroat trout, and, more generally, in terms of applied salmonid management.  iii  TABLE OF CONTENTS Abstract  ii  Table of Contents  iii  List of Tables  viii  List of Figures  ix  List of Symbols  xii  Acknowledgements  xiii  Dedication  xiv  CHAPTER 1: The origin and maintenance of genetic variation in small populations: coastal cutthroat trout (Oncorhynchus  clarkii clarki) as a model  system  1  1.1 Introduction  1  1.1.1 The relevance of genetic variation  1  1.1.2 ... and small population size  2  1.1.3 Small populations with unexpectedly "high" levels of diversity?  4  1.1.4 Some unanswered questions  6  1.1.5 Coastal cutthroat trout: a model paradox  7  1.1.6 Conservation implications  11  1.2 References  14  CHAPTER 2: Individual assignment-based assay of ambient hybridization between coastal cutthroat trout (Oncorhynchus  clarkii clarki) and rainbow  trout (O. mykiss): Implications for conservation  22  2.1 Introduction  22  2.1.1 The population-level effects of interspecific hybridization  22  2.1.2 The coastal cutthroat-rainbow trout system  24  2.2 Materials and Methods 2.2.1  Sampling and DNA extraction  26 26  2.2.2 Microsatellite amplification and scoring  27  2.2.3 Reference populations for species identification  28  2.2.4 Assignment of unknowns/Calculation of hybrid indices (HI)  29  2.2.5 Genetic diversity comparisons  30  iv  2.3 Results  31  2.3.1 Hybrid assignment  31  2.3.2 Effects of hybridization on levels of genetic variation  32  2.4 Discussion  33  2.4.1 Assignment and distribution of hybrids  33  2.4.2 Impact on levels of genetic diversity and population fitness  35  2.4.3 Residual hybridization or ancestral polymorphism?  38  2.4.4 Conservation implications and future directions  39  2.5 References  49  CHAPTER 3: The breeding system of coastal cutthroat trout (Oncorhynchus clarkii clarki) and the maintenance of genetic diversity in small populations  59  3.1 Introduction  59  3.1.1 The maintenance of genetic variation in small populations  59  3.1.2 Effective population size (N ) and the N / N ratio  60  3.1.3 The breeding system of coastal cutthroat trout  62  e  e  3.2 Methods and Materials  65  3.2.1 Study site and field sampling  65  3.2.2 Microsatellite amplification and basic genetic information  67  3.2.3 Parentage analysis and reproductive success  69  3.2.4 Estimates of the effective number of breeders (N ) and the N / N ratio b  b  70  Demographic estimates  71  Genetic estimates  72  3.3 Results  74  3.3.1 Number and composition of spawners  74  3.3.2 Microsatellite and summary genetic data  76  3.3.3 Parentage assignment  77  Female reproductive success.  77  Male reproductive success  78  3.3.4 Correlates of reproductive success and assortative mating  78  3.3.5 Effective number of breeders (Nb)  79  Demographic estimates  79  Genetic estimates  80  3.4 Discussion  80  3.4.1 Confidence in parentage analyses  81  3.4.2 Fluctuations in the number and composition of spawners  83  3.4.3 Inferred patterns of mating in coastal cutthroat trout  86  3.4.4 Variance in individual reproductive success  87  Female reproductive success  88  Male reproductive success  90  3.4.5 Effective number of breeders (N ) and levels of genetic diversity  92  3.4.6 Ecological/ evolutionary implications  95  3.4.7 Conservation implications  98  b  3.5 References  111  CHAPTER 4: Distance, sure... but timing may be everything: Genetic population structure in coastal cutthroat trout {Oncorhynchus  clarkii clarki)  from British Columbia, Canada  125  4.1 Introduction  125  4.1.1 The relevance of population structure  125  4.1.2 Estimating dispersal and gene flow in natural populations  127  4.1.3 Coastal cutthroat trout as a model paradox  129  4.2 Methods and Materials  133  4.2.1 Field sampling and DNA extraction  133  4.2.2 Microsatellite amplification and removal of hybrids  135  4.2.3 Genetic analysis  136  Basic statistics  136  Determination of population structure  137  Estimation of dispersal rates  138  4.2.4 Spatial and hierarchical partitioning of genetic diversity 4.3 Results  139 141  4.3.1 Levels of genetic variability  141  4.3.2 Genetic population structure  143  4.3.3 Estimates of dispersal rates  144  4.3.4 Spatial and hierarchical partitioning of genetic diversity  145  4.4 Discussion  147  vi  4.4.1 Levels of intrapopulation diversity  147  4.4.2 Population structure in coastal cutthroat trout  150  4.4.3 Why then, is there so little gene flow among anadromous cutthroat trout?.... 155 4.4.4 Implications for evolution  160  4.4.5 Conservation implications  161  4.4.6 Unanswered questions/ Future directions  164  4.5 References  176  C H A P T E R 5: Nuclear phylogeography reveals deep divergences and new glacial refuge for coastal cutthroat trout (Oncorhynchus  clarkii clarki) on the  central coast of British Columbia  191  5.1 Introduction  191  5.1.1 Pleistocene glaciations in western North America  191  5.1.2 Coastal cutthroat trout  192  5.1.3 Nuclear phylogeography of coastal cutthroat trout  194  5.2 Methods and Materials  196  5.2.1 Sampling and DNA amplification  196  5.2.2 DNA sequencing and alignment  197  5.2.3 Genetic Analyses  198  Phase estimation and population-level diversity  198  Genetic structuring and delineation of distinct groups  198  Spatial and hierarchical partitioning of genetic diversity  200  5.3 Results  202  5.3.1 DNA sequence diversity at GH2-iD  202  5.3.2 Genetic population structure  203  5.3.3 Spatial and hierarchical partitioning of within-population diversity  204  5.4 Discussion  206  5.4.1 Molecular systematics and the origins of coastal cutthroat trout  206  5.4.2 Evolution of the growth hormone genes and their application to phylogeography  208  5.4.3 Intrasubspecific variation and the Pleistocene history of coastal cutthroat trout  ,  5.4.4 Evidence for the use of multiple refugia during the Pleistocene  210 213  vii  5.4.5 Recolonization dynamics and levels of genetic variation  218  5.4.6 Conservation implications  221  5.4.7 Limitations and future directions  223  5.5 References  235  C H A P T E R 6: The origins and maintenance of genetic variation in small populations revisited: concluding remarks and implications f o r s a l m o n i d conservation  ...247  6.1 Nested approaches and the hierarchical nature of biological diversity  247  6.2 Small populations and the maintenance of genetic variation revisited  249  6.3 Summary of findings and implications for salmonid conservation  253  6.3.1 Minimizing the adverse influence of hatchery supplementation  254  6.3.2 Synergistic effects of habitat loss  257  6.3.3 Defining meaningful units for conservation and resource management  259  6.3.4 Application to coastal cutthroat trout in B C  262  6.4 Towards an evolutionary perspective on fisheries management and conservation  267  6.5 References  271  viii  LIST O F T A B L E S  Table 2-1. Summary data for populations sampled in the hybridization assay 41 Table 2-2. Microsatellite loci used in the study; data presented includes data for all samples within populations 42 Table 2-3. Taxon-level comparison of genetic diversity measures between coastal cutthroat trout (CCT) and rainbow trout (RBT) reference populations and the assigned hybrid group (HYB) 42 Table 2-4. Population-level comparison of genetic diversity parameters and hybridization levels by region 43 Table 2-5. Population-level Spearman rank testing of correlations between hybridization level (%HYB) and genetic diversity measures for populations containing hybrids 43 Table 3-1. Characterization of the seven polymorphic microsatellite loci in Chonat Lake coastal cutthroat trout across years (N = 479 different individuals) 102 Table 3-2. Summary genetic data for Chonat Lake spawners and offspring samples from 2001-2002 102 Table 3-3. Female reproductive success (Early vs. Late arriving and combined) 103 Table 3-4. Male reproductive success (fighter vs. sneaker and combined) 103 Table 3-5. Demographic estimates of the number of breeders (N ) 104 Table 3-6. Genetic estimates of the effective number of breeders (N ) and 9 5 % confidence intervals 104 b  b  Table 3-7. Harmonic mean estimates of the number of breeders (N ) and N l N b  b  ratios from 2001 to 2002 for all methods 105 Table 4-1. Summary data for coastal cutthroat trout populations included in the metapopulation-level assay 166 Table 4-2. Microsatellite loci used in the study 167 Table 4-3. Regional comparison of diversity values 167 Table 4-4. Summary of recent migration rates (m) among populations inferred from Bayesian assignment analyses within regions 168 Table 4-5. Hierarchical partitioning of multilocus microsatellite DNA variation based on A M O V A analyses under various hypothetical scenarios 169 Table 5-1. Sample locations and diversity statistics for populations included in the phylogeography assay for both the full GH2-iD sequence (627 nucleotides) and the base sequence (305 nucleotides) 225 Table 5-2. Haplotype definition and summary of the nucleotide diversity for GH2-iD (positions 1-627) 226 Table 5-3. Spatial autocorrelation summary for pairwise F and straight-line geographic distance (km) based on GH2-iD haplotypes for distance class sizes of 50, 100, 200, 400, 600 and 800 km 226 Table 5-4. Hierarchical partitioning of GH2-iD base sequence variation based on A M O V A analyses under various hypothetical scenarios 227 Table 6-1. Summary of the major findings of each thesis chapter as they pertain to the maintenance of genetic variation in small populations 268 st  ix  LIST O F F I G U R E S  Figure 1-1. Comparative analysis of the genetic divergence typical of anadromous salmonid populations with respect to geographic distance (modified from Hendry et al. 2004) 12 Figure 1-2. Overview of the nested design of the study. Ovals illustrate the geographic context for the study and the relative spatial scales involved for each thesis component 13 Figure 2-1. Overview of the sampling locations for coastal cutthroat trout and rainbow trout populations included in this study. Inset shows regional perspective and location of populations in Alaska and the Queen Charlotte Islands. Numbers correspond to those in Table 2-1... 44 Figure 2-2. Log-likelihood based results for coastal cutthroat trout (CCT) and rainbow trout (RBT) assignment, showing distribution of reference samples and assigned individuals. Hybrids have been removed for clarity 45 Figure 2-3. Distribution of Hybrid Index (HI) values for reference samples and assigned individuals; C C T - coastal cutthroat trout, RBT - rainbow trout, HYB - hybrid 46 Figure 2-4. Factorial correspondence analysis (FCA) of the multilocus genetic relations ships between assigned coastal cutthroat trout (CCT), rainbow trout (RBT) and hybrids (HYB) along the first two factorial axes 47 Figure 2-5. Distribution (%) of coastal cutthroat trout (CCT), rainbow trout (RBT), and hybrids (HYB) by sampled population and by region. Numbers on X-axis correspond to population codes in Table 2-1 48 Figure 3-1. Location of the study site and overview of the Chonat Lake system 106 Figure 3-2. Length frequency distribution of Chonat Lake spawners in 2001 and 2002 as measured by fork length (mm). S R M refers to stream resident males.... 107 Figure 3-3. Distribution of the number of inferred mates for Chonat Lake spawners by year. Mean values and the coefficient of variation (CV = s/ X) are shown. The approximate normal curve is given for reference 108 Figure 3-4. Reproductive success of male and female spawners with respect to fork length (mm). Triangles represent stream resident males, squares - fighter males, diamonds - early arriving females, and circles - late arriving females. Spearman r coefficients and corresponding p-values for combined correlations are shown 109 Figure 3-5. Scatterplot of size-assortative mating for inferred mating pairs (male vs. female) by year. Spearman r coefficients and corresponding p-values are shown 110 Figure 4-1. Overview of the sampling locations for the 42 'pure' coastal cutthroat trout populations included in this study. Population numbers correspond to those in Table 4.1 170 Figure 4-2. Factorial correspondence analysis of the multi-locus genetic relationships between coastal cutthroat trout populations along the first two factorial axes. Circles encompass regional groups. Population numbers correspond to site locations in Table 4-1 and Figure 4-1 171 Figure 4-3. Consensus maximum likelihood tree among coastal cutthroat trout samples based on 1,000 bootstrapped data sets in the C O N T M L module of PHYLIP. Bootstrap values greater than 50 percent are indicated 172  Figure 4-4. Pairwise isolation-by-distance correlations between either F t or CavalliSforza and Edward's chord distance (CSE) and shortest water distance between populations (km); for all populations, and excluding above-barriers populations 173 Figure 4-5. Pairwise isolation-by-distance correlations between Cavalli-Sforza and Edward's chord distance (CSE) and shortest water distance between populations (km) for individual regions; a) including all populations and b) excluding above-barriers populations 174 Figure 4-6. Multi-distance class analysis of spatial autocorrelation within each of the four primary regions: a) Clayoquot Sound; b) Strait of Juan de Fuca; c) Strait of Georgia (West); d) Strait of Georgia (East). Triangles represent the observed cumulative class-specific correlation coefficient (r) and stippled lines indicate the 9 5 % confidence intervals for estimates based on 10,000 random permutations under the null. Double asterix indicate marked statistical significance (one-sided p < 0.01), single asterix denote marginal statistical significance (0.01 > one sided p < 0.05) 175 Figure 5-1. Overview of location of sample sites for coastal cutthroat trout showing the locations of several putative refugia in the northern part of the range. Numbers correspond to site locations in Table 5.1 228 Figure 5-2. Relationships between the identified GH2-iD haplotypes; a) consensus U P G M A based on Kimura-2-parameter distances relative to the immediate outgroups of westslope cutthroat (WCT) and rainbow trout (RBT). Tree rooted by Atlantic salmon outgroup sequence from McKay et al. (1996); (b) Inferred median-joining haplotypic network for GH2-id haplotypes (i- full 627 nucleotide sequence and ii. base sequence of 305 nucleotides). The relative frequencies in north (white), central (black), and southern (grey) populations are also given. See text for explanation 229 Figure 5-3. Relative haplotypic frequencies of the four base haplotypes by population and by region. Spearman rank r values are given for the correlation between haplotype frequency and latitude. P-values are based on a permutation process (n = 1000) 230 Figure 5-4. Summary of Bayesian analyses of population structure; (a) Dendrogram representing distribution of posterior probabilities for population-level analyses using B A P S 2.1; (b) Modal value of A K (Evanno et al. 2005) for individual-level Bayesian analyses using S T R U C T U R E . 231 Figure 5-5. Factorial correspondence analysis (FCA) of the multi-locus genetic relationships between coastal cutthroat trout populations along first two factorial axes. Population numbers correspond to site location in Table 5-1. Relative location of the 4 base GH2-iD haplotypes are shown. Population clusters inferred from B A P S are circled and numbered with Roman numerals that correspond with the groupings in Figure 5-4a 232 Figure 5-6. Spatial autocorrelograms for associations between pairwise genetic (F ) and geographic distances (km) for (a) 50 m distance class width and (b) 200 m distance class width. Solid line indicates correlation value (r) as a function of geographic distance. Stippled lines indicate upper and lower 9 5 % confidence intervals based on permutation process under the null assumption 233 Figure 5-7. Hypothetical relationships between the GH2-iD haplotypes 234 s  st  xi  Figure 6-1. Overview of the possible consequences of habitat degradation for coastal cutthroat trout populations. 'LWD' refers to large woody debris, 'F' refers to female. Indication (-) suggests negative relationship 269 Figure 6-2. Possible delineation of major conservation groups for coastal cutthroat trout in BC under (a) the Evolutionarily Significant Unit (ESU) concept (Waples 1991, 1995) and (b) the Designatable Unit concept (Green 2005). Results for the method of Crandall et al. (2000) are not shown 270  xii  LIST O F S Y M B O L S  (Used in Chapter 3, in order of appearance) N  estimated total population size from mark-recapture statistics  N  harmonic mean of 2001 and 2002 N estimates  N  effective population size  e  g  salmon model parameter analogous to t; generation length defined as the average age at maturity weighted by age specific fecundities  N  effective number of breeders (calculated by various methods, see below)  b  N  harmonic mean of 2001 and 2002 N estimates  b  b  Nc  census number of spawners ( N  Nm  number of male spawners observed  Nf  number of female spawners observed  k  mean reproductive success (# Y O Y fry) for V R S estimates  cr  2  k  + Nf)  m  variance associated with mean reproductive success  Nbm effective number of males in V R S estimates N f  effective number of females in V R S estimates  F  Pollack (1983) estimator of the standardized variance in allele frequencies for  b  k  two-point genetic estimates of N b  b  salmon model parameter describing the number of generations between time points that accounts for the age structure in cohorts which make up the population  N  e  estimated value of N assuming grand mean estimate of N applies to the 3-4 e  cohorts making up the generation  b  xiii  ACKNOWLEDGEMENTS  I would like to offer my sincere thanks to the many people who helped me bring this doctoral thesis to fruition. First off, I would like to thank my supervisor Eric Taylor, for giving me the chance to see the "wild west coast" in a way few people ever get to do, and for giving me an opportunity to grow as a biologist and as a person. I would like to thank the members of my thesis committee: Mike Healey, Mike Doebeli, Sally Aitken, and Jordan Rosenfeld for their insight, encouragement and varied perspectives. I would especially like to thank the many friends and labmates who came along with me on my travels through field and laboratory, sharing silent moments and sometimes heated debate on all manner of topics. First and foremost, a big thanks to Patrick Tamkee for his constant help and infectious enthusiasm. "It's all about the question!" Thanks also to Mike Stamford, Ramona de Graaf, Aubrey Hawco, Elaine Goudie, Lisa Johnson, Matt Sniatynski, Emily Rubidge, Katriina lives, and Jen Gow for their encouragement and support. I would also like to acknowledge the many government biologists who assisted with sample collection and technical issues over the years, particularly Sue Pollard and Ted Down at the Fisheries Science Unit in Victoria for their continued support and interest in coastal cutthroat trout conservation. Special thanks to Bob Devlin and the staff at the DFO labs in West Vancouver for their assistance in early stages of this research. Thanks to Ron Ptolemy, Jim Roberts, Pat Slaney, and Rob Knight of the BC Ministry of Environment for sharing their thoughts on cutthroat trout, and to John Wenburg and Kim Hastings of the US Fish and Wildlife Service in Alaska for sharing some of their DNA samples. A big thanks to Debbie Adam and the staff at the N A P S Unit at U B C for technical assistance with what was a considerable amount of genotyping. Finally I would like to acknowledge the funding agencies that supported this research, including the National Science and Engineering Research Council (NSERC), the BC Habitat Conservation Trust Fund (HCTF), the American Fisheries Society (AFS), the University Of British Columbia (UBC) and finally, the Grant B. Culley, Jr. Foundation, which provided financial and logistical support for the field component of the mating system chapter.  xiv  Q^fyeatiheaiaia dedicated to- my wife, ^lainej and ou& dauyAteVj  2ily  Q/fa&cmda... fiw fihew fiatiencej 6acrifece and faye/eu au/ifawt t/wvu<jA good timet andludj at G^&aw- tAi& UHW/O fo ifo end. OS$llmy /owe...  1  C H A P T E R 1: T H E ORIGIN A N D M A I N T E N A N C E O F G E N E T I C VARIATION IN SMALL POPULATIONS: C O A S T A L CUTTHROAT TROUT  {ONCORHYNCHUS  CLARKII CLARKI) A S A M O D E L S Y S T E M  "... evolution depends on a certain balance among its factors. There must be gene mutation, but an excessive rate gives an array of freaks, not evolution; there must be selection, but too severe a process destroys the field of variability, and thus the basis for further advance; prevalence of local inbreeding within a species has extremely important evolutionary consequences, but too close inbreeding leads merely to extinction. A certain amount of crossbreeding is favourable but not too much. In this dependence on balance, the species is like a living organism. At all levels of organization, life depends oh the maintenance of a certain balance among its factors." -Wright (1932:365)  1.1  1.1.1  INTRODUCTION  The relevance of genetic variation...  Since the days of typological thinking and immutable species concepts, the variation existing among individuals has been seen as more than a mere imperfection of the physical world, but as a real and important characteristic of living things. Darwin and Wallace were the first to distill the significance of the variation among individuals, recognizing such variation as the raw material on which natural selection and evolution depend. Since that time, the heritable component of phenotypic variation in populations (i.e.- genetic variability) has been recognized as an essential part of a species' ability to respond to changes in the environment, be they naturally occurring or of an anthropogenic nature. Increased levels of genetic diversity appear to afford populations the flexibility to persist in specific habitats and may also allow them to colonize a larger range of habitats than might otherwise be possible (Vrijenhoek and Lerman 1982; Carvalho 1993; Ellstrand and Elam 1993; Vrijenhoek 1994). Beginning in the 1960's, the introduction of new biochemical and molecular techniques revolutionized the field of population genetics by suggesting that not all variation was necessarily adaptive in nature. The incredible amounts of variation  2  existing within populations (in terms of heterozygosities and number of alleles) were difficult to reconcile with the dominant contemporary paradigm of balancing or diversifying selection. Kimura (1968, 1983) was among the first to suggest the most of the genetic polymorphism being observed in populations had no adaptive significance whatsoever and that for the most part, alternate genotypes were functionally equivalent. Effectively, this shifted the emphasis away from selection and towards mutation and drift as the main determinants of genetic diversity in natural populations. Neutral DNA markers proved more amenable to mathematical and theoretical treatments than markers subject to selection and were able to describe a range of evolutionary processes (Weir 1996). Importantly, the levels of polymorphism at neutral genetic loci are expected to be proportional to the amount of variation existing for quantitative traits (i.e., - those characters more directly involved in population persistence; Frankham 1999; Moran 2002; but see Lynch 1996). Since attempts to measure the strength of selection and the levels of variation at quantitative trait loci remain exceedingly difficult to perform (Falconer 1981; Clegg 1997; Conner 2001), neutral markers are often used to gauge the effects of historical and contemporary processes on levels of genetic diversity in general (but see Pfrender et al. 2000; Reed and Frankham 2001). For the past 20 years, molecular techniques incorporating neutral DNA polymorphism have been used extensively in the fields of phylogeography, population genetics, and conservation biology as a tool for management and in attempts to understand the process of evolutionary change (reviewed b y A v i s e 1994; Haig 1998).  1.1.2 ...and small population size  A s noted, neutral theory shifts the emphasis away from selection and towards mutation and drift as the main determinants of genetic diversity in natural populations. New variation enters a population through mutation; the de novo generation of novel genetic combinations, or through immigration. It is lost from populations through genetic drift (the stochastic sampling of genotypes from populations of finite size). The rate at which genetic variation is lost is expected to be inversely proportional to its effective population size (Wright 1931; Crow and Kimura 1970). The effective population size (N ) represents the size of an idealized Wright-Fisher population (one e  3  of constant size and even sex ratio, discrete generations, random breeding, and random survivorship) that experiences the same losses of genetic variation as the population of interest (Wright 1931, 1938). It has a well-established theoretical basis and often allows for valuable insight into the factors most affecting the demography of populations in the wild. The concept is particularly relevant to the evolutionary ecology of small populations in that N is the single-most important factor influencing e  the amount of inbreeding or genetic drift and therefore, the main determinant of the rate at which genetic diversity is lost (AH ~ 1/2 N ). e  It follows, therefore, that small isolated populations will tend to carry less genetic variability, as they will be more profoundly affected by drift and inbreeding than larger populations (e.g., - Lacy 1987; Hedrick et al. 1996). In a meta-analysis of the allozyme data available at the time, Soule (1976) did, in fact, observe a strong positive correlation between levels of heterozygosity and logarithmic population size for a range of animal species. Frankham (1996) reached similar conclusions: genetic diversity is generally lower in small populations, in populations with restricted ranges, and in endangered species. The ultimate effects of this reduced genetic diversity in small populations are, however, less apparent. While it is generally expected that a positive relationship exists between genetic diversity and fitness (e.g., - Beacham and Withler 1985; Beacham 1991; Vrijenhoek 1994), reduced levels of genetic variability do not necessarily translate into reduced viability or persistence in the wild. Inbreeding depression has been shown to increase extinction rates in laboratory populations of plants, Drosophila, and mice (Waser and Price 1994; Frankham 1995) as well as affecting reproductive rates in captive mammals (Ralls and Ballou 1983), but there are few documented cases in nature of the kind of "extinction vortex" suggested by Caughley (1994); see also (Lande 1988; Hedrick et al. 1996; Saccheri et al. 1998). In many instances, apparently genetically impoverished species appear to be thriving or at least maintaining their numbers in the wild (O'Brien 1994; Hedrick 1995). In some cases, low levels of molecular variation have actually been shown to be associated with increased levels of quantitative genetic variation (e.g., Armbruster et al. 1998).  4  1.1.3  Small populations with unexpectedly "high" levels of diversity?  Amos and Harwood (1998) make an interesting argument that publication biases tend to inflate the amount of DNA polymorphism that is considered "normal" or "healthy". They argue that results for monomorphic loci in species often go unreported as researchers tend to focus on patterns of variation, not necessarily it's absence. This "file-drawer" effect (cf. Conner 2001) tends to bias the levels of reported polymorphism upwards. There are, in fact, numerous examples in the literature of small, isolated populations which appear to be carrying higher levels of genetic variation than might be expected based solely on their population size; e.g., giant kangaroo rats (Good et al. 1997); Atlantic salmon (Moran and Garcia-Vazquez 1998); and the Mauritius kestrel (Nichols et al. 2001). Are publication biases likely to apply in these cases or are these populations somehow able to counteract the effects of small population size and stochastic losses of genetic variability? Balancing selection has often been put forth as a factor contributing to the maintenance of genetic variation in small populations (Endler 1986; Gillespie 1991; Karl and Avise 1992; Hughes and Yeager 1998; Richman 2000). Under balancing selection, levels of polymorphism may be independent of population size, instead depending on the various selection gradients or the environmental factors promoting them. The evidence for balancing selection, however, is often weak and inconsistent (e.g. - Kreitman and Akashi 1995; Conner 2001) and in very small populations, even characters subject to selection should approach effective neutrally as drift becomes the dominant force (Kimura 1983). Mutation can certainly add new genetic variation to populations, but small, isolated populations are not often likely to persist long enough in evolutionary time for mutation to replenish the variation lost to drift. Furthermore, mutations may be difficult to detect over the timeframes involved in typical empirical studies. Kimura and Crow (1963) were among the first to point out the importance of genetic exchange between adjacent populations in maintaining genetic diversity. While high levels of gene flow could essentially cause separate populations to become panmictic (and its absence leave them vulnerable to extinction), occasional  5  gene flow between otherwise divergent populations could prevent local extinction as well maintain levels of genetic variability, both locally and throughout the range (Mills and Allendorf 1996; Vucetich and Waite 2000). Numerous theoretical and empirical studies have indeed shown that more alleles can be maintained in subdivided populations (Maruyama 1970; Good et al. 1997) and that occasional migration and gene flow between distinct populations can maintain higher heterozygosity levels (Lacy 1987; Hamrick and Godt 1996; Harrison and Taylor 1997). The levels of genetic variability within and between populations may, therefore, reflect differences in patterns of movement between populations, both those of contemporary origin and those occurring historically (Templeton et al. 1995; Walter and Epperson 2001). While not believed to generate diversity perse, this type of subdivision is often implicated in the maintenance df genetic variability between populations through its effect on effective population size (Whitlock and Barton 1997). Similarly, episodic or rare hybridization events between taxa (which are essentially another form of gene flow) may, in some cases, benefit local populations by introducing novel adaptations. Anderson and Stebbins (1954) hypothesized that by rapidly producing new recombinant genotypes, introgressive hybridization could allow for enhanced levels of genetic variability in populations. While hybridization has been demonstrated to occur in a wide range of species (Arnold 1997; Dowling and Secor 1997; Scribner et al. 2001; Arnold 2004; Mallet 2005), only recently have researchers considered the effects of hybridization events in the context of population-level genetic processes (e.g., - Ellstrand and Elam 1993; Burke and Arnold 2001). It appears that if successful hybridization events are episodic or rare in nature, then the infusion of new variation and reduction in inbreeding can sometimes be beneficial, especially for small isolated populations. If, on the other hand, hybridization is widespread and recurrent, then the breakdown of locally adapted gene complexes and resulting outbreeding depression may instead reduce population fitness; particularly in taxa that show marked adaptation to local conditions (Gharrett and Smoker 1991; Taylor 1991; Rhymer and Simberloff 1996). This points to the influence of breeding systems in maintaining the levels of genetic variation that exist in populations. The types of mating systems present in  6  nature vary incredibly and represent a continuum from the "chuck it and chance it" types (characterized by environmentally driven gamete dispersal) to those in which parental care may persist throughout an organism's life (reviewed by Reynolds 1996). At the very least, most species have evolved mechanisms to reduce the occurrence of inbreeding in populations (e.g. - dispersal, kin recognition) and often, further social or mating structure results in non-random mating (e.g. - Gross 1996). Theoretical considerations have suggested mating structures that are able to prevent reductions in effective population size in small populations (reviewed by Caballero 1994; Nunney 1999; Wang and Caballero 1999) and recent advances in molecular techniques have allowed researchers to test the generality of these predictions in wild populations (e.g. - Jones and Avise 1997; Taggart et al. 2001; Avise et al. 2002). In doing so, they have enhanced the study of mating systems and provided insights into population and family structuring, breeding strategies and individual reproductive success on a level never before available (Avise 1994; Moran and Garcia- Vazquez 1998; Streiff et al. 1999; Taggart e t a l . 2001).  1.1.4 Some unanswered questions Despite years of study and their obvious importance to evolution and conservation biology, the biological processes responsible for the maintenance of genetic variation in the wild are still poorly understood (Li 1997; Rainey et al. 2000). Mutation, drift and selection are clearly expected to be the major factors influencing the levels of genetic diversity in natural populations; however, historical factors, interactions with the environment and other species, and increasingly, anthropogenic effects also serve to influence the distribution of intraspecific diversity through their effects on population size and connectivity (Mayr 1963; Templeton et al. 1995; Angers et al. 1999; Hewitt 2001; Turgeon and Bernatchez 2001). A thorough understanding of the relative importance of both historical and contemporary factors in shaping extant patterns of geographic variation is ultimately required to answer any questions related to evolutionary change or the conservation of species biodiversity. Unfortunately, few studies to date have been designed to incorporate more than a couple of factors at a time because of the cost and work involved so that their effects cannot be determined relative to other such factors (but see Angers et al. 1999).  7  Fewer still have been able to incorporate both historical and contemporary factors in a single cohesive framework (but see Comes and Abbott 1998; McGlashan and Hughes 2000; Costello et al. 2003). In this thesis, I address several of these issues using coastal cutthroat trout (Onchorhynchus clarkii clarki) as a model system.  1.1.5  Coastal cutthroat trout: a model paradox  Cutthroat trout (Oncorhynchus clarkii, Salmonidae) is a polytypic species native to western North America that shows incredible amounts of phenotypic and life history variation across its range. Most taxonomists currently recognize 14 subspecies: four major subspecies showing substantial divergence; Coastal (O. c. clarkii), Westslope (O. c. lewisi), Lahontan (O. c. henshawi), and Yellowstone cutthroat (O. c. bouvieri), and ten minor subspecies of limited range or recent origin (Allendorf and Leary 1988; Behnke 1992). The most widespread of the subspecies, coastal cutthroat trout (O. clarkii clarki) inhabits coastal systems from the Eel River in northern California to Prince William Sound, Alaska; it is found in the majority of British Columbia's coastal waters, including Vancouver Island and the Queen Charlotte Islands. Coastal cutthroat trout possesses a number of unique characteristics that make it a particularly good model system in which to investigate the forces acting to promote and maintain genetic variability in small populations. A s noted, populations throughout their range show surprisingly high levels of genetic variation (in terms of heterozygosities and genetic subdivision) considering the small population sizes characteristic of the species (Utter et al. 1980; Campton and Utter 1987; Wenburg et al. 1998; Figure 1.1). Unlike the closely related species of Pacific salmon that have population sizes in the thousands to tens of thousands, population sizes in coastal cutthroat trout may be on the order of 10's to 100's and vary considerably between years (Sumner 1952, 1962; Trotter 1989). Wenburg et al. (1998) referred to it as the "cutthroat paradox": How, with such apparently small population sizes, are coastal cutthroat trout able to maintain the reported levels of genetic variability? Of course, several factors may be implicated. A s the most widely distributed of the cutthroat subspecies, the range of coastal cutthroat trout encompasses several putative glacial refugia. This may suggest the  8  possible influence of historical factors in shaping the levels and distribution of genetic variation in populations throughout the region, as it has for many other north temperate species (e.g.- Soltis et al. 1997; Bernatchez and Wilson 1998). If coastal cutthroat trout, for example, survived glaciation in multiple refugia, then it is possible that divergence and subsequent gene flow between refugial races upon glacial retreat may have been responsible for the origin of the high levels of genetic diversity observed in the subspecies. Over smaller spatial and temporal scales, coastal cutthroat trout also tend to exhibit extensive population structure (e.g. - Wenburg and Bentzen 2001), suggesting the possible relevance of genetic subdivision and gene flow among adjacent populations. Considering what is known about the genetic structuring of this subspecies, it seems possible that occasional gene flow on the order of "one-migrant-per-generation" ( O M P G , Wright 1931; Slatkin 1987) could be responsible for maintaining levels of genetic variation within populations throughout the range. The O M P G concept has often been applied in conservation biology and captive breeding programs as the theoretically optimal level of immigration for minimizing the occurrence of inbreeding and the loss of genetic diversity in small populations (Mills and Allendorf 1996; Vucetich and Waite 2000). While the breeding system of coastal cutthroat trout has been understudied relative to other Oncorhynchus species, populations appear to exhibit the complex polygynandrous breeding system typical of salmonids (reviewed by Fleming and Reynolds 2004). It may be possible, therefore, that the mating system of coastal cutthroat trout has evolved to counteract the three primary variables which reduce N  e  below the number of sexually mature adults in a population: fluctuations in population size, variance in individual reproductive success above random (binomial) expectations, and unequal sex ratios among breeders (Caballero 1994; Nunney 1999; Wang and Caballero 1999). Finally, while similarities in chromosome arm number and past gene duplication events permit many salmonid taxa to readily hybridize (Chevassus 1979; Allendorf and Thorgaard 1984; Ferguson et al. 1988), this is not generally the case for coastal cutthroat trout and its sister species, rainbow trout (O. mykiss). These species have an overlapping natural distribution in coastal watersheds that suggests a long co-evolutionary history in sympatry (Trotter 1987; Behnke 2002). While limited hybridization has been observed between naturally  9  sympatric populations (substantially more common in anthropogenically altered habitats), such events are believed to be episodic in nature and not generally progressing to the formation of hybrid swarms (Campton and Utter 1985; Young et al. 2001; Ostberg and Rodriguez 2004). A s such, the coastal cutthroat-rainbow trout system provides a good model in which to assess the impact of episodic hybridization on levels of genetic variation in natural populations. Using a nested approach (Figure 1.2), I document the hierarchical partitioning of genetic variation in this salmonid at various spatial and temporal scales to determine the influence of these mechanisms in determining levels of genetic variation in wild populations. Specifically I test four hypotheses: Chapter 2 - Limited occurrence of hybridization between coastal cutthroat and its sister species, rainbow trout are associated with higher levels of genetic diversity. In Chapter 2, species-specific microsatellite alleles and loglikelihood based assignment tests were employed to classify individuals as either pure coastal cutthroat trout, pure rainbow trout, or as hybrids. Comparisons are then made using a number of standard diversity metrics (number of alleles, heterozygosities, etc.) to determine the effect of inferred hybridization levels on measures of genetic diversity (1) at the species level (between assignment classes) and (2) for individual populations. Samples were collected from throughout Vancouver Island, the Georgia Basin, and the Queen Charlotte Islands. These areas have been subject to markedly different biogeographic histories and levels of contemporary anthropogenic disturbance. Chapter 3 - The mating system of coastal cutthroat trout compensates for small population size through some or all of the following means: 1) many individuals contribute to breeding; (2) sex ratios are equal; (3) variation in individual reproductive success is minimal and randomly (Poisson) distributed; (4) multiple mates or breeding events are common; and (5) breeding occurs between age-classes and divergent life-history forms. In Chapter 3, I investigate the relative influence of mating factors through demographic and genetic analyses of a representative coastal cutthroat  10  population over two successive years. Through genetic parentage analyses of the relationships between spawners and young-of-the-year (YOY) offspring, I describe the number and composition of successful spawners in the system as well as the effects of breeding patterns on N , N / N ratios, and the maintenance e  e  of genetic variation in the population. The site chosen for this study was Chonat Lake on the north end of Quadra Island, BC. Chapter 4 - High levels of genetic diversity are maintained by gene flow among adjacent populations on the order of one migrant per generation (OMPG).  In Chapter 4, I estimate levels of intrapopulation genetic diversity at microsatellite loci as well as long term and contemporary levels of gene flow among 42 spatially adjacent populations to determine: (1) whether populations in British Columbia exhibit similarly high levels of genetic subdivision and diversity as observed in other areas to the south; (2) whether populations are exhibiting appreciable gene flow on the order of O M P G ; (3) whether these parameters vary among four geographic regions chosen for their unique combination of physiogeographic characteristics; and (4) whether there are distinct groups or metapopulations in this area which could be used to define conservation units for the subspecies in Canada.  Chapter 5 - The high levels of genetic variation in coastal cutthroat trout originated with allopatric divergence in multiple Pleistocene refugia and the secondary contact between genetically divergent refugial races. In Chapter 5, I examine phylogeographic variation in populations of coastal cutthroat trout spanning several putative glacial refugia to (1) identify those glacial refugia used by coastal cutthroat trout during the Pleistocene and the refugial races which may have resulted; (2) reconstruct the recolonization history of the subspecies in Cascadia and the likelihood of secondary contact between refugial groups; and (3) determine whether higher levels of genetic variation in coastal cutthroat trout are correlated with secondary contact between refugial groups. I chose to forego the standard approach of sequencing mtDNA markers, and here pursue a nuclear phylogeography for the subspecies. To these ends, I sequenced a portion of the nuclear type 2 growth hormone (GH2); targeting  11  variation in the fourth intron (GH2-iD), the largest and most well studied of all the non-coding regions found in salmonid G H genes.  1.1.6  Conservation implications  Coastal cutthroat trout are a salmonid of increasing conservation concern in western North America. For several decades, there have been dramatic declines in both the number and distribution of populations throughout much of their range; the result of overharvesting, habitat loss and degradation, and other anthropogenic influences such as the introduction of non-native fishes (Johnson et al. 1999; Costello and Rubidge 2005). To better prioritize the allocation of conservation resources targeting coastal cutthroat trout, a thorough understanding of the historical and contemporary affiliations between groups of populations and their underlying subspecific biodiversity is required. The recognition of distinct groups with independent evolutionary histories is central to the conservation of biological diversity and may be particularly important for a subspecies like coastal cutthroat trout which exhibits extensive phenotypic variation over a wide geographic area (Bernatchez and Wilson 1998; Bowen 1999). Furthermore, insight gained into the means by which small populations in nature are able to maintain genetic diversity will be of interest and practical utility to any researchers working in the field of conservation genetics or with captive breeding populations of endangered species. The range of many species has become fragmented by development and habitat degradation and many once contiguous populations are now restricted to islands of suitable habitat (e.g., - Gerlach and Musolf 2000; Neraas and Spruell 2001). Many endangered species now exist only as a handful of captive individuals scattered among zoos and reserves around the globe. The survival of entire species may ultimately depend on the ability of researchers to identify and minimize the impact of factors that could cause any further loss of intraspecific biodiversity, as well as maximizing the evolutionary potential of that which remains (Fernandez and Caballero 2001; Lynch and O'Hely 2001; Moritz 2002; Woodworth et al. 2002).  Figure 1-1. Comparative analysis of the genetic divergence typical of anadromous salmonid populations with respect to geographic distance (modified from Hendry et 2004).  13  Figure 1-2. Overview of the nested design of the study. Ovals illustrate the geographic context for the study and the relative spatial scales involved for each thesis component.  14 1.2  REFERENCES  Allendorf, F. W., and R. F. Leary. 1988. Conservation and distribution of genetic variation in a polytypic species, the cutthroat trout. Conservation Biology 2:170 184. Allendorf, F. W., and G. H. Thorgaard. 1984. Tetraploidy and the evolution of  salmonid fishes Pp. 1-53 in I. Turner and J . Bruce, eds. Evolutionary genetics of fishes. Plenum Press, New York.  Amos, WV, and J . Harwood. 1998. Factors affecting levels of genetic diversity in natural populations. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 353:177-186. Anderson, E., and G. Stebbins. 1954. Hybridization as an evolutionary stimulus. Evolution 8:378-388. Angers, B., P. Magnan, M. Plante, and L. Bernatchez. 1999. Canonical correspondence analysis for estimating spatial and environmental effects on microsatellite gene diversity in brook charr (Salvelinus fontinalis). Molecular Ecology 8:1043-1053. Armbruster, P., W. Bradshaw, and M. Holzapfel. 1998. Effects of postglacial range expansion on allozyme and quantitative genetic variation of the pitcher-plant mosquito, Wyeamyia smithii. Evolution 52:1697 - 1704. Arnold, M. 1997. Natural Hybridization and Evolution. Oxford University Press, New York. Arnold, M. 2004. Natural hybridization and the evolution of domesticated, pest and disease organisms. Molecular Ecology 13:997-1007. Avise, J . C. 1994. Molecular markers, natural history and evolution. Chapman and Hall, New York. Avise, J . C , A. Jones, D. Walker, J . A. DeWoody, and e. al. 2002. Genetic mating systems and reproductive natural histories of fishes: Lessons for ecology and evolution. Annual Review of Ecology & Systematics 36:19-45. Beacham, T. 1991. Developmental stability, heterozygosity and genetic analysis of morphological variation in pink salmon (Oncorhynchus gorbuscha). Canadian Journal of Zoology 69:274-278. Beacham, T., and R. E. Withler. 1985. Heterozygosity and morphological variability of pink salmon (Oncorhynchus gorbuscha) from southern British Columbia and Puget Sound. Canadian Journal of Fisheries and Aquatic Sciences 27:571-579. Behnke, R. J . 1992. Native trout of western North America. American Fisheries Society Monograph 6.  15  Behnke, R. J . 2002. Trout and Salmon of North America. Simon and Schuster, New York. Bernatchez, L, and C. C. Wilson. 1998. Comparative phylogeography of nearctic and palearctic fishes. Molecular Ecology 7:431-452. Bowen, B. 1999. Preserving genes, species or ecosystems? Healing the fractured foundations of conservation policy. Molecular Ecology 8:S5-S10. Burke, J . , and M. Arnold. 2001. Genetics and the fitness of hybrids. Annual Review of Ecology & Systematics 35:31-52. Caballero, A. 1994. Developments in the prediction of effective population size. Heredity 73:657 - 679. Campton, D., and F. Utter. 1985. Natural hybridization between steelhead trout (Salmo gairdneri) and coastal cutthroat trout (Salmo clarki clarki) in two Puget Sound streams. Canadian Journal of Fisheries and Aquatic Sciences 42:110-119. Campton, D., and F. Utter. 1987. Genetic structure of anadromous cutthroat trout (Salmo clarki clarki) populations in the Puget Sound area: Evidence for restricted gene flow. Canadian Journal of Fisheries & Aquatic Sciences 44:573-582. Carvalho, G. 1993. Evolutionary aspects offish distribution: genetic variability and adaptation. Journal of Fish Biology 43:53-73. Caughley, G. 1994. Directions in conservation biology. Journal of Animal Ecology 63:215-244. Chevassus, B. 1979. Hybridization in salmonids: results and perspectives. Aquaculture 17:113-128. Clegg, M. T. 1997. Plant genetic diversity and the struggle to measure selection. Journal of Heredity 88:1-7. Comes, H., and R. Abbott. 1998. The relative importance of historical events and gene flow on the population structure of a Mediterranean ragwort, Senecio gallicus (Asteraceae). Evolution 52:355-367. Conner, J . 2001. How strong is natural selection? Trends in Ecology & Evolution 16:215-217. Costello, A. B., T. Down, S. Pollard, C. Pacas, and E. Taylor. 2003. The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: an examination of microsatellite DNA variation in bull trout, Salvelinus confluentus (Pisces: Salmonidae). Evolution 57:328-344. Costello, A. B., and E. Rubidge. 2005. C O S E W I C status report on coastal cutthroat trout (Oncorhynchus clarkii clarki). Draft Report. Pp. 91pp. Prepared for the  16  Committee on the Status of Endangered Wildlife in Canada, Ottawa. Crow, J . , and M. Kimura. 1970. An Introduction to Population Genetics Theory. Harper and Row, New York. Dowling, T., and C. Secor. 1997. The role of hybridization and introgression in the diversification of animals. Annual Review of Ecology & Systematics 28:593-619. Ellstrand, N., and D. Elam. 1993. Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology & Systematics 24:217-243. Endler, J . 1986. Natural Selection in the Wild. Princeton University Press. Princeton, NJ. Falconer, D. 1981. Introduction to quantitative genetics. Longman, London. Ferguson, M. M., R. G. Danzmann, and F. W. Allendorf. 1988. Developmental success of hybrids between two taxa of salmonid fish with moderate structural gene divergence. Canadian Journal of Zoology 66:1389-1395. Fernandez, J . , and A. Caballero. 2001. A comparison of management strategies for conservation with regard to population fitness. Conservation Genetics 2:121-131. Fleming, I. A., and J . Reynolds. 2004. Salmonid breeding systems. Pp. 264-294 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford. Frankham, R. 1995. Inbreeding and extinction: a threshold effect. Conservation Biology 9:792-799. Frankham, R. 1996. Relationship of genetic variation to population size in wildlife. Conservation Biology 10:1500-1508. Frankham, R. 1999. Quantitative genetics in conservation biology. Genetical Research 74:237-244. Gerlach, G., and K. Musolf. 2000: Fragmentation of landscape as a cause for genetic subdivision in bank voles. Conservation Biology 14:1066-1074. Gharrett, A., and W. Smoker. 1991. Two generations of hybrids between even- and odd-year pink salmon (Oncorhynchus gorbuscha): a test for outbreeding depression? Canadian Journal of Fisheries and Aquatic Sciences 48:1744-1749. Gillespie, J . H. 1991. The Causes of Molecular Evolution. Oxford University Press, New York. Good, S. V., D. F. Williams, K. Ralls, and R. C. Fleischer. 1997. Population structure of Dipodomys ingens (Heteromyidae): The role of spatial heterogeneity in  17  maintaining genetic diversity. Evolution 51:1296-1310. Gross, M. R. 1996. Alternative reproductive strategies and tactics: diversity within sexes. Trends in Ecology & Evolution 11:92-98. Haig, S. M. 1998. Molecular contributions to ecology. Ecology 79:413 - 425. Hamrick, J . , and M. Godt. 1996. Conservation genetics of endemic plant species. Pp. 281-304 in J . C. Avise and J . Hamrick, eds. Conservation genetics: case histories from nature. Chapman and Hall, New York. Harrison, S., and A. Taylor. 1997. Empirical evidence for metapopulation dynamics. Pp. 27-42 in I. Hanski and M. Gilpin, eds. Metapopulation Biology: Ecology, Genetics, and Evolution. Academic Press, San Diego. Hedrick, P. W. 1995. Elephant seals and the estimation of a population bottleneck. Journal of Heredity 86:232-235. Hedrick, P. W., R. C. Lacy, F. W. Allendorf, and M. E. Soule. 1996. Directions in conservation biology: Comments on Caughley. Conservation Biology 10:13121320. Hendry, A., V. Castric, M. Kinnison, and T. P. Quinn. 2004. The evolution of philopatry and dispersal: Homing versus straying in salmonids. Pp. 52-91 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford. Hewitt, G. M. 2001. Speciation, hybrid zones and phylogeography - or seeing genes in space and time. Molecular Ecology 10:537-549. Hughes, A., and M. Yeager. 1998. Natural selection at major histocompatability complex loci of vertebrates. Annual Review of Genetics 32:415-435. Johnson, O., M. Ruckelshaus, W. Grant, F. Waknitz, A. Garrett, G. Bryant, K. Neely, and J . Hard. 1999. Status review for coastal cutthroat from Washington, Oregon, and California. Pp. 292. National Marine Fisheries Service, Seattle, WA. Jones, A., and J . C. Avise. 1997. Microsatellite analysis of maternity and the mating system in the Gulf pipefish Syngnathus scovelli, a species with male pregnancy and sex-role reversal. Molecular Ecology 6:203-213. Karl, S., and J . C. Avise. 1992. Balancing selection at allozyme loci in oysters: implications from nuclear RFLP's. Science 256:100-102. Kimura, M. 1968. Evolutionary rate at the molecular level. Nature 217:624-626. Kimura, M. 1983. The neutral theory of molecular evolution. Cambridge University Press, Cambridge, UK.  18  Kimura, M., and J . Crow. 1963. The measurement of effective population number. Evolution 17:279-288. Kreitman, M., and H. Akashi. 1995. Molecular evidence for natural selection. Annual Review of Ecology & Systematics 26:403-422. Lacy, R. C. 1987. Loss of genetic diversity from managed populations: Interacting effects of drift, mutation, immigration, selection, and population subdivision. Conservation Biology 1:143-158. Lande, R. 1988. Genetics and demography in biological conservation. Science 241:1455-1460. Li, W. 1997. Molecular Evolution. Sinauer Associates, Sunderland, MA. Lynch, M. 1996. A quantitative genetic perspective on conservation issues. Pp. 471 501 in J . C. Avise and J . Hamrick, eds. Conservation genetics: case histories from nature. Chapman and Hall, New York. Lynch, M., and M. O'Hely. 2001. Captive breeding and the genetic fitness of natural populations. Conservation Genetics 2:363-378. Mallet, J . 2005. Hybridization as an invasion of the genome. Trends in Ecology and Evolution 20:229-237. Maruyama, T. 1970. On the fixation probability of mutant genes in a subdivided population. Gen. Res., Camb. 15:221-225. Mayr, E. 1963. Animal species and evolution. Harvard University Press, Cambridge, MA. McGlashan, D., and J . Hughes. 2000. Reconciling patterns of genetic variation with stream structure, earth history and biology in the Australian freshwater fish Craterocephalus stercusmuscarum (Antherinidae). Molecular Ecology 9:17371751. Mills, L. S., and F. W. Allendorf. 1996. The one-migrant-per-generation rule in conservation and management. Conservation Biology 10:1509-1518. Moran, P. 2002. Current conservation genetics: building an ecological approach to the synthesis of molecular and quantitative genetic methods. Ecology Freshwater Fish 11:30-55. Moran, P., and E. Garcia- Vazquez. 1998. Multiple paternity in Atlantic salmon: a way to maintain genetic variability in relicted populations. Journal of Heredity 89:551553. Moritz, C. 2002. Strategies to protect biological diversity and the evolutionary processes that sustain it. Systematic Biology 51:238-254.  19  Neraas, L, and P. Spruell. 2001. Fragmentation of riverine systems: the genetic effects of dams on bull trout (Salvelinus confluentus) in the Clark Fork River system. Molecular Ecology 10:1153-1164. Nichols, R. A., M. W. Bruford, and J . J . Groombridge. 2001. Sustaining genetic variation in a small population: evidence from the Mauritius kestrel. Molecular Ecology 10:593 - 602. Nunney, L. 1999. The effective size of a hierarchically structured population. Evolution 53:1-10. O'Brien, S. 1994. Genetic and phylogenetic analyses of endangered species. Annual Review of Genetics 28:467-489. Ostberg, C. O., and R. Rodriguez. 2004. Bi-parentally inherited species-specific markers identify hybridization between rainbow trout and cutthroat trout subspecies. Molecular Ecology Notes 4:26-29. Pfrender, M., K. Spitze, J . Hicks, K. Morgan, L. Latta, and M. Lynch. 2000. Lack of concordance between genetic diversity estimates at the molecular and quantitative-trait levels. Conservation Genetics 1:263-269. Rainey, P., A. Buckling, R. Kassen, and M. Travisano. 2000. The emergence and maintenance of diversity: insights from experimental bacterial populations. Trends in Ecology & Evolution 15:243-247. Ralls, K., and F. Ballou. 1983. Extinction: lessons from zoos. Pp. 164-184 in C. Schonewald-Cox, S. Chambers, B. MacBryde and W. Thomas, eds. Genetics and conservation: a reference for managing wild animal and plant populations. Benjamin-Cummings, Menlo Park, CA. Reed, D., and R. Frankham. 2001. How closely related are molecular and quantitative measures of genetic variation: a meta-analysis. Evolution 55:1095-1103. Reynolds, J . 1996. Animal breeding systems. Trends in Ecology & Evolution 11:6872. Rhymer, J . , and D. Simberloff. 1996. Extinction by hybridization and introgression. Annual Review of Ecology & Systematics 27:83-109. Richman, A. 2000. Evolution of balanced genetic polymorphism. Molecular Ecology 9:1953-1963. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Scribner, K. T., K. Page, and M. Bartron. 2001. Hybridization in freshwater fishes: a review of case studies and cytonuclear methods of biological inference. Reviews  20  in Fish Biology and Fisheries 10:293-323. Soltis, D., M. Gitzendanner, and D. Strenge. 1997. Chloroplast DNA intraspecific phylogeography of plants from the Pacific Northwest of North America. Plant Systematics and Evolution 206:353-373. Soule, M. E. 1976. Allozyme variation, its determinants in space and time. Pp. 60-77 in F. Ayala, ed. Molecular evolution. Sinauer Associates, Sunderland, MA. Streiff, R., A. Ducousso, C. Lexer, H. Steinkellner, J . Gloess, and A. Kremer. 1999. Pollen dispersal inferred from paternity analysis in a mixed oak stand of Quercus roburL. and Q. petraea (Matt.) Liebl. Molecular Ecology 8:831-841. Sumner, F. H. 1952. Migrations of salmonids in Sand Creek, Oregon. Transactions of the American Fisheries Society 82:139 - 150. Sumner, F. H. 1962. Migrations and growth of the coastal cutthroat trout in Tillamook County, Oregon. Transactions of the American Fisheries Society 91:77-83. Taggart, J., I. McClaren, D. Hay, J . Webb, and A. Youngson. 2001. Spawning success in Atlantic salmon (Salmo salarL.): a long-term DNA profiling-based study conducted in a natural stream. Molecular Ecology 10:1047-1060. Taylor, E. B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98:185-207. Templeton, A., E. Routman, and C. Phillips. 1995. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial DNA haplotypes in the tiger salamander, Ambystoma tigrinum. Genetics 140:767-782. Trotter, P. C. 1987. Cutthroat: Native trout of the West. Colorado Associated University Press, Boulder, C O . Trotter, P. C. 1989. Coastal cutthroat trout: a life history compendium. Transactions of the American Fisheries Society 118:463 - 473. Turgeon, J., and L. Bernatchez. 2001. Mitochondrial DNA phylogeography of lake cisco (Coregonus artedi): evidence supporting extensive secondary contacts between two glacial races. Molecular Ecology 10:987-1001. Utter, F., D. Campton, S, Grant, G. Milner, J . Seeb, and L. Wishard. 1980. Population structures of indigenous salmonid species of the Pacific Northwest. Pp. 285 - 304 in\N. J . McNeil and D. C. Himsworth, eds. Salmonid ecosystems of the North Pacific. Oregon State University Press, Cqrvallis. Vrijenhoek, R. C. 1994. Genetic diversity and fitness in small populations. Pp. 37-53 in V. Loeschcke, J . Tomiuk and S. Jain, eds. Conservation Genetics. Birkhauser Verlag, Basil.  21  Vrijenhoek, R. C , and S. Lerman. 1982. Heterozygosity and developmental stability under sexual and asexual breeding systems. Evolution 36:768-776. Vucetich, J . A., and T. A. Waite. 2000. Is one migrant per generation sufficient for the genetic management of fluctuating populations? Animal Conservation 3:261-266. Walter, R., and B. Epperson. 2001. Geographic pattern of genetic variation in Pinus resinosa: area of greatest diversity is not the origin of postglacial populations. Molecular Ecology 10:103-111. Wang, J . L, and A. Caballero. 1999. Developments in predicting the effective size of subdivided populations. Heredity 82:212-226. Waser, N . , and M. Price. 1994. Crossing-distance effects in Delpinium nelsonii: outbreeding and inbreeding depression in progeny fitness. Evolution 48:842-852. Weir, B. 1996. Genetic data analysis, Vol. II. Sinauer Associates, Sunderland, MA. Wenburg, J . K., and P. Bentzen. 2001. Genetic and behavioural evidence for restricted gene flow among coastal cutthroat trout populations. Transactions of the American Fisheries Society 130:1049-1069. Wenburg, J . K., P. Bentzen, and C. J . Foote. 1998. Microsatellite analysis of genetic population structure in an endangered salmonid - the coastal cutthroat trout (Oncorhynchus clarki clarki). Molecular Ecology 7:733-749. Whitlock, M . . C , and N . Barton. 1997. The effective size of a subdivided population. Genetics 153:427-441. Woodworth, L , M. Montgomery, D. Briscoe, and R. Frankham. 2002. Rapid genetic deterioration in captive populations: Causes and conservation implications. Conservation Genetics 3:277-288. Wright, S. 1931. Evolution in Mendelian populations. Genetics 16:97-159. Wright, S. 1932. The roles of mutation, inbreeding, crossbreeding, and selection in evolution. Proceedings of the Sixth International Congress of Genetics. 1:356-366. Wright, S. 1938. Size of populations and breeding structure in relation to evolution. Science 87:430-431. Young, W., C. O. Ostberg, P. Keim, and G. H. Thorgaard. 2001. Genetic characterization of hybridization and iritrogression between anadromous rainbow trout (Oncorhynchus mykiss irideus) and coastal cutthroat trout (O. clarki clarki). Molecular Ecology 10:921-930.  22  C H A P T E R 2: INDIVIDUAL A S S I G N M E N T - B A S E D A S S A Y O F AMBIENT HYBRIDIZATION B E T W E E N C O A S T A L C U T T H R O A T T R O U T CLARKII  CLARKI)  A N D RAINBOW T R O U T (O. MYKISS): CONSERVATION  (ONCORHYNCHUS  IMPLICATIONS F O R  1  "... the elements of an entirely foreign genetic adaptive system can be carried over into a previously stabilized one, permitting the rapid reshuffling of varying adaptations..." - A n d e r s o n a n d S t e b b i n s , Jr. (1954:378)  2.1 2.1.1  INTRODUCTION  The population-level effects of interspecific hybridization  An understanding of the relationships between closely related taxa is often critical when attempting to decipher evolutionary processes and the history of species. These relationships include not only historical affinities, such as shared ancestry or common biogeography (Bernatchez and Wilson 1998; Gaston 2000), but the contemporary processes through which sympatric taxa are inexorably linked (e.g. - competition, gene flow; Slatkin 1987; Sutherland 1996). In many closely related groups, hybridization events have played an important role in shaping the ecological and evolutionary trajectories of species. While the process has been most studied in plants (e.g., - Anderson and Stebbins 1954; Rieseberg 1997), introgressive hybridization (or the horizontal transfer and incorporation of genetic material from one species to the other) has been demonstrated to occur in a wide range of taxa, providing much information on the processes of genetic divergence and speciation at both macro- and microevolutionary scales (Arnold 1997; Dowling and Secor 1997; Scribner et al. 2001; Arnold 2004; Mallet 2005). While true hybrid taxa appear to be relatively rare among higher taxa (Arnold 1997; Dowling and Secor 1997), there are animal groups where hybrid individuals are relatively common; including birds (Grant and Grant 1992), anurans (Bogart 1980; Pagano et al. 2001), and fish (Hubbs 1955; Verspoor and Hammar 1991; Scribner et al. 2001). In these cases, hybridization  Aversion of this chapter will be submitted for publication. Costello, A. "Individual assignment-based assay of ambient hybridization between coastal cutthroat trout and rainbow trout". 1  23  events may occur naturally in geographically defined hybrid zones where formerly allopatric forms come into contact (Barton and Hewitt 1989; Krueger and May 1991; Taylor 2004), or may follow the anthropogenic introduction of non-local species (Rhymer and Simberloff 1996; Allendorf et al. 2001; Utter 2001). Hybridization events will, of course, affect population demography to varying degrees, depending on the intensity and duration of the event, as well as the relative fitness of hybrid individuals (Arnold and Hodges 1995; Barton 2001; Burke and Arnold 2001). In some cases, the effect may be negligible while in others, newly formed hybrid populations may eventually go on to function as entirely new species (e.g. Smith et al. 2003; Seehausen 2004). Only recently, however, have researchers considered the effects of hybridization events in the context of population level genetic processes; that is, with regards to the maintenance of intrapopulation diversity and its effect on population fitness (Ellstrand and Elam 1993; Burke and Arnold 2001). Failure to recognize the presence of hybridization, for example, may obscure evolutionary relationships between taxa or confound phylogeographic and metapopulation level study (Wilson and Bernatchez 1998; Redenbach and Taylor 2002; Carson and Dowling 2006). In the majority of cases, the general effect of hybridization appears to be negative; either through wasted reproductive output or the reduced fitness and intermediacy of hybrids at important ecological or physiological traits (reviewed by Rhymer and Simberloff 1996, Wang et al. 2002a). Furthermore, the erosion of species barriers between intensively hybridizing taxa and the formation of 'hybrid swarms' characterized by a random assortment of recombinant genotypes (Forbes and Allendorf 1991), may result in the 'genomic extinction' of locally adapted genetic combinations (Rhymer and Simberloff 1996; Epifanio and Philipp 2001), leaving hybrid populations of uncertain taxonomic or legal status, and unknown ecological value (Allendorf et al. 2001). That being said, episodic or rare hybridization events between taxa may, in some cases, benefit local populations by introducing novel traits that may show some adaptive advantage (Lewontin and Birch 1966). Anderson and Stebbins (1954), for example, hypothesized that by rapidly producing new recombinant genotypes, introgressive hybridization (in plants) could allow for enhanced levels of genetic  24  variability in populations, allowing organisms to more readily track environmental change and ultimately leading to increased rates of evolution. While an imperfect and understudied relationship exists, increased levels of genetic variability are generally associated with increased levels of quantitative variation for fitness related traits (i.e. - those which contribute to the adaptability and persistence of populations; Clegg 1997; Booy et al. 2000; Montgomery et al. 2000; Reed and Frankham 2001; Wang et al. 2002b; Spielman et al. 2004). Such contributions may, therefore, be particularly relevant to small, isolated populations which by their nature, are expected to be more inbred and genetically depauperate than larger populations, with fewer alleles and lower levels of heterozygosity (Ellstrand and Elam 1993; Frankham 1996; Montgomery et al. 2000).  2.1.2 The coastal cutthroat-rainbow trout system To investigate the possible "creative role" of hybridization in generating or maintaining levels of genetic variability in small populations, I examined the effect of hybridization on levels of genetic variation in small populations of coastal cutthroat trout (Oncorhynchus clarkii clarki). The coastal cutthroat trout is a salmonid species native to the west coast of North America that tends to occur in small (generally < 100 individuals) and geographically structured populations (Trotter 1987; Behnke 2002). Interestingly, coastal cutthroat populations contain unusually high levels of genetic diversity given their typically small population sizes (mean of ~10 alleles per microsatellite locus and expected heterozygosities averaging - 0.60 - 0.70; Zimmerman et al. 1997; Wenburg et al. 1998; Wenburg and Bentzen 2001); Chapter 4). While several historical and contemporary factors likely contribute to the maintenance of this diversity (see other chapters), the identification of natural hybridization between coastal cutthroat trout and the closely related rainbow trout (Oncorhynchus mykiss) in sympatry suggests that episodic or on-going hybridization between the species may be at least partly responsible for the generation of the observed diversity. Fishes are one of the animal groups where hybridization and introgression are relatively common as external fertilization and weak ethological isolating mechanisms frequently allow for heterospecific matings (Hubbs 1955; Verspoor and Hammar  25  1991; Scribner et al. 2001; Seehausen 2004). In many salmonid species, similarities in chromosome arm number and gene duplication, permit taxa to hybridize without major developmental incompatibilities (Chevassus 1979; Allendorf and Thorgaard 1984; Ferguson et al. 1988). Nearly all of the Oncorhynchus trout, for example, appear to have evolved allopatrically in isolated drainage basins and generally lack reproductive isolating mechanisms (Trotter 1987; Allendorf and Leary 1988; Behnke 2002) . A s such, the secondary contact between forms nearly always results in hybridization between native and introduced fish and in many cases, a complete breakdown of species barriers (Leary et al. 1984; Rubidge et al. 2001; Hitt et al. 2003) . This is not generally the case, however, for coastal cutthroat trout and rainbow trout, whose overlapping natural distribution in coastal watersheds suggest a long coevolutionary history in sympatry (Trotter 1987; Behnke 2002). While limited hybridization has been observed between natural populations of coastal cutthroat trout and rainbow trout, such events are believed to be episodic in nature and do not generally progress to the formation of hybrid swarms (Campton and Utter 1985; Young et al. 2001; Ostberg and Rodriguez 2004). The exact nature of the reproductive isolating mechanisms are uncertain, but are generally believed to involve spatial and temporal differences in spawning behavior which tend to minimize interbreeding (Trotter 1987; Ostberg and Rodriguez 2004). A s such, the coastal cutthroat-rainbow trout system provides a good model in which to assess the impact of episodic hybridization on levels of genetic variation in natural populations, on population fitness, and with respect to the maintenance of species integrity. To this end, I used species-specific microsatellite alleles and log-likelihood based assignment tests based on commonly used microsatellite loci to classify individuals as either pure coastal cutthroat trout (CCT), pure rainbow trout (RBT), or as hybrids (HYB). Comparisons were then made using a number of standard diversity metrics (e.g. - number of alleles, heterozygosities, etc.) to determine the effect of hybridization levels on measures of genetic diversity (1) at the taxon level (between C C T , RBT, and HYB assignment classes) and (2) at the sampled population level. This work will be directly relevant to efforts to conserve coastal cutthroat trout, a species of conservation concern in western North America. Contemporary levels of hybridization in coastal cutthroat populations are believed to be increasing as  26  anthropogenic habitat loss and non-native introductions have greatly contributed to population decline (Johnson et al. 1999; Costello and Rubidge 2005). Since hybridized populations are largely of unknown ecological or taxonomic status, their relevance to overall species biodiversity and conservation, has to date, been understudied (e.g. - Allendorf et al. 2003). The work has further implications for general evolutionary questions related to the maintenance of genetic variation and is one of the first to empirically address the link between hybridization events and intrapopulation genetic diversity in small populations.  2.2 MATERIALS AND METHODS 2.2.1 Sampling and DNA extraction Between 2001 and 2003, C C T tissue samples were collected from streams in the southwest portion of British Columbia, Canada as part of a conservation initiative focused on the study and conservation of C C T in the province (Figure 2-1). The area is located immediately north of Puget Sound/ Olympic Peninsula in Washington and includes much of Vancouver Island and British Columbia's southern mainland. The area has undergone tremendous growth and urban development in the past 100 years, resulting in substantial habitat loss and the decline of many native fish populations (reviewed by Costello and Rubidge 2005). Paralleling these declines, has been a proliferation of hatchery programs for many salmonid species (including C C T and RBT) that are now widely used to support or replace natural fish production in the region. A total of -1955 individuals were collected from 51 streams on Vancouver Island, the Sunshine Coast, and intervening Gulf Islands; nine of these streams contained primarily RBT and were included for species comparison (see Table 2-1). While many of the sampled streams were smaller than those typically inhabited by rainbow trout (i.e. - first and second order), naturally occurring RBT are present at some of the sampling sites. While the morphological similarity of the species does make field identification difficult, especially for juveniles (McPhail and Carveth 1993; Pollard et al. 1997), every effort was made to sample only C C T using the phenotypic characters suggested by Pollard et al. (1997) with one exception; I found median  27  dorsal parr-like marks to be present in both C C T and R B T fry (CCT generally had three or fewer of these markings whereas R B T typically had five). Importantly, all sites were believed to be free of significant hatchery introductions that are known to promote interspecific hybridization between these taxa (see Discussion). Inferred levels of hybridization should, therefore, be viewed as conservative. By targeting the smaller, typically unstocked streams inhabited by C C T (and often under-represented in such studies), this work serves to describe "background" or "ambient" levels of hybridization in otherwise "good" cutthroat trout populations. Finally, for comparison, a further six populations from the Queen Charlotte Islands and Alaska (-1000 km to the north) were included in the analysis. These areas have experienced less dramatic population growth and environmental degradation and have not been subject to the same degree offish stocking (Costello and Rubidge 2005). Collected adipose or pelvic fin samples were stored in 9 5 % ethanol until DNA could be isolated from ~ 5mg of tissue using the Puregene DNA isolation kit (Gentra Systems, Inc.). 2.2.2  Microsatellite amplification and scoring  A total of 13 microsatellite loci were screened for variability in a subset of individuals covering the study area. Six loci were eventually chosen for inclusion based on amplification quality when subjected to multiplex P C R and two triplex P C R amplifications based on those of Wenburg and Bentzen (2001) were developed (Table 2-2). One of the included-loci (Sfo8) has been shown to differentiate between RBT and C C T on the basis of allelic size distribution and formed the basis of initial species determination (see below). Polymerase chain reactions were carried out using fluorescently-labeled primers in 10ul volumes of 10mM Tris-HCl (pH 8.3), 2mM MgCI , 0.8mM dNTP's, and 0.4 units of AmpliTaq Gold polymerase in MJ P T C 100 2  thermocyclers using a basic cycle profile of: 1 cycle (95°C/ 3 min), 10 cycles (94°C /1 min, T / 30 sec, 72°C /20 sec), 20 cycles (94°C / 30 s, T / 30 sec, 72°C / 20 sec), A  A  and 1 cycle (72°C / 5min), where T is the annealing temperature (see Table 2.2). A  P C R products were electrophoresed. on ABI Genescan Gels (using ABI Prism 377 sequencers) with R O X labeled internal GS400HD size standards. Raw fragment analysis and allele binning was facilitated by the ABI Genotyper v3.7 utility program.  28  2.2.3  Reference populations for species identification  To assign individuals, it was first necessary to define reference populations that characterize the species. To do this, tentative identifications were made based on observations made during field sampling as well as the microsatellite locus Sfo8, which has been previously found to be useful in differentiating C C T and R B T on the basis of allelic size distribution (Wenburg et al. 1998). While C C T exhibit speciesspecific Sfo8 alleles from 192 - 244 bp and R B T from 228 - 315 bp, an area of overlap exists between 228 and 244 bp (at low frequency) which limits the usefulness of species identification based only on this one marker. A s such, assignment tests (Manel et al. 2005) incorporating the genotypic information for all six microsatellite loci were used to define reference populations of R B T and C C T . Log-likelihood based assignments were performed using the program G E N E C L A S S 2 (Piry et al. 2004) and the frequency-based approach of Paetkau et al. (1995). From a total of 1955 genotyped individuals, three pooled groups were defined for initial screening and self-assignment: CCT:  Sfo8* 192-226 bp  1465 individuals  RBT:  Sfo8* 245 - 315 bp  408 individuals  Overlap:  Sfo8* 226 - 245 bp  82 individuals  The groupings of RBT and C C T based on Sfo8 genotyping formed the " P a s s l " reference populations. While self-assignment of the P a s s l reference populations was generally good (> 9 5 % correct assignment), considerable overlap between two groups and numerous misassignments confirmed the limited usefulness of Sfo8 as a sole marker in distinguishing the species. In many cases, misassignments were simply related to incomplete genotyping of individuals in the P a s s l reference populations. To reduce the overlap between reference datasets, all individuals with incomplete genotypes were removed from the P a s s l datasets and a critical loglikelihood ratio (likelihood of individual being assigned RBT/ likelihood of being assigned C C T ) of 0.81 was chosen as a cut-off for the remaining individuals (Figure 2.2). A s well, to standardize the number of individuals in each reference group (there  29 are many more C C T than RBT), 320 individuals were randomly chosen from the reduced P a s s l reference populations. The clustering of this reduced dataset proved much tighter and individuals were self-assigned with 1 0 0 % accuracy. This 'Pass2' dataset was used for further assignment of "unknowns" (note that any further mention of "reference populations" refers to the Pass2 dataset).  2.2.4  A s s i g n m e n t of unknowns/ Calculation of hybrid indices (HI)  "Unknown" individuals were defined as all those individuals not included in the reference populations. A total of 1315 unknowns were assigned as either 'CCT' or 'RBT' based on their genotypic similarity to the reference populations. To identify and account for individuals of intermediate character (i.e. - hybrids), a simple hybrid index statistic (HI) based on the observed likelihood values (Hansen et al. 2001) was calculated to distinguish between pure types (RBT or CCT) and hybrids:  HI = 1-{log(pX)/[log(pX)  +  log(pY)]}  where pX denotes the likelihood of an individual being a C C T and pY is the likelihood of being a RBT. A HI score approaching "0" indicates increasingly "RBT-like" character and a HI score approaching " 1 " indicates increasingly "CCT-like" character. There was a clear separation between the HI scores for the Pass2 reference populations, with individual RBT ranging from HI = 0.23 - 0.47 and C C T from HI = 0.57 - 0.75 (Figure 2-3). A s a conservative measure, a HI value > 0.54 was used as the cut-off for C C T identification (anything above 0.50 could technically be considered 'CCT'). This represents a value only slightly less than the lower extent of the Pass2 C C T reference population HI distribution (-0.57). Consideration of the R B T reference distribution led to the definition of hybrids as individuals with intermediate HI scores (0.47 < HI < 0.54). To further confirm the statistical likelihood of the observed hybrid assignments, we performed Bayesian cluster analysis using the program S T R U C T U R E (Pritchard et al. 2000) to compute posterior probabilities that assignment was consistent with the structuring of the genetic data. Specifically, I tested whether the genetic data was best explained by K = 1, 2, or 3 groups which correspond to the "no differences between species", " C C T vs. RBT", and "CCT, RBT,  30  HYB" models, respectively. To allow for initial fluctuations and convergence of parameter values, the length of the burn-in period was set to 100,000 iterations followed by 500,000 iterations of the Monte Carlo Markov Chain (MCMC) simulation.  2.2.5  Genetic diversity comparisons  To assess the effect of hybridization on levels of genetic diversity, several standard metrics of genetic diversity were compared at two levels of resolution. The first involves comparisons between the reference populations (CCT and RBT) and the group of those fish assigned as hybrids (HYB, see previous section). This comparison targets variation between the "taxa" without regard to geographical sampling location. The second, finer scale comparison examines correlations between levels of genetic variation and the degree of hybridization within individual populations (measured by % H Y B , percentage of hybrid individuals per sample site). Allele frequencies, and the mean number of alleles per locus (MNA) were calculated using the "Microsatellite Toolkit" macro for Excel (Park 2001). Estimates of F t, Nei's (1987) global gene s  diversity statistics ( H and H ) and allelic richness (A ) were computed using F S T A T s  t  R  v2.9 (Goudet 2001). Diversity within individuals (or observed individual heterozygosity, H i) was calculated using G E N E P O P v3.3 (Raymond and Rousset 0  1995) following the equations in (Rousset 1996). This analysis takes the observed frequencies of identical pairs of alleles as estimates of corresponding probabilities of identity per locus and per sample: in individuals, and Q HYB). The value and  (1-Qinter),  in  ter  Q tra in  representing the probability of allelic identity  representing the value within each group (CCT, RBT, or  (1-Qintra),  therefore, represents the diversity within individuals (H j)  the diversity among individuals within groups (average  0  Hi 0  = H ). 0  Corresponding F estimates based on the equations in Rousset (1996), i.e. - F is  Qintra - Qinter  is  =  /1 - Qinter) were computed in M S Excel. To illustrate the relationships  between C C T , RBT, and HYB groups, factorial correspondence analysis (FCA) of the genetic relationships between the reference populations and assigned hybrids was projected onto the first two factorial axes using the program G E N E T I X 4.03 (Belkhir et al. 2001).  For comparisons at the "taxon" level, the significance of differences between  31  means diversity values for assignment classes was assessed using pairwise t-test statistics which assume unequal variance between samples using the statistical program S P S S ( S P S S , Inc.). I assumed (because group sample sizes are generally > 100) that the sampling distribution of assignment classes is normally distributed even if the distribution of genetic variables in sampled populations is not (Zar 1999). For population level comparisons, S P S S was also used to perform non-parametric Spearman Rank testing of correlations between level of hybridization (%HYB) and diversity measures for all members of those populations found to contain hybrids (N=29). Visual inspection of the % H Y B data revealed one outlier which was removed from analysis: Homesite Creek (population 47) was found to have a relatively high percentage of hybrids (55%) but particularly low levels of genetic diversity ( H = 0.26, 0  M N A = 1.8). The population is isolated above an impassable anadromous barrier and is likely characterized by small population size. A s such, it was removed from % H Y B correlation analysis.  2.3 2.3.1  RESULTS  Hybrid assignment  Of the 1315 "unknown" individuals, 1176 were assigned as C C T and 139 were assigned as R B T (Figure 2-2). Using the pre-defined HI criterion, marginally assigned individuals were reclassified so that 1115 of the 1315 (85%) were assigned as C C T (HI > 0.54), 113 were assigned as R B T (HI < 0.47), and 87 individuals were identified as hybrids (0.47 < HI < 0.54; Figure 2-3). Factorial correspondence analysis (FCA) of the genetic relationships between these groups serves to illustrate the heterogeneity of the reference populations (note, for example, the marked distinction of the above barrier Loss Creek RBT population) and the intermediacy of hybrid individuals (Figure 2-4). The scope of the variation is not surprising as the three groups are composed of individuals from numerous source populations (see Table 23). Furthermore, the hybrid group is expected to include a variety of F1 and post-F1 reciprocal hybrids with varying affinities to the pure parental species. Bayesian clustering analysis using the program S T R U C T U R E , however, confirms that the K = 3 group model (CCT, RBT, and assigned HYB) is a more likely explanation given the  32  genetic data than either the K = 2 group (CCT vs. RBT) or one group models (e.g. no structure), further supporting the validity of HI assignment (data not shown). Hybrids were distributed throughout the study area in 29 of the 57 populations, with percent hybridization values (%HYB) ranging from a high of 5 5 % in Homesite Creek to 2 % (Chinukundl and Maidenhair creeks; Table 2-1, Figure 2-5). Mean levels of hybridization do not appear to vary substantially within the Georgia Basin (Table 24). Qualitatively however, the percentage of hybridized streams and % H Y B per stream appear greater on the west coast of Vancouver Island (i.e. - Clayoquot Sound and Juan de Fuca) than in the Strait of Georgia. A s expected, southwestern B C had a significantly higher average % H Y B than the Queen Charlotte Islands (9.1 vs. 3.8 % , ttest p = 0.009; Mann-Whitney U test p = 0.05) as well as marginally higher diversity measures (Table 2-3). Interestingly, no R B T were assigned in half of the hybridized populations (14/ 29).  2.3.2  Effects of hybridization on levels of genetic variation  While C C T and RBT reference populations were generally not significantly different in terms of genetic diversity levels, a trend towards higher diversity levels in hybrids was evident for all diversity measures in taxon level comparisons (Table 2-3). The hybrid group had significantly higher observed heterozygosities than either C C T or R B T reference populations (p < 0.001). Hybrid individuals also tended to be less inbred; average F values within individuals were significantly less than reference is  C C T (p < 0.002). This represents a 1 7 . 5 % increase in H a n d 3 0 % reduction in F for 0  is  hybrid individuals relative to pure CCT. Trends towards greater allelic diversity in hybrid individuals (as measured by MNA and A ) were evident as well, although R  statistical tests were generally non-significant. Similarly, comparisons across the 28 sampled populations containing hybrids suggest a significant positive relationship between level of hybridization (%HYB) and increased heterozygosities (both H and e  H , p < 0.01; Table 2-5). Again, trends towards greater allelic diversity in populations 0  (as measured by MNA and A ) were evident although tests were non-significant. A s R  in taxon-based comparisons, a negative correlation was indicated with F| (although S  the relationship was non-significant).  33  2.4  2.4.1  DISCUSSION  A s s i g n m e n t and distribution of hybrids  The complimentary use of species-specific microsatellite alleles and multilocus genotypic assignments appears to have been quite successful in identifying the mixed species composition of my sampled populations. The assignment of hybrids was generally consistent with limited species-specific P C R testing (e.g., - Ostberg and Rodriguez 2004; data not shown) as well as with field observations of species occurrence. In no cases (with the exception of Homesite Creek) were hybrids assigned to known allopatric pops of C C T (i.e. - populations located above impassable barriers). Homesite Creek appears to be an outlier, containing the low levels of genetic variation typical of other above barrier populations (e.g. - Loss Creek; see also Griswold 1996; Costello et al. 2003), but with an unexpectedly large proportion of assigned hybrids. Rainbow trout were not conspicuously present at the time of sampling and relatively few sampled individuals were eventually assigned as R B T (3%). This may point to a recent hybridization event, perhaps resulting from unrecorded "backpack" or "bucket" introductions of a small number of R B T that have been known to occur in the region.  The finding of hybrids in nearly half of the sampled populations is not wholly unexpected; evidence of past hybridization events is common among western North American fish species. Dramatic environmental fluctuations in the past likely produced cycles of isolation and sympatry between otherwise allopatric species (Bernatchez and Wilson 1998; Montgomery 2000; Scribner et al. 2001; Jacobs et al. 2004). Far less dramatic environmental fluctuations over shorter time scales have been shown to promote similar patterns of heterospecific mating (Grant and Grant 1992, 1994). Furthermore, CCT/ RBT hybrids have been previously identified along much of the west coast (Campton and Utter 1985; Johnson et al. 1999; Young et al. 2001; Ostberg and Rodriguez 2004). In excess of one-third of all C C T populations in Washington and Oregon are now expected to contain hybrids (Johnson et al. 1999) and Spruell et al. (1998) suggested that pure C C T and R B T populations (lacking  34  evidence of hybridization) no longer coexist on the Lower Columbia River. Here in British Columbia, Bettles (2004) showed widespread evidence of hybridization in 29 of 30 sympatric trout populations on Vancouver Island with the frequency of hybridization (%HYB) ranging from 3 - 8 8 % in populations and the percent of introgressed alleles in individuals ranging from 2 - 5 4 % . Unlike Bettles (2004) and similar studies, however, the populations sampled here were generally expected to represent pure cutthroat trout (with the exception, of course, of the nine representative RBT populations). Many of the streams were believed to contain allopatric C C T populations and in those areas of sympatry, every effort was made to identify and sample only C C T . Furthermore, all streams were generally expected to be free of stocking. Interestingly, no RBT were assigned in half of the hybridized populations (14/ 29). This may suggest that field sampling was relatively efficient in avoiding obvious RBT or that RBT were not present in the sampled streams reaches. Evidence for spatial segregation between the two species in sympatry has been previously documented (Hartman and Gill 1968; Ostberg et al. 2004). The observation of hybrids in these systems, therefore, may gauge "background" levels of hybridization in the region or the residual effects of straying from other stocked systems, as hybrid fish are known to have altered migratory behavior (e.g., - Hindar et al. 1991; Krueger and May 1991; Reeves et al. 1997; Scribner et al. 2001; Docker et al. 2003). Alternatively, the result may point to historical episodes of hybridization and it is possible that some identified hybrids may represent the descendants of hybridization events that occurred during postglacial recolonization (cf. Redenbach and Taylor 2002). In any event, my data are in agreement with similar studies that indicate that hybridization in the southwestern BC is widespread (even in unstocked systems). The estimated background level of - 9 % likely underestimates true levels in the region. A s expected, mean % H Y B values were significantly lower in the Queen Charlotte Islands and may reflect the differing levels of anthropogenic influence between the regions. While both areas have been impacted by resource extraction and urbanization, the levels of anthropogenic habitat degradation and hatchery stocking in the Queen Charlotte Islands are substantially less than that observed in  35  southwestern B C (Costello and Rubidge 2005). While once viewed as a panacea for declining fish stocks, it is now clear that introductions of hatchery fish do not generally translate into increased viability for wild populations as the primary causes for population decline (i.e. - habitat loss, overharvesting) often go unaddressed (Krueger and May 1991; Hilborn 1992; Flagg et al. 2000; Utter 2004). The introduction of hatchery fish appears, in fact, to directly contribute to declines in wild populations, not only through the diversion of management resources from more relevant activities (such as habitat remediation), but through a number of adverse ecological (Krueger and May 1991; Reeves et al. 1997; Flagg et al. 2000; Einum and Fleming 2001; Molony et al. 2003) and genetic (Hindar et al. 1991; Ryman and Laikre 1991; Rhymer and Simberloff 1996; Waples 1999; Scribneretal. 2001; Utter2004) impacts. Ongoing habitat degradation and continued hatchery introductions along the west coast appear to act synergistically to increase the reproductive overlap between these taxa on the spawning grounds. Indeed, the predominance of F1 and early generation hybrids in several studies (Campton and Utter 1985; Young et al. 2001; Docker et al. 2003) suggest that hybridization is current and ongoing in regions where anthropogenic influence is pervasive. 2.4.2  Impact on levels of genetic diversity and population fitness  Regardless of the mode and timing of hybridization, my results show clear evidence for a significant impact of hybridization on the genetic characteristics of natural populations of C C T . In both taxon and population level assays, hybridization was associated with significantly higher levels of genetic diversity relative to the reference populations of either species, particularly in terms of heterozygosity and reduced levels of inbreeding. The lack of stronger correlations between hybridization and measures of allelic diversity is likely due to the effects of drift in small populations, as allelic diversity is expected to be more sensitive to population size effects than is heterozygosity (Widmer and Lexer 2001; Leberg 2002). Again, such trends are not unexpected given the nature of hybridization events and have been extensively documented (Hansson and Westerberg 2002; Coltman and Slate 2003; but see Balloux et al. 2004). While several studies have identified hybridization in C C T populations,.relatively few have attempted to place such results within the  36  context of fitness implications for hybridized populations. Of what significance is a 1 7 % increase in observed heterozygosity and 3 0 % decrease in inbreeding coefficients for hybrid C C T ? Perhaps, not a lot in the long run. Salmonids tend to exhibit extensive genetic variability and population differentiation over small spatial scales; attributable, at least in part, to the physical discreetness of aquatic habitats, and to natal stream fidelity which encourages reproductive isolation (Quinn and Dittman 1990; Taylor 1991). Natal philopatry appears to be particularly well developed in C C T (Wenburg and Bentzen 2001) and serves to limit gene flow between adjacent populations (see moderate F  st  values in Table 2-3; Chapter 4). Combined with their typically small  population sizes, such high levels of demographic independence likely predispose C C T populations to significant levels of drift and inbreeding (Allendorf and Waples 1996; Wang et al. 2002b). Indeed, 8 5 % of all populations in this study had positive F  is  values (see Table 2-1). The deleterious consequences of inbreeding have been extensively reviewed; as inbreeding serves to increase the level of homozygosity across entire genomes, it is expected to ultimately depress population fitness through the expression of partly recessive deleterious alleles and possible loss of heterozygote advantage (Frankham 1998; Saccheri et al. 1998; Charlesworth and Charlesworth 1999; Hedrick and Kalinowski 2000; Spielman et al. 2004). At the genetic level, the sheltering of deleterious recessive alleles in early generation hybrids may, in some cases, allow for short term increases in hybrid fitness as even a single generation of outcrossing will tend to reduce the linkage associations found under inbreeding (Waser and Price 1994; Burke and Arnold 2001; Hansson and Westerberg 2002). Whitlock et al. (2000) demonstrated that rare to moderately frequent outcrossing events between populations can lead to positive heterosis, especially when population sizes are small (n < 100). Continued introgression over extended periods of time, however, may disrupt important longterm adaptations in wild salmonid populations. As anadromous fishes (i.e. - migrating between fresh and salt water environments), C C T and RBT have numerous genetically based adaptations to their local environments and often do poorly when transplanted oroutcrossed (Taylor 1991; Moran 2002). Reductions in population  37  fitness, for example, have been observed to occur where artificially propagated hatchery fish interbreed with local native populations (Hindar et al. 1991; Lynch 1991, 1996; Wang e t a l . 2002b). While there is little evidence of endogenous selection acting to structure CCT/ RBT hybrid zones (Taylor 2004), the relative lack of adult hybrids and preponderance of early generation hybrids (Campton and Utter 1985; Young et al. 2001) suggest that some manner of exogenous selection is likely constraining gene flow between species, either in terms of reproductive or migratory ecology. Campton and Utter (1985) and Young et al. (2001) have suggested that selection likely acts during the marine migration stage of the anadromous life cycle or during the overwintering period. Anadromous forms of both species typically spend 2-3 years in freshwater before migrating to the ocean. Rainbow trout, however, may reside at sea for 1-3 years and migrate extensively in the open ocean, while C C T rarely stray far from their natal estuary and do not typically overwinter at sea (Trotter 1987; Pearcy 1997). Importantly, hybrids between C C T and RBT have been found to be intermediate morphologically and in terms of swimming performance (Hawkins and Quinn 1996; Hawkins and Foote 1998) and may, therefore, be unsuited for either parental strategy or typical habitat.  The "creative" potential of hybridization and its benefit to populations of C C T , therefore, appears to depend critically on a balance between the risks of inbreeding depression relative to the risks of impacting locally important adaptations. If successful hybridization events are episodic or rare in nature, then the infusion of new variation and reduction in inbreeding may be beneficial, especially for small isolated populations. If, however, hybridization is widespread and recurrent, then the breakdown of locally adapted gene complexes and resulting outbreeding depression may further reduce fitness in parental populations (Rhymer and Simberloff 1996; Epifanio and Philipp 2001). The dichotomy points to the importance of rare events in the evolutionary process (i.e., - beneficial mutations, long-distance migration, founder events) and the importance of equilibria between competing forces (Waples 1995; Burke and Arnold 2001). It is apparent, however, that purely genetic factors are likely to be less relevant when ecological factors immediately impact on population viability  38 (e.g., - M c E l h a n y  et al. 2 0 0 0 ) . F o r s m a l l p o p u l a t i o n s of C C T  inhabiting marginal  h a b i t a t s , g a m e t i c w a s t a g e a n d c o m p e t i t i o n with i n t r o d u c e d f i s h (or h y b r i d s ) m a y b e  of  m o r e i m m e d i a t e c o n s e q u e n c e , further p r e d i s p o s i n g p o p u l a t i o n s to s t o c h a s t i c a l l y " b l i n k i n g out" in r e s p o n s e to e n v i r o n m e n t a l c h a n g e ( C a u g h l e y 1 9 9 8 ) . I n d e e d , t h e g e n e r a l d e c l i n e in C C T  1994;  S a c c h e r i et al.  p o p u l a t i o n s t h r o u g h o u t t h e r a n g e p o i n t s to  t h e l i k e l i h o o d of s u c h a r e l a t i o n s h i p .  2.4.3  Residual hybridization or ancestral polymorphism? T h e f i n d i n g of " r a i n b o w t r o u t a l l e l e s " in p u t a t i v e l y p u r e c o a s t a l c u t t h r o a t  p o p u l a t i o n s is m o s t often attributed to t h e o c c u r r e n c e of h y b r i d i z a t i o n b e t w e e n two taxa ( C a m p t o n and Utter 1985;  B a k e r et a l . 2 0 0 2 ; O s t b e r g et a l . 2 0 0 4 ) .  the  Indeed,  a n t h r o p o g e n i c i n f l u e n c e s h a v e c e r t a i n l y l e d t o m a r k e d i n c r e a s e s in h y b r i d i z a t i o n b e t w e e n t h e s e two t a x a ( r e v i e w e d by J o h n s o n et al. 1 9 9 9 ; C o s t e l l o a n d 2005). Other factors, however, a l l e l e s in C C T .  rates  Rubidge  m a y e q u a l l y a c c o u n t for t h e f i n d i n g of h e t e r o s p e c i f i c  P a r a - or p o l y p h y l e t i c r e l a t i o n s h i p s a p p e a r to b e c o m m o n a m o n g f i s h  s p e c i e s ( F u n k a n d O m l a n d 2 0 0 3 ) a n d m a y involve a n u m b e r of factors, i n c l u d i n g hybridization between overlapping taxa (e.g., - R e d e n b a c h and Taylor 2002)  and  i n c o m p l e t e l i n e a g e s o r t i n g b e t w e e n s i s t e r s p e c i e s w h e n s p l i t t i n g e v e n t s in g e n e t r e e s predate speciation events (Pamilo and Nei 1988; p r e s e n c e o f h e t e r o s p e c i f i c a l l e l e s in s o m e C C T  M a d d i s o n 1 9 9 7 ) . If, f o r e x a m p l e ,  the  populations m a y also result from  i n c o m p l e t e l i n e a g e sorting of s h a r e d a n c e s t r a l p o l y m o r p h i s m s b e t w e e n t h e t w o s i s t e r species, then hybridization rates be mistakenly overestimated.  T h e r e is s u b s t a n t i a l e v i d e n c e to s u p p o r t the i d e a that C C T  and RBT  may  s h a r e d a n c e s t r a l p o l y m o r p h i s m s . B a k e r e t a l . (2002), for e x a m p l e , f o u n d  that  contain  putatively pure coastal cutthroat populations often contained heterospecific "rainbow t r o u t a l l e l e s " a t l o w f r e q u e n c i e s (1 - 6 % ) f o r s e v e r a l n u c l e a r l o c i b u t n o t w e e (except for o n e locus, p 5 3 O n c o g e n e ;  versa  P a r k et a l . 1 9 9 6 ) . C o s t e l l o et a l . ( 2 0 0 1 )  demonstrated paraphyletic patterns between apparently pure coastal cutthroat a n d r a i n b o w trout m t D N A . A s well, the u n e x p e c t e d finding of high l e v e l s of  trout  apparent  h y b r i d i z a t i o n o n t h e w e s t c o a s t of V a n c o u v e r Island (an a r e a w h i c h is relatively u n i m p a c t e d by h u m a n s a n d lacks a significant stocking history), m a y provide  another  39  example. Considering that the area appears to have colonized by multiple refugial groups during deglaciation (see Chapter 5), the possibility that CCT populations in the area contain ancestral polymorphisms associated with these different groups cannot be discounted. 2.4.4  Conservation implications and future directions While I have shown that hybrid individuals and hybridized populations may show  elevated levels of genetic variation, it is not expected that this will be of substantial benefit to CCT populations as persistent and ongoing hybridization can eventually erode interspecific diversity and lead to the collapse of pure parental populations (Rhymer and Simberloff 1996; Epifanio and Philipp 2001). While these two taxa have shared a long evolutionary history in sympatry and have likely co-evolved with some "natural" level of genetic interaction between them, it is evident that anthropogenic influences are upsetting this natural balance. All evidence suggests that the mechanisms which have allowed CCT and RBT to traditionally coexist in sympatry are susceptible to habitat loss and the introduction of hatchery-produced fish (Krueger and May 1991; Reeves et al. 1997; Scribner et al. 2001; Docker et al. 2003). The fact that hybrids were found in small, unstocked CCT systems suggests that hybridization may be even more widespread than previously thought. This is problematic for future conservation of CCT because the production of hybrids is unidirectional; that is, all the progeny of a hybrid will be hybrids (Allendorf et al. 2001). Anthropogenically hybridized populations therefore represent a unique and uncertain biological entity, both in terms of their legal definition and in terms of their ecological relevance. Neither Canada nor the United States currently has an official policy regarding the inclusion of hybrid populations under their respective endangered species legislation (Allendorf et al. 2003; Costello and Rubidge 2005). The consensus view appears to be that while extensively hybridized populations may be important in terms of aesthetic values or recreational potential, they are of little value to efforts to preserve pure parental species. Their continued existence, in fact, may pose a threat to remaining pure populations in terms of hybrid straying and further hybrid mating events (Rhymer and Simberloff 1996; Hitt et al. 2003).  40  As such, I believe that there is a need to develop of a workable hybrid policy and implementation program to quantify the scope and severity of the problem in BC and elsewhere; preliminary efforts, for example, are being undertaken by the Committee for the Status of Endangered Wildlife in Canada (COSEWIC; E. Taylor, pers. comm.). Such a program would provide the blueprint for prioritizing populations for conservation efforts. In the interim, a thorough review of current provincial stocking programs for cutthroat and rainbow trout should be initiated. A s well, the initiative should focus on resolving the "ancestral polymorphism vs. hybridization" debate, possibly through the development of a "congeneric phylogeography" for C C T and RBT (and perhaps other closely related forms) that would investigate the nature and geographic distribution of shared genetic polymorphisms. It may be possible that these taxa are exhibiting heterospecific alleles for reasons other than hybridization. Rather than representing hybridized populations (which often receive less stringent management and conservation consideration; Allendorf et al. 2003), such populations may, in fact, be harboring rare genetic polymorphisms which represent an important component of the evolutionary history of the subspecies and are deserving of heightened, rather than diminished protection (Meffe and Carroll 1994; Avise and Hamrick 1996).  41  Table 2-1. Summary data for populations sampled in the hybridization assay. Lat  Location  Clayoquot Sound 49.23 1. Meares Creek 2. Fortune Channel creek* 49.21 49.09 3. Kootowis Creek 4. Kennedy Lake tributary* 49.07 49.09 5. Staqhorn Creek 49.03 6. Sandhill Creek 49.00 7. Lost Shoe Creek" 48.98 8. Smith Creek 48.97 9. Thornton Creek Strait of Juan de Fuca 48.73 10. Black Lake tributary* 48.70 11. Klanawa River 12. Botanical Beach creek* 48.53 48.52 13. Tom Baird Creek 48.48 14. Loss Creek 48.46 15. Hoard Creek 48.45 16. Clinch Creek 48.45 17. Rosemond Creek 48.43 18. Maidenhair Creek 48.43 19. Uglow Creek 48.43 20. Second Creek 48.38 21. Tugwell Creek" Strait of Georgia (West) 22. Stella lake tributary 50.34 50.34 23. Pye Lake tributary 24. unnamed Elk Bay creek 50.28 50.14 25. Menzies Creek 50.29 26. Chonat Lake 50.30 27. Vance Creek 50.27 28. Ashlar Creek 50.24 29. Granite Bay Creek 49.47 30. Waterloo Creek 49.46 31. McNaughton Creek 49.46 32. Cook Creek" 49.46 33. Chef Creek 49.43 34. Thames Creek 49.42 35. Nile Creek 49.39 36. Annie Creek 49.35 37. French Creek" 49.30 38. Craig Creek Strait of Georgia (East) 39. Kelly Creek 49.77 49.78 40. Whittail Creek 49.59 41. Nine Mile Creek 49.58 42. Sechelt Inlet creek 49.51 43. Angus Creek 49.49 44. Cook Creek 49.64 45. Myers Creek 49.64 46. Kleindale Creek 49.52 47. Homesite Creek 49.51 48. Halfmoon Creek 49.72 49. Whiskystill Creek 49.74 50. Rumbottle Creek 49.64 51. Moat Creek Queen Charlotte Islands/ Alaska 52. Chinukundl Creek" 53.45 53. Copper River tributary* 53.13 54. Mamin River 53.60 55. Mayr Lake 53.69 56. Yakoun River 53.38 57. Makaka Creek 60.51 R  R  Long  n  %CCT %RBT %HYB H  125.80 125.75 125.73 125.59 125.60 125.69 125.64 125.57 125.56  21 32 63 20 22 21 88 43 36  0.95 0.91 0.00 1.00 0.91 0.81 0.15 0.67 1.00  0.00 0.00 0.95 0.00 0.00 0.00 0.82 0.23 0.00  0.05 0.09 0.05 0.00 0.09 0.19 0.03 0.09 0.00  0.66 0.72 0.72 0.70 0.76 0.73 0.71 0.79 0.51  0.68 0.76 0.61 0.69 0.81 0.72 0.66 0.68 0.47  5.0 6.3 11.7 5.7 6.8 6.0 13.5 10.0 4.3  4.8 5.5 7.5 5.4 6.4 5.8 8.1 7.7 3.6  -0.036 -0.046 0.148 0.025 -0.070 0.021 0.071 0.144 0.065  125.11 124.95 124.45 124.43 124.27 124.19 124.18 124.17 124.09 124.09 124.09 123.85  25 29 21 39 51 53 23 39 55 24 38 44  1.00 0.21 0.81 1.00 0.00 0.94 0.91 0.51 0.98 0.17 1.00 0.02  0.00 0.79 0.00 0.00 1.00 0.02 0.00 0.28 0.00 0.67 0.00 0.98  0.00 0.00 0.19 0.00 0.00 0.04 0.09 0.21 0.02 0.17 0.00 0.00  0.52 0.77 0.69 0.63 0.13 0.68 0.61 0.74 0.64 0.77 0.56 0.74  0.30 0.59 0.45 0.59 0.08 0.45 0.60 0.64 0.56 0.73 0.51 0.72  3.5 8.3 4.7 5.8 1.7 6.2 5.0 8.3 6.3 8.8 4.5 8.3  3.4 6.9 4.5 5.0 1.6 4.9 4.6 6.6 5.0 7.6 4.1 6.5  0.429 0.233 0.347 0.067 0.426 0.332 0.015 0.134 0.125 0.055 0.094 0.024  125.52 125.52 125.44 125.39 125.30 125.27 125.35 125.30 124.79 124.76 124.76 124.76 124.66 124.64 124.59 124.36 124.24  39 18 20 18 49 58 32 19 32 28 30 33 41 19 38 51 31  1.00 1.00 0.65 0.89 1.00 1.00 1.00 1.00 1.00 1.00 0.17 0.94 1.00 0.79 1.00 0.22 0.97  0.00 0.00 0.30 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.73 0.00 0.00 0.05 0.00 0.69 0.00  0.00 0.00 0.05 0.11 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.06 0.00 0.16 0.00 0.10 0.03  0.18 0.46 0.79 0.67 0.67 0.64 0.04 0.67 0.73 0.66 0.75 0.73 0.72 0.73 0.66 0.80 0.69  0.17 0.46 0.65 0.61 0.65 0.68 0.05 0.66 0.64 0.73 0.75 0.66 0.70 0.77 0.69 0.83 0.62  1.8 3.3 6.5 7.2 6.0 5.8 1.3 5.3 7.5 5.5 7.8 7.7 7.2 7.3 6.7 9.7 6.5  1.7 3.2 6.3 6.6 5.0 4.5 1.3 5.1 6.2 4.8 6.9 6.3 5.9 6.7 5.4 7.9 5.9  0.055 0.005 0.178 0.096 0.034 -0.057 -0.061 0.026 0.128 -0.111 0.003 0.104 0.028 -0.058 -0.039 -0.030 0.108  124.38 124.35 123.78 123.82 123.76 123.77 123.99 123.99 123.94 123.91 124.43 124.50 124.45  33 36 31 19 50 30 33 22 29 28 34 30 5  0.97 0.94 0.90 1.00 1.00 0.17 0.97 1.00 0.41 1.00 1.00 1.00 0.80  0.00 0.03 0.00 0.00 0.00 0.73 0.00 0.00 0.03 0.00 0.00 0.00 0.20  0.03 0.03 0.10 0.00 0.00 0.10 0.03 0.00 0.55 0.00 0.00 0.00 0.00  0.72 0.72 0.68 0.63 0.67 0.75 0.63 0.46 0.25 0.52 0.67 0.68 0.78  0.64 0.65 0.65 0.68 0.53 0.75 0.57 0.43 0.26 0.52 0.58 0.69 0.65  6.5 6.0 5.0 4.5 6.0 7.8 4.2 3.0 1.8 4.2 5.2 5.8 4.3  5.6 5.1 4.6 4.4 4.8 6.9 3.8 2.9 1.7 3.6 4.6 5.3 2.5  0.115 0.097 0.049 -0.085 0.204 0.003 0.084 0.065 -0.051 0.005 0.136 -0.021 0.019  132.26 131.81 132.29 132.05 132.28 146.30  54 39 42 8 55 39  0.00 0.87 0.50 1.00 0.91 0.97  0.98 0.10 0.50 0.00 0.02 0.00  0.02 0.03 0.00 0.00 0.07 0.03  0.65 0.68 0.75 0.60 0.67 0.63  0.49 0.63 0.58 0.58 0.63 0.61  6.8 6.5 7.2 3.3 6.7 6.5  5.1 5.3 6.1 12.3 5.0 5.3  0.239 0.072 0.222 0.035 0.059 0.028  e  H  0  MNA A  R  F  is  Lat and Long refer to latitude and longitude, respectively; n is sample size, %CCT - percent assigned as CCT, %RBT - percent rainbow, %HYB - percent hybrids; H and H are the expected and observed heterozygosity, MNA is the mean number of alleles, A - allelic richness, F - Wright's inbreeding coefficient. Underlined creeks are located above a putative dispersal barrier, * - unnamed creek, R - indicates predominately rainbow populations. e  R  is  0  42  Table 2-2. Microsatellite loci used in the study; data presented includes data for all samples within populations. Locus Set A  Ocl2 Ogo4 Ogo8  Reference  Label  TA(°C)  AT  Range (bp)  HEX FAM FAM  58 58 58  20 32 9  113-159 118-206 89-117  0.66 0.68 0.58  0.87 0.90 0.77  HEX FAM' FAM  56 56 56  28 10 60  96-154 137-157 190-339*  0.71 0.48 0.71  0.93 0.72 0.92  Condrey and Bentzen 1998 Olsen et al. 1998 Olsen et al. 1998  H  H  S  T  SetB  Omy77 Morris et al. 1996 Oneul 1 Scribneret al. 1996 Sfo8 Angers et al. 1995  T is the annealing temperature, A is the number of alleles observed, Range is the allelic size range in bp. H is average within-population genetic diversity and H is the overall gene diversity (Nei 1987); $ - Locus with species-specific allele distribution (Wenburg et al. 1998; see text) A  T  s  t  Table 2-3. Taxon-level comparison of genetic diversity measures between coastal cutthroat trout (CCT) and rainbow trout (RBT) reference populations and the assigned hybrid group (HYB). Group  n  SP  He  Ho  MNA  A  R  F  i s  CCT  320  51  0.82  0.57  15.8  13.0 0.30  RBT  320  16  0.78  0.57  17.3  13.6 0.26  HYB  87  29  0.84  0.67** 18.8  18.8 0.21*  n = sample size, S P = the number of source populations contained in each class, H = expected heterozygosity, H = mean observed heterozygosity across individuals; MNA = mean number of alleles/ locus; A = allelic richness (minimum # diploid genotypes =79); F = Wright's inbreeding coefficient (all values significantly greater than zero); ** H Y B significantly greater than either C C T or RBT, t-test p < 0.001; * H Y B significantly less than C C T , t-test p < 0.002 0  R  is  e  43  Table 2-4. Population-level comparison of genetic diversity parameters and hybridization levels by region. Region  ~~  # POPS %HYB  H  Ho  A^  0.73 0.69 0.75 0.69  0.68 0.56 0.71 0.63  4.1 3.7 4.1 3.5  0.07 0.18 0.04 0.09  e  St  Southwestern British Columbia  Clayoquot Sound Strait of Juan de Fuca Strait of Georgia (West) Strait of Georgia (East)  7 ( 7 8 % ) 8.4 6 ( 5 0 % ) 12.0 7 ( 4 1 % ) 8.7 4 ( 3 1 % ) 4.8  (SW B C Average) Queen Charlotte Is/Alaska  0.15 0.15 0.12 0.15  24(47%)  9.1*  0.71  0.65  3.9  0.09  0.17  4(67%)  3.8*  0.66  0.59  3.5  0.10  0.23  # P O P S indicates the number of streams found to contain hybrids (percentage of total in parentheses). % H Y B represents the average percentage of hybrid individuals per region calculated from those streams containing hybrids. H = expected heterozygosity, H = observed heterozygosity; MNA = mean number of alleles/ locus; A = allelic richness (minimum # diploid genotypes =79); F a n d F t = Wright's F coefficients; * - paired values significantly different (t-test p = 0.009; Mann-Whitney U test p = 0.05) e  0  R  is  s  Table 2-5, Population-level Spearman rank testing of correlations between hybridization level (%HYB) and genetic diversity measures for populations containing hybrids. Correlation with % H Y B  H  e  H  MNA  0  A  R  F  i s  Spearman's rho  Correlation coefficient  0.532  0.448  0.194  0.260  -0.277  p-value (1-tailed)  0.002  0.008  0.161  0.095  0.154  H = expected heterozygosity, H = observed heterozygosity, F = Wright's inbreeding coefficient; MNA = mean number of alleles/ locus; A = allelic richness (minimum # diploid genotypes =14). e  0  js  R  Figure 2-1.  -128*  -126'  -124'  -122°  -128°  -126  -124  -122  s  s  Overview of the sampling locations for coastal cutthroat trout and rainbow trout populations included in this study. Inset shows regional perspective and location of populations in Alaska and the Queen Charlotte Islands. Numbers correspond to those in Table 2-1.  Figure 2-2. Log-likelihood based results for coastal cutthroat trout (CCT) and rainbow trout (RBT) assignment, showing distribution of reference samples and assigned individuals. Hybrids have been removed for clarity.  R B T (n = 4 3 3 )  C C T (n =1435)  • Assigned Individuals •  •o  Reference Samples  80  HYB ZONE (n=87)  0 c  at  "~  n !•  60  E  20  rv." O-  L O-  c\Y  O-  cv. O  tN-  r\. O  mr O'  O-  <^ O^ O'  r& O' A O'  A O-  A O-  A O-  A O-  <& <ASO-  # O v  ^ O-  ^ Q'  JL  H y b r i d Index (HI)  Figure 2-3. Distribution of Hybrid Index (HI) values for reference samples and assigned individuals; C C T - coastal cutthroat trout, RBT - rainbow trout, HYB - hybrid. CD  3  L o s s Creek  2.5  RBT A  0  A A  A  •  A  RBT HYB CCT  1.5  <  TO o  -2  15  -15  Factorial Axis 1 (4.22%)  Figure 2-4. Factorial correspondence analysis (FCA) of the multilocus genetic relations ships between assigned coastal cutthroat trout (CCT), rainbow trout (RBT) and hybrids (HYB) along the first two factorial axes.  Figure 2-5. Distribution (%) of coastal cutthroat trout (CCT), rainbow trout (RBT), and hybrids (HYB) by sampled population and by region. Numbers on X-axis correspond to population codes in Table 2-1.  49  2.5  REFERENCES  Allendorf, F. W., and R. F. Leary. 1988. Conservation and distribution of genetic  variation in a polytypic species, the cutthroat trout. Conservation Biology 2:170 184.  Allendorf, F. W., R. F. Leary, N. Hitt, K. L. Knudsen, L. Lundquist, and P. Spruell. 2003. Intercrosses and the U.S. Endangered Species Act: Should hybridized populations be included as westslope cutthroat trout? Conservation Biology 18:1203-1213. Allendorf, F. W., R. F. Leary, P. Spruell, and J . K. Wenburg. 2001. The problems with hybrids: setting conservation guidelines. Trends in Ecology and Evolution 16:613-622. Allendorf, F. W., and G. H. Thorgaard. 1984. Tetraploidy and the evolution of salmonid fishes. Pp. 1-53 in I. Turner and J . Bruce, eds. Evolutionary genetics of fishes. Plenum Press, New York. Allendorf, F. W., and R. S. Waples. 1996. Conservation and genetics of salmonid fishes. Pp. 238-281 in J . C. Avise and J . Hamrick, eds. Conservation Genetics: C a s e Histories from Nature. Chapman and Hall, New York. Anderson, E., and G. Stebbins. 1954. Hybridization as an evolutionary stimulus. Evolution 8:378-388. Angers, B., L. Bernatchez, A. Angers, and L. Desgroseillers. 1995. Specific microsatellite loci for brook charr reveal strong population subdivision on a microgeographic scale. Journal of Fish Biology 47:177-185. Arnold, M. 1997. Natural Hybridization and Evolution. Oxford University Press, New York. Arnold, M. 2004. Natural hybridization and the evolution of domesticated, pest and disease organisms. Molecular Ecology 13:997-1007. Arnold, M., and S. Hodges. 1995. Are natural hybrids fit or unfit relative to their parents? Trends in Ecology and Evolution 10:67-71. Avise, J . C , and J . Hamrick. 1996. Conservation Genetics: Case Histories from Nature. Chapman & Hall, New York. Baker, J., P. Bentzen, and P. Moran. 2002. Molecular markers distinguish coastal cutthroat trout from coastal rainbow trout/ steelhead and their hybrids. Transactions of the American Fisheries Society 131:404-417. Balloux, F., W. Amos, and T. Coulson. 2004. Does heterozygosity estimate inbreeding in real populations? Molecular Ecology 13:3021-3031.  50  Barton, N. 2001. The role of hybridization in evolution. Molecular Ecology 10:551-568. Barton, N., and G. M. Hewitt. 1989. Adaptation, speciation and hybrid zones. Nature 341:497-503. Behnke, R. J . 2002. Trout and Salmon of North America. Simon and Schuster, New York. Belkhir, K., P. Borsa, N. Chikhi, N, Raufaste, and F. Bonhomme. 2001. G E N E T I X 4.03, logiciel sous windows T M pour la genetique des populations, in. Laboratoire Genome, Populations, Interactions, C N R S U M R 5000, Universite de Montpellier II, Montpellier, France. Bernatchez, L, and C. C. Wilson. 1998. Comparative phylogeography of nearctic and palearctic fishes. Molecular Ecology 7:431-452. Bettles, C. 2004. Interspecific hybridization between sympatric coastal cutthroat and coastal rainbow/steelhead trout on Vancouver Island, British Columbia: A conservation and evolutionary examination. M.Sc. thesis. Dept. of Biology. University of Windsor, Windsor, Ontario. Bogart, J . 1980. Evolutionary implications of polyploidy in amphibians and reptiles. Pp. 117-133 in W. Lewis, ed. Polyploidy: Biological Relevance. Plenum, New York. Booy, G., R. J . J . Hendriks, M. J . M. Smulders, J . M. Van Groenendael, and B. Vosman. 2000. Genetic diversity and the survival of populations. Plant Biology 2:379-395. Burke, J . , and M. Arnold. 2001. Genetics and the fitness of hybrids. Annual Review of Ecology & Systematics 35:31-52. Campton, D., and F. Utter. 1985. Natural hybridization between steelhead trout (Salmo gairdneri) and coastal cutthroat trout {Salmo clarki clarki) in two Puget Sound streams. Canadian Journal of Fisheries and Aquatic Sciences 42:110119. Carson, E., and T. Dowling. 2006. Influence of hydrogeographic history and hybridization on the distribution of genetic variation in the pupfishes Cyprinodon atrorus and C. bifasciatus. Molecular Ecology 15:667-679. Caughley, G. 1994. Directions in conservation biology. Journal of Animal Ecology 63:215-244. Charlesworth, B., and D. Charlesworth. 1999. The genetic basis of inbreeding depression. Genetical Research 74:329-340. Chevassus, B. 1979. Hybridization in salmonids: results and perspectives. Aquaculture 17:113-128.  51  Clegg, M. T. 1997. Plant genetic diversity and the struggle to measure selection. Journal of Heredity 88:1 -7. Coltman, D., and J . Slate. 2003. Microsatellite measures of inbreeding: a metaanalysis. Evolution 57:971-983. Condrey, M. J . , and P. Bentzen. 1998. Characterization of coastal cutthroat trout (Oncorhynchus clarki clarki) microsatellites and their conservation in other salmonids. Molecular Ecology 7:787-789. Costello, A. B., T. Down, S. Pollard, C. Pacas, and E. Taylor. 2003. The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: an examination of microsatellite DNA variation in bull trout, Salvelinus confluentus (Pisces: Salmonidae). Evolution 57:328-344. Costello, A. B., and E. Rubidge. 2003. C O S E W I C status report on cutthroat trout (Oncorhynchus clarki spp.). Pp. 73pp. Prepared for the Committee on the Status of Endangered Wildlife in Canada, Ottawa. Costello, A. B., and E. Rubidge. 2005. C O S E W I C status report on coastal cutthroat trout (Oncorhynchus clarkii clarki). Draft Report. Pp. 91pp. Prepared for the Committee on the Status of Endangered Wildlife in Canada, Ottawa. Costello, A. B., E. B. Taylor, S. M. Pollard, and R. H. Devlin. 2001. Cutthroat Trout in British Columbia: A Phylogenetic Assessment of Provincial Coastal and Westslope Populations. Pp. 249-254 in M. Brewin, A. Paul and M. Monita, eds. Bull Trout II. Trout Unlimited Canada, Canmore, A B . Docker, M., A. Dale, and D. Heath. 2003. Erosion of interspecific reproductive barriers resulting from hatchery supplementation of rainbow trout sympatric with cutthroat trout. Molecular Ecology 12:3515-3521. Dowling, T., and C. Secor. 1997. The role of hybridization and introgression in the diversification of animals. Annual Review of Ecology & Systematics 28:593-619. Einum, S., and I. A. Fleming. 2001. Implications of stocking: Ecological interactions between wild and released salmonids. Nordic J . Freshw. Res. 75:56-70. Ellstrand, N., and D. Elam. 1993. Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology & Systematics 24:217-243. Epifanio, J . , and D. Philipp. 2001. Simulating the extinction of parental lineages from introgressive hybridization: the effects of fitness, initial proportions of parental taxa, and mate choice. Reviews in Fish Biology and Fisheries 10:339-354. Ferguson, M. M., R. G. Danzmann, and F. W. Allendorf. 1988. Developmental  success of hybrids between two taxa of salmonid fish with moderate structural  52  gene divergence. Canadian Journal of Zoology 66:1389-1395. Flagg, T., B. Berejikian, J . Colt, W. Dickhoff, L. Harrell, D. Maynard, C. Nash, M. Strom, R. Iwamoto, and C. Mahnken. 2000. Ecological and behavioural impacts of artificial production strategies on the abundance of wild salmon populations a review of practices in the Pacific Northwest. N O A A Tech. Memo. NMFSNWFS-4. Pp. 98. Northwest Fish. Sci. Center, Seattle, WA. Forbes, S. H., and F. W. Allendorf. 1991. Associations between mitochondrial and nuclear genotypes in cutthroat hybrid swarms. Evolution 45:1332 - 1349. Frankham, R. 1996. Relationship of genetic variation to population size in wildlife. Conservation Biology 10:1500-1508. Frankham, R. 1998. Inbreeding and Extinction: Island populations. Conservation Biology 12:665-675. Funk, D., and K. Omland. 2003. Species-level paraphyly and polyphyly: Frequency, causes, and consequences, with insights from animal mitochondrial DNA. Annual Review of Ecology & Systematics 34:397-423. Gaston, K. 2000. Global patterns in biodiversity. Nature 405:220-227. Goudet, J . 2001. FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3). Available from http ://www. u n i I. ch/izea/softwa res/fstat. htm I. Grant, P., and B. Grant. 1992. Hybridization of bird species. Science 256:193-197. Grant, P., and B. Grant. 1994. Phenotypic and genetic effects of hybridization in Darwin's finches. Evolution 48:297-316. Griswold, K. E. 1996. Genetic and meristic relationships between coastal cutthroat trout (Oncorhynchus clarki clarki) residing above and below barriers in two coastal basins. Master's Dissertation. Pp. 83. Oregon State University, Corvallis. Hansen, M. M., E. Kenchington, and E. Nielsen. 2001. Assigning individual fish to populations using microsatellite DNA markers. Fish and Fisheries 2:93-112. Hansson, B., and L. Westerberg. 2002. On the correlation between heterozygosity and fitness in natural populations. Molecular Ecology 11:2467-2474. Hartman, G., and C. Gill. 1968. Distribution of juvenile steelhead and cutthroat trout (Salmo gairdneri and S. clarki clarki) within streams in southwestern British Columbia. Journal of the Fisheries Research Board of Canada 25:33-48. Hawkins, D. K., and C. J . Foote. 1998. Early survival and development of coastal cutthroat trout (Oncorhynchus clarki clarki), steelhead (Oncorhynchus mykiss), and reciprocal hybrids. Canadian Journal of Fisheries & Aquatic Sciences  53  55:2097-2104. Hawkins, D. K., and T. P. Quinn. 1996. Critical swimming velocity and associated morphology of juvenile coastal cutthroat trout (Oncorhynchus clarki clarki), steelhead trout (Oncorhynchus mykiss), and their hybrids. Canadian Journal of Fisheries and Aquatic Sciences 53:1487-1496. Hedrick, P. W., and S. T. Kalinowski. 2000. Inbreeding depression in conservation biology. Annual Review of Ecology & Systematics 31:139-162. Hilborn, R. 1992. Hatcheries and the future of salmon in the northwest. Fisheries 17:5-8. Hindar, K., N. Ryman, and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Canadian Journal of Fisheries and Aquatic Sciences 48:945957. Hitt, N., C. Frissel, C. Muhlfeld, and F. W. Allendorf. 2003. Spread of hybridization between native westslope cutthroat trout, Oncorhynchus clarki lewisi, and nonnative rainbow trout, Oncorhynchus mykiss. Canadian Journal of Fisheries and Aquatic Sciences 60:1440-1451. Hubbs, C. 1955. Hybridization between fish species in nature. Systematic Zoology 4:1-20. Jacobs, D., T. Haney, and K. Louie. 2004. Genes, diversity, and geological process on the Pacific Coast. Annual Review of Earth and Planetary Sciences 32:601652. Johnson, O., M. Ruckelshaus, W. Grant, F. Waknitz, A. Garrett, G. Bryant, K. Neely, and J . Hard. 1999. Status review for coastal cutthroat from Washington, Oregon, and California. Pp. 292. National Marine Fisheries Service, Seattle, WA. Krueger, C , and B. May. 1991. Ecological and genetic effects of salmonid introductions in North America. Canadian Journal of Fisheries and Aquatic Sciences 48 (Suppl. 1):66-77. Leary, R. F., F. W. Allendorf, S. R. Phelps, and K. L. Knudsen. 1984. Introgression between westslope cutthroat and rainbow trout in the Clark Fork River drainage. Montana. Proc. Mont. Acad. Sci 43:1-18. Leberg, P. 2002. Estimating allelic richness: Effects of sample size and bottlenecks. Molecular Ecology 11:2445-2449. Lewontin, R., and L. Birch. 1966. Hybridization as a source of variation for adaptation to new environments. Evolution 20:315-336. Lynch, M. 1991. The genetic interpretation of inbreeding depression and outbreeding depression. Evolution 45:622-629.  54  Lynch, M. 1996. A quantitative genetic perspective on conservation issues. Pp. 471 501 in J . C. Avise and J . Hamrick, eds. Conservation Genetics: Case Histories from Nature. Chapman and Hall, New York. Maddison, W. 1997. Gene trees in species trees. Systematic Biology 46:523-536. Mallet, J . 2005. Hybridization as an invasion of the genome. Trends in Ecology and Evolution 20:229-237. Manel, S., O. Gaggiotti, and R. S. Waples. 2005. Assignment methods: matching biological questions with appropriate techniques. Trends in Ecology and Evolution 20:136-142. McElhany, P., M. Ruckelshaus, M. J . Ford, T. Wainwright, and E. Bjorkstedt. 2000. Viable salmonid populations and the recovery of evolutionarily significant units. N O A A Tech. Mem. NMFS-NWFSC-42, Seattle, WA. McPhail, J . D., and R. Carveth. 1993. Field key to the freshwater fishes of British Columbia. Fish Museum, Dept. of Zoology UBC, Vancouver. Meffe, G., and 6. Carroll. 1994. Principles of Conservation Biology. Sinauer Associates, Sunderland, MA. Molony, B., R. Lenanton, G. Jackson, and J . Norriss. 2003. Stock enhancement as a fisheries management tool. Reviews in Fish Biology and Fisheries 13:409-432. Montgomery, D. 2000. Coevolution of the Pacific salmon and Pacific Rim topology. Geology 28:1107-1110. Montgomery, M., L. Woodworth, R. Nurthen, D. Gilligan, D. Briscoe, and R. Frankham. 2000. Relationships between population size and loss of genetic diversity: comparisons of experimental results with theoretical predictions. Conservation Genetics 1:33-43. Moran, P. 2002. Current conservation genetics: building an ecological approach to the synthesis of molecular and quantitative genetic methods. Ecology Freshwater Fish 11:30-55. Morris, D. B., K. R. Richard, and J . M. Wright. 1996. Microsatellites from rainbow trout {Oncorhynchus mykiss) and their use for genetic study of salmonids. Canadian Journal of Fisheries and Aquatic Sciences 53:120-126. Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. Olsen, J., P. Bentzen, and J . Seeb. 1998. Characterization of seven microsatellite loci derived from pink salmon. Molecular Ecology 7:1087-1089.  55  Ostberg, C. O., and R. Rodriguez. 2004. Bi-parentally inherited species-specific markers identify hybridization between rainbow trout and cutthroat trout subspecies. Molecular Ecology Notes 4:26-29. Ostberg, C. O., S. Slatton, and R. Rodriguez. 2004. Spatial partitioning and asymmetric hybridization among sym'patric coastal steelhead trout (Oncorhynchus mykiss irideus), coastal cutthroat trout (O. clarki clarki) and interspecific hybrids. Molecular Ecology 13:2773-2788. Paetkau, D., W. Calvert, I. Stirling, and C. Strobeck. 1995. Microsatellite analysis of population structure in Canadian polar bears. Molecular Ecology 4:347-354. Pagano, A., P. Crochet, J . Graf, P. Joly, and T. Lode. 2001. Distribution and habitat use of water frog hybrid complexes in France. Global Ecology and Biogeography 10:433-441. Pamilo, P., and M. Nei. 1988. Relationships between gene trees and species trees. Molecular Biology & Evolution 5:568-583. Park, L., P. Moran, and D. Dightman. 1996. A chinook salmon P C R - R F L P marker in the p53 locus. Animal Genetics 27:127-128. Park, S. 2001. Trypanotolerance in West African Cattle and the Population Genetic Effects of Selection. Ph.D. Dissertation. University of Dublin. Pearcy, W. 1997. The sea-run and the sea. Pp. 29-34 in J . D. Hall, P. A. Bisson and R. Gresswell, eds. Sea-run cutthroat trout: biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis. Piry, S., A. Alapetite, J . Cornuet, D. Paetkau, L. Badouin, and A. Estoup. 2004. G E N E C L A S S 2 : A software for genetic assignment and first-generation migrant detection. Journal of Heredity 95:536 - 539. Pollard, W., G. Hartman, C. Groot, and P. Edgell. 1997. Field identification of coastal juvenile salmonids. Pp.31. Harbour Publishing, Madeira Park, BC. Pritchard, J., M. Stephens, and P. Donnely. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945-959. Quinn, T. P., and A. Dittman. 1990. Pacific salmon migrations and homing: mechanisms and adaptive significance. Trends in Ecology and Evolution 5:174177. Raymond, M., and F. Rousset. 1995. G E N E P O P (version 1.2): Population genetics software for exact tests and ecumenicism. Journal of Heredity 86:248-249. Redenbach, Z., and E. B. Taylor. 2002. Evidence for historical introgression along a contact zone between two species of char (Pisces: Salmonidae) in northwestern North America. Evolution 56:1021-1035.  56  Reed, D., and R. Frankham. 2001. How closely related are molecular and quantitative measures of genetic variation: a meta-analysis. Evolution 55:1095-1103. Reeves, G., J . D. Hall, and S. V. Gregory. 1997. The impact of land-management activities on coastal cutthroat trout and their freshwater habitats. Pp. 138-144 in J . D. Hall, P. A. Bisson and R. Gresswell, eds. Sea-run cutthroat trout: Biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis. Rhymer, J., and D. Simberloff. 1996. Extinction by hybridization and introgression. Annual Review of Ecology & Systematics 27:83-109. Rieseberg, L. 1997. Hybrid origins of plant species. Annual Review of Ecology & Systematics 28:359-389. Rousset, F. 1996. Equilibrium values of measures of population subdivision for stepwise mutation processes. Genetics 142:1357-1362. Rubidge, E., P. Corbett, and E. Taylor. 2001. A molecular analysis of hybridization between native westslope cutthroat trout and introduced rainbow trout in southeastern British Columbia, Canada. Journal of Fish Biology 59:42-54. Ryman, N., and L. Laikre. 1991. Effects of supportive breeding on the genetically effective population size. Conservation Biology 5:325-329. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Scribner, K. T., J . Gust, and R. Fields. 1996. Isolation and characterization of novel salmon microsatellite loci: cross-species amplification and population genetic applications. Canadian Journal of Fisheries & Aquatic Sciences 53:833-841. Scribner, K. T., K. Page, and M. Bartron. 2001. Hybridization in freshwater fishes: a review of case studies and cytonuclear methods of biological inference. Reviews in Fish Biology and Fisheries 10:293-323. Seehausen, O. 2004. Hybridization and adaptive radiation. Trends in Ecology and Evolution 19:198-207. Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science 236:787-792. Smith, P., A. Konings, and I. Kornfield. 2003. Hybrid origin of a cichlid population in Lake Malawi: implications for genetic variation and species diversity. Molecular Ecology 12:2497-2504. Spielman, D., B. Brook, D. Briscoe, and R. Frankham. 2004. Does inbreeding and loss of genetic diversity decrease disease resistance? Conservation Genetics  57  5:439-448. Spruell, P., J . Pearce Smithwick, K. Knudsen, and F. W. Allendorf. 1998. Genetic analysis of rainbow and cutthroat trout from the lower Columbia River. Progress Report to Oregon Department of Fish and Wildlife. WTSGL98-103. Sutherland, W. 1996. From Individual Behaviour to Population Ecology. Oxford University Press. Oxford. Taylor, E. 2004. Evolution in mixed company. Pp. 232-263 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated. Salmon and their Relatives. Oxford University Press. Taylor, E. B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98:185-207. Trotter, P. C. 1987. Cutthroat: Native trout of the West. Colorado Associated University Press, Boulder, C O . Trotter, P. C. 1989. Coastal cutthroat trout: a life history compendium. Transactions of the American Fisheries Society 118:463 - 473. Utter, F. 2001. Patterns of subspecific anthropogenic introgression in two salmonid genera. Reviews in Fish Biology and Fisheries 10:265-279. Utter, F. 2004. Population genetics, conservation and evolution in salmonids and other widely cultured fishes: some perspectives over six decades. Reviews in Fish Biology and Fisheries 14:125-144. Verspoor, E., and P. Hammar. 1991. Introgressive hybridization in fishes: the biochemical evidence. Journal of Fish Biology 39 (Suppl A.):309-334. Wang, S., J . Hard, and F. Utter. 2002a. Genetic variation and fitness in salmonids. Conservation Genetics 3:321-333. Wang, S., J . Hard, and F. Utter. 2002b. Salmonid inbreeding: a review. Reviews in Fish Biology and Fisheries 11:301-319. Waples, R. S. 1995. Genetic effects of stock transfers offish. Pp. 51-69 in D. Phillips, ed. Protection of aquatic biodiversity. Proceedings of the World Fisheries Congress. Theme 3., New Delhi. IBH, Oxford. Waples, R. S. 1999. Dispelling some myths about hatcheries. Fisheries 24:12-21. Waser, N., and M. Price. 1994. Crossing-distance effects in Delpinium nelsonii:  outbreeding and inbreeding depression in progeny fitness. Evolution 48:842-852.  Wenburg, J . K., and P. Bentzen. 2001a. Genetic and behavioural evidence for  restricted gene flow among coastal cutthroat trout populations. Transactions of  58  the American Fisheries Society 130:1049-1069. Wenburg, J . K., P. Bentzen, and C. J . Foote. 1998. Microsatellite analysis of genetic population structure in an endangered salmonid - the coastal cutthroat trout {Oncorhynchus clarki clarki). Molecular Ecology 7:733-749. Whitlock, M . ' C , P. K. Ingvarsson, and T. Hatfield. 2000. Local drift load and the heterosis of interconnected populations. Heredity 84:452-457. Widmer, A., and C. Lexer. 2001. Glacial refugia: sanctuaries for allelic richness, but not for gene diversity. Trends in Ecology & Evolution 16:267-269. Wilson, C. C., and L. Bernatchez. 1998. The ghosts of hybridization past: fixation of arctic char (Salvelinus alpinus) mitochondrial DNA in an introgressed population of lake trout (S. namaycush). Molecular Ecology 7:127-132. Young, W., C , O. Ostberg, P. Keim, and G. H. Thorgaard. 2001. Genetic characterization of hybridization and introgression between anadromous rainbow trout (Oncorhynchus mykiss irideus) and coastal cutthroat trout (O. clarki clarki). Molecular Ecology 10:921-930. Zar, J . 1999. Biostatistical Analysis. Prentice Hall, New Jersey. Zimmerman, C., K. P. Currens, and G. Reeves. 1997. Genetic population structure of coastal cutthroat trout in the Muck Creek basin, Washington. Pp. 170-172 in J . D. Hall, P. A. Bisson and R. Gresswell, eds. Sea-run cutthroat trout: biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis.  59  C H A P T E R 3: T H E B R E E D I N G S Y S T E M O F C O A S T A L C U T T H R O A T T R O U T (ONCORHYNCHUS  CLARKII  CLARKI)  AND THE MAINTENANCE OF GENETIC  DIVERSITY IN S M A L L P O P U L A T I O N S  2  "Sexual conflicts are inevitable since males and females in most species differ in the cost and benefits of mating with one another." - R e y n o l d s (1996:72)  3.1 INTRODUCTION 3.1.1 The maintenance of genetic variation in small populations  The heritable component of phenotypic variation in populations has long been recognized as an essential part of a species' ability to respond to changes in the environment. Such variability not only affords populations the flexibility to persist in specific habitats, but may also allow them to colonize a larger range of habitats than might otherwise be possible (Vrijenhoek and Lerman 1982; Carvalho 1993; Ellstrand and Elam 1993; Vrijenhoek 1994). A primary goal of many conservation programs is, therefore, the preservation of this adaptive potential in populations and the maintenance of its heritable component, genetic variation. While three factors ultimately determine the standing amount of genetic variation'in populations (selection, drift and mutation), the relative influence of these factors vary over spatial and temporal scales in accordance with population size. For small populations (i.e. less than 100 adult individuals), the influence of drift and other stochastic factors tend to predominate, leaving small populations susceptible to increased losses of genetic variability and population decline (Lacy 1987; Newman and Pilson 1997; Saccheri et al. 1998). Indeed, meta-analyses of the empirical evidence tend to suggest that small and declining populations often carry less genetic variation than larger ones, appear more inbred, and may be more susceptible to extinction (Ellstrand and Elam 1993; Vrijenhoek 1994; Frankham 1995; Amos and Harwood 1998; De Rochambeau et al.  A version of this chapter will be submitted for publication. Costello, A. "The breeding system of coastal cutthroat and the maintenance of genetic diversity in small populations". 2  60  2000; Hedrick and Kalinowski 2000). The widespread decline of many species around the globe has generated much interest in conservation circles for practical ways of monitoring the effects of demographic trends on population viability. While it may be desirable to estimate demographic health in terms of population size alone (relatively easy to measure and often used as a key index of population viability), the relationship between census size and viability may not always be a straightforward one (e.g., - Geiger et al. 1997; McElhany et al. 2000). More appropriate, perhaps, is to make inferences about the demographics of a population of interest with reference to a standard benchmark such as its effective population size, a concept that has become central to the fields of evolutionary population genetics and conservation biology (Lande and Barrowclough 1987; Nunney 2000). The effective population size (N ) represents the e  size of an idealized Wright-Fisher population (one of constant size and even sex ratio, discrete generations, random breeding, and random survivorship) that experiences the same losses of genetic variation as the population of interest (Wright 1931, 1938). It has a well-established theoretical basis that often allows for valuable insight into the factors most affecting the demography of populations in the wild. The concept is particularly relevant to the study of small populations in that N is the is the single e  factor most influencing the amount of inbreeding or genetic drift and therefore, the main determinant of the rate at which genetic diversity is lost (A H = 1/2 N ). In terms e  of conservation, it is the ratio of the effective to census population sizes (N / N ) that e  c  is perhaps of more relevance; the ratio can be thought of as representing the proportion of a total population that effectively contribute to reproduction. In the majority of cases, N appears to be less than Nc (in some cases, by several orders of e  magnitude); recent reviews have suggested that N / N ratios are commonly < 0.50 e  c  and may average as low as ~ 0.11 (Nunney 1995; Frankham 1996) suggesting that the numbers of individuals actually supporting populations can be quite low.  3.1.2  Effective population size (N ) and the N / N ratio e  e  The types of breeding systems present in nature vary incredibly and represent a continuum from the "chuck it and chance it" types (e.g., - oysters, whose gametes are  61  broadcast and mixed by ocean currents), to those in which parental care may persist throughout most of the organism's life (Emlen and Oring 1977; Reynolds 1996). At the very least, most species have evolved mechanisms to reduce the occurrence of inbreeding in populations (dispersal, kin recognition; e.g. - Pusey and Wolf 1996) and further social or breeding structure often results in non-random breeding (Storz 1999). Accordingly, the breeding structure of a population is expected to have a major influence on its effective population size and the levels of genetic variation that it can carry. Three variables are primarily responsible for reducing N below the number of e  sexually mature adults in a population: fluctuations in population size, variance in individual reproductive success above random (binomial) expectations, and unequal sex ratios among breeders (Caballero 1994; Nunney 1999; Wang and Caballero 1999). While highly fecund species or those with variable juvenile survivorship are generally expected to have lower ratios than those species following stable, monogamous breeding strategies, the relative influence of these factors in the wild is not known with certainty. Frankham (1995), for example, suggested that fluctuations in population size was the largest variable reducing N / N in natural populations. e  Others have argued that high variance in reproductive success and unequal contribution to future generations may be the principal determinant (Nunney 1996; Storz et al. 2001). A better understanding of the factors most impacting N and N / N e  e  c  ratios is required if these concepts are to find greater utility in the conservation and management of declining species (Lande and Barrowclough 1987; Caballero and Toro 2000; Rieman and Allendorf 2001). Unfortunately, empirical evidence from wild populations is somewhat limited as the majority of studies to date have come from simulated data or from controlled demographic scenarios involving captive or managed populations and, therefore, may not necessarily be representative of the wild condition.  To address some of these uncertainties in a natural and inherently more complex demographic scenario, I investigated the breeding system of a wild population of coastal cutthroat trout {Oncorhynchus clarkii clarki), a salmonid fish native to western North America. Coastal cutthroat trout are interesting in that populations carry some of the highest levels of genetic variability reported for anadromous fishes (Utter et al. 1980; Wenburg and Bentzen 2001) while having sizes  62  that are much smaller than other salmonids (tens to hundreds as opposed to thousands to tens of thousands). Coastal cutthroat trout may, therefore, appear to have higher levels of genetic variation that one might predict based on their population sizes alone. Because breeding systems are so intimately tied to the forward transmittal of genetic material, it is possible that the breeding system of coastal cutthroat trout has evolved to counteract the effects of small fluctuating population size and may contribute significantly to the maintenance of genetic variation between generations (e.g., - Martinez et al. 2000; Garant et al. 2003).  3.1.3  The breeding system of coastal cutthroat trout  While the breeding system of coastal cutthroat trout has been understudied relative to other Oncorhynchus species, populations appear to exhibit the complex polygynandrous breeding system typical of salmonids (reviewed by Fleming and Reynolds 2004). Spawners return or "home" to small natal streams (Hasler and Scholz 1983) where females compete for preferred spawning areas and males compete for access to females. Fertilization is external and eggs are deposited in gravel substrates in nest complexes called redds. Females may show strong preference for certain spawning sites as better quality nest habitat determines, to a large extent, levels of egg mortality (Chapman 1988) and juvenile survival following emergence (Gibson 1993). Limiting amounts of quality spawning habitats can, therefore, lead to strong competition amongst females and high variance in individual reproductive success. Larger females may be at a selective advantage; not only because fecundity is known to increase with size (Giger 1972), but because they may be better able to exclude subdominant females from quality habitats. Furthermore, larger females may dig deeper redds which may be more resistant to disturbance by subsequent spawning events. Males contribute no parental care for eggs and so their reproductive success is determined primarily by their ability to gain access to females (Fleming and Reynolds 2004). Large "fighter" males establish breeding territories based on size and social dominance and attempt to mate with all females in the immediate area. Typically, this results in a relatively high variance in reproductive success among males with larger  63  individuals having greater success (e.g. - Fleming and Gross 1994). Alternate reproductive strategies are, however, available to circumvent direct competition. Sneaking tactics (where smaller males dart in next to a spawning pair to attempt fertilization of some of the eggs) are often employed by small stream resident male coastal cutthroat trout (Trotter 1989; Behnke 1992). These sneaking strategies have often been shown to be quite successful in other species (Garcia- Vazquez et al. 2001; Fleming and Reynolds 2004 and references therein) and may reduce the potential of certain fighter males to dominate successful matings in a population. A s noted, the number of spawners supporting coastal cutthroat trout populations are generally small (on the order of tens to hundreds) and may fluctuate widely from year to year (Sumner 1952, 1962). Few populations, however, are systematically monitored and actual trend data are quite limited. It is known that the age and size at maturity for coastal cutthroat trout is extremely variable across populations and life history types. In British Columbia, nonmigratory individuals typically mature at 2 - 3 years of age (and as small as 120 mm for precocial stream resident males) whereas anadromous or "sea-run" cutthroat trout rarely spawn before age 3 in even the most productive systems; an age of 4 is likely more typical (Slaney 2005). A s such, spawning populations in any one year may represent a mix of cohorts from several brood years. Unlike the Pacific salmon, coastal cutthroat trout are iteroparous (repeatspawning) so that some fish may reproduce every year or in alternate years. Energetic requirements and physical injury, however, often lead to considerable postbreeding mortality, especially for males (e.g., - Stearly 1992). The small and fluctuating population sizes of coastal cutthroat trout do not necessarily appear conducive to the maintenance of high levels of variation. High fecundities and intense breeding competition for both sexes should lead to high variance in reproductive success and further reductions in N / N ratios. What factors e  c  then are responsible for the high levels of genetic variation reported in coastal cutthroat trout populations? Many, of course, may be implicated, including gene flow between adjacent populations (Pannell and Charlesworth 2000; Consuegra et al. 2005) or historical effects that may not be readily apparent (Walter and Epperson 2001; Storz et al. 2002). These issues have been addressed separately (see other  64  chapters) and I restrict my discussion here to the possible contributions of the breeding system. In the simplest case scenario, coastal cutthroat trout populations may simply be larger than expected based on census sizes so that many individuals are available to contribute to breeding. Coastal cutthroat trout use very different habitat types at various life history stages and at different times of the year (reviewed by Trotter 1989). It may be that members of a population are spread out into a variety of different habitats so that counts in one area may miss a significant portion of the population at a given time. Populations may not, in fact, fluctuate that significantly when all life history components have been accounted for. Alternatively, it may be that the breeding system has evolved to compensate for the small number of spawners. For example, sex ratios among spawners should be roughly equal and variance in reproductive success not exceed random expectations to maximize N / N ratios (Wang 1997; De Rochambeau et al. 2000). There is some e  c  evidence to suggest that sex ratios in coastal cutthroat trout may favor females (Wenburg 1998 and references therein), although the presence of mature stream resident males may bring ratios closer to unity. Subordinate or "satellite" males in other salmonid species are often fairly successful in gaining access to females while the dominant male is distracted and may help reduce the variance in male reproductive success (Moran et al. 1996; Garcia- Vazquez et al. 2001). The multiple paternity which results (the eggs of an individual female are fertilized by more than one male) has been implicated as an important factor in the maintenance of high levels of genetic variation in small populations elsewhere (Moran and GarciaVazquez 1998; Martinez et al. 2000). Breeding between differentiated groups will tend to promote genetic diversity as different alleles (or different frequencies of the same alleles) are often maintained in reproductively isolated groups (Slatkin 1987). In the context of a closed breeding population of coastal cutthroat trout (evidence in Chapter 4 suggests that gene flow between adjacent populations is negligible), this may involve matings between differentiated cohorts or between different life history forms. Small, but significantgenetic differences have been observed between such groups in many salmonids, including coastal cutthroat trout (Campton and Utter 1987; Foote et al. 1989; Currens et al. 1990; Hindar et al. 1991; Wenburg et al. 1998).  65  To investigate the relative influence of these factors in a wild population, I studied the demographic and genetic characteristics of a representative coastal cutthroat population over two successive years. Through genetic parentage analyses of the relationships between spawners and young-of-the-year (YOY) offspring, I describe the number and composition of successful spawners in the system as well as the effects of breeding patterns on N , N / N ratios, and the maintenance of e  e  c  genetic variation in the population. If the breeding system has evolved to compensate for small variable population sizes, it may be expected that (1) many individuals contribute to breeding; (2) sex ratios are equal; (3) variation in individual reproductive success is minimal and randomly (Poisson) distributed; (4) multiple mates or breeding events are common; and (5) breeding occurs between age-classes and divergent lifehistory forms. A s noted, the breeding system of coastal cutthroat trout is poorly understood relative to other salmonids and to my knowledge, this is the first comprehensive assessment for a natural spawning population. I discuss the evolutionary and conservation implications of these findings with respect to the breeding system of two other salmonids: Atlantic salmon (Salmo salar), which is relatively well-studied and steelhead trout (Oncorhynchus mykiss), sister species to coastal cutthroat which is also poorly known but the subject of several recent studies  3.2 M E T H O D S A N D M A T E R I A L S 3.2.1 Study site and field sampling  The site chosen for the breeding study was Chonat Lake, located on the north end of Quadra Island, British Columbia (50° 17' N, 125° 16' W). It is a natural and fairly tractable system that allowed me to successfully monitor all spawning activity for a representative coastal cutthroat population. The lake itself is small (48 ha) with one main inlet stream and one outlet stream that enters the ocean at Chonat Bay in the Okisollo Channel (Figure 3-1). The inlet stream offers the only suitable spawning habitat in the system and is bounded by an impassable barrier (20 m falls) at -700 m from the creek mouth that prevents further upstream movement of fish. The fish fauna of the Chonat Lake system is a relatively simple assemblage of species dominated by  66  coastal cutthroat trout, prickly sculpin (Cottus asper), and three-spined stickleback (Gasterosteus aculeatus). The prickly sculpin is certainly the dominant littoral species in the lake and outlet in terms of numbers; at least 5 age classes are present. Threespined sticklebacks are present in large numbers in the lake and outlet, and appear to be planktivourous; in summer they can be seen schooling throughout the lake and likely provide a significant food source for lake resident coastal cutthroat trout. Coho (Oncorhynchus kisutch) and chum (O. keta) salmon have been observed to spawn in the inlet in small numbers but do not appear to have a large presence in the system. The Chonat Lake system has an apparently healthy population of coastal cutthroat trout comprising three different life history types: lake-resident (lacustrine), sea-run (anadromous), and stream-resident fish. The sexually mature stream resident males (which range in size from 100 -200 mm) are present only in the inlet stream and employ a "sneaking" strategy to reproduce (Trotter 1989; A. Costello, personal observation). To document the number and composition of potential spawners in the system, a small fish enumeration fence was installed at the mouth of the Chonat Lake inlet stream in January 2001. The fence was then monitored daily during the spawning periods of 2001 and 2002 (which typically peak in February and March). Fish passing the fence were anaesthetized with a mild sedative (MS-222) and tagged with visible implant tags (VI Alpha, Northwest Marine Technology, Inc.) in the anal fin. The VI tags have a unique alphanumeric code (e.g. - Y65) that allow individual fish to be identified upon recapture. Based on evidence collected in Chapter 4, migration into Chonat Lake is expected to be negligible and I consider the system effectively 'closed' for the purposes of this study. For each year, I estimated total adult population size in the system (N .including spawners and non-spawners) using markrecapture data and the corrected Lincoln-Peterson estimate (Pollock et al. 1990): ^  =  (n +1)(n +1) 1  2  (m +1) 2  where n-i is the number of marked individuals in the population, n is the total number 2  67  captured at time t (2001 or 2002), m is the number of recaptures of marked 2  individuals at time t; 9 5 % confidence limits were obtained using the following formula from (Pollock et al. 1990):  (3.2)  The harmonic mean of 2001 and 2002 estimates was taken as the average total population size (N). Before release, adult fish were sexed, measured for fork length (mm), and scale and fin tissue samples were collected. Note that only those fish actively expressing gametes (i.e. - sexually mature fish) were considered as potential parents in subsequent parentage analyses. The spawning streams typically inhabited by coastal cutthroat trout are generally quite small and susceptible to habitat degradation (Reeves et al. 1997; Rosenfeld 2001). In an effort to minimize my effect on the population, I chose not to disturb spawning gravels to sample full or half-sib families in redds (at the egg or alevin stage; e.g. - Thompson et al. 1998). Instead, I waited until fry had emerged from the substrate (early summer, typically during the month of May) and sampled a large number of young-of-the-year (YOY) progeny using pole seines. The small size of the Chonat Lake inlet stream, however, allowed for complete sampling of all available rearing habitat and ensured a good representative sample of Y O Y offspring (n = 197 and 156 in 2001 and 2002, respectively). All collected fin clips were stored in 9 5 % ethanol until DNA could be isolated from ~ 5 mg of tissue using the Puregene DNA isolation kit (Gentra Systems, Inc.).  3.2.2  Microsatellite amplification and basic genetic information  A total of 13 microsatellite loci were screened for variability in a subset of spawners. Seven loci were eventually chosen for inclusion based on amplification quality and two multiplex P C R based on those of Wenburg and Bentzen (2001) were developed (Table 3.1). Polymerase chain reactions were carried out using fluorescently-labeled primers in 10 ml volumes of 10 mM Tris-HCl (pH 8.3), 2 mM  68  MgCI , 0.8 mM dNTP's, and 0.4 units of AmpliTaq Gold Polymerase in MJ P T C 100 2  thermocyclers using a basic cycle profile of: 1 cycle (95°C/ 3 min), 10 cycles (94°C7 1 min, T / 30 sec, 72°C/ 20 sec), 20 cycles (94°C / 30 s, T / 30 sec, 72°C / 20 sec), A  A  and 1 cycle (72°C / 5min), where T is the annealing temperature (see Table 3.1). A  P C R products were electrophoresed on ABI Genescan Gels (using ABI Prism 377 sequencers) with R O X labeled internal size standards. Raw fragment analysis and allele binning was facilitated by the ABI G E N O T Y P E R v3.7 utility program. Characterization of the seven microsatellite loci (number of alleles, heterozygosities, polymorphic information content (PIC), exclusionary probabilities (Dodds et al. 1996), null allele frequencies, etc.) were performed using the parentage assignment program C E R V U S (Marshall et al. 1998). Across all samples, two of the microsatellite loci (Omy77 and Ogo8) were found to show significant heterozygote deficiencies (primarily in fry samples) which could suggest that non-amplifying (null) alleles are present at these loci. To identify their effect on parentage assignment, I manually recoded all hpmozygotes at these loci as null heterozygotes according to the procedure implemented in the parentage assignment program, P R O B M A X (Danzmann 1997). The correction did not significantly improve parentage assignment and was not used further in the analyses. Summary genetic information for spawners and Y O Y fry samples (observed and expected heterozygosity, mean number of alleles per locus and total alleles per sample) were calculated using the Microsatellite Toolkit macro for MS E X C E L (Park 2001). Tests for deviations from Hardy-Weinberg and linkage equilibria were performed using the Markov chain method implemented in G E N E P O P ver. 3.1 (Raymond and Rousset 1995) with default values. Inbreeding coefficients (F ) were is  calculated using F S T A T (Goudet 2001). Finally, because fluctuations in population size can dramatically effect diversity estimates and effective population size (e.g., Frankham 1995), the program B O T T L E N E C K (Piry et al. 1999) was used to test for any recent bottlenecks in Chonat Lake samples. The first test measures deviations from the expected "L-shaped" allele frequency distribution (Luikart et al. 1998) based on coalescent simulation. The second evaluates the significance of any excesses in expected heterozygosity relative to that expected at drift-mutation equilibrium using  69 both sign and Wilcoxon's signed-rank tests under the stepwise (SMM) and two-phase (TPM) mutation models (Luikart and Cornuet 1998).  3.2.3  Parentage analysis and reproductive s u c c e s s  A number of methods exist which will allow progeny to be assigned back to parents on the basis of genotypic allele frequencies (reviewed by Wilson and Ferguson 2002; Jones and Ardren 2003). Many methods, however, require complete sampling of potential parents or at least limited numbers of spawners and known parentage one sex (e.g., - Moran et al. 1996). Here, the study is performed under near-natural conditions with a large number of potential parents (particularly for 2001). A s well, redds were not directly sampled as in similar studies (Taggart et al. 2001) so that female parentage was unknown. Likelihood methods are, however, available to resolve parentage in cases when progeny are collected outside of aggregated family groups, when neither parent is known, and in cases of incomplete parental sampling (e.g. - Neff et al. 2000a, 2000b). I used the program C E R V U S (Marshall et al. 1998) for the assignment of parentage. The program utilizes a categorical allocation algorithm to assign parentage to individuals at a predetermined level of statistical significance based on a large number of randomly simulated parents drawn from observed Chonat Lake allele frequencies. In each year, the assignment process began with the assignment of Y O Y fry to the sex with the fewest number of potential parents (i.e., 2001 males, 2002 females). Those parents assigned with at least 8 0 % confidence were assigned as the 'known parent' and a second round of assignment attempted for the other sex. Again, only those matches exceeding the 8 0 % confidence level were assigned parentage. Similar analyses were performed using the 'most-likely' criteria offered in C E R V U S (i.e. most-likely parent for individual Y O Y fry assigned parentage regardless of the statistical significance). Although a greater proportion of Y O Y fry were assigned under the 'most-likely' criteria, results were generally congruent between the two analyses and so the latter will not be discussed at length. Note that (because I expected them to have been present for spawning events in both years) all stream resident males were considered as potential parents in both years of the study, regardless of the  70  year they were sampled. A s a conservative measure, I did not, however, assume that all possible parents had been sampled; other spawners (including unsampled stream resident males) may have returned to the stream prior to or following fence enumeration. Instead, I assumed to have sampled 8 0 % of all possible parents and allow for a low degree of typing error-mismatch between parent-offspring dyads ( 1 % error rate). Reproductive success was expressed as the number of Y O Y fry assigned to each spawner. A related indicator of success (the number of inferred mates), may be expected to have an influence on reproductive success and is included as well. Statistical description and analyses of these variables were performed using either M S E X C E L or S P S S statistical packages: non-parametric Mann-Whitney U tests were used to assess differences between group means (by sex within years, by year across sexes, etc.); Levene's test for variance homogeneity was used to assess the significant differences between variance components. To allow for comparison across sexes and years (as well as between studies), variance components were standardized to the coefficient of variation (CV) which is the standard deviation in reproductive success divided by its mean value (Zar 1999). To determine if reproductive success for either sex was correlated with size, Spearman rank testing was used to examine the significance of correlations between reproductive success (or the number of mates) and fork length (mm). Mann-Whitney U tests were used to compare the mean reproductive success (or number of mates) of females that differed in their entry date into the stream (i.e. - E A R L Y or LATE). This analysis was performed only for females as the majority of males either arrived early or were present for the entire sampling period (e.g. - stream resident males). Finally, to test for the presence or absence of any trends towards size-assortative breeding in spawners, Spearman rank correlations were performed between male fork length and female fork length for all breeding pairs inferred from parentage analyses. For all analyses, outliers were visually inspected and removed prior to analysis.  3.2.4  Estimates of the effective number of breeders (N ) and the N / N ratio b  b  A s noted, salmonids tend to show marked variation in the timing of maturation  71  within individual populations. Consequently, the breeders in any one year may come from several brood years, each of which contributes a certain proportion to reproduction. Genetic estimates of effective population size in Chonat Lake samples were therefore made under the "salmon model" described by Waples (1990a, 2005). A s the name implies, the model was developed for salmon, which are semelparous, but differs from the majority of discrete generation models in that it allows for variation in the age at maturity. In practice, the model is expected to be applicable to iteroparous species (like coastal cutthroat trout) where the frequency of repeat spawning is low (< 1 0 % ; Ardren and Kapuscinski 2003). Under this model, the effective population size (N ) has been found to be approximately equal to: e  N *gxN e  (3.3)  b  where g is the generation length defined as the average age at maturity weighted by age specific fecundities, and N is the harmonic mean of N estimates across years b  b  making up the generation (Waples 2005). In the absence of detailed age data for Chonat Lake spawners, I limit my interpretation to the effective number of breeders in a given year (N ). To determine the effective number of breeders contributing to fry b  production at Chonat Lake, both demographic and genetic methods were employed. Demographic estimates My demographic methods utilized data collected during fence enumeration regarding the number and composition of spawners. Three methods were used: the first was the census estimate which is essentially the number of mature spawners observed at the fish fence in each year:  Census N (or N ) = N + N b  where N  m  c  m  (3.4)  f  is the number of males (fighter males + stream resident males) and Nf is  the number of females enumerated. The second was the sex ratio estimate (SR). Variability in the sex ratio can reduce the effective number of breeder relative to the census number (Lande and Barrowclough 1987). The reduction in N caused by an b  72  unequal sex ratio is given as:  S  R  > = w r k  <-  N  3 5)  m  f  The third demographic method was the variance in reproductive success estimate (VRS). This estimate combines the effects of sex ratio and the variance in the individual reproductive success (as inferred from parentage analyses) on the effective number of breeders. N m and N f are the effective numbers of breeding males and b  b  females, respectively (Kimura and Crow  1963).  The effective number of males  N  b  m  was calculated as follows:  VRS N  b m  =  (3.6)  ( >» »i-V N  k  ( X m + ^ - l ) m  K  where N  m  is the census number of breeding males, k is the mean number of m  offspring produced per male and a  2  k m  is the variance in male reproductive success  (inferred from the parentage analysis). The effective number of female breeders ( N f ) b  was calculated using the same formula but with female-specific data. Finally, the V R S N  b  estimate was calculated as:  VRS N =  4  b  A  /  ^  (3.7)  Genetic estimates In many natural populations, the relevant demographic parameters used to estimate effective population size are often not available and so a variety of genetic methods for inferring the number of breeders have been developed (reviewed by Wang 2005). These methods indirectly estimate effective population size by comparing associations between alleles or the variance in allele frequencies over time with the assumption that effective population size is the dominant factor influencing drift in the population being studied. Two basic types of estimates exist:  73  one-sample methods, primarily the linkage disequilibrium (Hill 1981; Bartley et al. 1992) and the heterozygote-excess methods (Pudovkin et al. 1996; Luikart and Cornuet 1999); and two sample methods, the most common of which is the temporal moments based method (Waples 1989; Anderson 2005). Temporal estimates are based on the premise that any fluctuation in allele frequencies over time is primarily a function of drift acting in accordance with the number of offspring sampled. Both approaches provide independent and complementary information about effective population size and depending on the sampling design, may apply to different time frames (Waples 2005). Here I used the linkage disequilibrium method (LD) which measures departures from linkage equilibria in pairwise combinations of loci against those based on random breeding and binomial sampling. The LD (or inbreeding) effective number of breeders is given by (Hill 1981) as:  LD A / = 1 / 3 ( / - 1 / S )  (3.8)  2  b  where r is the correlation of allele frequencies (calculated using the program 2  NeEstimator; Peel et al. 2004) and S is the sample size. For Y O Y fry samples (Plan II samples under the salmon model), the LD method estimates the inbreeding effective number of breeders in the spawning population which contributed to it (e.g. - N for b  2001 spawners from 2001 Y O Y estimates). Linkage disequilibrium estimates based on spawner samples (Plan I) are not included here because spawning populations are composed offish from several cohorts in salmonids and LD methods instead estimate a number somewhere between N and N unless age data is available to b  e  partition the sample into individual cohorts (Waples 2005). A second set of genetic estimates were made based on the temporal changes in allele frequencies (or the rate of coalescence of alleles) between sample periods. Here, N is estimated from temporal changes in standardized allele frequency (Nei b  and Tajima 1981; Pollack 1983; Waples 1989) given as:  .  t  1  f  (x,-y,f  ( 3 9 )  74  where F is the Pollack (1983) estimator of the standardized variance in allele k  frequency change calculated using the program NeEstimator (Peel et al. 2004); A is the number of alleles at a locus, x and y are the frequencies of the /th alleles in the population at the two sample points. The weighted mean value of F is related to the temporal effective number of breeders by the formula:  Temporal N =—, h b  , 2(F-VS)  (3.10)  Where S is the harmonic mean of sample sizes at the two time points and b describes the nature of the age structure in the breeding population and the number of years between samples. In the absence of appropriate age data (and hence, b values), I have inserted t (number of generations between sample points = 1) which gives the relationship under the discrete generations model. For samples taken at time t < the generation length (3-4 years), a single year of spawners represents only a portion of a generation and comparing two years of spawners is equivalent to taking two samples of the same overall population (Waples 2005). For comparison, two sample N estimates were also made using a Bayesian approach based on b  coalescence using the TM3 program (Berthier et al. 2002) as well as through a pseudo-maximum likelihood method incorporated in M L N E (Wang 2001). For all methods (both demographic and genetic) and in each year, I compared N estimates b  to the observed number of spawners (N ) and identify the factors most responsible for c  reductions in N / Nc ratios. I further compared harmonic mean values for each type of b  estimate across years to the harmonic mean total population size estimated through mark-recapture data to give estimated N / N ratios for the population. b  3.3 3.3.1  RESULTS  Number and composition of spawners  In 2001, the Chonat Lake fish fence was monitored daily from January 31 to April 1 (61 days). During that time, a total of 80 spawners were enumerated at the Chonat Lake fence: 56 females + 24 males (includes only sexually mature fish). In  75  2002, the fence was monitored from January 31 to April 5 (65 days), during which time less than half that number were counted: 21 females + 18 males for a total of 39 spawners. In both years, females tended to be larger and less variable than males in terms of forklength (female mean = 373 mm, coefficient of variation (CV) = 0.10; male mean = 339 mm, C V =0.17; Figure 3-2). Based on the length frequency distribution, it is likely that at least three age classes/ cohorts of each sex are represented in each year. Over both years of sampling, an additional seven mature stream resident males were sampled: two in 2001 and five in 2002 (the higher number in 2002 is due to increased sampling effort, not a change in abundance). These males were smaller than fighter males enumerated at the fence (mean forklength = 172 mm, C V = 0.24) and are assumed to have been present during both spawning periods. Other than the stream resident males, no repeat spawning males were observed. Of the 56 female spawners encountered in 2001, just 5% (3/56) returned to spawn in 2002. Based on the combined mark-recapture data (spawners and non-spawners), total adult population size was estimated to be 451 ± 116 ( 9 5 % confidence interval) for 2001 and 617 ± 137 for 2002 (harmonic mean over both years = 526). In each year, the movement offish into the spawning creek came in two main runs (associated with increased flow events) which I define here as E A R L Y vs. LATE. In both years, males showed up first and were more plentiful than females for the first month of sampling with a sex ratio (SR = males/ females) of ~ 1.5, while the second run offish was predominantly female (SR = 0.19). Final ratios approached 0.86 in 2002 but were strongly skewed towards females in 2001 (SR ~ 0.43) although the presence of undetected stream resident males may have increased the ratio slightly. Spawners were present in the creek prior to monitoring and several were collected on their outmigration (others may have not been collected). In 2001, both males and females stayed in the creek.an average of ~ 5 days but this varied considerably (from 1 to 22 days). Interestingly, 2001 spawners made several movements in and out of the spawning stream with 3 4 % of males and 9 % of females making two or three upstream movements past the fence. Other individuals left and did not return during fence monitoring (e.g., - females leaving with full complement of eggs). This behavior was not observed among 2002 spawners.  76  3.3.2  Microsatellite and summary genetic data  The seven microsatellite loci showed moderate levels of genetic variation with the number of alleles ranging from 2 to 14 (mean = 7.3 across loci); observed heterozygosities ranged between 0.34 and 0.92 (mean = 0.65) while expected heterozygosities ranged between 0.41 and 0.89 (mean = 0.69). This is similar to the levels Wenburg and Bentzen (2001) reported for ten anadromous coastal cutthroat trout populations in the Puget Sound area of Washington State. At the four loci shared between both studies (Omy77, O n e u l 1, Sfo8, Ssa85), the average for Hood Canal populations was marginally higher than the values obtained for the pooled Chonat Lake samples in terms of the mean number alleles (10.3 vs. 8.0), H (0.75 vs. 0  0.71), and H (0.77 vs. 0.75). Across all loci, however, expected heterozygosity e  values were equal in both studies (0.69).The mean polymorphic information content of our marker set (PIC; Marshall et al. 1998) was 0.643, allowing for a total exclusionary power of 0.957 in parentage analyses when neither parent was known, and 0.995 when one parent had already been assigned. Consideration of HardyWeinberg equilibria found evidence for significant heterozygote deficiencies at Omy77 and Ogo8 (p < 0.001; Table 3-1). When considering individual samples, heterozygosity values were not significantly different between spawners and Y O Y offspring although 2002 fry showed a 5 % reduction in observed heterozygosity relative to spawners (0.63 and 0.66, respectively; Table 3-2). In no cases were unique alleles present in Y O Y offspring (i.e., - alleles that were not present in that year's spawning population). The converse was not true; only 47 of the 51 alleles present in 2002 spawners were detected in 2002 offspring and the mean number of alleles between these groups differed significantly (7.3 vs. 6.7; one-sided p = 0.036). In 21 pairwise comparisons between loci, no cases of linkage disequilibrium was observed in either spawner sample; four cases were observed in the 2001 fry sample and 11/21 cases in 2002 fry. A s expected, the degree of inbreeding was greatest in 2002 with F estimates for 2002 is  spawners and fry at 0.06 and 0.08, although only the latter was significantly greater than zero (Table 3-2). No evidence for recent bottlenecks were found in samples under the S M M model; under the T P M model (with an 8 0 % proportion of SMM), tests  77  for heterozygote excess relative to mutation-drift equilibrium were found in 2001 and 2002 fry samples (Wilcoxon sign-ranked tests; p = 0.027 and 0.039, respectively). Finally, all pairwise sample comparisons were significantly differentiated in terms of allele frequencies (Fisher's method) except for the 2001 and 2002 spawner comparison (p = 0.52). The lack of genotypic differentiation between 2001 and 2002 spawners is expected under the salmon model because both samples are essentially sampled from the same generation (Waples 2005). 3.3.3  Parentage assignment  Given that neither parent was known a priori, parentage assignment was fairly successful in assigning fry back to potential parents. Across years, 6 3 % of fry were assigned at least one parent at the 8 0 % confidence level, while parental pairs were assigned for nearly half of all fry (49%). Insufficient power to differentiate between the two most-likely spawners with the desired level of probability accounted for the majority of unassigned fry as there were no cases in which fry were incompatible with the sampled spawner population. Female reproductive success Female reproductive success varied markedly between years (Table 3-3). In 2001, Y O Y fry were assigned to just 40 of the 56 potential females (71 %). Females that were successful had a mean reproductive success of 2.2 fry produced (ranging from 1 - 8) and a relatively high coefficient of variation = 0.78. In 2002, 20 of the 21 sampled females (95%) were assigned fry (ranging from 1 - 1 3 ) with a significantly greater average reproductive success of 5.1 fry (Mann-Whitney U, p < 0.001). The coefficient of variation for reproductive success for 2002 females was 0.66 and suggests a significant difference in the distribution of reproduction success between years (Levene's test for variance homogeneity, p = 0.001). While females showed a similar range of inferred number of mates between years (1 - 8; Figure 3-3), the average for 2001 females was also less than in 2002 (2.0 vs. 3.8; Mann-Whitney U, p < 0.001) while C V values were greater (0.72 vs. 0.50 in 2002). The majority of 2001 females were inferred to have had one mate while the rest had 2-8 mates. In contrast, most 2002 females appear to have had 4 different mates. In each year, there were  :  7 8  considerably fewer early-arriving females present and so most fry production was assigned to late-arriving females (averaging ~ 7 5 % ; see also "Correlates..." below). Male reproductive success Male reproductive success, on the other hand, did not vary substantially between years (Table 3-4). In 2001, fry were assigned to 29 of the 31 potential males (94%).  Mean reproductive success averaged 3.8 per male, but ranged from 1 - 11. In  2002, all 25 males were assigned Y O Y fry with a similar mean reproductive success of 3.8 (range 1-13). A slight increase in C V values was observed for 2002 males but not significantly so. In terms of the number of mates, the mean averaged ~ 3 in each year with nearly equal variance (Figure 3-3). Both fighter and sneaker strategies appear to have been successfully employed by males at Chonat Lake. While reproductive success for fighters and sneakers did not vary significantly between years, fighters in any one year were consistently more successful in terms of the number of fry assigned (averaging -4.1 vs. 3.1 across years) and in the number of mates (3.3 vs. 2.5). When considering total individual success over the both years combined, however, stream resident sneaker males were at an advantage in that all were inferred to have been successful in both years (averaging - 6 fry and 6 different mates; Table 3-4). Post spawning mortality is expected to be high for fighter males (Fleming and Gross 1994; Fleming and Reynolds 2004) and no repeat spawning fighter males were observed at Chonat Lake; the lone fighter male captured in both 2001 and 2002 was immature and did not spawn in 2001.  3.3.4  Correlates of reproductive s u c c e s s and assortative mating  In 2001, male reproductive success was significantly correlated with body size for both fighter and stream resident males (combined Spearman r = 0.487, p = 0.011; Figure 3-4). In 2002, the reproductive success of stream resident males again showed a positive correlation with size, while fighter males showed a slight negative trend that was not significant (combined Spearman r = -0.01, p = 0.48). For females, a slight negative correlation was detected in 2001 and a stronger one in 2002 between reproductive success and size (r = -0.41, p = 0.04). The trend was particularly significant when both years were combined (r = -0.35, p = 0.004). In  79  terms of the number of mates, again there was a significant positive correlation with size for males in 2001 (r = 0.35, p = 0.04), but not in 2002. A negative correlation between number of mates and size existed for females over both years combined (r = -0.36, p = 0.003). There were no statistically significant differences between early and late-arriving females in; either year with respect to the mean number of fry (or mates) assigned, although early spawners appear to have had slightly greater success in 2001 (# Y O Y = 2.6 vs. 2.1; # mates = 2.5 vs. 1.8; Table 3-3). In 2002, the trends were reversed with late arrivals being assigned 5.8 vs. 3.3 fry and 4.1 vs. 2.8 mates. Interestingly, late-arriving females were larger in 2001 (averaging 382 vs. 341 mm fork length; p < 0.001), but had lower reproductive success than late females in 2002 (#YOY fry assigned = 2.2 vs. 5.8, p < 0.001; number of mates = 1.8 vs. 4.2, p = 0.001). Earlyarriving females did not differ significantly between years in terms of either size, reproductive success or the number of mates. For both sexes, reproductive success and number of mates were highly correlated (Spearman r = 0.964, p < 0.001). Finally, there was no evidence of size-assortative mating in inferred pairs from either 2001 or 2002 (Spearman r = 0.002 and -0.077, respectively; p > 0.40; Figure 3-5).  3.3.5  Effective number of breeders (N ) b  Demographic estimates The simplest estimate of the effective number of breeders (census number or Nc) is essentially the number of spawners enumerated each year = 87 in 2001 and 46 in 2002 (Table 3-5). Consideration of the final sex ratio in each year led to a reduction in sex ratio N estimates by 8 % for 2001 (to 79.8), but had little effect in 2002 (< 1% b  reduction to 45.7). When both sex ratio and the variance in individual reproductive success were accounted for (VRS N estimates), the effective number of breeders b  dropped by a similar amount in both 2001 and 2002 (a further reduction of 25 - 2 2 % to 58.1 and 35.6, respectively). Variation in reproductive success at Chonat Lake, therefore, appears to have had a much greater impact on N than did sex ratio. It is, b  however, interesting to look further at how variance in individual reproductive success affected the N / N ratios for each sex. For males, reproductive success did not b  significantly differ between years and so the effective reduction of N also similar between years; N m / N b  m  m  (to N m ) was b  = 0.81 in 2001 and 0.82 in 2002. Significant  differences were noted in female reproductive success between years (higher mean success and lower variance in 2002). The effect of this was to lessen the effective reduction of N (to N ) in 2002 by - 1 4 % relative to 2001 (the N / N ratio in 2001 = f  b f  b f  f  0.62 while the ratio in 2002 = 0.76). Harmonic mean demographic estimates of N  (N ) b  b  ranged from 44.15 (VRS) to 60.18 (census; Table 3-7).  Genetic estimates Genetic estimates of the effective number of breeders were generally greater than demographic estimates. Depending on the method used, genetic estimates ranged from 89.0 - 147.4 for 2001 spawners and from 24.9 - 49.3 for 2002 spawners (Table 3-6). Consequently, in some cases (particularly for 2001), estimated N / N b  c  ratios exceeded 1. The linkage disequilibrium estimate (based on Y O Y fry genotypes) generally provided the most precise estimates (i.e. - smallest confidence intervals). Harmonic mean genetic estimates of N were comparable with demographic b  estimates, ranging from 40.41 to 70.00 (Table 3-7). Comparison of the estimated number of breeders to my estimate of total population size (A/) suggested that N l N b  ratios ranged from 8 - 1 3 % at Chonat Lake, with a grand mean average of just 1 1 % . Finally, in the absence of specific age data for spawner samples, I assumed that my grand mean N = 56 applies equally across the 3-4 cohorts expected to comprise the b  generation. This gives a rough estimate of N (A/ ) from 169 - 226 over the e  e  generation and an N / N ratio of 32 - 4 3 % , which is lower than the value recently e  estimated for steelhead trout in Washington state (53%; Ardren and Kapuscinski 2003). 3 . 4 DISCUSSION Fish show a remarkable diversity of breeding behaviors (from monogamy and male pregnancy to cuckoldry and group spawning); diversity which has been characterized by a large body of natural history literature (e.g., - Breder and Rosen  81  1966) and more recently, by genetic investigation (Avise et al. 2002). The study of salmonid breeding systems in particular, has provided much insight into the ecology and evolution of sexual dimorphism, intrasexual competition and selection, and the development of alternative reproductive strategies (e.g., - Gross 1996b; Hendry and Stearns 2004; Stearns and Hendry 2004). Salmonid research has also made substantial contributions to the field of conservation genetics, including application of effective population size (Waples 1990b; Ardren and Kapuscinski 2003) to the genetic management of captively bred populations (Ryman and Laikre 1991; Duschene and Bernatchez 2002). My findings build on this work and illustrate how a salmonid breeding system may directly influence the maintenance of genetic diversity in small populations of coastal cutthroat trout. To my knowledge, this study represents the first attempt to describe the breeding system of this subspecies in the wild, but these findings should be applicable to many species characterized by small and fluctuating population sizes.  3.4.1  Confidence in parentage analyses Many of the findings made here, particularly those relating to breeding patterns  and reproductive success, were inferred from my interpretation of parentage analyses. Clearly, a perfect exclusion approach to assignment (where all but the one true parent can be excluded on the basis of genotype) is preferred. When one parent is known, exclusion of possible mates can often be accomplished with only one or two loci (DeWoody et al. 2000; Mehranvar et al. 2004). Unfortunately, as the number of potential parents increases (or when neither sex is known), the number of loci required for complete exclusion is often prohibitive (Marshall et al. 1998). Since I chose not to collect fry samples in aggregated family groups (i.e. - from redds), maternal parents were unknown and certain analytical approaches were, therefore, not feasible. Nonetheless, the likelihood methods I employed proved relatively successful given the near natural conditions of the study and the large number of potential parents. In a similar study performed on a natural steelhead trout population, Seamons et  82  al. (2004) used likelihood techniques to assign parentage for a total of 275 potential parents sampled over 4 years. Using 12 loci with an average heterozygosity of 0.90, the authors were able to assign 7 3 % of all offspring to at least one parent, while parental pairs were assigned to 4 0 % of offspring. Using fewer loci, I was able to assign at least one parent to 6 3 % of the fry, and parental pairs to nearly half of all fry with a high degree of confidence using the categorical allocation approach. Unassigned fry resulted primarily from a lack of power to distinguish between the two most likely candidates at the given confidence level. While a larger (or more polymorphic) marker set would likely have improved overall assignment success (Bernatchez and Duchesne 2000; Neff et al. 2000b), I found little appreciable difference between analyses at the 8 0 % confidence level and "most-likely" criterion. All fry were assigned under the most-likely scenario so that mean and variance components of reproductive success did increase for spawners. There was, however, little difference between C V values for reproductive variance and similar trends were apparent between years. Categorical allocation methods can, however, lead to an upward bias in estimates of variance for individual reproductive success (Jones and Ardren 2003). This occurs because the success of parents with many homozygous loci is overestimated relative to heterozygous individuals. This is not necessarily a flaw in the likelihood method as, biologically, a homozygous individual would be twice as likely to have contributed a particular allele than would a heterozygous individual. The presence of null alleles at some loci, however, may exacerbate the problem. Null alleles typically result from point mutations in the flanking regions of a genetic marker and are typically inferred to be present in loci that exhibit significant heterozygote deficiencies (e.g., - Callen et al. 1993). My approach (and the approach of the C E R V U S parentage program) was to treat possible null alleles as any other dyad mismatch or genotyping error, allowing for a low rate of mismatch between parent and offspring (1%). The success of likelihood methods appears to be fairly robust when such non-zero errors rates are allowed (Marshall et al. 1998; Wilson and Ferguson 2002). Finally, breeding assignments were generally consistent with demographic observations; spawners that left the spawning creek (or were found dead) soon after arrival were assigned few fry and inferred breeding pairs were  83  generally known to be sympatric at some point in the stream.  3.4.2  Fluctuations in the number and composition of spawners  Considering that the Chonat Lake spawning stream is only ~ 700 m in length, the relatively large number of spawners encountered (87 in 2001, 46 in 2002; harmonic mean ~ 60) was surprising. These numbers would be considered a fairly large spawning population for this subspecies in this area. While few trend data exist for individual populations, point counts and "best guess" estimates for several larger and more complex streams in the general area are often less than 20 breeding adults (e.g. - Scholten 1997; Slaney 2005). The reliability of these point counts, however, may be questionable as they are typically snorkel counts performed over a short time span and are uncorrected for viewer efficiency. Alternatively, it may be that the Chonat Lake system is more productive (or less impacted) than others in the area. The inlet stream is short but has relatively good amounts of gravels and instream structure that provide excellent pool and riffle habitats. The creek is further headed by a marshy area which maintains stable year-round flow and possibly nutrient input. Chonat Lake itself appears fairly productive in terms of invertebrate and planktonic prey as well as forage fish (sticklebacks, cottids) and may support a larger population than is typical of most streams in the area. Data presented in Chapter 4, for example, suggests that the expected effective population size at Chonat Lake may be at least an order of magnitude greater than the majority of populations sampled in that chapter. Equally notable is the marked variation in spawner numbers between years. Fluctuations have been observed in other areas (Sumner 1952, 1962; Wenburg 1998) and are not wholly unexpected. Heterogeneity between years in terms of environmental (temperature, precipitation) and anthropogenic factors (harvest, habitat disturbance, etc.) would certainly impact the number of returning spawners in any given year. I calculated the coefficient of variation for spawner number (i.e., standard deviation/ mean, CV) to be ~ 0.44. This falls within the range of estimates published by Fleming (1998) for 19 Atlantic salmon populations (range = 0 . 1 8 - 1 . 1 3 , median C V value = 0.51) but is less than values obtained for a natural steelhead trout  84  population in nearby Washington State. (0.70; Seamons et al. 2004). The value calculated at Chonat Lake, however, is based on only two years of data and may considerably underestimate the true degree of variation between years. Interestingly, it was primarily the number of females that varied between years (female C V = 0.64, male C V = 0.15) and in particular, larger females which were noticeably absent in 2002 (Figure 3-2). The different life history pathways employed by the sexes at Chonat Lake may partly account for the fluctuations in spawner number. There is considerable anecdotal evidence to suggest that female coastal cutthroat trout may make more extensive saltwater feeding forays than males (e.g. - Giger 1972; Massey 1984; Wenburg and Bentzen 2001). Anglers of coastal cutthroat trout, particularly those who frequent saltwater beaches and estuaries, have long suggested that there might well be two types of anadromous cutthroat, referring to them as 'beach trout' (smaller fish with yellow bellies; mostly males which stay near local bay or estuary); and true 'searun' trout (larger silvery fish which travel greater distances; mostly females). The majority of females in the system (> 95%) did appear to have been anadromous (i.e. - large body size, silver in appearance, often with attached marine ectoparasites). Fighter males at Chonat Lake tended to be smaller and more variable in size than females, rarely had a typically anadromous coloration, and may instead follow a primarily lacustrine life history type, feeding on sticklebacks and cottids in the lake. This contention is further supported by genetic evidence at Chonat Lake; although statistically nonsignificant, F values among Chonat Lake females are an order less st  than that among males as would be expected if dispersal was sex-biased (e.g. Blundell et al. 2002; Goudet et al. 2002). Interestingly, females also appear more inbred and may contain marginally lower levels of genetic diversity, suggesting reduced effective population size relative to males (F t: 0.001 vs. 0.010; F : 0.040 vs. s  is  -0.007; H : 0.65 vs. 0.69; A : 6.0 vs. 6.3, respectively). 0  R  The relative benefits and risks associated with anadromous movement have been well documented for salmonids (Northcote 1997; McDowall 2001; Hendry et al. 2004); females, which optimize their reproductive fitness by attaining large size and greater fecundities, would likely benefit to a greater extent from the benefits of rich  85  saltwater feeding areas than would males, even if such movements made them susceptible to increased mortality. Because anadromy is generally associated with a reduced survival or probability of repeat spawning in salmonids (Fleming 1998), it might be expected that the anadromous component of populations may fluctuate widely from year to year. I did, in fact, identify a high concentration of harbour seals (Phbca vitulina) in Chonat Bay in 2002 (5-7 visible at any one time). Seals are known to prey heavily upon anadromous salmonids and indeed (both this year and last), many fish were observed to have bite marks and scarring consistent with seal attack. It is possible that reduced ocean survival, perhaps due to high seal predation, may have contributed to the decline in large female spawners in 2002. Repeat spawning was generally low at Chonat Lake, with just 5% of 2001 spawning females returning to spawn in 2002. It may be that male coastal cutthroat trout numbers may vary more markedly in systems where freshwater adult feeding habitat (e.g., - lake habitat) is limited and more males pursue anadromous life histories. Seamons et al. (2004), for example, found that sex ratios in steelhead trout populations could fluctuate widely (from 2:1 male-biased to 2:1 female-biased), but tended to average ~1:1 over the four year study. Sex ratios in the Chonat Lake spawning stream varied markedly during each season and between years. A s with other salmonid species, males tended to arrive earlier than females; several were present in the spawning stream prior to fence monitoring and up to a month before late-arriving females. The bias towards females we observed in 2001 is consistent with previous studies that have suggested that sex ratios in coastal cutthroat trout populations can often favor females (Wenburg 1998 and references therein). In salmonids, however, operational sex ratios tend to favor males even when observed sex ratios do not. Salmonid males not only arrive earlier on the spawning grounds, but they may be sexually receptive for a longer period (up to a month compared to a few days for females; Fleming and Reynolds 2004). The smaller time frame for female receptivity reflects, in part, the need to deposit eggs soon after ovulation; delays can lead to over-ripening of eggs and reduced viability (de Gaudemar and Beall 1998). A s well, while males can immediately mate again with other females, female spawners generally require a 'timeout' period for recovery between successive matings. For Atlantic salmon, the time between successive  86  mating events can extend from 4 hours to 9 days but most females tend to complete their egg deposition in 5-6 days (Fleming and Reynolds 2004), similar to the estimated period of stream residency for females at Chonat Lake in 2001. 3 . 4 . 3 Inferred patterns of mating in coastal cutthroat trout  Parentage analyses suggest that Chonat Lake cutthroat trout exhibit a complex array of mating patterns, including monogamy, polygyny, polyandry, and polygynandry. Both males and females were inferred to have had up to 8 partners; which likely results primarily from opportunistic mating attempts by multiple males at a redd. This was corroborated with the lone spawning event observed in the system (spawning occurs primarily at night). A female was observed to spawn immediately downstream of a small pool formed by large woody debris (tree which had fallen across the stream) in ~10 cm of water. The female was being attended by a large fighter male (close and slightly behind) and a smaller subdominant fighter male (slightly further behind). Two or three stream resident males were also present, darting in and out from nearby cover. All were observed (at various times) to attempt to mate the female as she continued to deposit her eggs. The benefits of polygyny are clear for males which maximize their reproductive success by fertilizing the eggs of as many females as possible. For females, the benefits of multiple paternity are less clear but it may serve to ensure fertilization of eggs or provide genetic benefits for the resulting fry (Reynolds 1996; Fox and Rauter 2003). Unlike the Pacific salmon, coastal cutthroat trout females do not appear to defend nest sites once eggs have been deposited; the tactic may be overly risky in small streams and energetically expensive for an iteroparous species which may live to spawn in subsequent years. The lack of nest defense should, however, allow for an increased number of breeding events for females. Interestingly, while male patterns did not vary, female breeding patterns were clearly different in high and low density years. In the high density year (2001), nearly half the 40 successful females were inferred to have had only one mate, compared with only 1 0 % monogamy among males. High levels of female-female competition for spawning habitat at the higher density may have prevented these females from  87  attaining more mates and greater reproductive success (see also next section). In 2002, females showed considerably higher levels of polyandry with a mean of ~ 4 mates and only 1 0 % with a single mate. The higher level of multiple mating for females in 2002 may, therefore, represent an important form of compensation during periods of low number of breeders. Multiple paternity of this type has been implicated as an important factor in the maintenance of high levels of genetic variation in other salmonids, including small, isolated populations of Atlantic salmon in southern Europe (Moran and Garcia- Vazquez 1998; Martinez et al. 2000). I found little evidence for size-assortative breeding at Chonat Lake in either year, suggesting that, for the most part, spawners are not mating exclusively with similarly sized individuals. Size assortative mating may have been expected given the apparent benefits each sex gains from mating with larger individuals (reviewed by Fleming and Reynolds 2004). Considering the variety of mating behaviors observed, size assortative mating may well be present but overshadowed by opportunistic mating and sneaking strategies. While I do not have specific age data for the spawner samples, based on their length-frequency distribution at least 3 - 4 age classes are present for each sex and breeding between cohorts and different life history components is, therefore, very likely. Since different alleles (or different frequencies of the same alleles) are often present in cohorts and life history types (Campton and Utter 1987; Foote et al. 1989; Currens et al. 1990; Hindar et al. 1991; Wenburg et al. 1998), it is likely that interbreeding between cohorts further helps maintain allelic diversity. My data support this; while heterozygosity values were similar between years, the total number of alleles present in the 2001 and 2002 spawner samples differed slightly (49 vs. 51, respectively). 3.4.4 Variance in individual reproductive s u c c e s s  As noted, the observed correlation between reproductive success and number of mates does not necessarily suggest that the number of incurred mates is a significant determinant of reproductive success (although it may well be). Rather, the relationship may be an artifact of the low number of fry sampled per mating event (in many cases just one or two fry). A s such, my data should be viewed as a  88  conservative estimate of breeding statistics. Only a small percentage of eggs survive incubation to reach the fry stage (estimated at - 1 0 % , Slaney 2005) and so while the entire creek was sampled, some mating events may not have been detected. Furthermore, my sampling strategy only accounts for variance in the reproductive success among individuals through to the Y O Y fry stage. Ideally, sampling should be performed at the same life stage as post-enumeration mortality will effect mean reproductive success and variance values as the cohort ages (Crow and Morton 1955; Waples 2002). Female reproductive success Two factors appear to account for the majority of variance in reproductive success for female salmonids: fecundity and the survivorship of offspring. Egg production in coastal cutthroat trout increases linearly with body size so that larger females have the potential for greater reproductive success than smaller ones (Giger 1972; Mercer 1982). While female fecundity sets the upper limit on reproductive success, the ability of females to secure quality spawning habitat can lead to significant variation in individual success. The quality of spawning habitats is known to have a dramatic effect on egg survival and the future development of Y O Y fry. For instance, eggs deposited in well-oxygenated, high quality gravels suffer significantly lower levels of mortality during incubation (Chapman 1988; Reeves et al. 1997) and fry often emerge earlier and are able to access better rearing habitats (Gibson 1993). Larger body size is expected to allow larger females to displace smaller ones from preferred habitats. The benefits of spawning early, however, are countered by the risk of redd superimposition (or the overlap of nests from different females) by later spawning females, Again, larger females may have a competitive advantage in that they can dig deeper nests which may be less subject to nest destruction by subsequent spawning events (Fleming and Reynolds 2004). The wide fluctuation in female numbers between years led to markedly different breeding densities for females in the spawning stream, a factor which I believe to have had a significant impact on the distribution of individual reproductive success in the population. In 2001 (high density year), a smaller proportion of females were  89  found to have been successful ( 7 1 % vs. 9 5 % in 2002) and those that were successful had significantly lower success than 2002 females in terms of both the number of Y O Y fry and mates assigned. Two factors may account for the lower success in 2001. First, elevated densities in the creek may increase the number of unspawned eggs as only a certain proportion of females at any one time are able to secure the best available spawning habitats (Fleming and Gross 1994). Displaced females may be forced to use suboptimal habitats or to delay deposition of their eggs (Chebanov 1991). Importantly, once a female has ovulated, delayed deposition can quickly lead to over-ripening and reduced viability of eggs, delayed embryo development, and decreased competitive ability in the resulting fry (e.g., - Springate et al. 1984; de Gaudemar and Beall 1998). Instantaneous density effects appear to have most impacted late arriving females in 2001; these fish had the lowest mean reproductive success and fewest mates of any group (male or female). High levels of synchrony were observed in their arrival time with as many as 35 females arriving over the course of a two day flow event. Elevated levels of female-female competition, and perhaps a lack of suitable cover, appear to have forced many females to delay spawning or leave the stream entirely. Many late arriving females in 2001 were observed to make several movements up and down past the fish fence. Several, in fact, were observed to have left the system without having spawned and with a full complement of eggs. This behavior was not observed in the low density year (2002). Second, high spawner density (over the entire spawning period) appears to increase the likelihood of redd superimposition (e.g. - Curry and Noakes 1995). When spawning habitats are limited or there is a strong preference for certain habitats, repeat spawning at the same location by successive females can lead to the partial excavation of earlier spawning efforts even at relatively low spawner densities (Blanchfield and Ridgway 1997; Taggart et al. 2001). While I was not able to quantify the amount of redd superimposition in the Chonat Lake inlet stream, Essington et al. (1998) found high rates of redd superimposition by both brown trout, Salmo trutta (34%) and brook trout, Salvelinus fontinalis (53%) in a system similar in size to the Chonat Lake inlet stream (2  nd  order, mean width ~4m) with similar spawner numbers  (between 17 and 31 females). Redd superimposition has been found to have a significant impact on the distribution of reproductive success among females in other  90  salmonid species (e.g., - van der Berghe and Gross 1989). Hayes (1987), for example, found that the reproductive success of brown trout in a small New Zealand inlet stream was reduced by over 9 4 % due to redd superimposition by later-spawning rainbow trout (Oncorhynchus mykiss). While I found no statistically significant trends with respect to arrival time, late spawning females had nearly twice the mean reproductive success as early spawning females in the low density year (2002). The coefficient of variation for reproductive success among these late spawners, however, was the largest for any group at Chonat Lake (male or female, C V = 1.07), suggesting that other factors were contributing to high levels of variation between individuals. Surprisingly, I found a significant negative correlation between reproductive success (and the number of mates) and body size in females. It seems unlikely that males are avoiding larger females. Instead, I believe the negative relationship is rooted in the unique spawning habitats utilized by coastal cutthroat trout. Coastal cutthroat tend to spawn in some of the smallest streams inhabited by salmonids (generally 1 to 3 order and ~1 - 3 m in width; (Trotter 1989; Rosenfeld 2001). While st  rd  this is likely an adaptation which helps to reduce their interactions with larger salmonids such as steelhead trout (Behnke 1992; Johnson et al. 1999), their occupancy of smaller streams may make coastal cutthroat trout spawners especially vulnerable to predation by a number of riparian and aquatic predators (Heggenes and Borgstrom 1988; Lonzarich and Quinn 1995; Ruggerone et al. 2000). Quinn et al. (2001b), for example, found that when sockeye salmon (Oncorhynchus nerka) axe easily caught (as they are in small, shallow streams), predation is often size selective as larger fish have a higher energetic payoff. The authors found predation levels declined with increasing density but tended to be greater for early-arriving fish, especially if predators were satiated by or could utilize only a fixed number of fish. While I did not quantify predation rates with respect to size, I did observe multiple cases of large spawners that had been captured and left partially eaten beside the creek. Male reproductive success Male reproductive success in salmonids is expected to be primarily a function of  91  access to females, and more specifically, by their proficiency at fertilizing females. Relative body size appears to be the primary determinant of reproductive success for fighter males; large males aggressively compete and establish dominance hierarchies to gain access to females (Fleming and Reynolds 2004). Alternate breeding strategies, however, are known to exist which help circumvent direct male-male competition. In many cases, subordinate or "satellite" males are often fairly successful in gaining access to females while the dominant male is distracted. As well, there are the small, precocial males know as "jacks" or "sneaker" males which use their small size and maneuverability to sneak fertilization attempts with spawning females (Gross 1996; Fleming 1998). Precocial jack coho and sockeye salmon, as a group, are often as successful, if not more successful than large fighter males in fertilizing females while precocial Atlantic salmon parr may account for up to 6 5 % of all paternity in some populations (Fleming 1998 and references therein). I found that nearly all sampled males at Chonat Lake were assigned some level of paternity; over both years, 54/56 (96%) males were assigned Y O Y fry (the two fish which weren't successful either left the creek shortly after arriving or were found dead besides the creek). Both fighter and sneaker strategies were evident at Chonat Lake with stream resident males being assigned a significant portion of all fry (up to 2 3 % ) . In any one year, fighters were more successful than sneakers in terms of the number of fry and mates assigned. All stream resident males were, however, assigned fry in both years of the study while no repeat-spawning fighter males were observed. High levels of post-spawning mortality are expected for fighter males as intense male-male competition and higher levels of wounding or predation often prevent repeat spawning (Fleming and Reynolds 2004). Data on coastal cutthroat trout stream resident male survivorship and repeat spawning are limited but individuals are expected to be capable of breeding over several years and may, therefore, have considerably higher lifetime reproductive success than fighter males. While there may be a genetic component influencing the strategy pursued, the relative success of alternate strategies is expected to show negative density-dependence and show conditional requirements (i.e. - high early growth rates may allow for precocial maturation; reviewed by Fleming and Reynolds 2004).  92  Unlike females, I did not find male reproductive success to vary substantially between years at Chonat Lake. In both 2001 and 2002, male spawners were assigned an average of 3.8 fry and ~3.0 mates. As well, the variance in reproductive success among individual males (as a group, C V ~ 0.66) did not vary substantially between years and was similar to values reported for Atlantic salmon (CV = 0.77; Garant et al. 2001), brown trout (0.87; Dannewitz et al. 2004), and steelhead trout males (0.63 - 1.62; Seamons et al. 2004). While male reproductive success may be ultimately determined by their ability to fertilize females, it does not appear to be tied to number of females, or at least, the number of females was not limiting in either year at the spawning creek. Levels of male-male competition (and the resulting variance in reproductive success) should instead be tied to the number of competing males or sex ratios (Fleming and Reynolds 2004). This may be seen in the correlation analyses for male reproductive success. As expected, male reproductive success was positively correlated with body size for stream resident and fighter males in 2001, but only for stream resident males in 2002. The lack of a correlation between size and reproductive success for 2002 fighter males suggests that the benefits of large size in males may decrease as male sex ratio increases and more males are simultaneously competing for access to available females. Consequently, while mean success did not vary between years, its distribution appears to have varied slightly.  3.4.5 Effective number of breeders (N ) and levels of genetic diversity b  There was considerable variation between the various estimates of N , with b  genetic estimates generally exceeding ones based on demographic information. Several factors may account for this. First, each of the various methods I employed have their own strengths and potential for bias (Basset et al. 2001; Tallmon et al. 2004; Wang 2005; Waples 2006). For example, the algorithms used by T M 3 and M L N E appear to underestimate Ne when only one generation passes (Tallmon et al. 2004; although I did not find this). For this reason, I included many methods to get a sense of how robust the estimates may be. Second, demographic and linkage disequilibrium methods estimate the inbreeding  effective  size which relates to number  of parents and predicts the expected reduction in heterozygosity caused by inbreeding. In contrast, the temporal genetic methods estimate the variance  effective  93  size which relates to number of offspring sampled and predicts the variance in allele frequencies between generations due to finite population size. The two are only expected to be equal in a population of constant size (Crow and Denniston 1988) which would not seem to apply here. Third, my methods utilize information taken at the Y O Y fry stage and do not, therefore, account for the survival of offspring through to adulthood (estimated to be on the order of 1-2 % ; Slaney 2005). A s such, I would expect that genetic diversity and inferred effective population sizes may also decline with increased mortality as the cohort ages (Crow and Morton 1955; Waples 2002). Similarly, temporal estimates utilized spawner and offspring samples as time points for independent estimates in each year. While this is appropriate for our purpose of comparing genetic estimates of N with my demographic information, population b  samples should ideally be taken at the same life history stage (Waples 2002). Finally, it is entirely possible that I did not collect all possible spawners (especially in 2001); spawners were present in the creek prior to (and likely following) fence enumeration. I did not, however, find any fry which were incompatible with my sampled spawners (i.e. - there were no unique alleles in fry samples). Both demographic and genetic estimates suggest that a relatively large proportion of spawners contributed to fry production in any one year; i.e., - V R S estimates of N / N = 0.67 in 2001 and 0.77 in 2002 (mean = 0.73). This is somewhat b  c  higher than values reported for steelhead trout by Ardren and Kapuscinski (2003); mean = 0.61, range 0.41 - 0.67, and for several studies reviewed by McElhany et al. (2000). Variance in individual reproductive success had the largest impact on N , b  reducing the effective number of breeders by 3 3 % in 2001 and 2 3 % in 2002 relative to the census number. Sex ratio had a minor effect, even in 2001 when the ratio was highly female biased (reduction of ~ 8%). Interestingly, the increased N / N ratio in b  c  2002 (a year of low spawner density at Chonat Lake) suggests that some form of compensation may be acting to sustain the population at lower spawner abundance (i.e. - increased productivity at lower density). Compensatory mechanisms of this type have been described previously in sockeye salmon (Chebanov 1991), coho salmon (Fleming and Gross 1994), and more recently, steelhead trout (Ardren and Kapuscinski 2003). The study of Ardren and Kapuscinski (2003) appears to have been the first to document how fluctuation in breeding densities in a wild salmonid  population can impact N and N / N ratios (ratios increased during years with few b  b  c  spawners). In the absence of specific data on the breeding system of that population, however, the authors could only postulate that a reduction in V R S was the primary means of genetic compensation. I extend their finding by confirming with assigned parentage for individuals that, indeed, reduction of V R S at lower densities is (at least partly) responsible for the increased ratios. My data have shown that increased reproductive success and a reduction in the variance associated with that success among females at lower spawner densities helped offset reductions in the effective number of breeders. Just as the distribution of female reproductive success varied significantly between years, so too did the effective reduction of N to N f. In 2001, N / N was equal to 0.62 while in 2002, the f  b  bf  f  ratio equaled 0.76 (a difference of 1 4 % between years). The mechanisms most likely accounting for this are density-dependent and were discussed in the previous section, namely a reduction in female-female competition, less delay in the deposition of eggs, and a lower likelihood of redd superimposition. I also found that females showed considerably higher levels of polyandry in the low density year, a factor implicated in the maintenance of high levels of genetic variation in small populations of Atlantic salmon in southern Europe (Moran and Garcia- Vazquez 1998; Martinez et al. 2000). Exclusion of the multiple paternity contributed by stream resident males at Chonat Lake, for example, would reduce V R S estimates of N by 1 4 % in 2001 and by 1 7 % in b  2002. High N /Nc ratios and compensatory factors, however, were not able to buffer b  the effects of low spawner number in 2002 and prevent the loss of genetic variation in that year's fry population. As might be expected (Cornuet and Luikart 1997; Widmer and Lexer 2001), allelic diversity was more affected by low number of spawners than was heterozygosity levels; of the 51 alleles present in the 2002 spawner sample, only 47 were found in their offspring, representing an 8 % reduction in allelic diversity over one spawning season. Such reductions in allelic diversity may be only transient, however, if the other brood years making up the generation have greater allelic diversity. I did, after all, observe small differences in the number of alleles present in the spawner samples (ironically, more alleles were found in 2002 despite the smaller  95  sample size). Observed heterozygosity dipped marginally as well from 0.66 to 0.63 (~5%) with most loci showing significant departures from both Hardy-Weinberg and linkage equilibria. The significant linkage disequilibrium in fry (particularly the 2002 fry sample) is consistent with the sampling of full and half-sib fry families (Allendorf and Phelps 1981; Hansen et al. 1997) and the low number of spawners observed in 2002. It may further help explain the heterozygote deficiencies I observed at two of microsatellite loci. Importantly, the observed diversity levels in my Y O Y fry samples represent maximum values; the survivorship of fry to adulthood is expected to be low enough that further increases in V R S and losses of genetic diversity (particularly uncommon or rare alleles) may be expected as the cohort ages (Crow and Morton 1955; Waples 2002). 3.4.6  Ecological/ evolutionary implications  It is clear that the evolution of breeding systems represent a tradeoff between innumerable evolutionary forces. Fortunately, many of the relevant mechanisms and their theoretical underpinnings have been well characterized for salmonids. In the absence of other influences, sexual selection tends to favour the evolution of large size; larger males are able to obtain a greater number of mates and larger females have higher fecundity and obtain better nest sites. The evolution of large body size, however, is often constrained by increased energetic or physiological costs as well as increased predation risk (reviewed by Quinn et al. 2001a; Fleming and Reynolds 2004). It is the balance between these two forces (sexual vs. natural selection) that shapes, to a large extent,, the characteristics of salmonid breeding systems; namely the development of secondary sexual characteristics and degree of sexual dimorphism, the age and size at maturity, as well as the development of alternate reproductive tactics. While I did not explicitly set out to test the relative intensity of these selective forces, one can get a sense of their magnitude by examining the total variance in reproductive success among individuals. The index of total selection, (/; Crow 1958) provides a measure of the "opportunity" or "potential" for selection (natural + sexual) and the upper limit of for selection intensities (Arnold and Wade 1984; Fleming and Gross 1994).  96  Mathematically, / is related to CV, the coefficient of variation (/ = variance in reproductive success / mean or C V ) although the two measure different concepts. In 2  2  the majority of cases, the highest / values (and thus selection intensities) are found among males; in coho salmon (male mean / = 1.3, female mean / = 0.15; Fleming and Gross 1994) and Atlantic salmon (males 0.59, females 0.53; Garant et al. 2001). I found that selection intensities for females were as high or higher than males at Chonat Lake (0.60 vs. 0.41 in 2001, 0.45 vs. 0.48 in 2002). From the distribution of reproductive success one may infer that the effect of natural selection for smaller body size and later spawning in females was apparently countered by high levels of sexual selection among late-arrivals in 2001. I observed no real advantage of size but found early spawners to have experienced less selection and marginally higher reproductive success (early / = 0.34, late / = 0.73). In the low density year (2002), I found much less evidence of sexual selection. Instead, there was a significant negative correlation between body size and reproductive success for females as well as substantially increased success for late-arriving females which is what one would expect if the effects of predation and natural selection predominate (early / = 0.53, late / = 0.37). While the evolution and maintenance of life history variation among females has traditionally received considerably less attention, the high selective forces that I have proposed suggest that life history variation and alternate reproductive tactics should exist among females at Chonat Lake. Females do not tend to develop elaborate secondary characters (but see Fleming and Gross 1989), nor do they generally exhibit precocial maturation (likely constrained by requirements of egg size and storage). Instead, alternate breeding strategies among females may be cryptic, perhaps relating to run-timing or to characteristics related to egg deposition. In this study, these alternate tactics may involve 'fighters' (larger dominant females which arrive early and compete for access to the best habitats), or those adapted to wait it out (smaller, less dominant females which may delay spawning to avoid or lessen direct competition for spawning habitat). "Waiters", if they exist, may be expected to shown delayed ovulation or an extended period of egg viability. The evolution of runtiming is expected to be adaptive and I have previously detailed its relative cost and benefits for females. Importantly, there appears to be a genetic basis for run-timing in  97  many species which would allow for its maintenance between generations (Taylor 1991; Gharrett and Smoker 1993; Quinn et al. 2000; Hendry et al. 2002). While there appears, however, to be considerable variation in egg size among individuals in a population (reviewed by Einum et al. 2004), the relative influence of such factors on reproductive success remains untested. Regardless, the maintenance of life history variation and alternate reproductive strategies point to the advantages of bet-hedging or risk-dispersion mechanisms in small, variable populations. The breeding system of coastal cutthroat trout may, in fact, typify the extent to which risk-dispersal or minimizing mechanisms pervade salmonid breeding systems. Coastal cutthroat trout exhibit one of the broadest and most variable spectra of migratory behaviours of all the salmonids, owing perhaps to the diversity of life history strategies and habitats occupied by the species (Northcote 1997; Johnson et al. 1999). This 'generalist' strategy has the effect of not only spreading mortality or predation risk across a larger geographic area, but further allows for significantly different opportunities for growth among individuals. The variable age and size of resulting individuals at maturity (as well as low levels of iteroparity) provide for the overlap of different cohorts making up a generation and allow for the complex breeding patterns that we have observed. The breeding between life history forms and differentiated cohorts is expected to provide genetic benefits to future generations but also tends to buffer variation in the number of spawners in any given year. Stream resident male production, for example, may show a proportional increase in years when fighter males are scarce (e.g. - Moran and Garcia- Vazquez 1998). My results go further, suggesting that fluctuations in population size are countered by changes to the actual breeding patterns within a population. I found evidence that compensatory mechanisms operate at both high and low densities and may help offset the impacts of low spawner number as well as reducing the overrepresentation of stronger brood years in the next generation (e.g. Rose et al. 2001; Reichard et al. 2004). The implications for the evolution and conservation of populations are obvious (see also next section).  98  3.4.7 Conservation implications C o a s t a l c u t t h r o a t trout a r e a s p e c i e s of s o m e c o n s e r v a t i o n c o n c e r n in w e s t e r n N o r t h A m e r i c a . W h i l e h i s t o r i c a l l y w i d e s p r e a d , t h e r e h a v e b e e n d r a m a t i c d e c l i n e s in t h e n u m b e r a n d distribution of p o p u l a t i o n s a l o n g m o s t of t h e P a c i f i c c o a s t . P o p u l a t i o n s h a v e b e c o m e i n c r e a s i n g l y d i s p l a c e d from their preferred habitat of low gradient c o a s t a l s t r e a m networks, a r e a s that often s e r v e a s f o c a l points for d e v e l o p m e n t (Hall et a l . 1997;  J o h n s o n et al. 1999; C o s t e l l o a n d R u b i d g e  human 2005).  Efforts to c o n s e r v e t h e s p e c i e s a r e h i n d e r e d , not o n l y b y a l a c k of h a b i t a t of habitat p r o t e c t i o n , but b y t h e l a c k of better b i o l o g i c a l i n f o r m a t i o n for t h e  adequate  species,  w h i c h is understudied relative to other s a l m o n i d s . Effective m a n a g e m e n t  and  c o n s e r v a t i o n d e p e n d s o n having a predictive k n o w l e d g e of the effects of management  efforts or e n v i r o n m e n t a l c h a n g e o n p o p u l a t i o n s ( J o h n s o n  M c E l h a n y et al. 2 0 0 0 ) . Unfortunately,  et al.  1999;  m u c h of t h i s b a s i c b i o l o g i c a l i n f o r m a t i o n is  l a c k i n g for c o a s t a l cutthroat trout a n d this s t u d y r e p r e s e n t s t h e first a t t e m p t  to  d e s c r i b e t h e b r e e d i n g s y s t e m o f a n a t u r a l c o a s t a l c u t t h r o a t trout p o p u l a t i o n in t h e wild.  First, I f o u n d that the b r e e d i n g s y s t e m of c o a s t a l cutthroat trout is quite  complex;  a w i d e array of tactics w e r e s u c c e s s f u l l y e m p l o y e d by both m a l e s a n d f e m a l e s at Chonat  L a k e a n d I h a v e p r e s e n t e d e v i d e n c e of m a n y t y p e s of m a t i n g b e h a v i o u r  monogamy,  t h r o u g h to p o l y g y n a n d r y .  I f o u n d that mating o c c u r r e d b e t w e e n  from  cohorts  a n d life h i s t o r y t y p e s a n d t h a t b o t h f i g h t e r a n d s n e a k e r m a l e s c o n t r i b u t e d s i g n i f i c a n t l y to fry p r o d u c t i o n . T h i s c o m p l e x i t y h a s i m p o r t a n t r e p e r c u s s i o n s for o n e of t h e  more-  c o m m o n l y a p p l i e d c o n s e r v a t i o n m e a s u r e s for d e c l i n i n g s a l m o n i d p o p u l a t i o n s ,  namely  the c a p t i v e rearing of b r o o d s t o c k a n d h a t c h e r y s u p p l e m e n t a t i o n of wild p o p u l a t i o n s . The  n a t u r a l f e c u n d i t y of f i s h a n d t h e relative e a s e with w h i c h their r e p r o d u c t i v e c y c l e  c a n b e m a n i p u l a t e d , h a v e m a d e the h a t c h e r y p r o d u c t i o n of s a l m o n i d s a  common  r e s p o n s e to declining fish populations (Lichatowich 1999; M o l o n y et al. 2003). Hatchery fish, however,  a r e routinely i n t r o d u c e d to t h e wild w i t h o u t a n  understanding  of t h e e f f e c t i v e n e s s of t h e t r a n s f e r , t h e f a t e of t h e r e l e a s e d f i s h , o r its i m p a c t s o n w i l d populations (Kruegerand  M a y 1991;  Flagg etal. 2000;  Utter2004).  99 While captive breeding programs can reduce the variance in individual reproductive success and potentially reduce inbreeding (Caballero and Toro 2000; Duschene and Bernatchez 2002; Fernandez et al. 2004), the majority employ limited numbers of spawners which may not be representative of the population as a whole (Campton 2004). No current hatchery protocols would seem able to capture the diverse array of breeding patterns that I have observed and none are designed to include or account for the contribution of stream resident males. I have shown that stream resident males can have substantially higher lifetime reproductive success than fighter males and their widespread maintenance in coastal cutthroat trout populations is likely an important adaptation to buffer the effects of environmental stochasticity on spawner returns. Unfortunately, a wider recognition of the importance of resident components to the conservation of "anadromous populations" is lacking. This may, in part, result from the difficulties in characterizing resident populations or from previous misconceptions about the sneaking tactic. In the past, precocial maturation has been considered to be a detrimental characteristic of salmonid populations rather than a viable life history strategy (e.g. - Taggart et al. 2001). I believe that current hatchery practices for coastal cutthroat trout should, therefore, be re-evaluated and made more biologically realistic by emulating naturally occurring patterns where possible. Second, the dramatic fluctuations observed in the number of females observed between years provides some insight into the influence that density-dependent mechanisms have on reproduction and the relative influence of selection. Densitydependent mechanisms underlie many aspects of fisheries management but in many cases, the actual ecological processes involved may be often poorly resolved (Rose et al. 2001). I found evidence for increased reproductive success at low spawner densities; the result of an increase in levels of polyandry and reduction in reproductive variance among individuals. I interpret this as a form of compensation which may serve to sustain the population at lower spawner abundance. Conversely, high female densities reduced both the proportion of females contributing offspring, as well as mean reproductive success. This may seem counterintuitive, but the reduction in spawner success at high densities can itself be seen as a form of compensation; preventing, for example, the overrepresentation of a particular cohort in a future  100  generation. This reduction appears to be mediated, to a large extent, by the limitations of spawning habitat in the small streams utilized by coastal cutthroat trout. The habitat requirements of the species are such that populations typically inhabit cold water habitats with limited productivity and finite amounts of suitable spawning habitat. While residing in spawning tributaries, coastal cutthroat trout are located almost exclusively in habitat units formed by large woody debris, boulders, or bedrock. Such instream structure creates the necessary pool habitat to catch and retain spawning gravels as well as providing cover from predation; spawners are vulnerable to predation and experience high mortality when suitable riparian or instream cover is lacking (Behnke 1992). In fact, their reliance on quality habitats have led coastal cutthroat trout to be described as a "canary in a coal mine" (Reeves et al. 1997) or as a "sentinel species" for the health of aquatic ecosystems (Slaney 2005). My data add to the large body of data that support the use of these indicator species concepts. This study suggests that factors that increase female-female competition or reduce the amount of available spawning habitat may reduce population productivity through decreased reproductive success and reduced survival of embryos (Chapman 1988; Reeves et al. 1997; Rosenfeld 2001). Since populations are supported by relatively few spawners, any reduction in egg deposition, over-ripening and egg-wasting in unsuccessful females (not to mention reduced egg survival in the substrate) will have a significant impact on population growth and viability (McElhany et al. 2000). Importantly, one may expect a similar response in populations where hatchery supplementation artificially increases spawner density or otherwise influences the age structure of spawning populations (Ryman and Laikre 1991; Flagg et al. 2000; Molony et al. 2003). Finally, numerous recommendations have been made in the conservation literature regarding the minimum population sizes needed to avoid losses of genetic diversity (or other deleterious genetic effects) in small populations. Franklin (1980) and Soule (1980), for example, suggested that an population size of 50 was required to avoid inbreeding depression and that a size of 500 was necessary to retain genetic variation sufficient for long-term adaptive potential and population persistence. The so  101  called "50/500" rule has emerged as a generally accepted guideline for species management (e.g., - Mace and Lande 1991; Hilderbrand and Kershner 2000). The rule, however, makes several unrealistic assumptions which may limit its applicability to salmonids (primarily that all individuals find mates and have equal reproductive success). Waples (1990a) and Allendorf et al. (1997) have suggested that at least 100 effective breeders are required per year to maintain genetic variation in salmon populations over the short term. It may be questionable if typical cutthroat spawning streams can consistently support that level of breeders. By all accounts, the Chonat Lake system would seem more productive than the many in the region; the combination of a productive lake and stable stream flows should certainly allow for the maintenance of a larger population than would a small, ephemeral stream directly entering the open ocean. Yet even here I did not find that the number of effective spawners came close to the number proposed by these authors and even though a relatively large proportion of spawners contributed to fry production, I observed at least one year of significantly declining levels of genetic diversity in fry. While the effect may be transient, an extended run of low spawners returns could quickly erode allelic diversity in coastal cutthroat trout populations. A s Ardren and Kapuscinski (2003) point out, taxon-specific N / N or N / N ratios would be particularly useful to b  c  e  the management of wild populations where census data are typically the only demographic data available. My results (and those of Ardren and Kapuscinski 2003), however, suggest that such ratios should be used with caution. Ratios may not be constant over the short term scales required for management decisions and should be expected to vary between populations in accordance to local environmental conditions.  102 Characterization of the seven polymorphic microsatellite loci in Chonat Lake coastal cutthroat trout across years (N = 479 different individuals).  Table 3-1.  Locus  Reference  A  H  PIC  P 0  P1  Null freq  0.80 0.50 0.81 0.89  0.773 0.375 0.782 0.875  0.440 0.125 0.447 0.631  0.616 0.187 0.623 0.774  +0.076 +0.038 +0.029 -0.019  (Condrey and Bentzen 1998) 5 0.60 0.62 (Olsen et al. 1998) 8 0.79 0.78 (Olsen et al. 1998) 6 0.34 0.41  0.568 0.751 0.380  0.207 0.369 +0.013 0.400 0.579 -0.002 0.087 0.229 +0.088  7.3 0.65 0.69 51  0.643  T  0  H  e  E  E  Multiplex 1 (TA = 56 °C)  Omy77 Oneu11 Sfo8 Ssa85  (Morris et al. 1996) (Scribner et al. 1996) (Angers etal. 1995) (O'Reilly etal. 1996)  9 2 7 14  0.70 0.47 0.76 0.92  §  Multiplex 2 ( T = 58 °C) A  Ocl2 Ogo4 Ogo8 Mean Total  §  0.957 0.995  Parameters: A , number of alleles; He, expected heterozygosity; PIC, polymorphic information content; P 0, exclusionary probability when neither parent is known; P 1, exclusionary probability when one parent is known; Null freq, estimated null allele frequency; §, loci showing significant heterozygote deficiencies under Hardy-Weinberg expectations. T  E  E  Table 3-2. Summary genetic data for Chonat Lake spawners and offspring samples from 2001-2002. Population  n  2001 Spawners 87 2001 YOY fry 197 2002 Spawners 46 2002YOYfry 156  A  A  T  H (SD)  H (SD)  0.66 (0.02) 0.66 (0.01) 0.66 (0.03) 0.63 (0.01)  0.67 (0.07) 0.68 (0.07) 0.70 (0.06) 0.68 (0.07)  0  49 7.0 49 7.0 51 7.3 47 6.7  a  a  HW L D  e  F  i s  95% CI  1/7 0/21 0.02 (-0.03 - 0.06) 2/7 4/21 0.03 (-0.003 - 0.06) 1/7 0/21 0.06 (-0.01 - 0.10) 6/7 11/21 0.08 ( 0.04-0.12)  Within sample diversity parameters: n, number of samples; A , total number of alleles; A, mean number of alleles per locus; H , observed heterozygosity (standard deviation in parentheses); H , expected heterozygosity (standard deviation in parentheses); HW, proportion of loci deviating from Hardy-Weinberg equilibrium following sequential Bonferroni corrections; LD, proportion of pairwise comparisons between loci deviating from linkage equilibrium following sequential Bonferroni corrections; F , Wright's inbreeding coefficient and its associated 95% confidence interval; a - significantly different (p = 0.036, based on 1000 permutations in FSTAT) T  0  is  e  103  Table 3-3. Female reproductive success (Early vs. Late arriving and combined). Total fry Year 2001  2002  No. successful  # Y O Y assigned  females  Mean (CV)  assigned  N u m b e r of m a t e s M e a n (CV)  Early  26/89  (29%)  10/16  (63%)  2.6  (0.58)  2.5  Late  63/89  (71%)  30/35  (75%)  2.1 (0.85)  1.8  C  (0.79)  Combined  89/197*(45%)  40/56  (71%)  2.2  2.0  d  (0.72)  (86%)  3.3  Early  20/101 ( 2 0 % )  Late  81/101 ( 8 0 % )  Combined  101/156*(65%)  6/7  a  b  (0.78)  e  (0.73)  14/14 ( 1 0 0 % )  5.8 (1.07)  20/21 ( 9 5 % )  5.1 (0.66)  2.8  (0.61)  4.1 (0.45) c  a  b  (0.57)  e  3.8  d  (0.50)  Successful spawners were assigned at least 1 Y O Y fry. C V is the coefficient of variation = s / X . * percentage of total fry sample, a, b, c, d - paired means significantly differ (Mann-Whitney U, p < 0.001)  e - paired variances significantly differ (Levene's test for variance homogeneity, p = 0.001)  Table 3-4.  M a l e r e p r o d u c t i v e s u c c e s s (fighter v s . s n e a k e r a n d c o m b i n e d ) . Total fry  Year 2001  assigned  males  M e a n (CV)  94%)  4.0  (0.59)  3.4  (0.49)  (100%)  3.1  (0.86)  2.3  (0.41)  94%)  3.8  (0.63)  3.1  (0.52)  18/18 ( 1 0 0 % )  4.1  (0.73)  3.1  (0.55)  (1.00%)  3.1  (0.40)  2.7  (0.29)  25/25 ( 1 0 0 % )  3.8  (0.69)  3.0  (0.51)  87/109 ( 8 0 % )  22/24 (  Sneaker  22/109 ( 2 0 % )  7/7  109/197*(55%)  Fighter  73/95  (77%)  Sneaker  22/95  ( 2 3 % ) ..  Combined  95/156*(61%)  # Y O Y a s s i g n e d N u m b e r of m a t e s Mean (CV)  Fighter  Combined  2002  No. s u c c e s s f u l  29/31 (  7/7  Successful spawners were assigned at least 1 Y O Y fry. C V is the coefficient of variation = s / X . * percentage of total fry sample.  104  Table 3-5. Demographic estimates of the number of breeders ( N ) . b  Year  Method  N  2001  Census Sex Ratio VRS  31 31 31  Census Sex Ratio VRS  25 25 25  2002  m  N m b  25.1  20.4  Nbm/N  m  N  N  f  0.81  56 56 56  0.82  21 21 21  b f  34.5  15.9  Nb,/N  N  f  Nb/Nc  b  0.62  87 79.8 58.1  1.00 0.92 0.67  0.76  46 45.7 35.6  1.00 0.99 0.77  VRS is the variance in reproductive success method, N is the number of male spawners; N is the number of females, N or effective number of males under the VRS method; N is the effective number of females under the VRS method; N is the census number of breeders in each year (N + N,). m  f  bm  bf  c  m  Table 3-6. Genetic estimates of the effective number of breeders (N ) and 95% confidence intervals. b  N 2001  2002  b  9 5 % CI  N^ N  Linkage disequilibrium Moments based temporal Pseudo-likelihood temporal (MLNE) Bayesian temporal approach (TM3)  104.3 89.0 147.4 107.2  (83.8- 132.7) (35.3- 428.9) (69.6- 1231.8) (43.4 - 135.5)  1.23 1.02 1.69 1.23  Linkage disequilibrium Moments based temporal Pseudo-likelihood temporal (MLNE) Bayesian temporal approach (TM3)  41.7 49.3 45.9 24.9  (35.4(20.3(27.8(16.6-  0.91 1.07 0.99 0.54  49.3) 197.7) 100.7) 39.8)  c  Linkage disequilibrium estimates are based on YOY fry samples. For temporal estimates, spawners and YOY fry in any one year represent generations 0 and 1 respectively. N is the total census number of breeders in each year. c  105  Table 3-7. Harmonic mean estimates of the number of breeders (N ) and NJ N ratios from 2001 to 2002 for all methods. b  N  Method  b  NJ  N  Nl  c  N  b  NJ  N  D e m o g r a p h i c estimates Census  60.18  S e x Ratio V a r i a n c e in reproductive s u c c e s s  58.12  0.97  1.00  0.11  44.15  0.73  0.08  59.58  0.99  0.11  0.11  Genetic estimates Y O Y L i n k a g e Disequilibrium S p a w n e r to Y O Y temporal S p a w n e r to Y O Y Pseudo-likelihood ( M L N E ) S p a w n e r to Y O Y B a y e s i a n a p p r o a c h (TM3) Grand Mean  N  c  63.45 70.00 40.41  1.05  0.12  1.16  0.13  0.67  0.08  56.56  0.11  is the harmonic mean census number of breeders between years; N is the harmonic mean  estimated total adult population size = 526; N  e  estimate of N  b  is the estimated value of N assuming our grand mean e  applies equally to the 3-4 cohorts making up the generation (N  e  = 169 - 226).  Figure 3-1. Location of the study site and overview of the Chonat Lake system. o  CD  107  150  200  250  300  350  400  450  350  400  450  Forklength (m m )  20 18  2002  16  0) i  14  • Females  12  • Males • SRM  10  150  200  250  300 Fork length (mm)  Figure 3-2. Length frequency distribution of Chonat Lake spawners in 2001 and 2002 as measured by fork length (mm). S R M refers to stream resident males.  Males  Females 20  Mean = 3.1, CV = 0.52  Mean = 2.0, CV = 0.72  •  Numbei  2001 / .  . I  :  0  I " —1——.—J  # Mates  /  1  J  /  i  i - —  # Mates  Mean = 3.0, CV = 0.51  2002  i  Mean = 3.8, CV = 0.50  E  /  3  \  2  3  4  5  # Mates  6  7  8  # Mates  Figure 3-3. Distribution of the number of inferred mates for Chonat Lake spawners by year. Mean values and the coefficient of variation (CV = s/X) are shown. The approximate normal curve is given for reference.  -j  00  Males  Females  r = 0.49, p = 0.01  2001  r = -0.09, p = 0.29  <  < *  500  3  2  280  300  320  340  Forklength (mm)  360  380  400  420  440  420  440  Forklength (mm)  r= -0.01, p = 0.48  r = -0.41, p = 0.04 •  2002  <  < A  1  A  A  280  Forklength (mm)  300  320  340  360  360  400  Forklength (mm)  Figure 3-4. Reproductive success of male and female spawners with respect to fork length (mm). Triangles represent stream resident males, squares - fighter males, diamonds - early arriving females, and circles - late arriving females. Spearman r coefficients and corresponding p-values for combined correlations are shown.  110  500  r = 0.002, p > 0.40  2001 400  • :  •  i  s  o  0) CD  200  100 260  280  300  320  340  360  380  400  420  440  460  F e m a l e Forklength (mm)  500  2002 £  400  E  l  • 300  _CD CO ^  200  r = -0.07, p > 0.40  ••  •  • •  • •  • • • •  •  •  280  300  •  «  320  • •  • • •  « • • •  100 260  • •  •a  • •  •  • •  ••  •  • • •  • ••  340  360  380  400  • 420  440  460  F e m a l e Forklength (mm)  Figure 3-5. Scatterplot of size-assortative mating for inferred mating pairs (male vs. female) by year. Spearman r coefficients and corresponding p-values are shown.  111  3.5 REFERENCES Allendorf, F. W., D. Bayles, D. Bottom, K. P. Currens, C. Frissel, D. Hankin, J . A. Lichatowich, W. Nehlsen, P. C. Trotter, and T. Williams. 1997. Prioritizing Pacific salmon stocks for conservation. Conservation Biology 11:140-152. Allendorf, F. W., and S. R. Phelps. 1981. Use of allelic frequencies to describe population structure. Canadian Journal of Fisheries and Aquatic Sciences 38:1507-1514. Amos, W., and J . Harwood. 1998. Factors affecting levels of genetic diversity in natural populations. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 353:177-186. Anderson, E. 2005. An efficient Monte Carlo method for estimating Ne from temporally spaced samples using a coalescent-based Likelihood. Genetics 170:955-967. Angers, B., L. Bernatchez, A. Angers, and L. Desgroseillers. 1995. Specific microsatellite loci for brook charr reveal strong population subdivision on a microgeographic scale. Journal of Fish Biology 47:177-185. Ardren, W., and A. Kapuscinski. 2003. Demographic and genetic estimates of effective population size (N ) reveals genetic compensation in steelhead trout. Molecular Ecology 12:35-49. e  Arnold, S., and M. Wade. 1984. On the measurement of natural and sexual selection: theory. Evolution 38:709-719. Avise, J . C , A. Jones, D. Walker, J . A. DeWoody, and e. al. 2002. Genetic mating systems and reproductive natural histories of fishes: Lessons for ecology and evolution. Annual Review of Ecology & Systematics 36:19-45. Bartley, D., M. Bagley, G. Gall, and B. Bentley. 1992. Use of linkage disequilibrium data to estimate effective size of hatchery and natural fish populations. Conservation Biology 6:365-375. Basset, P., F. Balloux, and N. Perrin. 2001. Testing demographic models of effective population size. Proceedings of the Royal Society of London - Series B: Biological Sciences 268:311-317. Behnke, R. J . 1992. Native trout of western North America. American Fisheries Society Monograph 6. Bernatchez, L., and P. Duchesne. 2000. Individual-based genotype analysis in studies of parentage and population assignment: how many loci, how many alleles? Canadian Journal of Fisheries & Aquatic Sciences 57:1-12. Berthier, P., M. Beaumont, J . Cornuet, and G. Luikart. 2002. Likelihood-based  112 estimation of the effective population size using temporal changes in allele frequencies: A genealogical approach. Genetics 160:741-751. Blanchfield, P., and M. Ridgway. 1997. Reproductive timing and the use of redd sites by lake-spawning brook trout (Salvelinus fontinalis). Canadian Journal of Fisheries and Aquatic Sciences 54:747-756. Blundell, G., M. Ben-David, P. Groves, R. Bowyers, and E. Geffen. 2002. Characteristics of sex-biased dispersal and gene flow in coastal river otters: implications for natural recolonization of extirpated populations. Molecular Ecology 11:289-303. Breder, C , and D. Rosen. 1966. Modes of Reproduction in Fishes. T F H Publications, Jersey City, New Jersey. Caballero, A. 1994. Developments in the prediction of effective population size. Heredity 73:657-679. Caballero, A., and M. A. Toro. 2000. Interrelations between effective population size and other pedigree tools for the management of conserved populations. Genetical Research 75:331-343. Callen, D., A. Thompson, Y. Shen, H. Phillips, R. Richards, J . Mulley, and G. Sutherland. 1993. Incidence and origin of 'null' alleles in the ( A C ) microsatellite markers. American Journal of Human Genetics 52:922-927. n  Campton, D. 2004. Sperm competition in salmon hatcheries: the need to institutionalize genetically benign spawning protocols. Transactions of the American Fisheries Society 133:1277-1289. Campton, D., and F. Utter. 1987. Genetic structure of anadromous cutthroat trout (Salmo clarki clarki) populations in the Puget Sound area: Evidence for restricted gene flow. Canadian Journal of Fisheries & Aquatic Sciences 44:573-582. Carvalho, G. 1993. Evolutionary aspects offish distribution: genetic variability and adaptation. Journal of Fish Biology 43:53-73. Chapman, D. 1988. Critical review of variables used to define effects of fines in redds of large salmonids. Transactions of the American Fisheries Society 117:1-21. Chebanov, N. 1991. The effect of spawner density on spawning success, egg survival, and size structure of the progeny of the sockeye salmon, Oncorhynchus nerka. Journal of Ichthyology 31:101-106. Condrey, M. J . , and P. Bentzen. 1998. Characterization of coastal cutthroat trout (Oncorhynchus clarki clarki) microsatellites and their conservation in other salmonids. Molecular Ecology 7:787-789. Consuegra, S., E. Verspoor, K. D, and C. Garcia de Leaniz. 2005. Asymmetric gene  113 flow and the evolutionary maintenance of genetic diversity in small, peripheral Atlantic salmon populations. Conservation Genetics 6:823-842. Cornuet, J . M., and G. Luikart. 1997. Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:2001-2014. Costello, A. B., and E. Rubidge. 2005. C O S E W I C status report on coastal cutthroat trout (Oncorhynchus clarkii clarki). Draft Report. Pp. 91pp. Prepared for the Committee on the Status of Endangered Wildlife in Canada, Ottawa. Crow, C , and C. Denniston. 1988. Inbreeding and variance effective population numbers. Evolution 42:482-495. Crow, J . 1958. Some possibilities for measuring selection intensities in man. Human Biology 30:1-13. Crow, J . , and N. Morton. 1955. Measurement of gene frequency drift in small populations. Evolution 9:202-214. Currens, K. P., C. B. Schreck, and H. W. Li. 1990. Allozyme and morphological divergence of rainbow trout (Oncorhynchus mykiss) above and below barriers in the Deschutes River, Oregon. Copeia 3:730-746. Curry, R. A., and D. Noakes. 1995. Groundwater and the selection of spawning sites by brook trout (Salvelinus fontinalis). Canadian Journal of Fisheries and Aquatic Sciences 52:1733-1740. Dannewitz, J., E. Petersson, J . Dahl, T. Prestegaard, A. Lof, and T. Jarvi. 2004. Reproductive success of hatchery-produced and wild-born brown trout in an experimental stream. Journal of Applied Ecology 41:355-364. Danzmann, R. G. 1997. P R O B M A X : a computer program for assigning unknown parentage in pedigree analysis from known genotypic pools of parents and progeny. Journal of Heredity 88:333. de Gaudemar, B., and E. Beall. 1998. Effects of overripening on spawning behaviour and reproductive success of Atlantic salmon females spawning in a controlled flow channel. Journal of Fish Biology 53:434-446. De Rochambeau, H., F. Fournet-Hanocq, and J . V. Khang. 2000. Measuring and managing genetic variability in small populations. Annales de Zootechnie 49:7793. DeWoody, J . A., D. Walker, and J . C. Avise. 2000. Genetic parentage in large half-sib clutches: theoretical estimates and empirical appraisals. Genetics 154:1907-1912. Dodds, K., M. Tate, J . McEwen, and A. Crawford. 1996. Exclusion probabilities for pedigree testing farm animals. Theoretical and Applied Genetics 76:369-380.  114 Duschene, P., and L. Bernatchez. 2002. An analytical investigation of the dynamics of inbreeding in multi-generational supportive breeding. Conservation Genetics 3:4760. Einum, S., M. Kinnison, and A. Hendry. 2004. Evolution of egg size and number. Pp. 126-153 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford. Ellstrand, N., and D. Elam. 1993. Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology & Systematics 24:217-243. Emlen, S., and L. Oring. 1977. Ecology, sexual selection, and the evolution of animal mating systems. Science 197:215-223. Essington, T., P. Sorenson, and D. Paron. 1998. High rate of redd superimposition by brook trout (Salvelinus fontinalis) and brown trout (Salmo trutta) in a Minnesota stream cannot be explained by habitat availability alone. Canadian Journal of Fisheries and Aquatic Sciences 55:2310-2316. Fernandez, J . , M. A. Toro, and A. Caballero. 2004. Managing individuals' contributions to maximize the allelic diversity maintained in small, conserved populations. Conservation Biology 18:1358-1367. Flagg, T., B. Berejikian, J . Colt, W. Dickhoff, L. Harrell, D. Maynard, C. Nash, M. Strom, R. Iwamoto, and C. Mahnken. 2000. Ecological and behavioural impacts of artificial production strategies on the abundance of wild salmon populations - a review of practices in the Pacific Northwest. N O A A Tech. Memo. NMFS-NWFS-4. Pp. 98. Northwest Fish. Sci. Center, Seattle, WA. Fleming, I. A. 1998. Pattern and variability in the breeding system of Atlantic salmon (Salmo salar), with comparisons to other salmonids. Canadian Journal of Fisheries & Aquatic Sciences 55:59-76. Fleming, I. A., and M. R. Gross. 1989. Evolution of adult female life history and morphology in a Pacific salmon (coho: Oncorhynchus kisutch). Evolution 43:141157. Fleming, I. A., and M. R. Gross. 1994. Breeding competition in a Pacific salmon (coho: Oncorhynchus kisutch): measures of natural and sexual selection. Evolution 48:637-657. Fleming, I. A., and J . Reynolds. 2004. Salmonid breeding systems. Pp. 264-294 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford. Foote, C. J . , C. C. Wood, and R. E. Withler. 1989. Biochemicalgenetic comparison of sockeye salmon and kokanee, the anadromous and nonanadromous forms of  115  Oncorhynchus nerka. Canadian Journal of Fisheries and Aquatic Sciences 46:149 -158. Fox, C , and C. Rauter. 2003. Bet-hedging and the evolution of multiple mating. Evolutionary Ecology Research 5:273-286. Frankham, R. 1995. Effective population size/adult population size ratios in wildlife: a review. Genetical Research 66:85-107. Frankham, R. 1996. Relationship of genetic variation to population size in wildlife. Conservation Biology 10:1500-1508. Franklin, I. 1980. Evolutionary changes in small populations. Pp. 135-150 in M. E. Soule and B. Wilcox, eds. Conservation Biology: An Evolutionary-Ecological Perspective. Sinauer Associates, Sunderland, MA. Garant, D., J . Dodson, and L. Bernatchez. 2001. A genetic evaluation of mating system and determinants of individual reproductive success in Atlantic salmon {Salmo salarL.). Journal of Heredity 92:137-145. Garant, D., J . Dodson, and L. Bernatchez. 2003. Differential reproductive success and heritability of alternative reproductive tactics in wild Atlantic salmon (Salmo salarL.). Evolution 57:1133-1141. Garcia-Vazquez, E., P. Moran, J . L. Martinez, J . Perez, B. de Gaudemar, and E. Beall. 2001. Alternative mating strategies in Atlantic salmon and brown trout. Journal of Heredity 92:146-149. Geiger, H., W. Smoker, L. Zhivotovsky, and A. Gharrett. 1997. Variability of family size and marine survival in pink salmon (Oncorhynchus gorbuscha) has implications for conservation biology and human use. Canadian Journal of Fisheries and Aquatic Sciences 54:2684-2690. Gharrett, A., and W. Smoker. 1993. Genetic components in life history traits contribute to population structure. Pp. 197-202 in J . G. Cloud and G. H. Thorgaard, eds. Genetic Conservation of Salmonid Fishes. Plenum Press, New York. Gibson, R. 1993. The Atlantic salmon in fresh water: spawning, rearing, and production. Reviews in Fish Biology and Fisheries 3:39-73. Giger, R. 1972. Ecology and management of coastal cutthroat trout in Oregon. Rep. No.6. Pp. 61. Oregon State Game Comm., Corvallis. Goudet, J . 2001. FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3). Available from http://www.unil.ch/izea/softwares/fstat.html. Goudet, J., N. Perrin, and P. Waser. 2002. Tests for sex-biased dispersal using biparentally inherited genetic markers. Molecular Ecology 11:1103-1114.  11  Gross, M. R. 1996a. Alternative reproductive strategies and tactics: Diversity within sexes. Trends in Ecology and Evolution 11 :A92 - A98. Gross, M. R. 1996b. Alternative reproductive strategies and tactics: diversity within sexes. Trends in Ecology & Evolution 11:92-98. Hall, J . D., P. A. Bisson, and R. Gresswell. (Eds.).1997. Sea-run cutthroat trout: biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis. Pp.187. Hansen, M. M., E. E. Nielsen, and K. Mensberg. 1997. The problem of sampling families rather than populations: relatedness among individuals in samples of juvenile brown trout Salmo trutta L. Molecular Ecology 6:469-474. Hasler, A., and A. Scholz. 1983. Olfactory Imprinting and Homing in Salmon. Springer-Verlag, Berlin. Hayes, J . 1987. Competition for spawning space between brown trout (Salmo trutta) and rainbow trout (S. gairdneri) in a lake inlet tributary, New Zealand. Canadian Journal of Fisheries and Aquatic Sciences 44:40-47. Hedrick, P. W., and S. T. Kalinowski. 2000. Inbreeding depression in conservation biology. Annual Review of Ecology & Systematics 31:139-162. Heggenes, J . , and R. Borgstrom. 1988. Effect of mink, Mustela vision Schreber, predation on cohorts of juvenile Atlantic salmon, Salmo salarL., and brown trout, S. trutta L, in three small streams. Journal of Fish Biology 33:885-895. Hendry, A., T. Bohlin, B. Jonsson, and O. Berg. 2004. To sea or not to sea? Anadromy versus non-anadromy in salmonids. Pp. 92-125 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford, USA. Hendry, A., and S. C. Stearns. 2004. Evolution Illuminated: Salmon and Their Relatives. Pp. 520. Oxford University Press, Inc., Oxford. Hendry, M., J . K. Wenburg, K. Myers, and A. Hendry. 2002. Genetic and phenotypic variation through the migratory season provides evidence for multiple populations of wild steelhead in the Dean River, British Columbia. Transactions of the American Fisheries Society 131:418-434. Hilderbrand, R. H., and J . L. Kershner. 2000. Conserving inland cutthroat trout in small streams: How much stream is enough? North Am. J . Fish. Mgmt. 20:513520. Hill, W. G. 1981. Estimation of effective population size from data on linkage disequilibrium. Genetical Research 38:209-216. If  117  Hindar, K., B. Jonsson, N. Ryman, and G. Stahl. 1991. Genetic relationships among landlocked, resident, and anadromous brown trout, Salmo trutta L. Heredity 66:83 - 91. Johnson, O., M. Ruckelshaus, W. Grant, F. Waknitz, A. Garrett, G. Bryant, K. Neely, and J . Hard. 1999. Status review for coastal cutthroat from Washington, Oregon, and California. Report NMFS-NWFSC-37. Pp. 292. National Marine Fisheries Service, Seattle, WA. Jones, A., and W. Ardren. 2003. Methods of parentage analysis in natural populations. Molecular Ecology 12:2511-2523. Kimura, M., and J . Crow. 1963. The measurement of effective population number. Evolution 17:279-288. Krueger, C , and B. May. 1991. Ecological and genetic effects of salmonid introductions in North America. Canadian Journal of Fisheries and Aquatic Sciences 48 (Suppl. 1):66-77. Lacy, R. C. 1987. Loss of genetic diversity from managed populations: Interacting effects of drift, mutation, immigration, selection, and population subdivision. Conservation Biology 1:143-158. Lande, R., and G. Barrowclough. 1987. Effective population size, genetic variation, and their use in population management. Pp. 87-123 in M. E. Soule, ed. Viable Populations for Conservation. Cambridge University Press, New York. Lichatowich, J . A. 1999. Salmon Without Rivers. Island, Washington, DC. Lonzarich, D., and T. P. Quinn. 1995. Experimental evidence for the effect of depth and structure on the distribution, growth, and survival of fishes. Canadian Journal of Zoology 73:2223-2230. Luikart, G., F. W. Allendorf, J . Cornuet, and W. Sherwin. 1998. Distortion of allele frequency distributions provides a test for recent population bottlenecks. Journal of Heredity 89:238-247. Luikart, G., and J . Cornuet. 1998. Empirical evaluation of a test for identifying recently bottlenecked populations from allele frequency data. Conservation Biology 12:228-237. Luikart, G., and J . M. Cornuet. 1999. Estimating the effective number of breeders from heterozygote excess in progeny. Genetics 151:1211-1216. Mace, G., and R. Lande. 1991. Assessing extinction threats: toward a reevaluation of IUCN threatened species categories. Conservation Biology 5:148-157. Marshall, T. O , J . Slate, L. E. B. Kruuk, and J . M. Pemberton. 1998. Statistical confidence for likelihood-based paternity inference in natural populations.  118  Molecular Ecology 7:639-655. Martinez, J . L, P. Moran, J . Perez, B. de Gaudemar, E. Beall, and E. GarciaVazquez. 2000. Multiple paternity increases effective size of southern Atlantic salmon populations. Molecular Ecology 9:293 - 298. Massey, J . 1984. Beach Fishing. Pp. 47-54 in M. Aikens, ed. A Cutthroat Collection: A Guide to Understanding and catching the Mysterious Trout. Maclean Hunter Ltd., Vancouver. McDowall, R. 2001. Anadromy and homing: two life-history traits with adaptive synergies in salmonid fishes? Fish and Fisheries 2:78-85. McElhany, P., M. Ruckelshaus, M. J . Ford, T. Wainwright, and E. Bjorkstedt. 2000. Viable salmonid populations and the recovery of evolutionarily significant units. N O A A Tech. Mem. NMFS-NWFSC-42, Seattle, WA. Mehranvar, L., M. Healey, A. Farrell, and S. Hinch. 2004. Social versus genetic measures of reproductive success in sockeye salmon, Oncorhynchus nerka. Evolutionary Ecology Research 6:1167-1181. Mercer, S. 1982. Broodstock development and evaluation of enhancement techniques. Washington State Department of Fisheries Research Report F-68-R6. Pp. 45. Molony, B., R. Lenanton, G. Jackson, and J . Norriss. 2003. Stock enhancement as a fisheries management tool. Reviews in Fish Biology and Fisheries 13:409-432. Moran, P., and E. Garcia- Vazquez. 1998. Multiple paternity in Atlantic salmon: a way to maintain genetic variability in relicted populations. Journal of Heredity 89:551553. Moran, P., A. M. Pendas, E. Beall, and E. Garcia-Vazquez. 1996. Genetic assessment of the reproductive success of Atlantic salmon precocious parr by means of V N T R loci. Heredity 77:655-660. Morris, D. B., K. R. Richard, and J . M. Wright. 1996. Microsatellites from rainbow trout (Oncorhynchus mykiss) and their use for genetic study of salmonids. Canadian Journal of Fisheries and Aquatic Sciences 53:120-126. Neff, B. D., J . Repka, and M. R. Gross. 2000a. Parentage analysis with incomplete sampling of candidate parents and offspring. Molecular Ecology 9:515 - 528. Neff, B. D., J . Repka, and M. R. Gross. 2000b. Statistical confidence in parentage analysis with incomplete sampling: how many loci and offspring are needed? Molecular Ecology 9:529-539. Nei, M., and F. Tajima. 1981. Genetic drift and the estimation of effective population size. Genetics 98:625-640.  119  Newman, D., and D. Pilson. 1997. Increased probability of extinction due to decreased genetic effective population size - experimental populations of Clarkia pulchella. Evolution 51:354-362. Northcote, T. G. 1997. Why sea-run? An exploration into the migratory/ residency spectrum of coastal cutthroat trout. Pp. 20-26 in J . D. Hall, P. A. Bisson and R. Gresswell, eds. Sea-run cutthroat trout: biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis. Nunney, L. 1995. Measuring the ratio of effective population size to adult numbers using genetic and ecological data. Evolution 49:389-392. Nunney, L. 1996. The influence of variation in female fecundity on effective population size. Biological Journal of the Linnean Society 59:411-425. Nunney, L. 1999. The effective size of a hierarchically structured population. Evolution 53:1-10. Nunney, L. 2000. The limits to knowledge in conservation genetics - The value of effective population size. Evolutionary Biology 32:179-194. Olsen, J., P. Bentzen, and J . Seeb. 1998. Characterization of seven microsatellite loci derived from pink salmon. Molecular Ecology 7:1087-1089. O'Reilly, P. T., L. C. Hamilton, S. K. McConnell, and J . M. Wright. 1996. Rapid analysis of genetic variation in Atlantic salmon (Salmo salar) by P C R multiplexing of dinucleotide and tetranucleotide microsatellites. Canadian Journal of Fisheries and Aquatic Sciences 53:2292-2298. Pannell, J . R., and B. Charlesworth. 2000. Effects of metapopulation processes on measures of genetic diversity. Philosophical Transactions of the Royal Society of London - Series B: Biological Sciences 355:1851-1864. Park, S. 2001. Trypanotolerance in West African Cattle and the Population Genetic Effects of Selection. Ph.D. Dissertation. University of Dublin. Peel, D., J . Ovenden, and S. Peel. 2004. NeEstimator: software for estimating effective population size, Version 1.3. Queensland Government, Department of Primary Industries and Fisheries. Piry, S., G. Luikart, and J . M. Cornuet. 1999. B O T T L E N E C K : A computer program for detecting recent reductions in the effective population size using allele frequency data. Journal of Heredity 90:502-503. Pollack, E. 1983. A new method for estimating the effective population size from allele frequency changes. Genetics 104:531-548. Pollock, K., J . Nichols, C. Brownie, and J . Hines. 1990. Statistical inference for  120  capture-recapture experiments. Wildlife Monographs 107:1-97. Pudovkin, A. I., D. V. Zaykin, and D. Hedgecock. 1996. On the potential for estimating the effective number of breeders from heterozygote-excess in progeny. Genetics 144:383-387. Pusey, A., and M. Wolf. 1996. Inbreeding avoidance in animals. Trends in Ecology and Evolution 11:201-205. Quinn, T. P., A. Hendry, and G. Buck. 2001a. Balancing natural and sexual selection in sockeye salmon: interactions between body size, reproductive opportunity, and vulnerability to predation by bears. Evolutionary Ecology Research 3:917-937. Quinn, T. P., M. Unwin, and M. Kinnison. 2000. Evolution of temporal isolation in the wild: genetic divergence in timing of migration and breeding by introduced chinook salmon populations. Evolution 54:1372-1385. Quinn, T. P., L. Wetzel, S. Bishop, K. Overberg, and D. Rogers. 2001b. Influence of breeding habitat on bear predation and age at maturity and sexual dimorphism of sockeye salmon populations. Canadian Journal of Zoology 79:1782-1793. Reeves, G., J . D. Hall, and S. V. Gregory. 1997. The impact of land-management activities on coastal cutthroat trout and their freshwater habitats. Pp. 138-144 in J . D. Hall, P. A. Bisson and R. Gresswell, eds. Sea-run cutthroat trout: Biology, management, and future conservation. Oregon Chapter, American Fisheries Society, Corvallis. Reichard, M., C. Smith, and W. C. Jordan. 2004. Genetic evidence reveals densitydependent mediated success of alternative mating behaviours in the European bitterling (Rhodeus sericeus). Molecular Ecology 13:1569-1578. Reynolds, J . 1996. Animal breeding systems. Trends in Ecology & Evolution 11:6872. Rieman, B. E., and F. W. Allendorf. 2001. Effective population size and genetic conservation criteria for bull trout. North Am. J . Fish. Mgmt. 21:756-764. Rose, K. A., J . H. Cowan, K. O. Winemiller, R. A. Myers, and R. Hilborn. 2001. Compensatory density dependence in fish populations: importance, controversy, understanding and prognosis. Fish and Fisheries 2:293-327. Rosenfeld, J . 2001. Freshwater habitat requirements of anadromous cutthroat trout and implications for forestry impacts. Fisheries Management Report No. 113. Pp. pp. 77. Fisheries Research, Vancouver. Ruggerone, G., R. Hanson, and D. Rogers. 2000. Selective predation by brown bears (Ursus arctos) foraging on spawning sockeye salmon (Oncorhynchus nerka). Canadian Journal of Zoology 78:974-981.  121 Ryman, N., and L. Laikre. 1991. Effects of supportive breeding on the genetically effective population size. Conservation Biology 5:325-329. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Scholten, A. 1997. Vancouver Island anadromous coastal cutthroat trout: Synoptic survey. Report prepared for Fisheries Research Biodiversity Unit, Fisheries Branch, Ministry of Environment Lands and Parks. Pp. 8 + appendices, Vancouver, BC. Scribner, K. T., J . Gust, and R. Fields. 1996. Isolation and characterization of novel salmon microsatellite loci: cross-species amplification and population genetic applications. Canadian Journal of Fisheries & Aquatic Sciences 53:833-841. Seamons, T., P. Bentzen, and T. P. Quinn. 2004. The mating system of steelhead, Oncorhynchus mykiss, inferred by molecular analysis of parents and progeny. Environmental Biology of Fishes 69:333-344. Slaney, P. 2005. Coastal cutthroat trout as sentinels of Lower Mainland watershed health: A n action plan for conservation, restoration, and recovery. Draft report for Ministry of Water, Land and Air Protection, Lower Mainland (Region 2), Surrey, BC. Pp. 94. Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science 236:787-792. Soule, M. E. 1980. Thresholds for survival: maintaining fitness and evolutionary potential. Pp. 151-169 in M. E. Soule and B. Wilcox, eds. Conservation Biology: An Evolutionary-Ecological Perspective. Sinauer Associates, Sunderland, MA. Springate, J . , N. Bromage, J . Elliot, and D. Hadson. 1984. The timing of ovulation and stripping and their effects on the rates of fertilization and survival to eyeing, hatch, and swim-up in the rainbow trout {Salmo gairdneri R.). Aquaculture 142:139-148. Steady, R. F. 1992. Historical ecology of Salmoninae, with special reference to Oncorhynchus. Pp. 622-658 in R. Mayden, ed. Systematics, Historical Ecology, and North American Freshwater Fishes. Stanford University Press, Stanford, C A . Stearns, S. C , and A. Hendry. 2004. The salmonid contribution to key issues in evolution. Pp. 3-19 in A. Hendry and S. C. Stearns, eds. Evolution Illuminated, Salmon and their relatives. Oxford University Press, Oxford. Storz, J . , U. Ramakrishnan, and S. Alberts. 2002. Genetic effective size of a wild primate population: Influence of current and historical demography. Evolution 56:817-829. Storz, J . F. 1999. Genetic consequences of mammalian social structure. Journal of Mammalogy 80:553-569.  122 Storz, J . F., H. Bhat, and T. Kunz. 2001. Genetic consequences of polygyny and social structure in an Indian fruit bat, Cynopterus sphinx. II. Variance in male mating success and effective population size. Evolution 55:1224-1232. Sumner, F. H. 1952. Migrations of salmonids in Sand Creek, Oregon. Transactions of the American Fisheries Society 82:139 - 150. Sumner, F. H. 1962. Migrations and growth of the coastal cutthroat trout in Tillamook County, Oregon. Transactions of the American Fisheries Society 91:77 - 83. Taggart, J . , I. McClaren, D. Hay, J . Webb, and A. Youngson. 2001. Spawning success in Atlantic salmon (Salmo salarL.): a long-term D N A profiling-based study conducted in a natural stream. Molecular Ecology 10:1047-1060. Tallmon, D. A., G. Luikart, and M. Beaumont. 2004. Comparative evaluation of a new effective population size estimator based on approximate Bayesian computation. Genetics 167:977-988. Taylor, E. B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98:185-207. Thompson, O , W. Poole, M. Matthews, and A. Ferguson. 1998. Comparison, using minisatellite D N A profiling, of secondary male contribution in the fertilization of wild and ranched Atlantic salmon (Salmo salar) ova. Canadian Journal of Fisheries and Aquatic Sciences 55:2011-2018. Trotter, P. C. 1989. Coastal cutthroat trout: a life history compendium. Transactions of the American Fisheries Society 118:463 - 473. Utter, F. 2004. Population genetics, conservation and evolution in salmonids and other widely cultured fishes: some perspectives over six decades. Reviews in Fish Biology and Fisheries 14:125-144. Utter, F., D. Campton, S. Grant, G. Milner, J . Seeb, and L. Wishard. 1980. Population structures of indigenous salmonid species of the Pacific Northwest. Pp. 285-304 in W. J . McNeil and D. C. Himsworth, eds. Salmonid Ecosystems of the North Pacific. Oregon State University Press, Corvallis. van der Berghe, E., and M. R. Gross. 1989. Natural selection resulting from female breeding competition in a Pacific salmon (Oncorhynchus kisutch). Evolution 43:125-140. Vrijenhoek, R. C. 1994. Genetic diversity and fitness in small populations. Pp. 37-53 in V. Loeschcke, J . Tomiuk and S. Jain, eds. Conservation Genetics. Birkhauser Verlag, Basil. Vrijenhoek, R. O , and S. Lerman. 1982. Heterozygosity and developmental stability under sexual and asexual breeding systems. Evolution 36:768-776.  123 Walter, R., and B. Epperson. 2001. Geographic pattern of genetic variation in Pinus resinosa: area of greatest diversity is not the origin of postglacial populations. Molecular Ecology 10:103-111. Wang, J . 2001. A pseudo-likelihood method for estimating effective population size from temporally spaced samples. Genetical Research 78:243-257. Wang, J . 2005. Estimation of effective population sizes from data on genetic markers. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 360:1395-1409. Wang, J . L. 1997. More efficient breeding systems for controlling inbreeding and effective size in animal populations. Heredity 79:591-599. Wang, J . L., and A. Caballero. 1999. Developments in predicting the effective size of subdivided populations. Heredity 82:212-226. Waples, R. S. 1989. A generalized approach for estimating effective population size from temporal changes in allele frequency. Genetics 121:379-391. Waples, R. S. 1990a. Conservation genetics of Pacific salmon. II. Effective population size and the loss of genetic variability. Journal of Heredity 81:267-276. Waples, R. S. 1990b. Conservation genetics of Pacific salmon. III. Estimating effective population size. Journal of Heredity 81:277-289. Waples, R. S. 2002. Evaluating the effect of stage-specific survivorship on the Ne/ N ratio. Molecular Ecology 11:1029-1037. Waples, R. S. 2005. Genetic estimates of contemporary effective population size: to what time periods do the estimates apply? Molecular Ecology 14:3335-3352. Waples, R. S. 2006. A bias correction for estimates of effective population size based on linkage disequilibrium at unlinked gene loci. Conservation Genetics 7:167-184 Wenburg, J . K. 1998. Coastal cutthroat trout (Oncorhynchus clarki clarki): genetic population structure, migration patterns, and life history traits. Doctoral Dissertation. School of Fisheries. University of Washington, Seattle. Wenburg, J . K., and P. Bentzen. 2001. Genetic and behavioural evidence for restricted gene flow among coastal cutthroat trout populations. Transactions of the American Fisheries Society 130:1049-1069. Wenburg, J . K., P. Bentzen, and C. J . Foote. 1998. Microsatellite analysis of genetic population structure in an endangered salmonid - the coastal cutthroat trout (Oncorhynchus clarki clarki). Molecular Ecology 7:733-749. Widmer, A., and C. Lexer. 2001. Glacial refugia: sanctuaries for allelic richness, but  not for gene diversity. Trends in Ecology & Evolution 16:267-269. Wilson, A., and M. M, Ferguson. 2002. Molecular pedigree analysis in natural populations of fishes: approaches, applications, and practical considerations. Canadian Journal of Fisheries and Aquatic Sciences 59:1696-1707. Wright, S. 1931. Evolution in Mendelian populations. Genetics 16:97-159. Wright, S. 1938. Size of populations and breeding structure in relation to evolution. Science 87:430-431. Zar, J . 1999. Biostatistical Analysis. Prentice Hall, New Jersey.  125  C H A P T E R 4: D I S T A N C E , S U R E . . . B U T TIMING M A Y B E E V E R Y T H I N G : G E N E T I C P O P U L A T I O N S T R U C T U R E IN C O A S T A L C U T T H R O A T T R O U T (ONCORHYNCHUS  CLARKII  CLARKI)  F R O M BRITISH C O L U M B I A , C A N A D A  3  "The concept of population that we carry in our minds - a group of interbreeding individuals having little or no contact with other such groups - is far removed from what we actually see." - C a u g h l e y (1977: 4)  4.1 4.1.1  INTRODUCTION  The relevance of population structure  In sexual species, the 'population' serves as the fundamental biological unit at which most evolutionary and ecological dynamics are observed. It is, for example, the unit in which sexual selection and reproduction occur, thereby linking the often disparate fields of ecology, evolution, and conservation biology. The need to identify the causes and consequences of variation among populations (for genetic and fitness-related traits) further unite these fields; in terms of evolutionary change, such variation is essential to processes such as dispersal, adaptation, and ultimately speciation (Mayr 1982; Futuyma 1986; Avise 1994). In terms of ecology and conservation, this information is required to understand the demographic interdependence of otherwise distinct groups as well as in estimating key population parameters including migration rates, effective population sizes, and viability in the wild (Beissinger and Westphal 1998; Reed et al. 2002). A s Waples and Gaggiotti (2006) point out, populations are the natural focal units for conservation and management. Indeed, the preservation of adaptive diversity within species depends on knowledge of the scale at which population structuring occurs (Ricker 1954; Nielsen 1999). Unfortunately, as Waples and Gaggiotti (2006)  A version of this chapter will be submitted for publication. Costello, A. "Distance sure... but timing may be everything: population structure in coastal cutthroat trout".  3  126  and others also point out, the process of defining distinct population segments can often be hampered by the ambiguity surrounding the applicability of different 'population concepts' to individual groups (Schaefer 2006). It can be further hampered by our ability to infer the true nature of population subdivision in the wild. For example, the delineation of population structure may easily be made when populations are obviously disjunct and isolated or in cases where populations may be differentiated by distinguishing ecological characteristics. In continuously distributed species, however, the delineation of distinct groups often proves more difficult. In many cases, a lack of prior knowledge of species biology inhibits attempts to delineate meaningful population segments where population structuring does not coincide with obvious physical or ecological barriers. Perhaps the primary factor determining the nature of population structure in a species is the level of connectivity among adjacent populations (Hastings and Harrison 1994; Hanski 1999). Connectivity may itself be difficult to quantify given its own inherent complexity (Webster et al. 2002; Belisle 2005), but its importance to understanding population dynamics cannot be understated. In terms of population viability and conservation, demographic connectedness among adjacent populations may, for example, directly improve population viability if immigration helps maintain small or declining populations in stochastic environments where they might otherwise be extirpated (e.g. - Alderman et al. 2005). Extinction-recolonization dynamics, whereby locally extirpated populations are replaced from nearby source populations, have also been demonstrated to occur in the wild (Hanski 1999). In other cases, demographic connectedness may involve mating dispersal and levels of gene flow among adjacent but otherwise isolated groups. In terms of evolutionary processes, dispersal and gene flow may either inhibit or promote evolutionary change; while too much can essentially make populations panmictic and its absence predispose populations to genetic drift and other stochastic forces, low levels of gene flow may allow adjacent populations to diverge while preventing the loss of genetic variation in small populations (Kimura and Crow 1963; Maruyama 1970; Slatkin 1987). The 'one-migrant-per-generation' ( O M P G ; Wright 1931; Slatkin 1987) and similar genetic rescue concepts (e.g. - Ingvarsson 2001;  127 Tallmon et al. 2004) are perhaps the clearest applications of this phenomenon to conservation biology. Such concepts often find utility in captive breeding programs as the theoretically optimal level of outbreeding and gene flow to minimize the occurrence of inbreeding and the loss of genetic diversity in small populations (Hedrick 1995; Mills and Allendorf 1996; Vucetich and Waite 2000; Wang 2004). While not expected to generate diversity perse,  low levels of migration on the order  of O M P G (i.e. -1-10 immigrants per generation) are expected to promote the retention of adaptive variation among populations in the wild.  4.1.2  Estimating dispersal and gene flow in natural populations  In some cases, it may be possible to estimate the connectivity of distinct groups by directly observing the movement of individuals between adjacent populations. This may be further facilitated by the use of tagging or mark-recapture methodologies that have been used to great advantage in the past (Pollock et al. 1990; Schwarz et al. 1993; Bernard et al. 1995; Quinn 2005). For many species, however, direct observation of movement may be difficult and cost-prohibitive, giving at best only a snapshot of movement over typically small temporal and spatial scales. Furthermore, these methods have little ability to distinguish between migration (predictable feeding, breeding or life-history movements sensu Endler (1977)), and dispersal which involves interbreeding and successful reproduction in non-natal locales (i.e., - it is possible to have physical movement without gene flow, but not wee versa). A full understanding of population structure can, therefore, only be described over a series of spatial and temporal scales which are appropriate for the species under consideration. To address such questions, a number of complementary indirect approaches have been developed which typically rely on information collected from genetic allele frequency data (reviewed by Neigel 1997; Bohonak 1999; Pearse and Crandall 2004). Traditionally, indirect estimates of gene flow have been made using Wright's Fstatistics (Wright 1931, 1951; Weir and Cockerham 1984). Under Wright's island model of population subdivision, F t = 1/ 4 N m + 1 at drift-migration equilibrium, s  e  where N m (or simply Nm), is the effective number of migrants per generation. At Nm e  128 values < 1, genetic drift tends to be the predominant factor influencing patterns of genetic differentiation among populations (suggesting high levels of demographic independence). At Nm values > 1, it is the effect of dispersal and gene flow which most influences population structure. While the island model upon which this relationship is based involves numerous simplifying assumptions which are unlikely to be met under natural conditions (Bossart and Prowell 1998; Whitlock and McCauley 1999), the relationship between Nm and F t remains a useful method of inferring the s  average effects of migration on regional patterns of genetic subdivision over longer time frames (Slatkin and Barton 1989; Neigel 2002; Pearse and Crandall 2004). More recent advances in population genetic theory based on the coalescent (Kingman 1982; Tavare 1984; Hudson 1998; Beerli and Felsenstein 1999; Hey and Nielsen 2004) have allowed for the development of more sophisticated investigation of population structure. Even these methods, however, are currently based on relatively simple models which often suffer from unrealistic assumptions (e.g., - constant population size or migration rates). Furthermore, they often involve intensive computational requirements which may be prohibitive when more than a few populations are being considered (Pearse and Crandall 2004; but see Steele and Storfer 2006). An alternative approach is embodied in the various individual-based inference methods which attempt to assign individuals to their most likely population of origin based on individual multilocus genotypes and estimates of alleles frequencies in putative source populations (Paetkau et al. 1995; Rannala and Mountain 1997; Waser and Strobeck 1998; Cornuet et al. 1999; Davies et al. 1999; Pritchard et al. 2000; Wilson and Rannala 2003; Manel et al. 2005). Unlike summary methods based on Fstatistics, individual-based 'assignment tests' are able to capture detailed information about individual patterns of dispersal among populations while requiring relatively few assumptions about underlying dispersal models. They have been used widely to estimate various aspects of movement and gene flow between populations, including the composition and origins of migrants in admixed populations (Fraser et al. 2005; Taylor and Costello 2006) and whether, for example, dispersal is sex-biased (Blundell et al. 2002; Fraser et al. 2004). They have been further used in the context of parentage analysis to determine the reproductive success of immigrants in a  129  population (Jones and Ardren 2003; McLean et al. 2004) as well as the identification of interspecific hybrids (Anderson and Thompson 2002). One disadvantage of these assignment methods, however, is that they typically give data only about recent patterns of movement (i.e. - within the past few generations) and, therefore, provide complementary information to that inferred from F or coalescent approaches. Used st  in combination, however, these methods may allow for an understanding of the relative importance of both historical and contemporary gene flow; information that is ultimately required to answer any questions related to evolutionary change or those concerning the conservation of species biodiversity.  4.1.3  Coastal cutthroat trout as a model paradox  A useful species for addressing some of these general issues is the coastal cutthroat trout {Oncorhynchus clarkii clarki). The coastal cutthroat trout is a salmonid species native to the west coast of North America which tends to occur in small (generally < 50 -100 individuals) and geographically structured populations (Trotter 1987; Behnke 2002). They occupy a wide range of habitats and can be found from small streams directly entering the sea, to estuaries, larger rivers and their tributaries, to sloughs, ponds and lakes, as well as isolated headwater streams where their relatively small size at maturity allows them to utilize smaller stream reaches than other salmonids along the coast (Trotter 1989). Correspondingly perhaps, coastal cutthroat trout exhibit one of the broadest and most variable spectra of migratory behaviors of all the salmonids (Northcote 1997; Johnson et al. 1999). During their lifetime, coastal cutthroat trout will undergo a series of different types of movement: seasonal movements (feeding, overwintering), spawning runs, and those associated with ontogenic life history shifts. Fluvial (river-migratory), adfluvial (migrating between lakes and rivers), and resident forms (non-migratory) are present throughout their range (often within the same population), while anadromous or diadromous forms (migrating to and from saltwater) exist along the coast where access to the sea is available.Most available information is limited to the anadromous component of populations (the component primarily discussed throughout this chapter).  Like many other salmonids, coastal cutthroat trout demonstrate a well-developed  130  ability to 'home' or return to their natal stream to breed (Quinn 1993; Dittman and Quinn 1996). Unlike salmon and other anadromous trout, coastal cutthroat trout do not make extensive use of rich offshore feeding areas. Instead, marine migration appears limited to 2-3 months during the summer when feeding forays may be largely confined to bays and estuaries in the vicinity of their natal stream, especially for males (Giger 1972; Behnke 1992). There is considerable evidence to suggest that females are the more migrant sex (Giger 1972; Massey 1984; Wenburg and Bentzen 2001; Chapter 3). Female salmonids, which often maximize their reproductive fitness by attaining large size and greater fecundities, would likely benefit to a greater extent from extended saltwater feeding than would males (Gross 1987; Northcote 1997; McDowall 2001; Hendry et al. 2004a). In Puget Sound, Washington, anadromous coastal cutthroat trout feed and migrate along the beaches, mostly in water less than 3 m deep. They do not appear to cross large bodies of open water but travel quickly along nearshore intertidal areas, moving only into open coastal waters where extensive inlets and fiords are absent (Johnston 1982; Pearcy 1997; Trotter 1997). This does not necessarily imply that they are unable to make longer migrations. Tagging studies in Oregon and Alaska, for example, suggest that anadromous movements in coastal cutthroat trout may regularly occur to ~ 70 km (Sumner 1952, 1962; Giger 1972; Jones 1977) and fish have been documented to move by as much as 290 km from where they had been tagged (Pearcy et al. 1990). Individuals are known to feed and overwinter in non-natal streams, however, relatively few studies have addressed reproductive straying rates (i.e. - dispersal) among coastal cutthroat trout (but see Wenburg and Bentzen 2001). Fewer still have investigated the genetic population structure of the subspecies. Despite apparently ample opportunity for gene flow, distinct population segments may form over smaller geographic distances than in other salmonid species and there is little evidence of actual gene flow among even adjacent populations. One of the first studies to examine population structure in coastal cutthroat trout surveyed allozyme variation in 21 anadromous populations from the Hood Canal and North Puget Sound in Washington. Campton and Utter (1987) found that coastal cutthroat trout populations were highly structured genetically with adjacent streams  131 (separated by as little as a few kilometers) supporting reproductively isolated groups. While nearly 9 5 % of the total variation was contained within individual populations, strong genetic differences between different regions of Puget Sound suggested some manner of limited gene flow between regions. Wenburg et a l , (1998) reached similar results in the same general area. Examining variation at six microsatellite loci in 13 populations from Washington, the authors found populations to be ultimately structured at the level of individual streams, with each being differentiated in terms of allele frequencies. Significant regional differences were found to exist, but the study was limited in the number of populations considered and perhaps performed on too broad a geographic scale (adjacent populations were not closer than 30 km). Subsequent work by Wenburg and Bentzen (2001) targeted populations separated over a smaller geographic range (2 - 70 km) in the Hood Canal, Washington. Again, using microsatellite loci, populations separated by as little as 2 km were found to be highly differentiated. Several minor studies (e.g. - Zimmerman et al. 1997; Johnson et al. 1999; Williams 2004) have reached similar conclusions. This is a much finer degree of population subdivision than observed in other salmonids. Pacific salmon and anadromous rainbow trout (O. mykiss) populations, for example, show little or no significant variation among the tributaries of larger river systems, at least at allozyme loci (Johnson et al. 1999). To date, there remains some question as to how this level of population subdivision is maintained. It may be that natal philopatry or the tendency to 'home' is particularly well developed in coastal cutthroat trout. Coastal cutthroat trout do spend more time in freshwater than other salmonids (up to several years before going to sea, although 2-3 years is more typical in BC). In that homing precision expected to be inversely related to S W residency (Quinn 2005), it is possible that homing is particularly well developed in this salmonid. Alternatively, it may be that selection may be acting to counter reproductive straying between different groups. Significant adaptive differences often exist between adjacent salmonid populations in terms of important life history characteristics such as size at maturity and the timing of reproduction (Ricker 1972; Taylor 1991). In some cases, assortative mating or possible mismatch in the timing or nature of breeding between homing fish and strays does appear capable of presenting a significant barrier to gene flow between adjacent populations (e.g. - Tallman and Healey 1994;  132 Hendry and Day 2005). Interestingly, coastal cutthroat trout populations contain unusually high levels of genetic diversity given their apparent demographic independence and typically small population sizes (mean of -10 alleles per microsatellite locus and expected heterozygosities averaging 0.60 - 0.70; Zimmerman et al. 1997; Wenburg et al. 1998; Wenburg and Bentzen 2001). While several historical and contemporary factors likely contribute to the origin and maintenance of this diversity (see other chapters), considering what is known about the genetic structuring of this subspecies, it seems possible that strong genetic subdivision and occasional gene flow on the order of "one-migrant-per-generation" (OMPG) could be partially responsible for maintaining levels of genetic variation within populations. To address this idea and the impacts this would have on population structure, I sampled numerous coastal cutthroat trout populations from southwestern British Columbia, Canada (a relatively unstudied part of their range) to better delineate the scope of a typical coastal cutthroat trout population and the forces most influencing population structure in the wild. Using information contained in six polymorphic microsatellite loci, I estimate levels of intrapopulation genetic diversity as well as long term and contemporary levels of gene flow among 42 adjacent coastal cutthroat trout populations. Specifically, the goals of this study are to determine: (1) whether populations in British Columbia exhibit similarly high levels of genetic subdivision and diversity as observed in other areas to the south; (2) whether populations are exhibiting appreciable gene flow on the order of O M P G ; (3) whether these parameters vary among four geographic regions chosen for their unique combination of physio-geographic characteristics; and (4) whether there are distinct groups or metapopulations (cf. Hastings and Harrison 1994) in this area which could be used to define conservation units for the subspecies in Canada. The recognition of distinct groups is central to the conservation of biological diversity and is particularly important for a subspecies like coastal cutthroat trout which exhibits extensive phenotypic variation over a wide geographic area (Bernatchez and Wilson 1998; Bowen 1999). The coastal cutthroat trout is a salmonid of increasing conservation concern in western North America (Johnson et al. 1999; Costello and Rubidge 2005) and a better understanding of what  133 exactly constitutes a 'typical' coastal cutthroat trout population is urgently needed to better prioritize the allocation of conservation resources targeting this salmonid.  4.2  4.2.1  METHODS AND MATERIALS  Field sampling and DNA extraction  Between 2001 and 2003, I collected coastal cutthroat trout D N A samples from streams in the southwest portion of British Columbia as part of a conservation initiative targeting coastal cutthroat trout in the province. To address an information gap, sampling was conducted over smaller geographic scales than previous studies (targeting populations separated by 0.1 km to ~ 150 km within regions) in a loosely hierarchical fashion (i.e. -streams within bays, bays within regions). My sampling scheme was designed to include sites that differ in terms of their physical and hydrological regimes, focusing on areas that were separated by varying depths of water and by different degrees of shoreline complexity. Four primary areas were sampled around Vancouver Island and the BC mainland: (1)  Clayoquot S o u n d / Ucluelet Inlet - located on the west coast of  Vancouver Island near Tofino; generally a low elevation area with a convoluted shoreline characterized by extensive shallow mudflats or sandy beaches that are punctuated by numerous rocky islands and headlands. Streams in this area are often low gradient and meandering with extensive anadromous sections (i.e. - stream sections accessible to anadromous fishes). (2)  Strait of Juan de Fuca - located south of Barkley Sound on the west coast of Vancouver Island; has a relatively straight, uniform shoreline that is primarily rocky but punctuated by small pocket beaches of coarse, cobble type materials. There are few estuarine environments and the majority of creeks empty directly into the open Strait over rocky intertidal platforms. Anadromous stream sections are often short as the narrow coastline is  134  backed by cliffs up to 20 m in height. (3)  Strait of Georgia (West) - includes the highly-populated east coast of Vancouver Island between Nanaimo and Campbell River; semi-protected estuarine environment with low gradient streams extending over broad lowland plateaus. The shoreline in this area is relatively straight and uniform with extensive estuarine and shallow nearshore habitats. Several other streams to the immediate north along Discovery Passage were also included; that area has been less developed and is characterized by relatively deep but narrow channels subject to tidal forces.  (4)  Strait of Georgia (East) - includes populations on the BC mainland between Gibsons and Powell River as well as the east coast of Texada Island. The area is characterized by a fjord-like character with high-relief landforms and a convoluted shoreline of inlets, headlands and glacially deepened channels.  To give a regional perspective to the distribution of genetic variation among these areas, several streams were included from the Queen Charlotte Islands and Alaska (1000 - 2000 km to the north) for comparison. In all cases, sampling was performed on small stream systems (primarily 1 to 3 order) lacking a history of st  rd  hatchery stocking. In the majority of cases, sampling was targeted towards juveniles (young-of the year, parr, etc.) which are more likely to be in their natal streams. In some cases, however, adult or spawner samples were used (Sandhill Creek, Chonat Lake, Yakoun River, and Makaka Creek). Generally, all sampled populations were believed to be anadromous (and able to exhibit gene flow) with the exception of several streams located above migration barriers (e.g. - Staghorn Creek and an unnamed Kennedy lake tributary; Stella Lake, Ashlar Creek, Myers and Kleindale creeks, as well as Halfmoon Creek; see Table 4-1). A total of ~1955 individuals were collected from 57 streams. Collected adipose or pelvic fin samples were stored in 9 5 % ethanol until DNA could be isolated from ~ 5mg of tissue using the Puregene DNA isolation kit (Gentra Systems, Inc.).  135 4.2.2  Microsatellite amplification and removal of hybrids  A total of 13 microsatellite loci were screened for variability in a subset of individuals covering the study area. Six loci were eventually chosen for inclusion based on amplification quality when subjected to multiplex P C R and two triplex P C R amplifications based on those of Wenburg and Bentzen (2001) were developed (Table 4-2). Polymerase chain reactions were carried out using fluorescently-labeled primers in 10ul volumes of 10mM Tris-HCl (pH 8.3), 2mM MgCI , 0.8mM dNTP's, and 2  0.4 units of AmpliTaq Gold polymerase in MJ P T C 100 thermocyclers using a basic cycle profile of: 1 cycle (95°C/ 3 min), 10 cycles (94°C /1 min, T / 30 sec, 72°C 120 A  sec), 20 cycles (94°C / 30 s, T / 30 sec, 72°C / 20 sec), and 1 cycle (72°C / 5min), A  where T is the annealing temperature (see Table 4-2). P C R products were A  electrophoresed on ABI Genescan Gels using ABI Prism 377 sequencers and R O X labeled internal GS400HD size standards. Raw fragment analysis and allele binning was facilitated by the ABI G E N O T Y P E R v3.7 utility program. Throughout the study area, coastal cutthroat trout are known to coexist with sympatric populations of their sister species, rainbow trout (O. mykiss).  The two  species are similar morphologically and are often difficult to distinguish as juveniles (e.g., - McPhail and Carveth 1993; Pollard et al. 1997). Furthermore, both forms have been shown to hybridize under compromised environmental conditions (degraded habitat, introduced fishes; Campton and Utter 1985; Young et al. 2001; Docker et al. 2003). While many of the sampled streams were smaller than those typically inhabited by rainbow trout, naturally occurring rainbow trout were present at some of the sample sites. To avoid contamination of my sampling efforts, a series of morphological and genetic traits were used to remove any misidentified rainbow trout and putative hybrids prior to genetic analysis (see Chapter 2). Having removed any questionable samples and limiting the analysis to populations with a minimum of 15 coastal cutthroat trout, I present data here for a reduced subset of 1257 individuals in 42 'pure' coastal cutthroat trout stream samples (Table 4-1).  136 4.2.3  Genetic analysis  Basic statistics  Basic descriptive statistics of the genetic variation present at individual microsatellite loci and within populations were compiled using either the "Microsatellite Toolkit" macro for Excel (Park 2001) or the program F S T A T v2.9 (Goudet 2001). Diversity values included observed and expected heterozygosities, Nei's within-sample (H ) and total genetic diversity (H ), the total number of alleles s  t  (A ) and allelic richness (A ) determined using the rarefaction method of El Mousadik T  R  and Petit (1996), Wright's F-statistics (F and F ) and their associated confidence is  st  intervals were determined based on a permutation process using the programs G E N E T I X (Belkhiretal. 2001) or A R L E Q U I N ver. 3.1 (Excoffier et al. 2005). Tests for deviations from Hardy-Weinberg equilibrium were performed for each locuspopulation combination in G E N E P O P v3.3 (Raymond and Rousset 1995) using an exact test in which p-values were estimated using a Markov chain method. Tests for genotypic linkage disequilibrium for all combinations of locus pairs within a population were also made using a Markov chain method with G E N E P O P default values. Sequential Bonferroni corrections (Rice 1989) were applied to these calculations to maintain a nominal table-wide a = 0.05. Finally, the effective population size (N ) for sampled populations was estimated e  using the linkage disequilibrium method of Bartley et al. (1992) as implemented in the program NeEstimator v1.3 (Peel et al. 2004) with the empirical bias correction suggested by Waples (2006). A s the number of cohorts sampled varied among streams, I attempted to standardize estimates to reflect an average four cohort generation typical of cutthroat trout (see Chapter 3). Under the "salmon model" described by Waples (1990a, 2005), the effective population size (N ) has been found e  to be approximately equal to:  N *gxN e  b  (4.1)  where g is the generation length defined as the average age at maturity weighted by  137  age specific fecundities and N is the harmonic mean of N estimates across years b  b  making up the generation (Waples 2005). In the absence of data for the spawning populations contributing to my cutthroat samples, I assume equal representation across the 4 cohort generation to give the expected effective population size E(N ) as: e  E(N )K-g*e  (4.2)  where N is the effective number of breeders given by the linkage disequilibrium b  method and C is the number of cohorts sampled. Determination of population structure  Tests for population differentiation between all possible pairs of populations were performed for each locus and over all loci combined using Fisher's method in G E N E P O P with default values. I further used the Bayesian model-based clustering method of Corander et al. (2003) as implemented in the program B A P S v3.2 (Corander et al. 2004). The program treats both population allele frequencies and the number of distinct groups as random variables which can be modeled in a Bayesian framework. Like similar methods, the program assumes population clusters will exhibit Hardy-Weinberg and linkage equilibria as well as low migration rates, but differs from other such programs in that it can consider populations as the basic sampling unit (rather than individuals) and estimates which populations have differentiated allele frequencies rather than simply partitioning individuals into HardyWeinberg populations (reviewed by Pearse and Crandall 2004). Using a stochastic optimization algorithm, B A P S examines the posterior mode of the sampling solutions to determine the partition (i.e. - the number of distinct population clusters) with the greatest likelihood given the data.  A visual representation of the multilocus genetic variation existing among populations is provided by the Factorial Correspondence Analysis (FCA) plotting algorithm in GENETIX. A s well, a consensus maximum likelihood tree was constructed using the C O N T M L module of the PHYLIP genetics package (Felsenstein  138 1993) . C O N T M L estimates phylogenies from allelic frequency data under a model in which all divergence is due to genetic drift in the absence of new mutations. The unrooted consensus C O N T M L tree presented is based on 1000 bootstrapped datasets generated in the PHYLIP S E Q B O O T module. Bootstrap values greater than 5 0 % are indicated. Estimation of dispersal rates  A s noted, long-term dispersal rates between populations have traditionally been approximated using standard F-statistics (Weir and Cockerham 1984). At migrationdrift equilibrium, the effective number of migrants per generation (N m) is e  approximately equal to 1 - F / 4 in the equilibrium island model (Wright 1931, 1965). st  Here, estimates were made from F values calculated in A R L E Q U I N . Long-term gene st  flow was also estimated using the private alleles method of Slatkin (1985) and Barton and Slatkin (1986) as implemented in G E N E P O P . This method equates the average frequency of alleles that are found in only one population, [p(1)], to Nm through the analytically-derived equation l o g [p(1)] ~ a log (Nm) + b where a and b are 10  10  constants dependent on sample size. Both methods have been found to be robust and give reasonably good estimates of long term patterns of gene flow under a wide range of scenarios (e.g. - Slatkin and Barton 1989). Estimates of recent migration rates (over the past few generations) were made using the method of Wilson and Rannala (2003) as implemented in the program B A Y E S A S S v1.3. This method has more in common with assignment-based analyses in that it seeks to identify recent immigrants and their descendents based on the transient linkage disequilibrium observed in admixed populations (e.g. - Briscoe et al. 1994) . B A Y E S A S S simultaneously estimates migration rates, inbreeding coefficients, and population allele frequencies (having removed inferred immigrants prior to calculation). While it is limited to estimating m (the proportion of migrant individuals in each generation) rather than Nm (the total number which may have immigrated), the method makes far fewer assumptions about underlying population structure than previous methods and is one of the first non-equilibrium approaches that does not require loci to be in Hardy-Weinberg equilibrium. Using default settings, I performed  139 the B A Y E S A S S analysis for each of the four primary regions separately, averaging the mean migration rate over 3 successive runs from different seed values. I allowed for a burn-in period of 1 million iterations followed by 3 million iterations with a thinning frequency of 2000 (data collection). Statistical significance of the inferred immigration rates was determined against the 9 5 % confidence intervals generated by B A Y E S A S S under the null assumption. 4.2.4  Spatial and hierarchical partitioning of genetic diversity  To complement Bayesian analyses, I also examined the spatial partitioning of genetic variation in coastal cutthroat trout populations at variously nested levels. First, using the analysis of molecular variance (AMOVA) approach of Excoffier et al. (1992) as implemented in the program A R L E Q U I N , I determined the percentage of the total genetic variation explained by genetic differences within populations  (VIP),  among populations within groups (VIG), and by differences between groups ( V B G ) under a variety of relevant hypotheses. For example, I grouped populations according to geographical region (Clayoquot Sound vs. Strait of Juan de Fuca, etc.), according to the clustering suggested by Bayesian analysis, and by barriers. Previous work I have done in bull trout (Salvelinus confluentus) has shown that the location of barriers may play an important role in structuring genetic diversity among salmonid populations (Costello et al. 2003). I separated populations isolated above and below migration barriers, both collectively, and by individual barrier. Finally, because this subspecies is of conservation concern, I partition the distribution of genetic variation by conservation/ management units to determine how well they coincide with natural population structuring in coastal cutthroat trout. Here, I group populations by provincial management units (Regions 1 - Vancouver Island, 2 - Lower B C mainland, and 6 - Skeena/ North Coast) as well as by tentative designatable units under Canada's Species at Risk Act (SARA; see Costello and.Rubidge 2005).  Second, the tendency towards natal philopatry and limited anadromous movement in coastal cutthroat trout suggests that anadromous movement is likely restricted to adjacent populations and follows a stepping stone model of dispersal (Kimura and Weiss 1964). Such a pattern of dispersal is expected to lead to a pattern  140  of 'isolation by distance' which is characterized by increasing genetic differentiation among populations with increasing geographic distance (Wright 1943; Slatkin 1993). To test for isolation by distance, the Mantel test option in the M S Excel macro G e n A l E X v.6 (Peakall and Smouse 2005) was used to compare the two distance matrices (genetic and geographic), giving a correlation coefficient (r) and its statistical significance based on a permutation process. At migration-drift equilibrium, scatter (i.e., residuals from a fitted linear regression line) should increase with increased geographic separation as drift, rather than gene flow, becomes the dominant force. To determine whether populations have yet reached a drift-migration equilibrium, I applied the approach of Hutchison and Templeton (1999) whereby, subsequent to a significant Mantel test result between genetic and geographic distances, a second Mantel test was performed using residuals from the initial fitted line (calculated using MS E X C E L ) against geographic distance. Two genetic distances were analyzed: pairwise F t values were computed in s  A R L E Q U I N and Cavalli-Sforza and Edward's (1967) chord distances (CSE) values were generated in the G E N D I S T program of the PHYLIP ver. 3.5. The latter metric is purely drift-based and should outperform measures such as Nei's D (which are based on mutational processes) in recently diverged populations (Goldstein et al. 1995; Paetkau et al. 1997). Three different geographic distances (km) between each pair of sample sites was determined using the geographic information system program ArcView ver. 3.2 (Environmental Systems Research Institute): (1) Euclidean (straightline distance); (2) shortest water distance; and (3) shortest water distance limited to depths less than 300 m. It has been suggested that depths > 300 m may act as a barrier to anadromous movement by coastal cutthroat trout (Johnston 1982; Pearcy 1997). The shortest water distance (SWD) between populations was found to give the best fit and will be the only geographic distance measure discussed at length. Finally, I employed spatial autocorrelation analyses to further analyze patterns of genetic variability in coastal cutthroat trout. Essentially, the concept is similar to the Mantel tests described above. Autocorrelation, however, provides a better description of spatial patterns than do traditional Mantel statistics by partitioning geographic distances into discrete classes which allow for the calculation of different correlation  141 coefficients (r) at different geographic distances. The method has been shown to be particularly useful in delineating the scope of intraspecific conservation units among continuously distributed populations (Diniz-Filho and de Campos Telles 2002). Under restricted gene flow and in the absence of selection, adjacent populations will often be characterized by positive spatial genetic autocorrelation at short distance classes; with increasing geographic distance, the relationship will subsequently decline through zero near the limits of dispersal (or sampling limits) and often become negative at the highest distance classes (Sokal and Wallenberg 1983; Epperson and Li 1997). A s such, autocorrelation analysis provides an estimate of the geographical extent over which dispersal or metapopulation processes may be operating (Peakall e t a l . 2003). Using the same distance matrices described for Mantel testing, I employed the spatial autocorrelation method of Smouse and Peakall (1999) as implemented in the Excel macro GenAlEx v.6 to perform a multi-distance class analysis (MDC) of the relationships between genetic and geographic distances. Using a basic distance class width of 10 km, this analysis combines successive distance classes in cumulative groups when determining correlation coefficients (i.e., from 0-10 km, from 0-20 km, etc.). The statistical significance of correlations in each distance class was determined using a one-tailed probability test based on 10,000 random permutations. Spatial autocorrelation was performed for each region independently to determine if the spatial scale of population subdivision differs among the four primary regions.  4.3 4.3.1  RESULTS  Levels of genetic variability  Relatively high levels of genetic diversity were observed at the six microsatellite loci (Table 4-2). Across all populations, the number of alleles at individual loci ranged from 7 (Oneul 1) to 27 (Ogo4); Nei's gene diversity (i.e. - heterozygosity) averaged 0.62 within populations and 0.83 across all populations. Tests for conformity to Hardy-Weinberg equilibrium indicated no instances of heterozygote excess at any of  142 the microsatellite loci; tests for heterozygote deficit were statistically significant in 24 of 252 cases across loci (9.5%) following sequential Bonferroni adjustment (nominal a = 0.05). Ten of these cases were observed at O n e u l 1, 6 at Omy77, 5 at Sfo8, and one case at each of the remaining loci. Tests for genotypic linkage disequilibrium rejected the null hypothesis of independence in 28 of 1260 comparisons (2.2%) but significant results were not restricted to any single locus pair. Interestingly, the Hoard Creek population (Location 11) showed significant linkage disequilibrium at all 15 pairwise comparisons and heterozygote deficiencies at 4 of the 6 loci. This may be consistent with a limited number of spawning events or "family sampling" as all samples were of the same cohort from a single pool (Hansen et al. 1997). Levels of variation within individual populations were also generally high, but varied considerably between sites. Observed heterozygosity, for example, averaged 0.58 but ranged from a low of 0.05 in the Ashlar Creek population (Location 20) to a high of 0.79 in Staghorn Creek (Location 4; Table 4-1). Similarly, expected heterozygosity ranged from 0.04 to 0.76 with a mean of 0.62. Allelic diversity was moderate with an average of 31 total alleles per population across the 6 loci (ranging from 8 to 40). To account for uneven sample sizes, allelic richness (determined for a minimum number of 15 individuals) averaged 4.6 alleles per locus (ranging from 1.3 to 6.3). Significant inbreeding coefficients (F ) were generally positive, but varied on a is  case-by-case basis. Four cohort estimates of expected effective population size, E(N ), varied considerably from less than 1 to in excess of 400. Particularly high e  values were noted for the Chonat Lake and Chef Creek samples; the former being consistent with mark-recapture estimates made in Chapter 3. Clearly, measures of within-population genetic diversity appear to differ among the sampled regions. For example, both allelic diversity and heterozygosity were highest in Clayoquot Sound (Table 4-3). Allelic diversity was lowest in populations on the east side of the Strait of Georgia while the level of inbreeding appeared greatest in the Strait of Juan de Fuca samples. The degree of population subdivision measured by F t was moderate and s  similar in all regions, ranging from 0.16 in Clayoquot Sound to 0.25 along the western side of the Strait of Georgia.  143 4.3.2 Genetic population structure  All populations sampled here were found to be significantly differentiated from one another in terms of microsatellite allele frequencies (Fisher's contingency method, p < 0.0001). Even populations in adjacent streams or those sharing a common confluence to saltwater (e.g. - Kleindale and Myers creeks) were significantly differentiated. The discriminatory power of individual microsatellite loci, however, varied across the 861 pairwise population tests, from a high of 9 9 . 9 % for Omy77 to only 7 5 . 4 % at Oneul 1. The distinct nature of individual streams was further supported by Bayesian clustering analysis and the majority of the sampled coastal cutthroat trout streams (32/42) cluster individually. There were, however, indications of some population clustering: the unnamed Kennedy Lake tributary, Staghorn Creek, and Sandhill Creek (Locations 3-5) formed a distinct cluster as did the Botanical Beach and Tom Baird creeks (Locations 9, 10); Waterloo, Chef, and Thames creeks (Locations 22, 24, 25) formed a distinct population cluster as did the Mamin and Yakoun rivers (Locations 40, 41). With the exception of Sandhill Creek, populations in each of these clusters are geographically adjacent to one another. Despite the marked differences between individual populations, all appear to group according to geographical proximity at the regional level. The factorial correspondence analysis (FCA) of the multilocus genetic relationships between coastal cutthroat trout populations appears in Figure 4-2. Displayed are the first two factorial axes which together account for 19.82% of the variation among populations (the 3 and 4 factorial axes contain only slightly less variation). Populations in rd  th  Clayoquot Sound appear to be the most distinct with a tight grouping of the populations and slightly more distant grouping of populations from Ucluelet Inlet (Locations 6 and 7). These populations cluster more closely to the Queen Charlotte Island populations than they do to the more geographically proximate Juan de Fuca or Strait of Georgia populations. The east and west sides of the Straight of Georgia appear to cluster as a single geographic group, but with some subtle subgroupings. Many of the populations along Discovery Passage (Locations 16-21) are separate from the main distribution of populations and appear more widely scattered. This is consistent with the drift-based consensus maximum likelihood tree (Figure 4-3). While  144  bootstrap support was generally jow, treelopology very closely parallels geographic location. Again, populations in Clayoquot Sound cluster more closely with populations in the Queen Charlottes than they do to the more proximate Vancouver Island locations. Bootstrap support for this particular grouping was high (occurring in 9 8 % of the 1000 ML trees).  4.3.3  Estimates of dispersal rates  Regional estimates of long-term dispersal among populations (Nm; Table 4-3) indicated that, in most cases, both the Fst-derived value and the value calculated by the private-allele method were compatible; the former ranged from 0.8 effective migrants per generation on the west side of the Strait of Georgia to 1.3 in Clayoquot Sound, the latter ranged from 0.4 in the Strait of Juan de Fuca to 1.8 on the west side of the Strait of Georgia. Private allele estimates of Nm were generally less than or equal to the F t-derived value with the exception of the Strait of Georgia (West) s  populations where the value was more than twice the F  st  estimate. While average  regional levels of Nm appear on the order of one migrant per generation, pairwise Nm estimates between individual populations showed considerably more variation, as might be expected given the higher variance at low F  st  values (e.g., - Whitlock and  McCauley 1999). Across all regions, pairwise values between populations were generally less than 2, but ranged as high as 11 -14 effective migrants per generation (between Chef, McNaughton, Thames creeks; and between Kennedy Lake tributary and Sandhill Creek; data not shown). Estimates of recent migration rates provided by Bayesian individual assignment generally suggest low levels of contemporary migration within regions. Consistent with the levels of population structure and differentiation described in the previous section, there was evidence of recent immigration in only 11 of the 38 populations (29%; Table 4-4). The percentage of streams showing signs of immigration was greatest in Clayoquot Sound where four of the seven populations show signs of recent admixture (Meares Creek, the unnamed Kennedy Lake tributary, Staghorn Creek, and Sandhill Creek). This is followed by three of the eight populations along the Strait of Juan de Fuca (Botanical Beach, Tom Baird, and Hoard Creek, although  145 the finding of immigrants in Hoard Creek may be explained by the high level of linkage disequilibrium observed in that population). Four of twelve populations on the west side of the Strait of Georgia and none of the eleven populations on the east side of the Strait of Georgia showed signs of recent immigration. Actual migration rates, however, appear to be greatest along the western side of the Strait of Georgia (as high as 2 6 % from Stella Lake into Pye Lake and 1 8 % between Thames and McNaughton creeks). In some cases, the migration rates are roughly symmetrical between populations (e.g. - Botanical Beach and Tom Baird creeks) while in others, the rates are highly asymmetrical (e.g., - Stella Lake and Pye Lake where upstream migration from Pye Lake into Stella Lake is prevented by a one-way barrier to migration).  Spatial and hierarchical partitioning of genetic diversity  4.3.4  Results from A M O V A partitioning of the genetic variability found in these populations suggest that regional differences, while significant, explain a relatively minor proportion of the variation existing among populations (V p = V G + V| Table 4A  B  G;  5). Grouping populations by the five regions defined in Table 4-1 explained just 3 0 % of V p (~8 % of the total genetic variation). In fact, the variation existing within A  geographical regions is nearly three times as much as that existing between them. Combining populations along the Strait of Georgia with those along the Strait of Juan de Fuca (these areas appear to have been colonized by the same Columbia source population during deglaciation; see Chapter 5) marginally improved the explanatory power of the regional grouping. More variation, however, was explained by simply partitioning populations into a collective anadromous group versus populations isolated above different migration barriers (39% of V A P ) . Similarly, groupings based on provincial fisheries regions and proposed Designatable Units under SARA captured very little of the genetic diversity present in populations (just 14 and 24 % of V P , A  respectively). The grouping scenario which explained the greatest proportion of the variation among populations was that suggested by Bayesian clustering analysis in B A P S ( 8 3 % of V p ) . Under this scenario, the majority of populations are grouped A  singly, further illustrating the distinctiveness of each coastal cutthroat trout stream population.  146  Simple geographic distance between populations appears to offer a better description of their population structure than does regional affiliation. Pairwise genetic distances (F  st  and CSE) both showed a significant positive relationship with the  shortest water distance between populations (SWD in km) across the four primary regions (Figure 4-4) The relationship between F t and S W D improves significantly s  once above barrier populations are removed, but C S E continued to show a greater degree of correlation. Across the four regions, r = 0.62 between C S E and shortest water distance and r= 0.67 once above barrier populations were removed (both p < 0.001). Similar positive relationships were apparent for analyses performed on the four regions individually (Figure 4-5). It is clear, however, that each region differs in the degree to which isolation by distance has developed (i.e. - the slope and yjntercept for linear regression lines clearly differ between regions). The most notable difference exists between the Strait of Juan de Fuca populations and other regions. Populations in this area show a higher level of divergence over small distances and the increase in genetic divergence with geographic distance is 4-8 times that observed in the other regions. In no case, however, (within regions or across all regions) did I find evidence that populations had reached migration-drift equilibrium. Residual scatter from a linear regression line did not increase with distance as predicted at equilibrium (e.g. - (Slatkin 1993; Hutchison and Templeton 1999).  While all regions show strong positive spatial autocorrelation up to 20 km (and to at least 30 km in the Strait of Juan de Fuca), multi-distance class analysis using a 10 km distance class width suggests that the spatial extent of autocorrelation differs between the regions (Figure 4-6). First, the correlation coefficient r at small geographic distances (0-10 km) differs between regions (r= 0.27 in Clayoquot Sound, -0.20 in the Strait of Juan de Fuca and the Strait of Georgia (East), and 0.1 in Strait of Georgia (West). Second, in both the Strait of Juan de Fuca and the Strait of Georgia (East) populations, detectable correlation between genetic and geographic distances extends over approximately 30 - 40 km. In Clayoquot Sound, detectable correlation extends up to 60 - 70 km although one distance class (0-40 km) was marginally insignificant. Finally, while there is a lack of data for populations in the 60 to 100 km distance classes among the Strait of Georgia (West) populations, positive  147  correlation in the 0-100 km distance class suggests that detectable autocorrelation in that region may extend up to 110 km.  4.4 DISCUSSION  4.4.1  Levels of intrapopulation diversity  Levels of genetic variation in the coastal cutthroat trout populations sampled here (mean number of alleles per locus = 5.1, H = 0.62) are marginally lower but on e  par with the values identified in previous studies. Wenburg et al. (1998), for example, found the mean number of alleles per microsatellite locus ranged from 8-12 (and heterozygosity averaged 0.67) among 13 populations in Washington. Subsequent work by Wenburg and Bentzen (2001) over smaller spatial scales in the same general area (populations 2 - 100 km distant) found similar amounts of genetic diversity (average of 9 alleles per locus per population, H = 0.69). These levels are e  intermediate between the mean values reported for a number of freshwater and marine fish species (at least in terms of heterozygosity) and are consistent with a species showing limited anadromous dispersal. DeWoody and Avise (2000), for example, found expected heterozygosity in a number of freshwater fish species to average 0.54 while the value was 0.68 in marine species (mean number of alleles averaged 9.1 and 10.8, respectively). The greater diversity among marine fish likely results from the greater potential for gene flow and increased effective population sizes associated with marine residency. The discrete and often patchy nature of freshwater habitats often lead to a high degree of reproductive isolation and population structure among freshwater fishes with respect to marine environments where suitable habitats may be continuously distributed (e.g., - Mitton and Lewis 1989). The slightly lower values in B C (particularly in terms of allelic diversity) may reflect differences in the relative sizes of the creeks surveyed or the life history stage at which sampling occurred between studies. Most creeks sampled here were small (1 and 2 st  n d  order) in which juveniles cohorts (young-of-the-year and parr) were the  148 life-history stage most often sampled. Streams sampled by the cited Wenburg studies tended to be somewhat larger and it was the smolt stage (i.e., - stage associated with initial entry into saltwater) which was most often sampled. A s smolting in coastal cutthroat trout may occur at a range of ages (Giger 1972; Trotter 1989), smolt samples would likely include a greater number of cohorts and possibly draw on several distinct subpopulations (increasing the effective population size over which sampling occurred). Alternatively, the lower allelic diversity values for BC populations may also reflect the historical effects of postglacial dispersal from glacial refugia. Most currently occupied habitats in British Columbia were unavailable to fish populations during the late Pleistocene so that the majority of extant populations in this area are descended from postglacial immigrants which survived glaciation elsewhere (McPhail and Lindsey 1986; see Chapter 5). Population bottlenecking and the founder-flush cycles experienced during recolonization would tend to reduce allelic diversity to a greater extent than it would heterozygosity levels (Wade and McCauley 1988; Bernatchez and Wilson 1998) and has been well documented for fish species in this region (e.g. - McCusker et al. 2000; Costello et al. 2003).  It is clear that substantial variation exists between populations in terms of the levels of genetic diversity exhibited. Some of this may be explained by the presence of migration barriers between populations. The above-barrier Ashlar Creek population, for example, had heterozygosity levels that were an order of magnitude less than levels exhibited by below-barrier populations, as well as substantially reduced allelic diversity ( A = 1.3 vs. average of 4.6). Similarly, Stella Lake (isolated R  upstream of Pye Lake by a migration barrier) showed substantially lower levels of diversity relative to the latter (AR = 1.7 vs. 3.2; H = 0.17 vs. 0.46). The reduction in 0  genetic diversity in isolated populations has been well demonstrated for many species, including coastal cutthroat trout; the result is typically one of reduced effective population size and the heightened influence of genetic drift. Not all abovebarrier populations, however, showed reduced levels of genetic diversity. For example, Staghorn Creek and the unnamed Kennedy Lake tributary (isolated above Kenn Falls on the Kennedy Lake outlet river) showed some of the highest levels of diversity observed in this study (H = 0.79 and 0.69; A = 6.3 and 5.5, respectively). 0  R  Kennedy Lake is the largest natural lake system on Vancouver Island and likely  149  contains a sufficient diversity of habitats to support a large coastal cutthroat trout population, reducing the effects of drift and the loss of allelic diversity. This points to the possible influence of habitat characteristics (e.g. - patch size, heterogeneity) on population size and the maintenance of genetic variation in small populations (e.g., Johnson et al. 1992; Angers et al. 1999). While regional differences do exist among populations, they generally explain only a relatively minor portion of the total variation existing among populations (just 27 % of V P or ~8 % of the total variation examined in this study). Campton and Utter A  (1987) and Williams (2004), found similar values for allozyme loci; regional differences were able to account for 21 % of V P in Puget Sound (1.24 % of th