Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Molecular evidence of current and historical introgressive hybridization between bull trout (Salvelinus… Redenbach, Zoë 2000

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

Item Metadata

Download

Media
831-ubc_2000-0543.pdf [ 7.87MB ]
Metadata
JSON: 831-1.0089603.json
JSON-LD: 831-1.0089603-ld.json
RDF/XML (Pretty): 831-1.0089603-rdf.xml
RDF/JSON: 831-1.0089603-rdf.json
Turtle: 831-1.0089603-turtle.txt
N-Triples: 831-1.0089603-rdf-ntriples.txt
Original Record: 831-1.0089603-source.json
Full Text
831-1.0089603-fulltext.txt
Citation
831-1.0089603.ris

Full Text

Molecular evidence of current and historical introgressive hybridization between bull trout (Salvelinus confluentus) and Dolly Varden (S. malma) by Z O E R E D E N B A C H B . S c , University of British Columbia, 1997 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Department of Zoology) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A August 2000 © Zoe Redenbach, 2000 In p resen t i ng this thesis in partial fu l f i lment of the requ i remen ts for an a d v a n c e d d e g r e e at the Univers i ty of Brit ish C o l u m b i a , I agree that the Library shal l m a k e it f reely avai lable fo r re fe rence and s tudy. I fur ther agree that p e r m i s s i o n for ex tens ive c o p y i n g o f th is thes is fo r scho lar ly p u r p o s e s m a y b e g ran ted by the h e a d o f m y depa r tmen t o r by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g o r pub l i ca t i on of this thesis for f inancial gain shall no t be a l l o w e d w i t hou t my wr i t ten p e r m i s s i o n . D e p a r t m e n t of T h e Univers i ty of Brit ish C o l u m b i a V a n c o u v e r , C a n a d a D E - 6 (2/88) A B S T R A C T Natural hybridization, once thought to be relatively rare, has been widely observed in both plants and animals. Studies of species that hybridize at low rates across wide ranges, however, are still relatively rare. Dol ly Varden (Salvelinus malma) and bull trout (S. confluentus) have widely overlapping ranges and have been shown to hybridize at two localities. This thesis is a molecular analysis of Dol ly Varden / bull trout hybridization over a range of spatial scales. Large-scale m t D N A phylogeography revealed that Dol ly Varden survived the last glaciation in a previously unsuspected refuge south of the ice sheet, which resulted in continuous sympatry of a Dol ly Varden refugial population with bull trout over the last 100,000 years. Discordant mitochondrial and nuclear phylogenies revealed historical introgression of bull trout mitochondrial D N A into Dol ly Varden. Population samples revealed widespread local sympatry and current hybridization throughout the overlapping range, with hybrids consisting of 0 to 25% of the local samples. A detailed analysis of hybridization within a single watershed in north-central British Columbia revealed that individuals of hybrid origin form 9% of the juvenile population of char (0.5% F i , 0.5% F n , and 8% backcross genotypes). Initial interspecific hybridization is unidirectional, Do l ly Varden male by bull trout female, likely attributable to a 'sneaking' mating strategy used by smaller Dol ly Varden males. The F i hybrids were fertile and backcrossed with both parental species. The introgression of nuclear and mitochondrial alleles was asymmetrical, with bull trout m t D N A and Growth Hormone 2 introgressing into Dol ly Varden at higher levels than both the three other nuclear loci and the reverse direction introgression of Dol ly Varden alleles into bull trout, suggesting that the strength of selection can vary across nuclear loci and between species. A s Dol ly Varden and bull trout appear to be distinct species in sympatry, despite introgressive hybridization, a balance between gene flow and selection must be in place. I suggest that selection acts primarily against the adult hybrid population, as the morphological and behavioural intermediacy of hybrid and backcross individuals as adults would affect their potential reproductive success and their ability to succeed in either of the two alternate life-histories bull trout and Dol ly Varden adopt in sympatry (adfluvial vs. stream-resident). i i TABLE OF CONTENTS Abstract ii Table of Contents iii List of Tables v List of Figures vii Acknowledgements viii Chapter 1: General Introduction 1 Chapter 2: Phylogeography of Dolly Varden: evidence for a southern glacial refuge and historical introgression of bull trout mitochondrial DNA 11 Introduction 12 Materials and Methods 16 Results 23 Discussion 37 Chapter 3: Widespread local sympatry and evidence for hybridization between Dolly Varden and bull trout throughout British Columbia 47 Introduction 48 Materials and Methods 50 Results and Discussion 52 Chapter 4: Directional hybridization, bi-directional backcrossing, and the asymmetrical introgression of alleles in a Dolly Varden / bull trout hybrid zone 62 Introduction 63 Materials and Methods 66 Results 78 i i i Discussion 97 Chapter 5: General Discussion 115 References 124 Appendix 1: Mitochondrial DNA ND1 sequences 134 Appendix 2: Growth Hormone 2 Intron C sequences.... 136 Appendix 3: Ribosomal DNA ITS-1 sequences 137 Appendix 4: Genotypes by age class and by tributary 138 Appendix 5: Bull trout allele frequencies by age class and by tributary 139 Appendix 6: Hybrid genotypes by age class and by tributary 140 iv LIST OF TABLES Table 1: Salvelinus sample sites for phylogenetic and hybridization analysis .....17 Table 2: Molecular markers ; 20 Table 3: P C R conditions 21 Table 4: Sequence divergence between Salvelinus m t D N A haplotypes, and haplotype and nucleotide diversity within and nucleotide divergence between bull trout and Dol ly Varden populations 31 Table 5: Sequence divergence between Salvelinus Growth Hormone 2 intron C sequences and ribosomal D N A first internal transcribed sequences 35 Table 6: Hybridization, species proportions, m t D N A clade by sample site 53 Table 7: Inbreeding deficits for char samples from across B . C . 60 Table 8: Genotypes of juvenile char collected in tributaries to Thutade Lake Watershed 79 Table 9: Visual versus genetic identification of juvenile char 81 Table 10: B u l l trout allele frequency and percentage of hybrid origins, by tributary ...82 Table 11: Inbreeding deficits by tributary 84 Table 12: Gametic phase disequilibria by tributary 85 Table 13: Cytonuclear disequilibria by tributary 86 v Table 14: Introgression of nuclear and mitochondrial alleles by tributary 88 Table 15: Static and dynamic cohort analysis 92 vi LIST O F FIGURES Figure 1: Parapatric species distributions of Dol ly Varden (5. malma) and bull trout (5. confluentus) 7 Figure 2: Phylogenetic tree of Salvelinus m t D N A sequences 24 Figure 3: Geographic distribution of Dol ly Varden and bull trout m t D N A clades 27 Figure 4: Phylogenetic tree of Salvelinus Growth Hormone 2 intron C sequences 33 Figure 5: Phylogenetic tree of Salvelinus ribosomal D N A first internal transcribed sequences .36 Figure 6: Distribution of hybridization amongst sympatric sample sites, and location of Thutade Lake watershed 55 Figure 7: Thutade Lake watershed sample sites and distribution of hybrids 67 Figure 8: Kemess Creek sample sites and distribution of groundwater upwelling (preferred Dol ly Varden spawning sites) 69 Figure 9: Introgression of nuclear and m t D N A alleles between Dol ly Varden and bull trout 90 Figure 10: Distribution of Dol ly Varden, bull trout and hybrid fry and juveniles in Kemess Creek 95 vi i A C K N O W L E D G E M E N T S M y supervisor, Dr. Eric Taylor, guided me through my undergraduate Honour's thesis so skillfully that I knew a Master's degree spent in his lab would be time very well spent. I was right. His door was always open and his enthusiasm unparalleled. His financial and moral support allowed me the unforgettable experience of presenting my research at a European conference. Perhaps most importantly, he always gave thoughtful consideration to my questions and queries. It is thanks to him that my time as a graduate student was so intellectually satisfying. M y supervisory committee, Lee Gass, JD McPhai l , Eric Taylor, and M i k e Whitlock, hashed over my proposal and made many insightful suggestions over the drafts of this thesis. They helped to guide my research and create the final thesis that is before you. M y too-short field season was spent in the company of Gordon Haas and John Hagen. Gordon's knowledge of char was an oft-visited source of information. His help in tracking down references and his assistance in a search for char gametes were greatly appreciated. John's study about the ecology of sympatric char provided the ecological framework for my genetics and laid the foundation for my hypothesizing. I spent three years surrounded by a wonderful, supportive lab group, Alon Altman, Dawn Cooper, Janelle Curtis, Jason Ladell, Steve Latham, Derek Louie, Megan McCusker, Jen McLean, Dave O'Brien, Mike Stamford, and Patrick Tamkee. Thanks to them the lab was never lonely and no problem ever went unsolved. I had some fantastic times. Finally, I received tissue samples and molecular markers from various sources. In particular, Dave Bustard sent many samples and shared much knowledge about Thutade Lake watershed. Chris Foote and Brian Urbain at the University of Washington gave me the Alaskan and Kur i l Island samples with which I put the last touches on my zoogeographical study. Paul Moran, also from the University of Washington, kindly shared seventeen primer sets developed in his lab. vi i i Chapter 1: General Introduction 1 Interspecific hybridization is a topic of great interest to both evolutionary and conservation biologists. It is a natural process by which new species can evolve (e.g. Meagher & Dowling 1991; Quattro et al. 1991, 1992; Abbott 1992; DeMarais et al. 1992; Bul l in i 1994) and by which rare or endangered species can be driven extinct (Arnold 1997). Introgression, the movement of genes across a species barrier, can provide the genetic diversity required to adapt to changing environments, but it can also lead to the dilution or disruption of genomes already well-adapted (Arnold 1997). Studies of the ecological and evolutionary conditions leading to hybridization or reproductive isolation, as well as the effects on the species concerned, are of great value in understanding the processes of evolution. Hybridization is important in biological theory, and many interesting studies have investigated cases of hybridization in both plants and animals. Botanists and zoologists have traditionally taken very different approaches to hybridization studies, due primarily to the perceived importance of hybridization in their respective taxa: for years hybridization has been recognized as widespread in plants, while in animals, hybridization was thought until fairly recently to be extremely rare (Hubbs 1955). The fundamentally different ways in which botanists and zoologists have perceived hybridization led to a theoretical gulf between botany and zoology in hybridization studies (Arnold 1997). Botanists have tended to study hybridization in terms of its creative role in evolution and have observed the patterns that reticulate evolution has produced in phylogenies. Zoologists, on the other hand, with the Biological Species Concept to guide their evolutionary theory, have tended to see hybridization as a negative force, breaking down the genetic diversity created by divergent evolution. If zoologists accorded hybridization any importance in evolution, it 2 was only to reinforce burgeoning reproductive isolation, by leading to selection against the misled individuals who foolishly strayed from the local gene pool. The result of this focus was that zoologists have studied process in current hybridization, generally using well-defined hybrid zones, but have typically ignored the patterns hybridization produces in phylogenies. These hybrid zones, where hybridization is geographically localized and often occurs at high levels, are living laboratories where hybridization can be studied in great detail. The well-studied Bombina toad (e.g. MacCal lum et al. 1998), Mercenaria clam (Bert & Arnold 1995), Mus mouse (e.g. Ferris et al. 1983) and Chorthippus grasshopper (e.g. Hewitt 1993) natural hybrid zones (as well as the plant Iris hybrid zone; e.g. Cruzan & Arnold 1994) have led to many insights into reproductive isolation and the types and levels of selection against (or sometimes for!) hybrids. Man-made hybrid zones, such as the experimental Gambusia zone in Biosphere 2 (Scribner & Avise 1994a & 1994b) and the hundreds of zones created by the introduction of non-native plant and animal species have also led to some interesting conclusions regarding the genetic swamping of species by hybridization (e.g. Echelle & Connor 1989; Wilde & Echelle 1992). Little emphasis, however, has been placed on diffuse hybridization between species naturally sympatric over large areas, with no clearly defined hybrid zones. Large spatial scales and low-level hybridization can make such studies daunting. They are important, however, because in many cases, hybridization does not occur in neatly pocketed, finely distributed hybrid zones. Freshwater fish, with their necessarily limited . distributions, can minimize these practical problems for scientists. The investigation of pattern in hybridization and evolution, i.e. the use of phylogenetics to investigate the role of hybridization in the evolution of a species or 3 genus (Arnold 1997), is also sorely lacking in animal studies. M y thesis examines both pattern and process in the natural hybridization of a freshwater, sympatric species pair whose entire genus (Salvelinus) appears to have an evolutionary history riddled with interspecific hybridization events. Hybridization, Evolution and Char Among vertebrates, hybridization is most commonly observed between fish species (Hubbs 1955). Within fish, as within most groups, levels of hybridization vary greatly between taxa (Hubbs 1955; Arnold 1997). Char, Salvelinus spp., provide plentiful evidence for current and historical hybridization. Molecular evidence of current and/or historic hybridization has been found between naturally sympatric populations of bull trout (S. confluentus) and Dol ly Varden (S. malma; Baxter et al. 1997), brook trout (S. fontinalis) and lake trout (5. namaycush; e.g. Berst et al. 1980), Arctic char (5. alpinus) and brook trout (Hammar et al. 1991; Bernatchez et al. 1995; Glemet et al. 1998), and Arctic char and lake trout (Hammar et al 1989; Wilson & Hebert 1993; Wilson & Bernatchez 1998). Brook trout have been introduced all over the world and have since begun to hybridize with naturally allopatric white-spotted char (S. leucomaenis; Kitano et al. unpubl.) and bull trout (Kitano et al. 1994). Lest it be said that hybridization is so commonly observed in char because char, as sports fish, are so well studied, it is worth pointing out the relative paucity of evidence for hybridization in the sister genera Oncorhynchus (salmon) and Salmo (trout). In fact, confusion has pervaded char taxomomy for decades. Geographic variation of morphology within species has lead to uncertain designations as "species complexes" (e.g. McPhai l 1961; Behnke 1980). Multiple molecular studies, using 4 different nuclear and mitochondrial markers, have failed to resolve the phylogeny within the genus and often yield contradictory results (e.g. Grewe et al. 1990; Crane et al. 1994; Phillips et al. 1994). I propose that the long history of confusion surrounding Salvelinus "species complexes" is at least partly due to current hybridization between sympatric species pairs, which masks differences between species and blends the line between interspecific differences and intraspecific geographic variants. The confusion of char systematics is also easily explained (if not deciphered) if the genus has been subject to hybridization and introgression throughout its history and evolution (Arnold 1997). Indeed, the study of phylogenetic pattern has long been a key tool in the repertoire of plant hybridization studies but is rarely used in animal hybridization (Arnold 1997). Sadly, the discussion of hybridization as an explanation for conflicting phylogenetic results is nearly absent in the char literature (but see Phillips et al. 1995, 1999). While phylogenetic studies may be able to shed some light on the presence of past hybridization events, ecological and population genetic studies of currently hybridizing populations are required to understand how natural hybridization affects the species involved, and their evolutionary future. This type of study, investigating the reproductive or habitat characteristics leading to hybridization, the types of selection against hybrids, and the evolution of reproductive isolation (or not!) between congeneric species, requires naturally sympatric populations with current hybridization. Bu l l trout (S. confluentus) and Dol ly Varden (S. malma) provide an excellent opportunity for such a study. 5 Dolly Varden and Bull Trout B u l l trout (S. confluentus) and Dol ly Varden (S. malma) were for many years considered to be geographic variations within the Arctic char complex. In 1961, McPhai l ' s morphological analysis described a southern form within this complex, which became known as the "S. malma species complex." In 1978, Cavender showed that within the S. malma of North America were two forms distinct enough to be called species, Dol ly Varden distributed along the coast and bull trout found inland, with a significant zone of range overlap (Figure 1). A decade later, the first in-depth study of the two species determined their ranges in more detail and succeeded in devising a morphological method, based on three characters, which could identify individuals by species throughout their North American ranges (Haas & McPha i l 1991). Further, the morphological analysis of 1580 individuals from 310 sites did not find any conclusive evidence of natural hybridization (Haas & McPhai l 1991). Consequently, the status of Dol ly Varden and bull trout as distinct species received more recognition. Cavender (1978) had identified two morphologically intermediate individuals from the area of range overlap, and suggested that hybridization may be occurring between the two species in areas of sympatry. Haas & McPhai l (1991) showed, through experimental crosses, that interspecific fertilization is possible in both directions and that there did not appear to be any biased mortality in the offspring produced. The F i hybrid fry were found to be morphologically intermediate and to have intermediate growth rates to pure parental offspring (Haas 1988). 6 Figure 1 : T h e parapatric s p e c i e s distributions of Dolly V a r d e n (S. malma) a n d bull t rou t (5. confluentus). 7 The first good evidence of natural hybridization between the two species came recently, from the Skagit River (McPhail & Taylor 1995) and the Thutade Lake watershed in north-central B . C . (Baxter et al. 1997). The presence of a post-Fi hybrid or backcross in Thutade Lake revealed not only that natural hybridization occurred in sympatry, but also that F i hybrids, in addition to being viable, are fertile. Unfortunately, although there are years of records of bull trout and Dol ly Varden in the literature, nothing prior to 1978 (and even later) can be firmly established to belong to one species or the other, unless the record or study was from a known area of allopatry (e.g. Leggett 1980). From areas of range overlap, or potential range overlap, the legitimacy of records is certainly in doubt, particularly as the two species are so difficult to identify morphologically. As recently as this year, private consultants and B . C . Ministry of Fisheries officials have sent samples to me for genetic species identification, often identifying the samples as "char" or "Dolly Varden" when in fact both species were present in the sample. The morphological similarity between Dol ly Varden and bull trout means that identifying hybrids in a sympatric population by morphology alone would be difficult, if not impossible. Further, any study relying solely on morphological identification of hybrids is obtaining only circumstantial evidence of hybridization, because of the a priori assumption that hybrids are of intermediate morphology (Campton 1987) which is certainly not always the case (e.g. Leary et al. 1983). To have any hope of studying hybridization between bull trout and Dol ly Varden would require the use of molecular markers to both identify hybrids and to determine their genotype class (e.g. F i genotype, Dol ly Varden backcross genotype, etc.). Lucki ly, the popularity of molecular tools for 8 the study of salmonid species means that many potential markers were available to begin the study. Perhaps even more fortuitously, bull trout and Dol ly Varden do not appear to be sister species with the Salvelinus genus and have diverged enough genetically that markers can be found which are fixed for alternate alleles in the two species (McPhail & Taylor 1995; Baxter et al. 1997; Leary & Allendorf 1997). Molecular markers are the primary tool that I wi l l use in my thesis. Study Objectives M y thesis comprises three distinct parts. First, due to the relatively recent recognition of bull trout and Dol ly Varden as distinct species, an important first step in studying their interspecific hybridization is to determine their phylogenetic and phylogeographical histories. Such a study has already been done for bull trout (Taylor et al. 1999), but the knowledge regarding relationships between Dol ly Varden populations in B . C . is slim. In addition to giving important information about the species, the phylogeography w i l l hopefully also yield additional support for the specific status of Dol ly Varden and bull trout. Perhaps the most interesting information that such a study could yield, in combination with the published bull trout phylogeography, is evidence for historical introgression, the "pattern" type of study described by Arnold (1997). This study is described in Chapter 2. Second, evidence of hybridization between bull trout and Dol ly Varden to date comes from two watersheds, Thutade Lake watershed (Baxter et al. 1997) and the Skagit River (McPhail & Taylor 1995). Genetic evidence for local sympatry of bull trout and Dol ly Varden extends only one river further, to the Quinault River on the Olympic 9 Peninsula, W A (Leary & Allendorf 1997). With such a paucity of information regarding the presence of the two species in local sympatry, I wanted to find out how often they are actually locally sympatric (i.e. found at the same place at the same time), and how widespread hybridization is within that locally sympatric range. This study is described in Chapter 3. Third, hybridization studies for bull trout and Dol ly Varden to date have been relatively small scale (McPhail & Taylor 1995; Baxter et al. 1997). A n in-depth look at the process of hybridization within a population of char would yield useful information regarding the dynamics of the interspecific hybridization, the fate of the hybrids, whether the two species can maintain their specific status in the face of hybridization, and if so, how. Such a study was done in the Thutade Lake watershed in north-central B . C . , and is described in Chapter 4. A n d finally, on a more general note, bull trout and Dol ly Varden are a species pair with interesting ecological life history roles in sympatry, distinct from the roles generally adopted in allopatry. This means that hybrids are in an interesting ecological position. Also, bull trout and Dol ly Varden are a species pair with an enormous area of range overlap, rather than a small, geographically defined hybrid zone. This means that they provide a good example for the currently sparse literature about widespread, low-level hybridization between animal species. More practically, they are both sports fish and bull trout is listed as a threatened species in B . C . and an endangered species in the United States. As such a large part of their respective ranges are sympatric, knowledge of their behaviour in sympatry is an important part of any management plan. 10 Chapter 2: Phylogeography of Dolly Varden: evidence for a southern glacial refuge and historical introgression of bull trout mitochondrial DNA 11 Introduction Bull Trout and Dolly Varden Phylogeography Phylogeography, the linking of phylogeny with geographic distribution, is important background information for analysis of hybridization between species. This is particularly true for species that inhabit recently deglaciated areas, because phylogeography can distinguish secondary contact of allopatric refugial populations (Hewitt 1996) from simultaneous range expansion of sympatric refugial populations. Mitochondrial D N A (mtDNA) is one of the most commonly used molecular tools in phylogeography (e.g. Bermingham & Avise 1986; Meyer et al. 1990; Taylor et al. 1997; Redenbach & Taylor 1999). Mitochondrial D N A is particularly useful for population level studies because it accumulates mutations rapidly and has an effective population size of 25% of that of nuclear D N A (Gyllensten & Wilson 1986). These characteristics mean that evolution over short time spans can be detected, and that small, isolated populations w i l l diverge rapidly in m t D N A haplotype due to genetic drift. The phylogeography of bull trout has recently been analyzed in detail using restriction fragment length polymorphism (RFLP) and sequence analysis of m t D N A . Analysis of samples from throughout the range of bull trout showed there to be two major lineages of bull trout (Taylor et al. 1999). These two clades mapped closely onto the geography of their range, with a 'coastal' clade predominating to the west of the coastal mountain ranges and an 'interior' clade predominating to the east of the coastal mountain range. The authors concluded from this phylogeography that bull trout survived the last Pleistocene glaciation in two refugia south of the ice sheet, the Chehalis refuge on the coast and the Columbia refuge in the interior (Taylor et al. 1999). 12 The phylogeography of Dol ly Varden in British Columbia, however, is a much murkier picture. In fact, the known eastern and southern boundaries of Do l ly Varden continue to expand, as genetic evidence finds Dol ly Varden in watersheds further south (in the Olympic Peninsula; Leary & Allendorf 1997) and east (headwaters of the Peace River; Baxter et al. 1997; Chapter 3 of this thesis) of the currently described range. Dol ly Varden have been described as two subspecies in North America, the northern form, S. m. malma, distributed from Alaska across the Bering Strait to Asia, and the southern form, S. m. lordi, distributed from Alaska to the southern limit of the Dol ly Varden range. There has been no genetic study of the phylogeography of Dol ly Varden within S. m. lordi, the subspecies present in British Columbia. The morphological studies of bull trout and Dol ly Varden across B . C . did not reveal any obvious differences within Dol ly Varden, although the focus of those studies was to show interspecific differences (Cavender 1978; Haas & McPhai l 1991). Most suppositions about the refugial origins of Dol ly Varden were published prior to the recognition of Dol ly Varden and bull trout as separate species. A common basis of refugial designations, however, is that if the species' current range encompasses a known refugial area, it is assumed to have survived the Wisconsinan glaciation there (McPhail & Lindsey 1970). The value of this belief as applied to Dol ly Varden is questionable. Dol ly Varden has recently been shown to be present in a single river south of the maximum extent of glaciation (Leary & Allendorf 1997). The vast majority of southern Dol ly Varden populations are present only in recently deglaciated areas. Working from current distributions, and the lack of Dol ly Varden in all but the extreme headwaters of 13 interior drainages, current information strongly suggests that Dol ly Varden colonized B . C . from a northern refuge, likely the Bering Refuge. The question of whether Dol ly Varden survived in a second refuge south of the ice sheets is crucial because it determines whether Dol ly Varden and bull trout have been in contact continuously over the last 100,000 years or whether the current sympatry is a secondary zone of contact created in the last 14,000 years. Different refugial populations could also have different histories of contact, with Columbian bull trout and Beringian Dol ly Varden isolated, but Chehalis bull trout and Dol ly Varden (if there were any) in continuous contact. As reproductive isolation is generally thought to be reinforced by contact and hybridization (e.g. Rundle & Schluter 1998) areas colonized by these different refugial populations could have significantly different levels of current hybridization. As a additional benefit, showing that bull trout and Dol ly Varden have different refugial histories and phylogeographical distributions also serves to support their specific status. Historical Introgression A second potential use for m t D N A analysis in Dol ly Varden is the inherent ability of m t D N A to record historical introgressive hybridization events. When m t D N A in a population of one species is more closely related to the m t D N A of a second species than to that of the original species, it is commonly concluded that interspecific introgression has occurred. This is most commonly revealed by discordancy of m t D N A with nuclear or morphological results (Arnold 1997). 14 Mitochondrial D N A has been used to show historic hybridization in many different species, such as mice (Ferris et al. 1983), frogs (Spolsky & Uzze l l 1984), and gobiid fishes (Mukai et al. 1997). It has also been used to show hybridization in char: between Arctic char (S. alpinus) and lake trout (S. namaycush) (Wilson & Bernatchez 1998) and between Arctic char and brook trout (S. fontinalis) (Bernatchez et al. 1995; Glemet et al. 1998). A l l of these studies have noted geographical localization of the introgressed populations, and most have proposed a causal relationship with glaciation. A s there is evidence of hybridization between bull trout and Dol ly Varden (McPhai l & Taylor 1995; Baxter et al. 1997; Chapters 3 & 4), evidence of introgression in the Dol ly Varden m t D N A phylogeny would not be unexpected. Objectives The original goal for this portion of my thesis was to produce an m t D N A phylogeny for Dol ly Varden, in order to obtain a phylogeographical history of the species for comparison to bull trout. Preliminary results, however, revealed paraphyly (i.e. evidence for historical introgression) in the Dol ly Varden m t D N A phylogeny. To confirm that this paraphyly was indeed discordant with other systematic traits, I sequenced two nuclear loci in several individuals with critical m t D N A haplotypes. These nuclear results could also provide molecular evidence in support the specific status of bull trout and Dol ly Varden within their sympatric range. 15 M a t e r i a l s a n d M e t h o d s Sample Collection Dol ly Varden samples were obtained from throughout their range, from Washington State to the Kur i l Islands in the western Pacific (Table 1). Individuals were presumed to be Dol ly Varden due to morphology and/or their collection locality. This supposition was confirmed by nuclear D N A analysis. Other Salvelinus species within the genus were either collected and sequenced or the published sequences were obtained from the literature (as per Table 1). Molecular Methods Molecular markers sequenced for this study were Growth Hormone 2 intron C (GH2), the first internal transcribed spacer region of r D N A (ITS-1), and the mitochondrial N A D H - 1 gene (mtDNA) (Table 2). G H 2 and ITS-1 are fully described in Chapter 4 (pp 71-72). N A D H - 1 is the N A D H dehydrogenase subunit 1 gene on the mitochondrial genome. The primer used to sequence N A D H - 1 (i.e. N D 1 - R , Table 2) is complementary to the mitochondrial t R N A - G l n gene and produced a sequence that included part of the t R N A - G l n gene (bp 1-36), the tRNA-Ile gene (bp 39-111), and the beginning of the N A D H - 1 coding gene (119-503). The mitochondrial N A D H 5 / 6 gene was amplified for restriction fragment length polymorphism (RFLP) analysis (described in Chapter 4, p 72-73). P C R conditions (Table 3) and laboratory methods for D N A extraction, R F L P analysis and product visualization were not described here, as they are thoroughly dscribed in Chapter 4 (pp 71-73). Sequencing reactions were performed exactly as per Redenbach & Taylor (1999). 16 Table 1: Sample sites and haplotype/sequence labels for Dol ly Varden, bull trout, and other Salvelinus species for m t D N A , G H 2 , and ITS-1 sequencing and population studies. Sample Drainage N ID on maps and/or haplotype in phylogenies mtDNA Sequencing Udobnaya River Kur i l Islands 1 D V - 2 Klutina River Cook Inlet, Alaska 1 D V - 1 Iliamna Lake Bristol Bay, Alaska 2 D V - l , D V - 3 Lower Togiak Lake Bristol Bay, Alaska 1 D V - 1 Taku River North Coast, B . C . 3 D V - l , D V - 6 , D V - 1 0 Tahltan River Stikine River 1 D V - 6 Chutine River Stikine River 1 D V - 9 Iskut River Stikine River 1 D V - 1 Zolzap River 1 upper Nass River 10 Clade N & S ( R F L P ) 1 Thutade Lake watershed upper Peace River, B . C . 348 Clade N (RFLP) Omineca River upper Peace River, B . C . 1 Clade S (seq. incompl.) Ecstall River Skeena River 1 Clade N (seq. incompl.) Ayton Creek Skeena River 1 D V - 6 Goathorn Creek upper Skeena River 6 Clade N (RFLP) Noyes Sound lower Skeena River 1 D V - 8 Brent Creek Queen Charlotte Islands 1 D V - 1 Aero River Queen Charlotte Islands 1 D V - 1 Honna River Queen Charlotte Islands 1 D V - 5 Ogden Channel Central Coast, B . C . 1 D V - 1 1 Kumelon Creek Central Coast, B . C . 1 D V - 1 Noosneck River Central Coast, B . C . 1 D V - 4 Dallery Creek Central Coast, B . C . 1 D V - 1 Misty Lake Queen Charlotte Islands 1 D V - 1 O'Connell Lake Vancouver Island 1 D V - 1 Claninick River Vancouver Island 1 Clade N (seq. incompl.) Zeballos River Vancouver Island 1 D V - E 17 Elk River Vancouver Island 1 Clade N (seq. incompl.) Thelwood Creek Vancouver Island 1 D V - C Phillips Creek Vancouver Island 1 D V - 7 Cowichan River Vancouver Island 1 D V - 1 Southgate River South Coast, B . C . 1 D V - B Toba River South Coast, B . C . 6 Clade S (RFLP) Mamquam River South Coast, B . C . 1 D V - B M i l l Creek South Coast, B . C . 1 D V - D Capilano River South Coast, B . C . 1 Clade S (seq. incompl.) Seymour River South Coast, B . C . 1 Clade S (seq. incompl.) Silverhope River South Coast, B . C . 1 D V - B Skagit River 2 southern B . C . & W A 81 Clade N & S ( R F L P ) 2 Quinault River Puget Sound, W A 1 D V - A D V / B T F i (Thutade) upper Peace River, B . C . 1 BT-Ia S. confluentus 'Interior' populations 15 BT-Ia , BT-Ib, BT-Ic S. confluentus3 'Coastal' populations 6 B T - C a , B T - C b , B T - C c S. sp. 4 K u r i l Islands 1 S. sp. 4 S. namaycush upper Peace River, B . C . 1 S. namaycush S. fontinalis upper Columbia R, B . C . 1 S. fontinalis GH2 Sequencing Udobnaya River K u r i l Islands 1 D V - 2 Taku River North Coast, B . C . 1 D V - 1 0 Ayton Creek Skeena River 1 D V - 6 Cowichan River Vancouver Island 1 D V - 1 Southgate River South Coast, B . C . 1 D V - B M i l l Creek South Coast, B . C . 1 D V - D Salmo River upper Columbia River 1 B T - I Mamquam River South Coast, B . C . 1 B T - C S. namaycush5 Moberly Lake, B . C . 1 S. namaycush ITS-1 Sequencing Udobnaya River K u r i l Islands 1 D V - 2 18 Taku River North Coast, B . C . 1 D V - 1 0 Cowichan River Vancouver Island 1 D V - 1 Southgate River South Coast, B . C . 1 D V - B M i l l Creek South Coast, B . C . 1 D V - D S. m. krascheninnikovi6 Sakhalin Island, Russia 1 S. m. kraschen. S. m. malma6 Kamchatka, Russia 1 S. m. malma S. m. lordi6 Kenai Peninsula, A K 1 S. m. lordi Howell Creek up. Flathead River, B . C . 1 B T - I Elwha River Strait of Juan de Fuca 1 B T - C S. namaycush Upper Peace River, B . C . 1 S. namaycush S. alpinus Nauyuk Lake, N W T 1 S. alpinus S. leucomaenis Japan 1 S. leucomaenis •7 S. fontinalis Wisconsin 1 5. fontinalis Populat ion Level Studies D V m t D N A R F L P Haplo . Tahltan River Stikine River 9 Clade N (n = 6) Chutine River Stikine River 19 Clade N (n = 3) Iskut River Stikine River 47 Clade N (n = 1) Thutade Lake upper Peace River 990 Clade N (n = 348) Omineca River upper Peace River 27 Clade S (n = 6) Goathorn River upper Skeena River 91 Clade N (n = 6) Southgate River South Coast, B . C . 28 Clade S (n = 6) Little Toba River South Coast, B . C . 10 Clade S (n = 6) Mamquam River South Coast, B . C . 18 Skagit River 2 southern B . C . / Wa 101 Clade N & S (n = Quinault River 8 Olympic Peninsula, W A 25 Clade S (n = l , b y 1 RFLP results from Taylor et al. (2000), 9 Clade N, 1 Clade S (or the indistinguishable bull trout) 2 RFLP results from McPhail & Taylor (1995), where mtDNA analysis of nuclear Dolly Varden produced 27% Clade N and 73% "bull trout" mtDNA. Given the southern locality of the Skagit River, the "bull trout" mtDNA is just as likely to be Dolly Varden Clade S mtDNA. 3 sequences and/or RFLP results obtained from Taylor et al. (1999) 4 this individual was morphologically identified as Dolly Varden, but since the mtDNA sequence divergence from S. malma was high and S. leucomaenis is sympatric in the Kuril Islands, the species designation has been omitted 5 sequence obtained from McKay et al. (1996) 6 sequences obtained from Phillips etal. (1999) 7 sequences obtained from Pleyte et al. (1992) 8 allozyme results from Leary & Allendorf (1997) 19 CD CJ IH 3 O I CO bfj c c o e bo 1 w CD CJ c CD CT1 , u 00 c s 03 Cs Os cy o O o VO > cd ca oo CJ ca O DO * r-O K O cs CD c o fl o I >H o o m II H o ca 03 o 03 t3 bfj bo S o bfj 03 cd O i n o u c s m O N Os 00 O O r-II > 03 bfj o ta o CJ bO u bO o bfj ca i CO H CO H i — i < a; Q 1 o O 2 o cs °3 o Os Ti-ll H CQ bfj ca ca bo bfj ca bfj 03 +-4 CJ 03 CJ bfj » o o 03 ca bfj •4—» bfj bfj ca +-< bfj CJ o 4—» -4—< o bfj 03 03 03 03 03 — i * i CS bfj 03 -i—» 03 73 3 & 3 CD M 03 pq o cs i n II > Q bfj 03 o3 03 03 bfj ca 03 •4-> bfj -4-4 -4-4 bfj ca o CJ ca CJ BP "ca pq H CD C '£ o o "cd -4-4 CD ca CD co oo ca o o o o m oo r-- ^H II II II H > H pq Q PQ bfj bfj bfj ca CJ bfj ca CJ ca bfj bfj ca ca bfj bfj CJ •*—* bfj oo Os Ov ca T3 03 6 ca bfj ca CJ *^  bfj -4-4 bO 03 bfj cj -4—* 03 bfj bfj ca CJ bfj CJ ca i cn cs cs M o PH ca bfj ca 4—» ca CJ CJ CJ ca ca CJ CJ o bfj 03 cn cs o PH oo Os Os c CD co '53 55 o i n 3 o o cs > Q bfj bfj 03 bfj bfj 4—> ta CJ CJ ta cj 4—4 o bfj 03 ta ca PH I VO in VO i n VO >n o >n cn cs II H pq ca o to ca ca cj bfj BP bb bfj o 03 ca % ca vo m oo Os Os c CD oo o o cs o m cn > CJ bfj ca ca ca bfj CJ cj cj bf) bO bfj ta -4—> bfj bfj PH O o cn c f o i n H PQ CO > Q ca o ca ca ca ca ca CJ % -4-» -4—> 15) o o bO CJ 4—4 CJ CJ bfj 20 Table 3: P C R conditions for molecular markers used in D N A analysis of Do l ly Varden and bull trout. Locus Annealing Temp. & Cycles M g C l 2 Primers Total d N T P Total Volume (cycles @ °C) (mM) (/xM) (MM) (Hi) G H 2 40 @ 55 1.5 0.25 200 25 ITS-1 1 @ 52, 5 @ 54, 35 @ 55 1.5 0.6 800 25 M T B 40 @ 56 1.5 0.08 320 20 F O K 35 @ 55 1.5 0.6 800 25 ND5/6 40 @ 54 2.5 0.6 800 25 N D 1 35 @ 54 2.0 0.6 800 50 21 Phylogenetic Analysis I analyzed all three sequenced molecular markers using the P H Y L I P computer analysis package (Felsenstein 1995). Data was subjected to bootstrapping (1000 replicates, S E Q B O O T ) , and analyzed by a distance tree-building program ( N E I G H B O R , using neighbor-joining method and randomizing input order). The 1000 resulting N E I G H B O R trees were combined using majority rule consensus tree analyses ( C O N S E N S E ) , with specified outgroups, as described in the Results. Bootstrap values greater than 70% were considered to be strong support, as they generally correspond to a greater the 95% probability that the clade is real (Hillis & B u l l 1993). I calculated sequence divergences using the D N A D I S T program of P H Y L I P applying the Kimura 2-parameter distances model of nucleotide substitution (Kimura 1980). I calculated haplotype and nucleotide diversities and nucleotide divergences within and between m t D N A haplotype clades and char species using the D A program of R E A P ( M c E k o y et al. 1992). I estimated haplotype diversity as per Ne i (1987) and nucleotide diversity and nucleotide divergence according to Nei & Tajima (1981). I scanned m t D N A sequences for restriction enzyme recognition sites using a web-based program, Webcutter 2.0 (copyright 1997 Max Heiman, U R L : www.firstmarker.com/cutter/cut2.htrnl). Several point mutations diagnostic for various m t D N A clades were identified using the phylogenetic analysis, but unfortunately none of these diagnostic mutations fell within the restriction enzyme recognition sites identified. 22 Results Mitochondrial DNA Phylogeography Thirty-one Dol ly Varden individuals, sampled from throughout the Dol ly Varden range from Washington to the Kur i l Islands, were sequenced for 503 base pairs of the mitochondrial t R N A - G l n , tRNA-Ile , and N A D H - 1 genes (Table 1). This sequencing yielded 16 different Dol ly Varden mitochondrial haplotypes (labeled D V - 1 to D V - 1 1 and D V - A to D V - E ; Table 1, Appendix 1). Most of these haplotypes were present in only one individual, with the exceptions of D V - 1 , D V - 6 , and D V - B , which were found in 12, 3 and 3 individuals, respectively. A second individual from the K u r i l Islands, visually identified as a Dol ly Varden, was sequenced, and produced a very divergent haplotype. I could not be certain of the visual species identification because the Asian white-spotted char (S. leucomaenis) is also found in the Kur i l Islands. No S. leucomaenis m t D N A sequence was available for comparison, so the individual was been designated S. sp. (Table 1, Figure 2). Lake trout (S. namaycush) and a Dol ly Varden/bull trout hybrid with an F i genotype and bull trout m t D N A (by R F L P ) from Thutade Lake were also sequenced for this study. The sequences for bull trout and brook trout, S. fontinalis, were obtained from the literature (Taylor et al. 1999; Table 1). Phylogenetic analysis revealed that Dol ly Varden m t D N A haplotypes are paraphyletic. The haplotypes fall into two clades, named Clade N (for 'northern', see below) and Clade S (for 'southern') (Figure 2). Unfortunately, few of the bootstrap values obtained are greater than 70%, generating uncertainty regarding the phylogenetic tree produced. The diagnostic point mutations identified by sequencing, however, support each of the major clade distinctions in the tree (Appendix 1). 23 Figure 2: Phylogenetic tree of Salvelinus mitochondrial D N A based on neighbor-joining analysis of Kimura 2-parameter distance inferred from 503 base pairs of the t R N A - G l n (36 bp), tRNA-I le (73bp), and N A D H - 1 (453 bp) genes. Bootstrap support out of 1000 resamplings is marked. Branch lengths do not approximate relative distances between haplotypes. Vertical hatch-marks represent mutations occurring once, and diagonal hatch-marks represent convergent mutations. Haplotypes are as per Table 1, and sequences as per Appendix 1. S. confluentus (BT-I 'interior,' and B T - C 'coastal') and S. fontinalis sequences were obtained from Taylor et al. (1999). S. fontinalis was designated as an outgroup in consensus tree-building. 24 38 35 16 52 42 100 tt 61 30 I j DV-1 41 DV-2 DV-3 i 5 i r 1A 39 DV-4 DV-5 DV-6 DV-7 — DV-8 — DV-9 DV-10 DV-11 M I N N I E S. sp. BT-lc I l l l l l l l l l l l l l l l l l l l h W ^ S. fontinalis 25 Clade N was monophyletic, contained 11 haplotypes obtained from 24 individuals, and had 52% bootstrap support on 1000 resamplings. The Dol ly Varden Clade S is not monophyletic, as it contains a bull trout mitochondrial D N A haplotype (BT-Ca; Figure 2). Dol ly Varden haplotype D V - B was in fact identical in sequence to B T - C a , one of the three coastal bull trout haplotypes (Taylor et al. 1999). Bootstrap support for the clade encompassing coastal bull trout and Clade S was 42%, despite a diagnostic point mutation (a T—>C transition). Support for Clade S was also low (30%), as the haplotypes had only the lack of a diagnostic mutation to group them together (a C—>T transition present in B T - C b and BT-Cc) . Unfortunately, there was no derived mutation shared by all Clade S Dol ly Varden haplotypes, to make them distinct from coastal bull trout haplotypes. B u l l trout haplotypes (from Taylor et al. 1999) are paraphyletic as well, with the 'interior' clade (BT-I) monophyletic and the 'coastal' clade paraphyletic with Clade S Dol ly Varden, as described above. A Dol ly Varden/bull trout hybrid with an Fi genotype from Thutade Lake had a m t D N A sequence haplotype identical to BT-Ia . There was 80% bootstrap support for the interior clade, but less than 50% support (46%) for the clade encompassing interior bull trout, coastal bull trout, and Clade S Dol ly Varden, despite a diagnostic point mutation (a G—>A transition). Haplotypes from Clade S were much more geographically localized than those of Clade N (Figure 3). Clade S was generally found only in the south of British Columbia and in Washington. The two exceptions were the newly discovered population of Dol ly Varden in the upper Omineca River headwaters of the Peace River (sequence) and the Zolzap River (RFLP) , a tributary to the Nass River in northern B . C . (Taylor et al. 2000). 26 Figure 3: Geographic distribution of Dol ly Varden m t D N A clades. Empty circles represent Clade N , darkened circles represent Clade S. Numbers and letters within circles represent haplotypes (Table 1, Figure 2). Circles without specific haplotype designations represent results obtained from incomplete sequences or through R F L P analysis. The solid line represents the approximate limits of the 'coastal' (to the west) and 'interior' (to the east) bull trout clades (as per Taylor et al. 1999) 27 28 Clade N was found from the southern tip of Vancouver Island to the K u r i l Islands in the western Pacific, encompassing almost the entire range of Dol ly Varden. A single northern haplotype, D V - 1 , was found in 12 individuals distributed from Vancouver Island to Bristol Bay, Alaska. Several sample sites were not sequenced, or produced poor sequences that were only used to observe diagnostic sites. While these samples could not be assigned a particular haplotype, R F L P analysis could distinguish Clade N Dol ly Varden from bull trout and Clade S Dol ly Varden. A diagnostic point mutation identified by sequencing had an associated restriction enzyme recognition site (Hae U l , which only cuts Clade N Dol ly Varden haplotypes; Table 2). However, an N A D H 5 / 6 Hind III restriction site previously identified (Baxter et ai. 1997) served to make this same distinction (because m t D N A does not recombine). As the Hind in site had been used to distinguish Clade N Dol ly Varden and bull trout/Clade S Dol ly Varden m t D N A in other parts of this thesis (Chapters 3 & 4), it was also used to complement the sequencing in this study. Analysis of m t D N A by R F L P analysis, unfortunately, could not be used to distinguish Clade S Dol ly Varden from coastal bull trout, or coastal bull trout/Clade S Dol ly Varden from interior bull trout. Sample sites that had been identified as Clade S by R F L P alone (e.g. Zolzap River, Table 1) should therefore be regarded with caution, because the result could in fact be true bull trout m t D N A ( B T - C or BT-I) in Dol ly Varden individuals (i.e. current introgressive hybridization, as in Chapter 4), instead of Clade S Dol ly Varden m t D N A in Dol ly Varden individuals (historical introgression, see below). There was approximately the same amount of sequence divergence between bull trout haplotypes (0.2% - 1.6%) as between Clade N Dolly Varden haplotypes (0.2 -29 1.4%; Table 4). A notable difference between the two species is that only six haplotypes were found in 22 bull trout individuals collected from throughout the bull trout range, while 11 haplotypes were found in 24 Dol ly Varden individuals (if Clade S is included, 16 Dol ly Varden haplotypes were found in 31 individuals with an increased maximum sequence divergence of 2.2%). As expected, haplotype diversity is smaller in bull trout (0.5751 ± 0.0802) than Dol ly Varden (0.7340 ± 0.0643). Even Clade S Dol ly Varden m t D N A , which falls within the coastal bull trout clade phylogenetically, has higher haplotype diversity than bull trout (0.7912 ± 0.0861; Table 4). Sequence divergence and nucleotide diversity within Clade N (0.2% - 1.4%, 0.00383) were greater than those within Clade S (0.2% - 0.6%, 0.00267) (Table 4). The haplotype diversities for the two clades, however, were very similar (0.7340 ± 0.0643 and 0.7912 ±0 .0861 ) . The Salvelinus sp. haplotype was identified as a sister group to Clade N with 61% support. There was a high level of sequence divergence between Salvelinus sp. and the Clade N Dol ly Varden that was also collected in the K u r i l Islands ( D V - 2 , 3.27%). It is worth noting that if S. sp. was in fact a Dol ly Varden individual, it would increase the percent sequence divergence, haplotype diversity and nucleotide diversity within both Clade N Dol ly Varden and Dol ly Varden as a whole and would increase the maximum percent sequence divergence between Dol ly Varden and bull trout (Table 4). It would not, however, affect the directionality of any of the trends described above. Lake trout (5. namaycush) was grouped with the Dol ly Varden / Salvelinus sp. with poor support (42%). Brook trout (S. fontinalis) was designated as the outgroup in consensus tree-building, because it is generally accepted to be the most divergent 30 c CD CD < 5 2 ° Q c Q X l £ *£ C oo cfl * « 53 <=* 2 > a > -s ^ - O / i \ oo c > CD X i O "S C - D .5 03 w tfl ^ CD .g fa 03 03 r -£ O « 8 c « CD -S lo . CD CD CD T 3 C 03 u I H © CD XI 03 -t—» tfl 03 O CD CD - C •*-» O c CD • > Os =3 Cfl CD « »H t~> 03 W CD 3 O 03 H oo O o3 00 J= 2 xt O T3 CD CD CD cfl o -c — > Q 00 o <4H CD i-03 oo CD O c CD CD ~* 3 « 1 3 iJ § 8 I jj 3 u CD _c 00 O 'S 53 o - ^ CD — i - CD CD 8 £ m 3 O CD CD" e S CD > xl X) 03 0 § X I ^ s *1 a ? CD 03 X l *H 03 3 O .CD X) ^8 CD U oo 00 cd 8 £ T T H o a> R — 03 3 H g r , O o ^ C 3 V, ~ 2 S I S -^ p So .2 "83 > 03 CD CD CD O S3 CD 00 SH 2 CD CD e CD 3 cr CD oo -*-> C CD a CD PH CD X J O 13 3 VO CN CN VO VO VO q VO co CN CN VO O oo - vd vd vd i n vd oo oo VO q K CN CN CN cn ^H 03 co O T t 4 T t T t CN CN CN CN cfl CO H PQ U H PQ i > Q I > Q CD O C CD 00 •H CD > cn cn m r-^  cn T f T t cn vo 6) o \ i n CN cn cn cn > Q T l -CN i CN CN CN i T j " CN O 00 > q ^H I o VO d i CN d U H PQ H PQ a. oo CO C CO H PQ 03 VO CN vq vo CN - H - H d CN Tt" 4 CN ^H d d d VO d t T t d CN CN 00 o O > P o o ^H o d i n T t o o o o CN VO T t O N VO r - 00 cn o O o o o o d d d o r-o o T t OO i n o o d r--o o oo ^H cn o o CN ON VO cn m CN 00 T t m CN cn T t o CN *-H o r-H o o O o O d d d d d H pq U l H pq oo > I > Q CN H O m oo o ° +l _ T t cn CN 1 - 1 T t cn o ri © ° +1 O N IO cn o cn o ri © ° +1 v^g ri © ° +1 CN vo ^ 00 O N O 2 T t N O cn o 00 I > Q u I H PQ H pq > Q H pq CD > IQ CD B I "Hi 31 Salvelinus species (Grewe et al. 1990; Crane et al. 1994; Phillips et al. 1995) and because the sequence divergence between it and the other Salvelinus species is much larger than between any of the other species (Table 4). Ribosomal DNA ITS-1 and GH2 Phytogenies Dol ly Varden and bull trout were sequenced at two nuclear loci, the first internal transcribed spacer region of r D N A and intron c of Growth Hormone 2. The individuals chosen for sequencing were specifically chosen for their m t D N A haplotype and collection location, so that the two Dol ly Varden and two bull trout m t D N A clades would be well represented (Table 1). These markers have been popular in char systematics studies, so I was able to obtain several sequences from elsewhere in the Salvelinus genus for inclusion in phylogenetic analysis. For r D N A ITS-1,1 have included sequences for three presumed Dol ly Varden subspecies (Phillips et al. 1999) and for Arctic char (5. alpinus), white-spotted char (S. leucomaenis), and brook trout (S. fontinalis) (Pleyte et al. 1992). Lake trout was sequenced at rDNA-ITS-1 for this study. For G H 2 , the lake trout sequence was obtained from M c K a y et al. (1996). Phylogenetic analysis of 484 base pairs of G H 2 showed Dol ly Varden and bull trout to be reciprocally monophyletic, with significant bootstrap support (Figure 4). Bootstrap support for the bull trout clade, grouping coastal and interior bull trout, is 100%. Bootstrap support for the Dol ly Varden clade is 94%, grouping northern (Clade N) and southern (Clade S) individuals from as far apart as southern B . C . and the K u r i l Islands. In fact, Dol ly Varden individuals with four different m t D N A haplotypes ( D V - 1 , D V - 1 0 , D V - B , and D V - D ) sampled from across B . C . yielded identical G H 2 sequences 32 94 82 tt 91 DV-2 DV-6 DV-1, 10, B, D 100 BT-C BT-I -HHH* S. namaycush Figure 4: Phylogenetic tree of Salvelinus Growth Hormone 2, based on neighbor-joining analysis of Kimura 2-parameter distance inferred from 484 base pairs of intron C. Bootstrap support out of 1000 resamplings is marked, and branch lengths do not approximate relative distances between haplotypes. Vertical and horizontal hatch-marks represent mutations occurring once, and diagonal hatch-marks represent convergent mutations. Haplotypes are as per Table 1, and sequences as per Appendix 2. S. namaycush sequence was obtained from M c K a y et al. (1996) and was designated as an outgroup in consensus tree-building. 33 (Appendix 2). Overall sequence divergence within the Dol ly Varden clade ranged from 0 to 0.42% (Table 5). Within bull trout sequence divergence was similarly small (0.21%), while sequence divergence between Dol ly Varden and bull trout ranged from 4.31 to 5.00%. Lake trout was used as the outgroup, although according to percent sequence divergence it is much closer in sequence to Dol ly Varden (1.47 - 1.90%) than to bull trout (4.56 - 4.79%) (Table 5). In the analysis of the 410 base pair r D N A ITS-1 sequence, Dol ly Varden and bull trout were again reciprocally monophyletic with significant bootstrap support (Figure 5). Every individual sequenced was different for at least one base pair. The Dol ly Varden clade (76% support) contained two sub-clades, which correspond to both sub-specific and geographic delineations. S. m. krascheninnikovi and S. m. malma, collected in Russia (Phillips et al. 1999), group with D V - 2 , collected in the K u r i l Islands, with 86% support. S. m. lordi, the 'southern' North American subspecies, grouped with B . C . individuals with 60% bootstrap support ( D V - 1 , D V - 1 0 , D V - B , and D V - D ) . The Dol ly Varden m t D N A Clade S ( D V - B and D V - D ) and Clade N (DV-1 and DV-10) individuals did not group separately within the Dol ly Varden clade. Percent sequence divergence within the Dol ly Varden clade ranged from 0 to 1.51% (Table 5). Coastal and interior bull trout grouped together with 100% support (Figure 5), and had 0.75 % sequence divergence between them (Table 5). Brook trout was used as the outgroup, as it is commonly accepted to be the most divergent of Salvelinus species and had the greatest percent sequence divergence with the most other species (Table 5). Arctic char grouped with Dol ly Varden with 91% support, while white-spotted char 34 Table 5: Percent sequence divergence between Growth Hormone 2 intron C sequence (above the diagonal) and between r D N A ITS-1 sequences (below the diagonal). Where identical headings intersect, the values shown represent variation within a group. In G H 2 , the S. namaycush sequence was obtained from M c K a y et al. (1996), in ITS-1 the S. alpinus, S. leucomaenis and 5. fontinalis sequences were obtained from Pleyte et al. (1992). r D N A ITS -1 Percent G H 2 Percent Sequence Divergence Sequence Divergence D V B T S. n. D V 0-1.51 0-0.42 4.31-5.00 1.47-1.90 D V B T 4.12-6.08 0.75 0.21 4.56-4.79 B T S. a. 0.99-2.53 5.19-6.08 n/a S. 1. 5.51-6.64 4.65-5.52 4.96 n/a S. n. 4.38-4.94 4.91-5.51 4.93 5.77 n/a S.f 4.91-5.76 7.13-7.78 4.64 6.59 5.46 D V B T S. a. S. 1. S. n. 35 ,60 100 ,91 76 86 m— DV-2 S. m. kraschen. •S. m. malma 56 60 1/ DV-B — — S. m. lordi | DV-1 89 86 ft/—DV-D —DV-10 | S. alpinus .65 46 . . .,100 •BT-C llllllll// BT-I llllll// S. leucomaenis l l l l l l l l l /^ S. namaycush • S. fontinalis MINIMUM//^  Figure 5: Phylogenetic tree of Salvelinus ribosomal D N A internal transcribed spacer region (rDNA ITS-1) based on neighbor-joining analysis of Kimura 2-parameter distance inferred from 410 base pairs of ITS-1. Bootstrap support out of 1000 resamplings is marked, and branch lengths do not approximate relative distances between haplotypes. Vertical and horizontal hatch-marks represent mutations occurring once, and diagonal hatch-marks represent convergent mutations. Haplotypes are as per Table 1, and sequences as per Appendix 3. S. m. krascheninninovi, S. m. malma, and 5. m. lordi sequences were obtained from Phillips et al. (1999), 5. alpinus, S. leucomaenis and S. fontinalis sequences were obtained from Pleyte et al. (1992). 5. fontinalis was designated as an outgroup in consensus tree-building. 36 grouped with bull trout with 65% support. Lake trout grouped with bull trout and white-spotted char, with less than 50% support (46%). A comparison of the three markers shows that m t D N A had a much smaller percent sequence divergence between Dol ly Varden and bull trout (1.2 to 2.6%; Table 4) than did either G H 2 (4.31 to 5.0%) or ITS-1 (4.12 to 6.08%) (Table 5). In both G H 2 and ITS-1, the percent sequence divergence between species was much greater than within species (5 to 10 times greater), whereas in m t D N A , the range of percent sequence divergences between species overlapped that of within species (Table 4). Discussion Two Distinct Species Phylogenetic analysis of two nuclear loci, Growth Hormone 2 intron C and the r D N A first internal transcribed spacer region, revealed that bull trout and Dol ly Varden from British Columbia are reciprocally monophyletic with significant bootstrap support and diagnostic point mutations. British Columbian Dol ly Varden grouped with Alaskan, Russian, and K u r i l Island Dol ly Varden with significant bootstrap support (94% in G H 2 , 76% in ITS-1), rather than grouping with bull trout. As wi l l be discussed below, the two species also have different refugial histories. Both of these results support the conclusion derived from morphological (Cavender 1978; Haas & McPhai l 1991) and ecological (Hagen 2000) research that bull trout and Dol ly Varden are two distinct species. The two species also had different levels of haplotype diversity. Percent sequence divergence due to mutation within bull trout was 0.2 to 1.6% and within Do l ly Varden is 0.2 to 1.4% (northern clade only, as the divergence produced by Clade S was due to both 37 introgression and mutation, discussed below). Haplotype diversity, however, was much lower in bull trout (n = 6 haplotypes out of 22 individuals) than in Do l ly Varden (n = 16 haplotypes out of 31 individuals, including both Clade S and Clade N haplotypes). This is particularly noticeable in the fact that the Dol ly Varden Clade S was 'younger' than the bull trout clade (i.e. it had a smaller nucleotide diversity and percent sequence divergence) and was much more restricted geographically, but already had almost as many haplotypes as bull trout do throughout their entire range. As both species appear to have survived the Wisconsinan glaciation in two refugia (discussed below), and as a large part of the bull trout range sampled for these haplotypes was in areas south of the maximum extent of glaciation, it appears that bull trout have maintained much less genetic variation than Dol ly Varden over time. Whether this lack of variation is explained by population bottlenecks or some other reason, it indicates a difference in the genetic diversity of Dol ly Varden and bull trout. Refugial Histories and Patterns of Recolonization. The presence of two distinct Dol ly Varden m t D N A clades with two distinct geographical distributions yielded valuable information about the refugial history of Dol ly Varden during the last Pleistocene glaciation. Clade N is distributed from the Kur i l Islands in the western Pacific through Alaska to southern British Columbia. It is, therefore, almost certain that Clade N was present in the Bering Refuge. It is also likely that the Bering Refuge provided much of the m t D N A diversity present in the populations with which British Columbia was recolonized, as Clade N is the only m t D N A type present in most of the northern two-thirds of B . C (Figure 3). 38 Clade S has a very limited geographical distribution, present in only the southern corner of B . C . and Washington State (with the exception of the Omineca River and the possible exception of the Zolzap River, discussed below). This narrow distribution suggests that there was a second Dol ly Varden refuge located to the south of the Cordilleran ice sheet. As Dol ly Varden is a coastal species, with a primarily anadromous life history, this refuge was almost certainly coastal. Given the current limited distribution of Dol ly Varden in the northwestern Unites States, the Chehalis Refuge on the Olympic Peninsula is the most likely location for this refuge. B u l l trout have also been hypothesized to have survived the Wisconsinan glaciation in two refugia, the Chehalis and the Columbian refugia, both south of the ice sheets (Taylor et al. 1999). If true, this would lead to a very important conclusion for hybridization studies: bull trout and Dol ly Varden were not isolated during the Wisconsinan. Although Beringian Dol ly Varden and Columbian bull trout may have been isolated, allopatric populations, in the Chehalis Refuge the two species were sympatric. This means that the current sympatric range in B . C . consists of two semi-distinct areas: a zone of secondary contact between isolated species populations (northern through mid-B.C.) and a zone of range expansion for two continuously sympatric species populations (southern B .C . ) . Between these two zones is an area where interior bull trout met southern Dol ly Varden and where northern Dol ly Varden met coastal bull trout. A l l four possible combinations of populations have occurred. The distribution of current hybridization with respect to phylogenetic history is further explored in Chapter 3. As Clade S likely originated prior to the Wisconsinan glaciation and Clade N is most certainly the original Dol ly Varden m t D N A type (both discussed below), it is not 39 possible to say with certainty that Clade S was not found in the Bering Refuge, nor that Clade N was not found in the Chehalis. In fact, the presence of Clade S in the Omineca River (and perhaps in the Zolzap River) suggests that Clade S was found in the Bering Refuge: the extreme headwaters of the Peace River (and Liard and Fraser rivers) were more likely colonized by Dol ly Varden via headwater transfer from coastal drainages than by migration up the Fraser and/or Columbia rivers and transfer to the Peace River via glacial lakes (Lindsey & McPhai l 1986; McPha i l & Lindsey 1986) with subsequent eradication from all but the headwaters of these interior systems. The possible presence of Clade N in the Chehalis Refuge cannot be supported or contested by the current distribution. Clade N is distributed right to the southern tip of Vancouver Island and into the Skagit River. However, the long-distance colonization potential of Dol ly Varden is certainly great, given their impressive migration records (e.g. deCicco 1992). In fact, of greater surprise than Bering Refuge Dol ly Varden reaching southern B . C . is the fact that Chehalis Refuge Dol ly Varden only colonized midway up Vancouver Island and along the adjacent mainland coast (Figure 3). If the Chehalis Dol ly Varden population was very small, it might explain this lack of colonization success. However, the Chehalis Refuge managed to support five distinct Clade S haplotypes. If the population was very small, genetic drift would have been expected to narrow this field substantially. So, why did the Chehalis Dol ly Varden not recolonize further north? I wi l l propose two separate but not mutually exclusive hypotheses to explain this limited northward spread. First, throughout most of their range, Dol ly Varden are anadromous. However, populations of Dol ly Varden in the southern half of Vancouver 40 Island and the south coast of the mainland may have a greater tendency to be stream or lake-resident (i.e. non-anadromous), while in the Fraser and Squamish rivers the anadromous species is bull trout (McPhail & Taylor, 1995). As McPha i l & Taylor (p. 8, 1995) state, "this apparent reversal in anadromous behaviour is confusing and ... has led some researchers to doubt whether they are really two species." This reversal can be explained by the fact that studies of populations of Dol ly Varden and bull trout have shown that, in sympatry, Dol ly Varden tend to assume a resident life history, while bull trout maintain a generally adfluvial migratory life history (McPhail & Taylor 1995; Baxter et al. 1997; Hagen 2000). It is likely that the Chehalis Dol ly Varden also had resident, non-anadromous life histories. As they existed in sympatry with bull trout in a relatively small population for a relatively long time (at least 80,000 years), it would not be surprising that this life history would become firmly established in the local Dol ly Varden population. A resident life history also explains how five haplotypes were maintained in the southern refuge: as stream resident Dol ly Varden populations would be isolated in adjacent coastal streams, genetic drift could allow fixation of different haplotypes in different populations, maintaining genetic variation. As the ice sheets began to retreat, and recolonization by salt-water routes became available, the resident Chehalis Dol ly Varden would be much slower to respond than the anadromous Dol ly Varden from the Bering Refuge. This slow response appears also to have been true of coastal bull trout: even though coastal bull trout are salt-water tolerant and include some anadromous populations, the bull trout in coastal drainages of B . C . north of the Squamish River have 'interior' type m t D N A (Taylor et al. 1999). For these drainages within such close marine 41 proximity to have all been colonized by bull trout from the interior via headwater transfer rather than by a relatively short marine bull trout migration (EB Taylor, pers. comm.) is odd. Even more mysterious is the complete lack of bull trout on Vancouver Island, located only 40 km from the mainland. The Chehalis Refuge Dol ly Varden at least made it that far. A second hypothesis is that selection favours Clade S m t D N A in southern populations of Dol ly Varden. Clade S is introgressed bull trout m t D N A (discussed below), but as it has been in Dol ly Varden long enough to diverge and radiate, it can now be considered to be 'true' Dol ly Varden mtDNA. B u l l trout have a more southern distribution than Dol ly Varden (stretching as far south as Nevada), and it is not unthinkable that bull trout m t D N A would have some selective advantage over Dol ly Varden m t D N A in a southern climate. Positive selection has in fact been proposed as an explanation for the fixation of Arctic char m t D N A in populations of lake trout (Wilson & Bernatchez 1998) and brook trout (Glemet et al. 1998). Clade S m t D N A may also be selected against further north, limiting the spread of Chehalis Dol ly Varden up the B . C . coast. A s an aside, the Skagit River is likely to contain both Clade N and Clade S Dol ly Varden m t D N A . In a sample of 81 'nuclear' Dol ly Varden, 27% contained Clade N m t D N A and 73% contained what was reported as "bull trout m t D N A " (McPhai l & Taylor 1995). Neither interior nor coastal bull trout m t D N A , however, is distinguishable from Clade S Dol ly Varden m t D N A by R F L P analysis. As 73% m t D N A introgression due to current hybridization is very high compared that observed in the northern Thutade Lake watershed (5.7%, Chapter 4), and as the Skagit River is well within the 42 geographical range of Clade S Dol ly Varden, it is likely that some of these Dol ly Varden had Clade S m t D N A . If they were not Clade S, however, but rather elevated levels of current introgression, then this elevation yields further evidence for selection for bull trout-like m t D N A in southern climates. Historical mtDNA Introgression G H 2 and ITS-1 phylogenies suggested that B . C . Dol ly Varden and bull trout are distinct species, but the results from the m t D N A phylogeny were discordant. A n entire clade of geographically localized Dol ly Varden m t D N A (Clade S) is more closely related to bull trout m t D N A than to the remainder of Dol ly Varden (Clade N) . The Clade S Dol ly Varden even shared a haplotype in common with coastal bull trout. How can this paraphyly be explained? It has been shown that Dol ly Varden and bull trout can and do hybridize (McPhail & Taylor, 1995; Baxter et al. 1997; Chapters 3 & 4). It has also been shown that this hybridization is unidirectional (Dolly Varden male by bull trout female), as revealed by the asymmetric introgression of m t D N A across the species barrier. Further, it is known that m t D N A can introgress permanently across a species barrier (e.g. Ferris et al. 1983; Spolksy & Uzze l l 1984), and that this introgression is often the best explanation for discordancies between mitochondrial and nuclear results (Arnold 1997). I propose that this Dol ly Varden paraphyly is best explained by historical introgression and fixation of coastal bull trout m t D N A into Dol ly Varden. Similar introgression has been observed within several pairs of char species (Bernatchez et al. 1995; Glemet et al. 1998; Wilson & Bernatchez 1998). In these cases, 43 however, the introgressed populations were currently allopatric, so introgression was easily shown to have occurred historically, presumably during temporary bouts of sympatry as deglaciated areas were recolonized. In this case, however, the introgressed Dol ly Varden are currently sympatric with bull trout, and geographically interspersed with 'pure' Clade N Dol ly Varden. In this case, how can one distinguish whether introgression is historic (and has resulted in the permanent presence of bull trout-like m t D N A in Do l ly Varden) or due to current hybridization? There are several factors that indicate that Clade S originated historically. First, the best evidence for historic introgression and fixation is that all Clade S haplotypes appear to be descended from a single ancestral coastal bull trout haplotype (Figure 2). If the introgression were due to current hybridization, it would be expected that other haplotypes present in the coastal bull trout population (i.e. B T - C b and B T - C c ) , or derivatives of these haplotypes, would also be present in Dol ly Varden. There is no evidence, however, of paraphyly within Clade S. Second, and related to the first point, the level of sequence divergence within Clade S suggests that i f a single introgression event occurred, it occurred prior to the Wisconsinan. Although calibration of the mitochondrial molecular clock is dubious at best, commonly used rates of 1% sequence divergences per M y r (e.g. Smith 1992) or 2% per M y r (e.g. Becker et al. 1988; Billingdon et al. 1990; Bernatchez & Dodson 1991) yield estimated divergence times of 100,000 to 400,000 years for the haplotypes within Clade S. As these estimates suggest that the introgression of Clade S predates the Wisconsinan glaciation, historical introgression with fixation is again supported over current hybridization. 44 Finally, Clade S Dol ly Varden are more widely distributed than coastal bull trout. If Clade S had originated due to current introgression, one would expect its distribution to match that of the coastal bull trout population. Clade S Dol ly Varden, however, were found in the Southgate River, which had interior clade bull trout. If current hybridiation had resulted in the introgression of bull trout m t D N A into Southgate Dolly Varden, one would expect the m t D N A to be closely related to that of the local bull trout (i.e. interior bull trout), which is not the case. For example, in a known case of current hybridization an F i genotype hybrid containing bull trout m t D N A by R F L P was shown by sequencing to have a haplotype identical to interior bull trout (BT-Ia), the local bull trout clade (Thutade Lake). Perhaps more substantial geographical evidence against current hybridization is that Clade S Dol ly Varden were found on Vancouver Island, which does not have any bull trout populations, precluding current hybridization as an explanation. It is also worth noting that the sequence divergence between bull trout and Dol ly Varden is much greater for nuclear markers (GH2: 4.31-5.00%, ITS-1: 4.12-6.08%) than for m t D N A (1.2-2.6% between Clade N and bull trout). It is possible that this difference is due to the fact that the nuclear sequences are from an intron and an ITS while the m t D N A region sequenced contains coding gene sequences. It is an interesting conjecture, however, that a hybridization event between Dol ly Varden and bull trout prior to the creation of Clade S resulted in this decreased level of mitochondrial sequence divergence. Further, if the Salvelinus sp. individual is in fact a Dol ly Varden, it is almost as distantly related to Clade N Dol ly Varden as to bull trout, suggesting that most Dol ly Varden m t D N A haplotypes in North America originated in bull trout. 45 In summary, nuclear phylogenetic evidence supported the specific status of Dol ly Varden and bull trout in British Columbia. Mitochondrial D N A phylogeography produced evidence that Dol ly Varden survived the Wisconsinan glaciation in a refuge south of the ice sheet in addition to the commonly understood Bering Refuge. Mitochondrial D N A phylogeny was discordant with nuclear phylogeny, providing evidence for historical introgression between Dol ly Varden and bull trout. In Chapter 3,1 wi l l attempt to determine the frequency of sympatry and hybridization throughout the zone of range overlap of bull trout and Dol ly Varden. 46 Chapter 3: Widespread local sympatry and evidence of hybridization between Dolly Varden and bull trout throughout British Columbia 47 Introduction In Chapter 2,1 described the phylogeny and phylogeography of bull trout and Dol ly Varden. However, one missing piece to the background of the bull trout / Dol ly Varden story is that there is very little known about local sympatry and natural hybridization in the area of Dol ly Varden and bull trout range overlap (Figure 1). In 1978, Cavender found Dol ly Varden and bull trout to be sympatric in a limited number of drainages: the Skeena, Taku, and Liard rivers in northern British Columbia and Puget Sound in Washington State. He reported that their morphological distinctness held in sympatry, but did not report any evidence of hybridization or local sympatry (i.e. individuals of both species present at the same place at the same time). A subsequent morphological study looked at the species' ranges in greater detail, including better coverage of the mid to southern coast of B . C . (Haas & McPhai l , 1991). This study found the two species to be sympatric all along the mainland coast of B . C . and in the headwaters of two interior drainages, the Liard River and the Fraser River. Haas & McPha i l (1991) assumed that the sample from the Fraser River was a museum error, because all other samples from this area were bull trout. The recent discovery of Dol ly Varden in the headwaters of the Peace River, however, another interior drainage with no previously identified Dol ly Varden populations (Baxter et al. 1997; below), supports the authenticity of their sample. Haas & McPhai l (1991) reported ten sites with local sympatry (in the Stikine, Nass, Skeena, Fraser, and Tahtsa rivers), but did not find any conclusive evidence of natural hybridization. Genetic evidence of sympatry has shown Dol ly Varden and bull trout populations to coexist in three watersheds to date. The Skagit River in southwestern B . C . (McPhail & 48 Taylor 1995), the Quinault River in Washington (Leary & Allendorf 1997), and the Thutade Lake watershed in north-central B . C . (Baxter et al. 1997) all contain both bull trout and Dol ly Varden individuals, as identified by allozymes and nuclear and mitochondrial D N A . Evidence of natural hybridization between the two species was found at two of those sites (Skagit River and Thutade Lake). It was noted in 1997 that all evidence of sympatry and hybridization between Dol ly Varden and bull trout is in recently deglaciated areas (Baxter et al. 1997), but a non-hybridizing sympatric population on the Olympic Peninsula (Leary & Allendorf 1997) is a recently discovered exception. In Chapter 2,1 discussed phylogeographic evidence that showed that Dol ly Varden did survive glaciation in a refuge, the Chehalis River valley, south of the ice sheets. The current Washington population, therefore, has presumably been sympatric with bull trout since before the Wisconsinan glaciation. The lack of current hybridization in this historically sympatric population supports Baxter et al.'s (1997) hypothesis regarding the cause of hybridization: that contact between previously allopatric populations in recently deglaciated areas led to hybridization because reproductive isolation was as yet incompletely evolved. In fact, reproductive isolation has been shown to increase in sympatry (e.g. Albuquerque et al. 1996), and in one case, even to increase with respect to sympatric refugial history over the last glaciation (Dowling et al. 1997) The purpose of the research I present in Chapter 3 of my thesis was threefold. Firstly, I wanted to determine the prevalence of local sympatry within the sympatric range of Dol ly Varden and bull trout. This characteristic is associated with the potential for interspecific hybridization: i f bull trout and Dolly Varden are not found together in 49 the same place at the same time, they cannot hybridize. Secondly, I wanted to determine how widespread hybridization is in the locally sympatric sites. Is it a rare phenomenon, or is it commonly associated with sympatry? Finally, i f levels of hybridization varied between sympatric sites, I wanted to determine whether there was an association with other characteristics, particularly the phylogeographic history of the local char populations. I have employed molecular assays of two nuclear loci with fixed allelic differences between Dol ly Varden and bull trout to screen for local sympatry and for hybridization in populations spread throughout the area of range overlap. Mater ia ls and Methods Sample Locations Samples were collected from nine river systems within the sympatric range of Dol ly Varden and bull trout, throughout British Columbia (Table 1). These samples ranged in size from 9 to 990 individuals. I obtained the nine samples opportunistically from seven different sources (private consultants, B . C . Ministry of Fisheries officials, my own collections) so sampling methods are primarily unknown and likely varied. A s most fin clips came from smaller individuals, however, it is presumed that electro shocking was a common method. Samples consisted of fin clips, with or without length and size information. Field identification by consultants typically identified individuals as 'char' at a minimum, or identified the entire sample as either bull trout or Dol ly Varden. Only in Thutade Lake watershed was consistent effort made in the field to identify older juveniles by species (this sample site is described in great detail in Chapter 4). 50 Molecular Methods Molecular markers (Table 2) and laboratory methods for D N A extraction, P C R conditions (Table 3), R F L P analysis and product visualization are not described here, as they are thoroughly described in Chapter 4 (pp 71-73). I scored two nuclear loci for all individuals from the nine sites, Growth Hormone 2 intron C (GH2) and Metallothionine ( M T B ) . I scored a further sub-sample of Dol ly Varden individuals from each site (as identified by nuclear markers) for the Hind III restriction site in the N A D H 5 / 6 m t D N A gene. Hybrid Identification and Genetic Analysis A s both G H 2 and M T B are nuclear loci with fixed, species-specific differences in P C R product fragment size (Table 2, pp 71-72), hybrids were easily identified as any individual containing alleles from both species, in whatever proportion. I tested samples that contained 'pure' individuals from both species (i.e. locally sympatric populations) for deviation from expected proportions of homozygotes and heterozygotes under random association of gametes. This was done using the H W exact test (Weir 1990), as calculated by G E N E P O P (Raymond & Rousset 1995; further described in Chapter 4, pp 75-78). As a measure of heterozygote deficit (Jiggins & Mallet 2000), Weir & Cockerham's (1984) estimate of the inbreeding coefficient (Fi S) was calculated, also using G E N E P O P . A sequential Bonferroni's correction (Rice 1989) was used to prevent artificial inflation of a due to multiple tests (6 samples by 2 loci =12 tests). I used mitochondrial D N A analysis of Dol ly Varden to determine whether a population contained southern (Clade S) or northern (Clade N) Dol ly Varden m t D N A (as 51 described in Chapter 2). Unfortunately, Clade S Dol ly Varden m t D N A cannot readily or dependably be distinguished from bull trout m t D N A with R F L P analysis alone. This m t D N A analysis could not, therefore, distinguish current introgressive hybridization (i.e. Dol ly Varden backcrosses with bull trout mtDNA) from the historic introgression that produced Dol ly Varden Clade S. Analysis of m t D N A by R F L P , therefore, shows with certainty only whether or not Clade N was found in my sample. Results and Discussion Within the overlapping ranges of Dol ly Varden and bull trout, two-thirds of the sites sampled (six of nine) were locally sympatric (Table 6). 'Local sympatry' is defined as having 'pure' individuals from both species present in my sample. (If sympatry is instead measured at the level of the allele, eight of nine sites contained alleles from both species.) This widespread local sympatry suggests that, within the overlapping ranges of bull trout and Dol ly Varden, the two species are often locally sympatric for at least part of their life history. The one site that was not locally sympatric by either criterion was the Mamquam River, for which my sample contained only Dol ly Varden. Both bull trout and Dol ly Varden are found in the Mamquam River (Taylor et al. 1999; Chapter 2), but my sample sites were located upstream of two waterfalls (10 and 20 metres), either of which would prevent upstream migration. It is not surprising that a reach of stream isolated above a barrier would support a resident population of only one of the two species, given expectations of competitive exclusion and the fact that, when in sympatry, bull trout and Dol ly Varden adopt alternate life histories (migratory adfluvial and stream resident, 52 Table 6: Distribution of hybridization throughout the sympatric range of Dol ly Varden and bull trout. Sample sites are listed from north to south (as in Figure 6). A l l eleven sites fall within the sympatric range, although only eight were locally sympatric (i.e. 'pure' individuals from both species were present). The nine sites analyzed in this study were scored using the Growth Hormone 2 and Metallothionine nuclear loci, and a hybrid is defined as any individual containing alleles from both species. The frequency of bull trout alleles refers to the number of bull trout alleles in the population divided by 4n (i.e. 2 diploid nuclear loci). Mitochondrial D N A (mtDNA) clade refers to the type of Dol ly Varden m t D N A ( N = northern, S = southern, as per Chapter 2) and the type of bull trout m t D N A (I = interior, C = coastal, as per Taylor et al. 1999) present at a site. In the hybrid genotypes, b is used to indicate homozygosity for bull trout alleles, v to indicate homozygosity for Dol ly Varden alleles, and H to indicate heterozygosity. Sample Site # # # % Freq. M t D N A Hybrid genotypes B T D V Hyb. Hyb. B T clade ( G H / M T B ) Tahltan River 7 0 2 22.2 0.778 N , I 2 H / H Chutine River 1 15 3 15.8 0.105 N , I b/v, 2 H/v Iskut River 13 33 1 2.1 0.282 N , I H/v Thutade Lake 1 580 354 56 5.7 0.614 N , I 17 H/v, 5 v /H , 9 H/b, 12b /H, 11 H / H , 2 b/v Omineca River 2 25 0 0 0.074 S , I -Goathorn River 47 44 0 0 0.516 N , I -Southgate River 11 10 7 25.0 0.482 s,i 2 H / H , 4 H/v, 1 b/v Toba River 0 9 1 10.0 0.025 s,i4 H/v Mamquam R. 0 18 0 0 0 ??, c -Skagit River 2 8 81 12 11.9 0.137 N + S', p 12 H / H (GH1/GH2) Quinault River 3 5 20 0 0 0.20 s, c -1 These results contain only the hybrids identified using GH2 and M T B , for better comparison to the other sample sites. Complete results for Thutade Lake watershed are described in Chapter 4. 2 McPhail & Taylor (1995), using 2 independent Growth Hormone loci 3 Leary & Allendorf (1997), using 5 diagnostic allozymes 4 Interior clade by geographic locality (E.B. Taylor, pers. comm.) 5 Mitochondrial D N A analysis of nuclear Dolly Varden observed 27% Clade N and 73% "bull trout" mtDNA. Given the southern locality of the Skagit River, the "bull trout" mtDNA is just as likely to be Dolly Varden Clade S mtDNA. 53 respectively) (McPhail & Taylor 1995; Baxter et al. 1997; Hagen 2000). A n important condition for the sympatric coexistence of Dol ly Varden and bull trout might, therefore, be the capacity of the local habitat to allow for these two alternate life histories (resident and migratory). Evidence of hybridization, defined as the presence of both bull trout and Dol ly Varden nuclear alleles within a single char individual, was observed throughout B . C . , from the Skagit River on the B.C./Washington border to the Stikine River in Northern B . C . (Table 6, Figure 6). Evidence of hybridization was seen in four of the six sample sites found to be locally sympatric in this study, or in five of eight sites when the results from two previous molecular studies are included (McPhail & Taylor 1995; Leary & Allendorf 1997; Table 6, Figure 6). Evidence of hybridization was also found at two of the three sites that contained 'pure' individuals from only one species (only bull trout in the Tahltan River and only Dol ly Varden in the Toba River; Table 6). Hybridization was not observed at all locally sympatric sites and levels of hybridization varied greatly between sites, from a low of 2.1% in the Iskut River to a high of 25% in Southgate River (Table 6). This begs the question of why levels of hybridization (or perhaps simply evidence of hybridization) should vary so widely between sample sites. M y sample sizes varied greatly between sites (from n = 9 to n = 990), which may have influenced my ability to characterize levels of local hybridization. However, more thoroughly sampled sites did not produce consistently higher (or lower) percentages of hybrids in the population. Nor did my detailed analysis of Thutade Lake watershed (Chapter 4) show any evidence of distributional "clumping" of juvenile hybrids such that small sample sizes might be expected to hit (high bias) or miss (low 54 f%| - Dolly Varden 9 - Possible areas of overlap ~J - Bu l l Trout 0 - Hybr id izat ion ^ - A reas of Over lap Q • N o hybr id izat ion Figure 6: Geographic distribution of sympatric sites with (darkened circles) and without (empty circles) nuclear evidence of hybridization. The results for Site 1 are from McPhai l & Taylor (1995), for Site 2 from Leary & Allendorf (1997). The location of the Thutade Lake watershed, the study site in Chapter 4, is also indicated. 55 bias) hybrids. The most obvious biological answer for the variation in percentages of hybrids found is that some sites simply do not provide the necessary ecological or habitat conditions for hybridization, for whatever reason. Unfortunately, very little information is available about these sample sites regarding local habitat, ecology, distribution of spawning sites, etc., so any explanations would be very speculative. There are, however, several large-scale characteristics that could be compared to find any patterns to explain this variation in hybridization. Firstly, in hybridization between Lepomis spp., it has been proposed that levels of hybridization are associated with the relative abundance of the parental species in the system (Avise & Saunders 1984). In this case, however, hybridization does not change with the abundance of either parental species in the system. With increasing frequency of bull trout alleles (p) in the population, % hybrids in the populations were 10.0, 0, 15.8, 2.1, 25.0, 0, 5.7, and 22.2% (Table 6), and regression analysis found no significant change in % hybrids as p increased (He,: (3 = 0, t = 0.84, p = 0.4331). Nor does hybridization increase as the relative abundance of the parental species changes: as the relative abundances diverge from 50/50 towards 100/0, hybridization changes from 0, 25, 5.7, 2.1, 22.2, 15.8, 0, and 10.0% (Table 6; non-significance under regression analysis, HQ: (3 = 0, t = -0.07, p = 0.8673). A similar lack of association between parental abundance and levels of hybridization was obtained from a detailed study of five tributaries to Thutade Lake (comparing levels of hybridization between tributaries to the lake with different local proportions of bull trout and Dol ly Varden; Table 10 in Chapter 4). 56 Hybridization did not vary with latitude: from north to south, % hybrids in the populations are 22.2, 15.8, 2.1, 5.7, 0, 0, 25.0, and 10.0%. Nor did it differ significantly between the interior (0 and 5.7%) and coastal (0, 2.1, 10.0,15.8, 22.2, and 25.0%) drainages that I sampled (t = 1.239, p = 0.2617). A potential source for coastal/interior differentiation is that in the interior drainages, Dol ly Varden are known only from extreme headwater populations. These Dol ly Varden populations are geographically isolated from the remainder of the Dol ly Varden species (present in the coastal drainages). A well-known model regarding hybrid zones suggests that zones are maintained by a balance between the dispersal of parental individuals into the hybrid zone and subsequent endogenous selection against the hybrids produced (the Tension Zone Model ; Barton 1979; Barton & Hewitt 1985). Under this model, these interior Dol ly Varden populations should collapse into a hybrid swarm because 'pure' Do l ly Varden cannot migrate into the system to replace the Dol ly Varden 'lost' to hybridization. Dol ly Varden most likely colonized the interior drainages 14,000 years ago via headwater transfer (McPhail and Lindsey 1970) and still remain genetically and ecologically distinct from bull trout in the interior drainages (Baxter et al. 1997; Hagen 2000; Chapter 4). Reproductive isolation would have to have been rapidly and firmly established to allow Dol ly Varden to persist for this length of time, thus resulting in lower expected levels of hybridization in interior drainages. Unfortunately, the primarily coastal distribution of Dol ly Varden limited the number of interior sympatric populations that could be sampled to test this hypothesis. Whether hybridization is higher for a phylogenetic group with an allopatric refugial history (i.e. using the m t D N A clade as an indicator of allopatric vs. sympatric 57 refugial history) is more difficult to answer, due to the small number of sample sites and the geographic distribution of the two clades (Figure 3). As both Dol ly Varden clades were represented within my study, I could test for differences in hybridization between clades. I saw no evidence for a difference in mean proportion of hybrids between Dol ly Varden Clade N (0, 2.1, 5.7, 15.8, and 22.2%) and Clade S (0, 10.0, 25.0%) (t = -0.323, p = 0.7575). Unfortunately, all of the sympatric sites from this study fall within the range of interior bull trout (Taylor et al. 1999), so no test for differences between bull trout clades could be done. The Skagit River (McPhail & Taylor 1995) and the Quinault River (Leary & Allendorf 1997), however, fall within the geographical distribution of the coastal bull trout clade and contained 11.9% (Clade N & possibly Clade S Dol ly Varden) and 0% (Clade S Dol ly Varden) hybrids respectively. If the hypothesis of hybridization via secondary contact were supported, we would expect that populations with coastal bull trout and/or southern Dol ly Varden (i.e. from the sympatric Chehalis Refuge, Chapter 2) to have lower levels of hybridization than those with interior bull trout and/or northern Dol ly Varden. N o difference in level of hybridization was shown between the bull trout clades (due to lack of samples) or between the Dol ly Varden clades (also possibly due to small sample sizes). In terms of interspecific clade pairs, however, the only clade pair that did not produce any evidence of hybridization was Clade S Dol ly Varden with coastal bull trout. Unfortunately, as only one sample site contained this clade pair (Quinault River; Leary & Allendorf 1997) and as there is great variation in levels of hybridization between sites, the hypothesis of variation in reproductive isolation due to phylogenetic history is impossible to test with my data. 58 Regardless of how much the levels of hybridization vary across the province, certain characteristics of the hybrids revealed very interesting trends. First, Growth Hormone 2 is more commonly heterozygous in hybrids than Metallothionine. In the combined sample, there were 49 individuals heterozygous for G H 2 and only 32 heterozygous for M T B . Although the pattern holds across five of the six sites ( G H 2 / M T B : 37/28, 2/0, 1/0, 6/2, 1/0), the difference is not significant (pooled %2 = 3.16, p ~ 0.07). The one exception to the trend, the Tahltan River, had equal heterozygosities for the two loci (2/2). I wi l l discuss this disparity in heterozygosity between loci in detail in Chapter 4. Second, expected values calculated for sympatric populations under Hardy-Weinberg Equilibrium expectations yielded a fairly consistent pattern: fewer heterozygotes were observed than expected, yielding positive FTS values ranging from 0.506 to 1.0 (Table 7). The exact H W test (Weir 1990) showed significant differences from expected values at all sites and loci but one (using the sequential Bonferroni correction; Rice 1989). The single exception, the Chutine River at the M T B locus, did not have any heterozygotes at all, but rather failed to deviate from expected values due to the unequal proportions of parental individuals (Table 7). These Fis and exact H W test results show that, although hybridization is widespread, significantly fewer heterozygotes (i.e. hybrids) are observed than expected i f the char formed a randomly mating population (i.e. F i S = 0). Even the Southgate River, with the highest observed proportion of hybrids (25.0%), contained a non-randomly mating char population. 59 Table 7: Inbreeding coefficient (Fis; as per Weir & Cockerham 1984), a measure of heterozygote deficit, for Growth Hormone 2 and Metallothionine in six locally sympatric sample sites. Fis values can extend f rom-1 to +1, with +1 being heterozygote deficit, -1 heterozygote excess, and 0 equivalent to expected under random mating. The p-values listed are from the exact Hardy Weinberg test (Weir 1990), which was done for each locus in all tributaries. In all cases but one (marked with an *, a = 0.025), the p-value was less than the associated Bonferroni sequentially corrected a (Rice 1989), indicating significant deviation from Hardy Weinberg expectations (i.e. random mating). Sample Site G H 2 M T B Fis p-value Fis p-value Chutine River 0.621 0.0340 1.0 0.0270* Iskut River 0.949 0.0000 1.0 0.0000 Thutade Lake 1 0.914 0.0000 0.934 0.0000 Omineca River 1.0 0.0011 1.0 0.0011 Goathorn River 1.0 0.0000 1.0 0.0000 Southgate R. 0.581 0.0027 0.859 0.0000 These results contain only the hybrids identified using GH and MTB, for better comparison to the other sample sites. Complete results for Thutade Lake watershed are described in Chapter 4. 60 In conclusion, my opportunistic sampling has provided evidence that local sympatry is common across the sympatric range of Dol ly Varden and bull trout. Hybridization is also fairly common, but apparently varies greatly between populations. However, at no sample site have the two species 'collapsed' into a single randomly mating hybrid swarm, implying that some factor(s) must be acting to maintain two distinct gene pools. In Chapter 4,1 wi l l look at the Thutade Lake watershed char population in detail, to examine the dynamics of a hybridizing population. Such a study is important in order to understand how the two species are maintained in sympatry. 61 Chapter 4: Directional hybridization, bi-directional backcrossing, and the asymmetrical introgression of alleles across a species barrier in a Dolly Varden / bull trout hybrid zone 62 Introduction Barriers to Reproductive Isolation Barriers to natural hybridization fall into two major categories: prezygotic and postzygotic (or pre- and postmating; Arnold 1997). Prezygotic barriers, i.e. those that prevent hybridization from occurring, typically involve mating behaviour and/or gamete recognition. Mating behaviour can prevent, or limit, hybridization by several mechanisms, such as assortative mating or the timing or location of mating. Gamete recognition can also play an important role, and in some cases has been shown to almost completely counter the effects of random mating between species (e.g. Howard et al. 1998; Rieseberg et al. 1998). A recent review of the hybrid zone literature points out that prezygotic barriers to hybridization are most frequently associated with bimodal hybrid zones, zones in which both parental genotypes predominate over the less common hybrid (i.e. intermediate) genotypes (Jiggins & Mallet 2000). The few hybrids produced in these cases are, presumably, subject to some sort of postzygotic barriers. Postzygotic barriers take the form of endogenous and/or exogenous selection against hybrids (Arnold 1997). These two types of selection, and their relative importance, have influenced much of the debate regarding the maintenance of hybrid zones over time. The commonly applied Tension Zone Model (Barton 1979; Barton & Hewitt 1985) assumed that hybrid zones were maintained by a balance between endogenous selection against hybrids, and the dispersal of parental individuals into the zone. This model assumes that all genotypic classes of hybrids were less fit than parental individuals, irrespective of environment. The Bounded Hybrid Superiority Mode l (also known as the Ecotone Model ; Moore 1977) assumed that selection against hybrids was 63 environmentally dependent, and that hybrids could be more fit than parentals, but only in the intermediate 'ecotonal' habitat that delimits the hybrid zone. More recently, scientists have begun to acknowledge that the actions of exogenous and endogenous selection are not mutually exclusive (Arnold 1997). Studies have also begun to reveal the action of both types of selection in maintaining particular hybrid zones, using cohort analysis (i.e. the measurement of proportions of hybrids in a cohort as it ages) (e.g. Bert & Arnold 1995). Studies have also shown that not all hybrids are alike and that the various genotypic classes of hybrids can be more, less, or equally fit to the parental species and that relative fitness can vary with respect to environment (reviewed in Arnold & Hodges 1995). Sympatric Biology of Dolly Varden & Bull Trout Dol ly Varden and bull trout have been shown to hybridize throughout the sympatric portion of their parapatric ranges (Chapter 3). Mitochondrial D N A evidence has shown that introgressive hybridization has also taken place historically (Chapter 2). Little information, however, is available regarding the dynamics of contemporary interspecific hybridization between Dol ly Varden and bull trout, nor regarding the fate of hybrids. This chapter recounts the results of a detailed population study in Thutade Lake watershed, in north-central British Columbia. In allopatry, both Dol ly Varden and bull trout have primarily migratory life histories, typically anadromous and adfluvial, respectively. Anadromous Dol ly Varden mature at large sizes (30 - 75 cm), although they are still slightly smaller than mature adfluvial bull trout (40 - 90 cm). Both species have an inherent phenotypic plasticity 64 regarding life history, however, and produce stream resident populations of small body size that are often associated with a physical barrier to migration. This type of intraspecific life history subdivision is common in many species of temperate fishes (Schluter 1996). The Thutade Lake watershed is located north central B . C . in the headwaters of the Peace River, an interior tributary to the Mackenzie River Arctic drainage system (Figure 6). As in most systems where bull trout and Dol ly Varden co-exist, the two species adopt alternate life histories in Thutade Lake watershed. Bu l l trout are migratory, with an adfluvial life history. As adults, they are piscivorous, feeding on kokanee (Oncorhynchus nerka) in Thutade Lake (Hagen 2000). They mature at large body sizes (40 to 87 cm), and each summer they migrate up and hold in tributary streams before spawning in late August through early September. Dol ly Varden are invertebrate feeders residing permanently in tributaries to the lake. They mature as small as 12 cm in length, grow to a maximum of 21 cm, and are also repeat spawners, mating in late September/early October. Juveniles of both species are invertebrate feeders in the tributaries and have not been found to have any significant ecological differences (Hagen 2000). B u l l trout juveniles, however, are opportunistically piscivorous and begin a size-dependent downstream migration to the lake by around three years of age. These alternate life histories have some interesting implications for population dynamics and hybridization in the watershed. As Dol ly Varden are stream resident, and have not been captured in the lake, it is likely that the Thutade population of Dol ly Varden is sub-divided between tributaries. Migratory bull trout, however, could potentially form a panmictic population in the watershed. Some evidence of homing is 65 available for bull trout (Latham 2000 suggested that there is subdivision in the Arrow Lakes, B . C . ) , but the strength of homing instinct and abilities are unknown. M y study is the first detailed look at natural hybridization between Dol ly Varden and bull trout, in an attempt to describe the dynamics of the hybridizing population. I make very little use of morphological analysis, because other than adult body size, Dol ly Varden and bull trout are very similar morphologically, and generally require a set of three characters and a linear discriminant function to consistently distinguish between them (Haas & McPha i l 1991). Molecular markers with fixed, diagnostic allelic differences between species allow easy and certain identification of hybrids. The nuclear markers w i l l allow distinctions to be seen between the various genotypic classes of hybrid present in the population, while the use of mitochondrial D N A (mtDNA) wi l l allow information to be gleaned as to directionality of interspecific crosses and any post-Fi mating. M y overall objective is to determine whether Dol ly Varden and bull trout maintain themselves as distinct species while sympatric and, i f so, to clarify the mechanisms preventing disintegration into a local hybrid swarm. Mater ia ls and Methods Study Site Thutade Lake watershed is located in the southwestern headwaters of the Mackenzie River system in north-central B . C . (Figure 6). It connects to the Mackenzie River drainage via the Finlay and Peace rivers. Thutade Lake is a large lake, about 40 km long, fed by several tributary rivers and streams (Figure 7). It was essentially pristine 66 Thutade Watershed to Finlay f A River I "S Attycelley River Figure 7: Thutade Lake watershed. Sample sites are marked by a cross, while the sites which produced hybrid individuals are encircled. The boxed area contains 17 sample sites and is shown in greater detail in Figure 8. 67 until 1997, when an open-pit mine was built near Kemess Creek and South Kemess Creek was dammed for use as a tailing pond. Environmental requirements have resulted in efforts to ensure that sedimentation into Kemess Creek is minimal while waters from South Kemess Creek have been diverted around the dam to ensure that overall flow in Kemess Creek is not reduced. Sampling was done on two geographic scales. On a large scale, sample sites were spread throughout the watershed, in five tributaries to the lake (Figure 7). On a fine scale, Kemess Creek and its tributaries were sampled in great detail in order to examine species and hybrid distributions within a tributary (Figure 8). Juvenile Collection Sampling took place over a two week span from the middle to end of August in both 1997 and 1998. Sampling was done by electroshocking and included two successive double-passes (moving upstream then downstream) of a 30 to 100 meter long section of stream, blocked at both ends by nets. In 1997, all fry (i.e. young of the year, defined as < 55 mm in length) caught at a site were sacrificed and stored whole in 95% E t O H . In 1998, non-lethal fin clips were taken from most fry and all juveniles caught at each site (caudal or adipose fin, depending on the size of the individual), again stored in 95% E t O H . Attempts were made to visually identify juveniles (> 55mm) collected in 1998 by species. The primary character used was the approximate ratio of upper jaw length to standard body length, known to be larger in bull trout than Dol ly Varden (Haas & McPhai l 1991). The presence (Dolly Varden) and absence (bull trout) of dorsal parr 68 Figure 8: Kemess Creek. Sample sites are marked by crosses, and the solid bar represents the location of the dam for the tailings pond, built in 1997. Areas of groundwater upwelling (preferred Dol ly Varden spawning sites) are marked by cross-hatching. The sites that contained hybrids are shown in Figure 10. 69 marks was also considered (J. Hagen, pers. comm.). Individuals for which the characteristics yielded uncertainty regarding species designations were labeled hybrids. Branchiostegal ray counts, the most diagnostic morphological trait between Dol ly Varden and bull trout (Haas & McPhai l 1991), were not taken, due to time constraints and potential injurious effects on the juveniles. Samples were grouped into size class categories. Fieldwork over the summer had shown that fry and juveniles from different age classes fell fairly neatly into particular size categories (J. Hagen, pers. comm.). The age classes and associated size categories (fork length in mm) are: fry/0+ = <55 mm, 1+ = 55-80 mm, 2+ = 81-110 mm, 3+ = 110-140 mm, and >3+ = >140 mm. The 3+ and >3+ age class Dol ly Varden individuals are large enough that they may have included mature individuals, but due to the lack of secondary sexual characteristics in Thutade Dol ly Varden such a distinction could not be made. Similar sized bull trout would not be mature, and the age/size of maturation of hybrids is unknown. Due to small collection sizes, I grouped the 2+, 3+ and >3+ age classes into a single age class, >1+, for analysis. Adult Collection Adult bull trout were angled from holding pools on Attichika River over the summer and early fall of 1998 (n = 29). Individuals were measured by length, sex was recorded i f obvious, and adipose fin clips taken and stored in 95% E t O H . In early summer 1998, adults were also netted from Thutade Lake (n = 15). Each of these adult individuals was scrutinized to be certain that it was not a Dol ly Varden, and none had any obvious intermediate morphological characteristics to imply that they might be hybrids. 70 Adult Dol ly Varden were probably collected during electroshocking, but due to the lack of obvious secondary sexual characteristics to indicate sexual maturity and due to the life-long stream resident life history, they could not be distinguished from juvenile Dol ly Varden. For this reason, any adults that may have been taken have been grouped with the juvenile Dol ly Varden. DNA Extraction The D N A from the 1997 samples and some of the 1998 samples was extracted from approximately 20 mg of adipose or caudal fin tissue by overnight digestion at 37°C in a 0.5% Sarcosyl/0.2M E D T A solution with Pronase, followed by a single phenol/chloroform extraction (as described in Taylor et al. 1996). D N A was precipitated in isopropanol, washed in 70% E t O H , and the D N A pellet was either dried in a vacuum-condenser for 5 minutes or left overnight to air dry. The pellet was resuspended in 75 to 100 |xL T E buffer (pH 8.0) and stored at -20°C. Most 1998 extractions were done using the G E N T R A Puregene D N A Extraction Ki t . The only differences from the above procedure were that I digested tissue at 55 °C in Puregene Ce l l Lysis solution with 10 u L of 1 mg/mL Proteinase K solution, and removed protein using the Puregene Protein Precipitation Solution. Genetic Markers Codominant nuclear markers with diagnostic species differences were used to identify hybrid individuals (Table 2). Growth Hormone 2 (GH2) is a coding gene, but the G H 2 primers amplify an intron, intron C (McKay et al. 1996). Metallothionine 71 ( M T B ) is also a coding gene, and again, the primers used amplify an intron to the coding sequence (J. Baker, unpubl. data, U . Wash., Seattle). In both of these markers, a significant deletion and/or insertion mutation has occurred in the intron, producing P C R amplified fragments with visibly different lengths for the two species (Table 2). The ribosomal D N A primers amplify the first internal transcribed sequence (ITS-1) of the r D N A , a transcribed region of spacer D N A between r D N A genes. The ITS-1 locus has a diagnostic Sma I restriction site present only in bull trout (Phillips et al. 1995). A short interspersed repetitive element (SINE) from the Fok family ( F O K ) of S INE ' s (retroposons originating from t R N A , with random reincorporation into the genome via a c D N A intermediate) has an insertion site present only in Dol ly Varden. This insertion produces a diagnostic difference in P C R amplified fragment lengths (Table 2). These nuclear markers were species diagnostic (i.e. bull trout and Dol ly Varden were fixed for alternate alleles at all loci), which was demonstrated by sampling and genetic analysis throughout the allopatric ranges of both species (n = at least 25 per species, per locus). G H 2 and ITS-1 were previously tested for diagnosticity (McPhai l & Taylor 1995; Baxter et al. 1997), while I tested M T B and F O K for diagnosticity for this thesis. No alternative alleles for any of these loci were observed within the 990 juveniles and 44 adults sampled from Thutade Lake watershed for this study. A mitochondrial marker was used to identify the mitochondrial cytotype of individuals (McPhail & Taylor 1995; Baxter et al. 1997). Only a single mitochondrial marker is required as m t D N A is maternally inherited and non-recombinant. The mitochondrial N A D H 5 / 6 dehydrogenase coding gene (mtDNA) has a Hind III restriction site present only in Dol ly Varden. More recent work on the biogeography of Dol ly 72 Varden, however, has shown that the lack of a Hind IH restriction site is not diagnostic for bull trout (Chapter 2). Southern populations of Dol ly Varden contain m t D N A closely related to bull trout m t D N A , which Hind III restriction analysis identifies as bull trout. This geographically localized paraphyly is likely caused by historical introgression of bull trout m t D N A into Dol ly Varden. Thutade Lake watershed, however, is outside the geographic range of this introgressed mtDNA, and so this paraphyly did not affect analysis of local, current hybridization. PCR and RFLP Analysis P C R reactions were run with varying conditions for the five markers (Table 3). A typical reaction consisted of a total volume of 25 jxL with 800 u M of total dNTPs, 0.6 ,uM of each primer, 1.5 m M M g C k , 1 x G i b c o B R L Taq D N A polymerase buffer, 3.75 units of Taq D N A polymerase, and 1 pJL of whole D N A . Denaturing steps were done at 95°C for 1 cycle, and at 94°C or 92°C for the remaining steps, extension steps were done at 72°C, and annealing temperatures varied (Table 3). Restriction digests were performed as per manufacturer's instructions (New England Biolabs), overnight, using 6 ,uL of P C R product in a total volume of 20 ,uL. The P C R amplified and restriction length differences were visualized using 1.5 % agarose gels stained with EtBr . Identification of Hybrid Classes A l l individuals collected from the Thutade Lake watershed in 1997 and 1998 (n = 1034) were analyzed using four nuclear markers (GH2, ITS-1, M T B , and F O K ) and a 73 single mitochondrial D N A (mtDNA) marker. Individuals were identified as being S. malma, S. confluentus, or hybrid using both the nuclear and m t D N A markers. Individuals were considered to be 'pure' S. malma (DV) or S. confluentus (BT) i f they contained only alleles diagnostic for the parental species at all loci, nuclear and mitochondrial. Individuals were considered to be hybrids if they contained any mix of alleles from the two parental species. The hybrids were subdivided into Fi, F n , backcross (BC) or mitochondrially introgressed (IG) genotype classes. F i genotype individuals are heterozygous at all nuclear loci. F n genotype individuals (or reverse backcrosses) are homozygous for alternate species alleles for at least two nuclear loci, with any background for remaining loci. Backcross (BC) genotypes are heterozygous at one to three nuclear loci, with the remaining nuclear loci homozygous for one parental species. These homozygous loci are S. confluentus in a bull trout backcross genotype ( B C B T ) or S. malma in a Dol ly Varden backcross genotype (BCnv)- Mitochondrially introgressed individuals (IG) have a 'pure' nuclear background, with the other species' mitochondrial D N A (i.e. there is no nuclear evidence of hybridization). These four hybrid classes are intended to describe the genotype of individuals, but do not necessarily reflect their parentage. For example, an individual classified as a F i genotype could actually be the result of a cross between a true F i and a parental individual, which would be expected to produce offspring heterozygous at all four nuclear loci 6.25% of the time. On the same note, an individual classified as a backcross genotype could be anything from a first generation backcross to an n t h generation backcross (i.e. essentially introgressed nuclear alleles). Hereafter, the term 'hybrid' w i l l 74 refer any/all of these genotypic classes, while more specific terms wi l l be used in reference to specific genotypic classes. Because only four nuclear markers were used, this type of identification could result in overestimates of numbers of pure parental and F i individuals and underestimates of numbers of backcross individuals. Individuals identified as F i using four nuclear loci could actually be later generation hybrids or backcrosses. This may result in an over-estimation of interspecific hybridization rates. As later generation backcrosses would be expected to have fewer heterozygous loci, an analysis of four nuclear alleles would also be expected to miss some individuals with later-generation hybrid genomes. These individuals would be misidentified as 'pure' parentals, thereby underestimating the total percentage of hybrid individuals in the population. Scoring additional nuclear markers would reduce this source of error (for example the percentage of first generation backcrosses heterozygous at all loci is 6.25, 3.12, 1.56, and 0.78% for 4, 5, 6, and 7 nuclear loci). At some point, however, the increase in precision ceases to justify the effort required in obtaining the information. Population Genetic Analyses I performed population genetic analyses for all loci for each tributary in which both parental species were found. Nuclear evidence of interspecific hybridization was found in all six tributaries as well as evidence that the F i hybrids had backcrossed in both directions. This last fact allowed me to apply population genetic concepts to the char population of Thutade Lake watershed (making hybrids the equivalent to migrants between two semi-isolated populations). 75 I calculated Weir & Cockerham's (1984) estimate of the inbreeding coefficient (Fis) as a measure of heterozygote deficit (Jiggins & Mallet 2000) using G E N E P O P Version 3.Id (Raymond & Rousset 1995). Fis is equivalent to 1 - (PAa/2pAPa), where P A 3 is the observed and 2p A p a the expected frequency of heterozygotes. Potential Fis values extend f rom-1 to +1, with positive values indicating heterozygote deficit, negative values heterozygote excess, and 0 equivalent to expected under random mating. I tested the statistical significance of the Fis using an exact Hardy Weinberg test (Weir 1990) with the HQ: "random union of gametes." This test was performed for each locus in all tributaries, using the complete enumeration method (Louis & Dempster 1987) in G E N E P O P (Raymond & Rousset 1995). I calculated nuclear gametic disequilibria (D 1) for all six permutations of nuclear loci pairs as per Hartl & Clark (1997), with the exception that I ignored the genotypic class of 'double heterozygotes' in calculating D , as it is such a small proportion of the population, and the loss of information was therefore not great (Avise & Van Den Avyle 1984; Campton 1987). D ' is D , the observed disequilibrium, divided by D m , the theoretical maximum disequilibrium. D ' can vary from -1 to +1, with positive values indicating an association between intraspecific alleles at the two loci. I considered D ' values to be statistically significant when there were significant deviations between observed proportions of genotypic classes and the expected proportions under a model of no linkage (physical or otherwise) between genotypes at the two loci (i.e. random mating in the char population with respect to these loci and/or no selection against particular genotypic classes). This was tested using G E N E P O P (Raymond & Rousset 1995), via a 76 contingency table for the probability test (or Fisher's exact test) and a Markov Chain (dememorization 2000, 200 batches, 1500 iterations). I calculated cytonuclear disequilibria (cD) as per Asmussen et al. (1987) for all four possible pairs of nuclear locus and mtDNA. The four cD values calculated are labeled c D i , c D 2 , c D 3 , and cD. c D i through C D 3 refer to the difference between observed individuals in genotypic classes versus those expected given nuclear and mitochondrial allele frequencies. In this case, a positive cD] means that more bull trout m t D N A alleles were associated with bull trout nuclear homozygotes than expected given random association, a positive c D 2 means that more bull trout m t D N A alleles were associated with nuclear heterozygotes than expected, and a negative c D 3 means that more Dol ly Varden m t D N A alleles were associated with Dol ly Varden nuclear homozygotes than expected (Asmussen et al. 1987). cD measures allelic disequilibria, and in this case, a positive cD means that more bull trout nuclear alleles were associated with bull trout m t D N A alleles than expected given random association. As with D ' , I considered cD, values to be statistically significant when there were significant deviations between observed proportions of genotypic classes and the expected proportions under a model of no association between nuclear and mitochondrial alleles (i.e. random mating in the char population with respect to these loci and/or no selection against particular genotypic classes). This was tested using the same method as for D' , but using the R X C computer program (no reference) to calculate a metropolis algorithm to obtain an unbiased estimate of the exact p-value. For all of these tests, I used the sequential Bonferroni correction method to compensate for tests on multiple loci and multiple tributaries, to ensure a maximum Type 77 I error rate of a = 0.05 (Rice 1989). As there were 20 tests for heterozygote deficit, 30 tests for linkage disequilibria, and 20 tests for cytonuclear disequilibria, this produced a total of 70 tests, and a corrected minimum a of 0.05/70 = 0.000714 for comparison to the smallest p-value, with a increasing to 0.05/n as the reverse ranked p-values (i.e. smallest to largest) were sequentially compared to a (Rice 1989). Cohort Analysis Cohort analysis was used to test whether the proportion of hybrids in a cohort changed over time. Both dynamic cohort (following a single cohort over time) and static cohort (looking at several cohorts in a 'snap shot' of time) analyses were done using contingency analysis. A Fisher's exact test was used to test whether the ratio of hybrid to Dol ly Varden individuals changed between age classes (again using the R X C program). Results Morphology O f the 990 juvenile char collected from six tributaries to Thutade Lake, 89 individuals were found to be hybrids, across five genotypic classes, yielding evidence for a bimodal distribution of genotypes in the watershed (Table 8, Appendix 4). O f the 44 adult S. confluentus sampled (species designated because of the associated life history), all 29 angled from Attichika River and 14 of the 15 netted from Thutade Lake were 'pure' bull trout, while a single adult (sampled from the lake) had a F i genotype and bull trout m t D N A . This individual was one of the smaller individuals and had been morphologically identified as a bull trout. 78 Table 8: Genotypes of all juvenile individual collected in tributaries of the Thutade Lake watershed in 1997 and 1998. Genotypes consist of four nuclear loci (GH2, ITS-1, M T B , F O K ) and the mitochondrial haplotype. 'Pure' S. malma and S. confluentus are designated as D V and B T , respectively. Hybrid genotypic classes are designated as F i (heterozygous at all nuclear loci), F n (homozygous for alternate alleles at two or more nuclear loci), backcrosses ( B C B T or B C D V : at least one heterozygous nuclear locus in a 'pure' nuclear background), or m t D N A introgressed ( I G D V : bull trout m t D N A in a 'pure' Dol ly Varden nuclear background). These classes designate genotype only, and are not intended to reflect ancestry. For tributary locations, see Figure 7. Genotypes broken down by age class are included in Appendix 4. Tributary # Sites1 N B T BCBT Fi F 2 1 n BCDV 2 IGDV D V Kemess Cr. 13/17 598 316 9 2 2/1 9/16 4 239 South Pass R. 2/2 64 47 0 0 0 0/2 5 10 Tributary 4 3/3 87 61 5 0 1 0/1 9 10 Attichika R. 1/2 88 74 4 3 3 0 0 0 7 Niven R. 4/5 94 75 10 0 1 1/0 2 5 Attycelley R. 2/2 59 0 0 0 0 1/1 0 57 Total 25/31 990 573 28 5 4/1 11/20 20 328 # / #: number of sample sites with hybrids / total number of sites 2 # / #: individuals with B T mtDNA / individuals with D V mtDNA 3 one of these individuals was heteroplasmic (contained both B T and D V mtDNA) 79 Morphological identification varied in efficacy for the juvenile samples. For fry (young-of-the-year, < 55 mm in length), a rapid morphological identification could not reliably distinguish the species, much less identify hybrids. For older juveniles (> 55 mm), a morphological identification was relatively reliable to species, but failed to detect most hybrids (Table 9). There were some genetically 'pure' parental individuals misidentified: two bull trout and two Dolly Varden misidentified as hybrids and five bull trout misidentified as Dol ly Varden. Only two of the 38 hybrids were correctly identified morphologically. Twenty-six of the 28 hybrids classified as Dol ly Varden were Dol ly Varden backcrosses and 7 of the 8 hybrids classified as bull trout were bull trout backcrosses. Hardy-Weinberg, Gametic Disequilibria and Cytonuclear Disequilibria A l l six tributaries contained hybrids, which formed from 3.4% to 18.4% percent of the tributary population samples (Table 10, Appendix 5). Tributaries also varied considerably in the relative proportions of the two parental species. In terms of allele frequencies, bull trout nuclear D N A (averaged over four loci) ranged from 0.425% to 89.49% of the D N A within a tributary. The four nuclear loci gave similar results within each tributary. The proportion of bull trout m t D N A within a tributary (which ranged from 1.69% to 94.68% within a tributary) was consistently higher than the proportion of bull trout nuclear alleles within the same tributary. There was no obvious relationship between the proportions of the two parental species and the number of hybrids present in a tributary. 80 Table 9: Visual versus genetic identification of juvenile char. Only char greater than 55 mm in length from the 1998 collection were visually identified. Visual identification consisted of a rapid morphological examination, focusing on the ratio of upper jaw length to standard length (shorter, blunter nose in S. malma; Haas & McPhai l 1991) and the presence (S. malma) or absence (5. confluentus) of dorsal parr marks (J. Hagen, pers. comm.). If an individual was intermediate or consisted of a mosaic with regard to these two characters, it was visually identified as a hybrid. B T Visual Identification D V Hybrid B T 114 5 2 Genotypic D V 0 168 2 Identification Hybrid 8 1 28 2 2 3 1 included 7 bull trout backcross genotypes and 1 mitochondrially introgressed Dolly Varden. 2 included 26 Dolly Varden backcross genotypes, 1 F[ genotype, and 1 bull trout backcross genotype. 3 included 1 Fj genotype and 1 bull trout backcross genotype. 81 Table 10: B u l l trout allele frequency and percentage of the sample with hybrid origins. Presented by tributary, with allele frequencies for four nuclear loci and 1 mitochondrial locus. As the two species were fixed for single alternative alleles at each locus, the frequency of Dol ly Varden alleles is 1 - (bull trout frequency). Allele frequency broken down by age class is included in Appendix 5. Tributary S. confluentus allele frequency % G H 2 ITS-1 M T B F O K m t D N A Hybrids Kemess 0.5619 0.5460 0.5468 0.5477 0.5719 7.19 South Pass 0.7500 0.7344 0.7422 0.7422 0.8125 10.94 Tributary 4 0.7644 0.7529 0.7586 0.7471 0.8735 18.39 Attichika 0.9034 0.9034 0.8920 0.8807 0.9195 7.95 Niven 0.8723 0.9043 0.8936 0.8883 0.9468 14.89 Attycelley 0.0085 0 0 0.0085 0.0169 3.39 82 In the Attycelley River, only three of the 531 alleles scored were bull trout, and thus the only 'pure' parental individuals were Dol ly Varden (Table 8). As hybridization at this site is, therefore, limited due to non-sympatry, Attycelley River w i l l be excluded from most of the following analyses. The exact Hardy Wienberg tests (Weir 1990) revealed a significant deviation from H W E expectations at all four nuclear loci (p < 0.0001) for all five sympatric tributaries. In every case, there was a deficiency of heterozygotes, producing positive F i S values ranging from 0.621 to 1.0 (Table 11). Gametic disequilibria were calculated for all six nuclear locus pairs in the five sympatric tributaries (Table 12). Contingency testing showed significant deviation from expected values under the FL,: "genotypes at one locus are independent from genotypes at the other locus" for all locus pairs in all tributaries (p < 0.0001). Calculated D ' values were all positive and ranged from 0.7328 to 1.0 (Table 12). Cytonuclear disequilibria were calculated for all four nuclear loci in the five sympatric tributaries (Table 13). Contingency testing revealed significant deviations from expected values produced under a H 0 : "genotypes at the nuclear locus are independent from cytotype" for all locus pairs in all tributaries (p < 0.000001). Cytonuclear disequilibria (Table 13) varied much more between tributaries than did F ] S (Table 11) and linkage disequilibria (Table 12). A l l three genetic statistics, however, showed similar results between loci within a tributary. 83 Table 11: Inbreeding coefficient (FK ; Weir & Cockerham 1984), a measure of heterozygote deficit, for four nuclear loci in five tributaries to Thutade Lake. Potential values extend f rom-1 to +1, with positive values indicating heterozygote deficit, negative values heterozygote excess, and 0 equivalent to expected under random mating. A n exact Hardy Weinberg test (Weir 1990) was performed for each locus in all tributaries, using the complete enumeration method (Louis & Dempster 1987) in G E N E P O P (Raymond & Rousset 1995). In all cases an exact p-value of less than 0.0001 was obtained, indicating significant deviation from Hardy Weinberg expectations (i.e. random mating). Attycelley River was excluded from analysis due to non-sympatry. Tributary G H 2 ITS-1 Fis M T B F O K Kemess South Pass 0.918 1.0 0.960 0.960 Tributary 4 0.968 0.847 0.938 0.940 Attichika 0.807 0.807 0.708 0.625 Niven 0.621 0.757 0.778 0.734 84 Table 12: Linkage disequilibria (D') for all pairs of nuclear loci, in all sympatric tributaries. The D ' values listed are the observed disequilibrium divided by the theoretical maximum disequilibrium. A l l pairs in all tributaries were contingency tested with a probability test using a Markov chain in G E N E P O P (Raymond & Rousset 1995), under the Ho: "Genotypes at one locus are independent from genotypes at the other locus." Ho was rejected for all pairs in all tributaries, with p < 0.0001 in every case. Attycelley River was excluded from analysis due to non-sympatry. Tributary GH/ ITS G H / M T B G H / F O K D ' I T S / M T B I T S / F O K M T B / F O K Kemess 0.9855 0.98191 0.9717 0.9792 0.9760 0.9826 South Pass 0.9999 0.9892 0.9892 1.000 1.000 0.9902 Tributary 4 0.9353 0.9680 0.9909 0.9273 0.9067 0.9913 Attic hika 0.8660 0.8658 0.8655 0.8658 0.8655 0.8012 Niven 0.8705 0.7328 0.7727 0.7710 0.8330 0.8131 85 Table 13: Cytonuclear disequilibria for all four nuclear loci, in all sympatric tributaries. cD is the gametic disequilibrium and c D i - CD3 are the genotypic disequilibria (as defined in Asmussen et al. 1987). A l l locus pairs in all tributaries were contingency tested with a probability test using a Markov chain, under the H 0 : "Genotypes at the nuclear locus are independent from cytotypes." H 0 was rejected for all pairs in all tributaries, with p < 0.000001 in every case. Attycelley River was excluded from analysis due to non-sympatry. Tributary Nuclear c D i c D 2 c D 3 cD Locus Kemess G H 2 0.23099 -0.0043 -0.2267 0.22883 ITS-1 0.229796 0.00453 -0.23433 0.232061 M T B 0.228364 0.00811 -0.23647 0.232419 F O K 0.2298 0.00429 -0.23409 0.23194 South Pass G H 2 0.1377 -0.0254 -0.1123 0.125 ITS-1 0.137695 0 -0.1377 0.137695 M T B 0.137695 -0.012695 -0.125 0.131348 F O K 0.13769 -0.01269 -0.125 0.131347 Tributary 4 G H 2 0.09592 0.00145 -0.0974 0.09664 ITS-1 0.091558 -0.004228 -0.08733 0.089444 M T B 0.094464 0.002907 -0.09737 0.095918 F O K 0.09301 0.00291 -0.09592 0.09446 Attichika G H 2 0.07214 0.00185 -0.074 0.7306 ITS-1 0.07214 0.00185 -0.07399 0.07306 M T B 0.070287 0.003699 -0.07399 0.072136 F O K 0.06844 0.00555 -0.074 0.07121 Niven G H 2 0.04414 0.004527 -0.048665 0.0464011 ITS-1 0.046967 0.002263 -0.04923 0.048099 M T B 0.046401 0.002263 -0.04866 0.047533 F O K 0.04584 0.00283 -0.0487 0.04725 86 The female parent in an interspecific cross is identified by the m t D N A of F i hybrids. In this case, five of the six F i genotype hybrids contained bull trout m t D N A (the single exception was heteroplasmic, containing both Dol ly Varden and bull trout mtDNA) . A binomial test indicates that interspecific hybridization is significantly unidirectional (n = 5, p=0.5, P(X=5) = 0.03125): Dol ly Varden male by bull trout female. This unidirectionality is supported by the m t D N A in the other genotypic classes of hybrid (discussed below). Introgression O f the 89 juvenile hybrids from the entire Thutade Lake watershed, there were only five F i genotypes and five F n genotypes (bvITHb, bvvvv, vvvbb, bHHvb, and bbvHb; as per Appendix 6), limiting the size of any potential hybrid swarm (i.e. pool of interbreeding hybrid individuals; Table 8). Backcross genotypes (n = 59) and mitochondrial intragressions (n = 20) were the most common hybrid classes (Table 8). Introgression, the movement of alleles between species, could therefore have a significant impact on the population genetics of the char species in Thutade Lake. In the six tributaries, introgression of Dol ly Varden nuclear alleles into bull trout ranged from 0 to 2.64%, while introgression of bull trout nuclear alleles into Dol ly Varden ranged from 0 to 3.12% (Table 14). Different tributaries had different predominant directions of introgression, although the directionality does not correspond to the prevalence of the two parental species at the site, which implies that the variation was due to chance or sampling error. The watershed average was 1.45% introgression of 87 Table 14: Introgression of nuclear and mitochondrial alleles between Dol ly Varden and bull trout. Individuals listed as B T or B C B T genotypes in Table 8 are here pooled into B T , while those listed as D V , I G D V , or B C D v genotypes are pooled into D V . A l l F i (n = 5) and F n (n = 5) were excluded from this analysis. A s four nuclear loci were analyzed, the percent introgressed nuclear alleles is calculated as number of alien alleles / 8n. Tributary Species n Introgressed Nuclear Alleles # alien % alleles introgression Introgressed Mitochondrial Alleles # alien % alleles introgression Kemess B T 325 17 0.65 0 0 D V 268 35 1.63 13 4.85 South Pass B T 47 0 0 0 0 D V 17 4 2.94 5 29.40 Tributary 4 B T 66 7 1.33 0 0 D V 20 1 0.62 9 45.00 Attichika B T 78 6 0.96 0 0 D V 7 0 0 0 0 Niven B T 85 18 2.64 0 0 D V 8 2 3.12 3 37.5 Attycelley B T 0 0 0 0 0 D V 59 2 0.42 1 1.69 Total B T 601 48 1.00 0 0 D V 379 44 1.45 31 8.18 88 bull trout nuclear alleles into Dol ly Varden and 1.00% introgression of Dol ly Varden nuclear alleles into bull trout (Table 14). Independent scrutiny of the nuclear loci revealed that bull trout alleles for the four nuclear loci introgress into Dol ly Varden to different degrees (based on the number of foreign alleles, H 0 : equal introgression, %2= 12.3636, p = 0.0062). Introgression of bull trout Growth Hormone 2 into Dol ly Varden occurs at 2.7 the average rate of the other three loci (21 of 758 alleles scored), each of which introgressed at approximately the same rate (7, 9, and 7 alleles for ITS-1, M T B , and F O K ) (Figure 9). In bull trout, however, Dol ly Varden alleles for the four loci were not found to introgress at significantly different rates (observed introgression of 11, 10, 12, and 15 out of 1202 scored alleles for G H 2 , ITS-1, M T B , and F O K , x 2 = 1.1667, p = 0.7610) (Figure 9). It is worth noting that if the number of introgressed G H 2 alleles in Do l ly Varden is reduced to the average of the other three loci (i.e. 7.67), then the overall rate of nuclear introgression into Dol ly Varden (1.02%) is equivalent to the rate of nuclear introgression into bull trout (1.00%). Mitochondrial introgression appeared to be unidirectional, as there was no introgression of Dol ly Varden m t D N A into bull trout (Table 14). Introgression of bull trout m t D N A into Dol ly Varden ranged from 0% to 45% in the six tributaries. Of 89 hybrids, 67 had bull trout mtDNA, 21 had Dol ly Varden m t D N A , and one was heteroplasmic (Table 8). A l l F i genotypes contained bull trout m t D N A , as did all bull trout backcross genotypes, all mitochondrially introgressed individuals, most F n genotypes, and many Dol ly Varden backcross genotypes. The only hybrid individuals with Dol ly Varden m t D N A were Dol ly Varden backcross genotypes (20 of the 31 B C D V ) 89 • bull trout alleles into Dolly Varden • Dolly Varden alleles into bull trout G H 2 ITS-1 MTB FOK Molecular Marker mtDNA Figure 9: Introgression of species-specific alleles across the species barrier at four nuclear loci and one mitochondrial locus. Percent introgression refers to the number of alien alleles present divided by the total number of alleles (i.e. 2n for nuclear, n for mitochondrial), where for Dolly Varden n = 379 (328 D V + 51 B C D V ) and for bull trout n = 601 (573 BT + 28 BC D v) (Table 8). 90 and a single F n genotype (bvvvv, as per Appendix 6). As this F n genotype was likely the result of a cross between two Dol ly Varden backcrosses it was an exception that helped to prove the rule. Cohort Analysis Cohort analysis can be used to observe the changes in the hybrid proportion of the population over time, as an indirect measure of how much selection is acting on the hybrids (e.g. Bert & Arnold 1995). The best method for this test this would be to observe how the proportion of different genotypic classes of hybrids changed over time, because different genotype classes (with different morphologies and behaviours) would be subjected to varying levels of selection pressure. However, as hybrids formed such a small proportion of the population (9.0%), subdividing hybrids by genotypic class would not allow large enough sample size for statistical significance. Therefore, the cohort analysis w i l l group all hybrid genotypes together. In both Kemess Creek and in the other tributaries to Thutade Lake, the proportion of bull trout in the population decreases in older age classes (Table 15). This is likely due to early migration of juvenile bull trout into the lake or perhaps migration from the sample sites into more appropriate stream habitat. In either case, a loss of individuals from one of the two parental populations could lead to artificial "increases" in the proportion of the hybrid population. To correct for this problem, I did contingency testing for cohort analysis using only hybrids and Dol ly Varden. This correction may not take into account any potential migration of hybrids, but unfortunately this source of error cannot be corrected for. 91 Table 15: Cohort analysis. Percentages of hybrids, Dol ly Varden and bull trout are listed by age class. Contingency test analysis were used to compare age classes in different ways: 'Dynamic' compares 1997 fry to 1998 1+ age class (i.e. following the same cohort over time, a dynamic cohort analysis), 'Fry ' compares 1997 fry to 1998 fry (i.e. are proportions of genotypic classes constant from year to year? In support of the validity of a static cohort analysis), and 'Static' compares the three 1998 age classes (fry, 1+ and >1+, a static cohort analysis which assumes that birth proportions of hybrids is constant). These analyses only use hybrids and Dol ly Varden because older bull trout migrate downstream. Year, Age Class n % hybrids % Dol ly Varden % bull trout % H / H + D V Test p-value Kemess 1997, fry 185 4.9 26.5 68.7 15.5 Dynamic 0.1625 1998, fry 211 7.6 35.1 57.3 17.8 Fry 0.8220 1998, 1+ 122 4.1 53.3 42.6 7.1 Static 0.0602 1998, >1+ 80 16.2 63.58 20.0 16.2 Total 598 7.2 40.0 52.8 7.2 Thutade* 1997, fry 131 8.4 6.1 85.5 57.9 Dynamic 1.0 1998, fry 112 12.5 5.4 82.1 70.0 Fry 0.5118 1998, 1+ 35 22.9 17.1 60.0 57.1 Static 0.3404 1998, >1+ 55 20.0 21.8 58.2 47.8 Total 333 13.2 9.6 77.2 57.9 * South Pass Creek, Tributary 4, Attichika River, and Niven River 92 Dynamic analysis, following a single cohort through time, is limited due to the fact that only two years of information are available: 1997 fry and 1998 1+ age classes. No significant difference in the proportion of hybrids was found in this single cohort over the two sample years, in either Kemess Creek (p = 0.1625) or in the remainder of Thutade (p= 1.0) (Table 15). For static cohort analysis (i.e. looking at several biological cohorts in a 'snapshot' in time), a key assumption is that the proportion of hybrids born into the population is relatively constant between cohorts. Contingency analysis revealed no significant difference between fry captured in 1997 and 1998 in either Kemess Creek (p = 0. 8220) or in the remainder of Thutade (p = 0.5118), supporting this assumption (Table 15). Static cohort analysis (using 1998 fry, 1998 1+ and 1998 >1+ age classes) revealed no significant difference between age classes, in either Kemess Creek (p = 0.0602) or the remainder of Thutade watershed (p = 0.3404) (Table 15). The p-value for the static cohort analysis of Kemess Creek was close to a (p = 0.0602), but no real trend was evident, as 1998 fry and >1+ age classes had approximately equal proportions of hybrids (17.8 and 16.2%, respectively), but the 1998 1+ age class (the intervening year) had fewer hybrids (7.1%) (Table 15). Kemess Creek Hybrid Zone Kemess Creek was sampled in detail, at 17 sites over approximately 16 km (Figure 8). Beginning downstream (southwest), sample sites included five mainstem and three side channel sites interspersed along Kemess Creek proper, one site in E l Condor Creek (a small tributary creek), two sites in South Kemess Creek, and six sites in North 93 Kemess Creek. South Kemess Creek was impounded in 1997 for the tailings pond of an open-pit mine, resulting in the loss of 8 km of bull trout habitat upstream. Environmental protection measures required that the waters from South Kemess Creek be diverted around the tailings pond, to maintain water flow levels in Kemess Creek. No sampling was done upstream of the tailings pond, due to the inability of bull trout to migrate past the diversion for spawning. The Dol ly Varden, bull trout and hybrids in Kemess Creek have been divided by age class, into fry and older juveniles, for analysis of distribution (Figure 10). As fry are distributed near to where they emerged, they are an indicator of adult spawning distribution. Older juveniles are mobile, however, and can have expanded their range throughout the tributary. The distribution of both fry (1997 and 1998) and juveniles (1+ and >1+ age classes) in Kemess Creek differed between the two char species (Figure 10). Dol ly Varden fry were primarily localized near sites of groundwater up welling in both 1997 and 1998 (preferred spawning sites for Dol ly Varden) (Figure 8). These sites are located in the headwaters of North Kemess (n = 61 fry) and in and around E l Condor Creek (n = 58). A few Dol ly Varden fry (n = 4) were also located near the confluence of North and South Kemess creeks, where there may have been a small suitable Dol ly Varden spawning site. B u l l trout fry were distributed throughout Kemess Creek (n = 248), with the exception of the single site in E l Condor Creek (very shallow waters) and the uppermost site in North Kemess Creek (Figure 10). Unlike Dol ly Varden, the bull trout fry distribution changed between 1997 and 1998. In 1998, bull trout fry extended 3 k m 94 95 further upstream than 1997 (to the total range shown). It is likely that this increase in bull trout spawning range was caused by the loss of spawning habitat in the newly impounded South Kemess Creek, forcing bull trout to migrate further upstream in North Kemess in search of spawning sites. Juveniles of the two species are distributed more similarly. Dol ly Varden juveniles were found at all sites but one (n = 116). The 1+ age class Do l ly Varden (n = 65) were more closely linked to the two primary Dol ly Varden spawning areas, with some spread downstream from each. The >1+ age class Dol ly Varden (n = 51) were spread throughout the system (n = 23 of these had lengths > 110 mm, and so may actually have been sexually mature Dol ly Varden). B u l l trout juveniles had similar distributions for both age classes, but with the 1+ age class (n = 52, to 3 r d N K site) extending further upstream than the >1+ age class (n = 16, to 2 n d N K site). The sharp decrease in the number of juvenile bull trout in older age classes is likely due to the size-dependent migration of bull trout juveniles downstream. The downstream movement of the upper limit of the 1+ and >1+ age class distributions could, therefore, be due to downstream migration or migration to more suitable stream habitat (such as a site on the Attichika River at which 11 bull trout with lengths ranging from 117-180 mm were collected). Hybrids were distributed throughout the system. Small sample sizes and the presence of different genotypic classes of hybrids ( F i , F n , B C B T , B C D V , IGDV) made it difficult to distinguish patterns from chance occurrence. Both F i genotype individuals (n = 2, 86 mm and 143 mm in length) were found within the Dol ly Varden spawning areas, but because older juveniles can migrate, this distribution is not particularly revealing. 96 The F n genotypes (n = 3), two of which were likely the offspring of a cross between two Dol ly Varden backcrosses (bvvvv, vvvbb, as per Appendix 6), were found in the headwaters. B u l l trout backcross genotypes (n = 9) were all found downstream of the North/South Kemess confluence, while Dol ly Varden backcross genotypes and mitochondrial introgressions (n = 29) were found in the Dol ly Varden spawning areas as fry, and throughout the system as juveniles. Both of these backcross genotype patterns are similar to the distributions of the parental species to which they are more closely related. Discussion Interspecific Hybridization Dol ly Varden and bull trout have recently become recognized as separate species on the basis of detailed morphological analysis (Cavender 1978; Haas & McPha i l 1991) and molecular evidence that they are not sister-species within the Salvelinus genus (e.g. PleyteefaZ. 1992; Crane et al. 1994; Phillips et al. 1995). However, molecular evidence of localized interspecific hybridization (McPhail & Taylor 1995; Baxter et al. 1997) has led to concerns about their ability to coexist as distinct entities within the sympatric portion of their largely parapatric distributions. In the Thutade Lake watershed, however, my population genetic analyses showed that Dol ly Varden and bull trout form two nearly separate, non-randomly intermating populations. There was significant heterozygote deficiency (Fis) and significant nuclear and cytonuclear disequilibria (D' and cD), despite the fact that hybrids form 9% of the char population. As ecological research in the watershed has found that the two species 97 have distinct life histories and reproductive behaviours (Baxter et al. 1997; Hagen 2000), we can say that their specific status is upheld in sympatry. The question then becomes, how are they doing it? The extremely bimodal distribution of genotypes in the watershed suggests that prezygotic isolation plays an important role (Jiggins & Mallet 2000). In this case, assortative mating is likely a large part of the answer. First generation hybrids are rare in the hybrid population: only 5 out of 990 juveniles (0.5%) sampled had a Fi genotype. For three reasons, low levels of interspecific hybridization (i.e. prezygotic isolation) best explain this low percentage, as opposed to selection against young F i hybrids (postzygotic selection). First, there is not much intrinsic selection against young F i hybrids: hybrid individuals do not have higher mortality than pure parentals as either eggs or fry when raised in controlled conditions (Haas & McPha i l 1991; E . B . Taylor, pers.comm.). Secondly, a careful analysis revealed no significant ecological differences between the two species as stream-resident juveniles in the watershed (Hagen 2000). This makes strong extrinsic selection at a young age unlikely. Finally, I saw F i genotype hybrids throughout the juvenile age classes sampled, although they were rare (two 1997 fry, one 1998 fry, two 1998 >1+ age class juveniles), revealing no evidence of the effects of either intrinsic or extrinsic selection in reducing F i numbers over time. Prezygotic barriers to gene exchange must play a large role in maintaining distinct populations. In this case, gametic incompatibility appears to be unimportant. Reciprocal artificial crosses have been made (Haas & McPhai l 1991), although never with the potential for competition between homo- and hetero-specific sperm types. Neither are 98 there physical barriers to hybridization, because, as with most fish, fertilization is external. Thutade bull trout and Dol ly Varden, however, differ in several key ways in terms of their spawning behaviour and timing. Firstly, Thutade bull trout are much larger at maturity (40-90 cm) than Dol ly Varden (12-21 cm). Char and other salmonids generally form size-assortative mating pairs for within species spawning (e.g. Sigurjonsdottir & Gunnarsson 1989; Maekawa et al. 1994). Since both Dol ly Varden and bull trout typically form size-assortative pairs, a consequence of the large body-size difference wi l l be non-random mating between species. Second, bull trout and Dol ly Varden have different spawning behaviours in allopatric areas (Armstrong & Morrow 1980), a difference which may hold in sympatric areas. Also, mature bull trout in Thutade have very obvious secondary sex characteristics while mature Thutade Dol ly Varden do not (J. Hagen, pers. comm.). Once again, this type of interspecific difference wi l l promote intraspecific matings. The third major difference is that Dol ly Varden and bull trout have different spawning site preferences: bull trout dig large redds (1.6 m 2) in deep (27-30 cm), fast moving (38-40 cm/s) water while Dol ly Varden dig small redds (0.1 m 2 ) in shallow (5-10 cm), slow moving (1-16 cm/s) water (Baxter et al. 1997). Further, while bull trout spawn in stream reaches with good cover throughout the tributaries, Dol ly Varden redds are localized in areas of groundwater seepage. Mating pairs w i l l therefore be unlikely to be in close proximity and wi l l be concentrated in different areas. Finally, Thutade bull trout spawn from mid-August to early September while Dol ly Varden spawn from early 99 September to mid-October, so there are temporal differences in spawning patterns (Hagen 2000). Mitochondrial D N A evidence yields a clue to the source of the permeability of the interspecific barrier. Because of the mode of inheritance of mitochondrial D N A (maternal, non-recombining), it can be used to resolve the directionality of an interspecific cross by identifying the species of the female parent. In this case, the F i genotype hybrids all contained bull trout mtDNA, indicating that the interspecific cross is Dol ly Varden male by bull trout female (a single F i genotype exception was heteroplasmic, indicating paternal leakage as opposed to the reciprocal interspecific cross). This unidirectionality is supported by the m t D N A in the other genotypic classes of hybrid: Dol ly Varden m t D N A is only found in Dol ly Varden backcross genotypes, while all other hybrid genotype classes contain bull trout m t D N A . A similar pattern was observed in Dol ly Varden / bull trout hybridization in the Skagit River (McPhai l & Taylor 1995). Such directionality is quite common in interspecific hybridization (e.g. Lamb & Avise 1986; Kitano et al. 1994) and is actually expected in this case for two reasons. Firstly, 'sneaking' often provides an exception to the size-pairing rule among salmonids (e.g. Gross 1985; Svedang 1992; Maekawa et al. 1993,1994). Sneaking is a common 'parasitic' mating strategy amongst salmonid species, whereby a smaller male wi l l rush into the redd of a larger mating pair of the same species and sneak a fertilization (Taborsky 1998). Both bull trout and Dol ly Varden have been reported to use sneaking as a strategy for intraspecific matings. In the Thutade Lake watershed, as in most sympatric sites, Dol ly Varden is by far the smaller of the two species. A hybrid mating, 100 therefore, would be expected to occur when a Dol ly Varden male sneaks on a bull trout spawning pair. In fact, this strategy has been proposed as an explanation for interspecific hybridization in other char and salmonid crosses (e.g. McGowan & Davidson 1992; Kitano et al. 1994). A second, but probably less influential factor in the directionality of the cross is the difference in spawning times. The overlap in spawning times occurs at the end of the bull trout spawning period. Salmonid males generally ripen earlier than females, and females remain ripe later i f not spawned (e.g. Kitano 1996). It is therefore more likely that a Dol ly Varden male would become ripe in time to encounter a bull trout female (i.e. it requires less spawning time overlap) than that a bull trout male would remain ripe long enough to encounter a ripe Dol ly Varden female. It is worth noting that a potential alternative explanation for an observation of unidirectionality could be biased mortality of one of the two reciprocal interspecific crosses (e.g. Rand & Harrison 1989). Reciprocal bull trout / Dol ly Varden laboratory crosses reared in controlled conditions, however, have identical mortalities, equal to those seen in pure parental crosses (Haas & McPha i l 1991). It is also unlikely that we are simply observing the results of single or very few crosses, as many different hybrid genotypes, in different age classes, in different tributaries, all show the same m t D N A pattern. A s for the location of interspecific crosses, it was expected that F i hybrid fry would be observed near areas of groundwater seepage, where spawning bull trout may have been within tempting reach of ripe Dol ly Varden males. This did not turn out to be the case. A l l three F i genotype fry caught in the watershed (out of 661 fry total) were 101 collected from the same sample site, in the Attichika River. Although a sample size of three is too small to draw firm conclusions, the fact that these fry are from two cohorts (two fry were collected in 1997, one in 1998) lends some credence to the observation. This site is located right in the middle of a stream reach heavily used in bull trout spawning (Bustard 1995, 1996, 1997, 1998). There was no obvious suitable habitat for Dol ly Varden spawning nearby (J Hagen, pers. comm.), although a small number of Dol ly Varden fry were captured at the site both years. This stream reach has habitat complexity, however, with a great deal of woody debris, providing ideal cover for sneakers. B u l l trout males are typically very aggressive during spawning (Leggett 1980; McPhai l & Murray 1979; Sexauer 1994), making such cover an advantage, if not a necessity, for sneakers. Habitat complexity and cover may therefore be a key requirement for the occurrence of interspecific hybridization. The only other juvenile F i genotype hybrids seen in the watershed were found in this same tributary to the lake, although they were in the Kemess Creek portion rather than the Attichika River. A n 86 mm F i genotype individual, likely a three-year-old, was found in the extreme headwaters of Kemess Creek, upstream of the range of bull trout juveniles. M y initial hypothesis regarding interspecific hybridization predicted the upstream limit of bull trout spawning would be prone to hybridization due to the rarity of bull trout at these sites. This prediction was supported by a single snorkel observation of a bull trout female building a redd hundreds of meters upstream of the nearest male bull trout, with a male Dol ly Varden lurking a few meters away (J. Hagen, pers. comm.). A n 86 mm char individual is well capable of migration, however, and so its locality cannot be assumed to be anything other than chance. A 143 mm F i genotype individual was 102 found in the mainstem of Kemess Creek, again not providing much information regarding the location of hybridization. A single F i genotype adult was observed and had been collected in the lake. Besides the obvious life history information (discussed below), the migration of F i hybrids into the lake could allow very localized hybridization events to affect other tributaries. It is unknown how much migration, i f any, there is between the Dol ly Varden populations in different tributaries, nor how much straying occurs between bull trout populations, which typically display some degree of stream fidelity/homing (Spruell et al. 1999; Latham 2000; Costello & Taylor, unpubl. data). Introgressive Hybridization First generation hybrids are at least sometimes fertile and have reproduced, as evidenced by the 84 hybrids seen with non-F] genotypic classes. The backcross offspring of F i hybrids by both parental species also appeared to be fertile and to reproduce, as the presence of several presumed later generation backcrosses indicates. Unfortunately, genotype alone, particularly when relatively few nuclear loci are being scored, is a very poor indicator of ancestry (Rieseberg & L i n der 1999). For example, when a hybrid has a genotype 'pure' for one species except for a single foreign allele (nuclear or mitochondrial), it is impossible to tell how many generations ago the ancestor was an F i (i.e. is the allele essentially introgressed, or is the individual a second, or even first, generation backcross?). Fifty-two of the 89 hybrids fall into this category. When two or three nuclear loci are heterozygous, it is more likely that the F i is a recent ancestor, but again, it is not certain. The unfortunate consequence of this uncertainty is that estimates 103 cannot be made about the reproductive success of F i hybrids, nor can any estimate be made regarding the speed at which introgressed alleles are weeded out of the two char species by natural selection. One thing that is clear, however, is that a population of randomly interbreeding hybrids (hybrid swarm, unimodal distribution) has not been produced. Only five F n genotypes were observed (defined as being homozygous for alleles from different species at two or more nuclear loci). The scarcity of F n genotype hybrids is not surprising, given how few F i hybrids are present in the reproductive pool. Indeed, given the rarity of F i (0.5% of char), it is surprising to see any F n produced at all i f mating is entirely random, which suggests that some type of non-random mating is occurring. Morphologically, F i hybrids appear to be intermediate to the two parental species (Haas & McPha i l 1991; McPha i l & Taylor 1995; Baxter et al. 1997). Such morphological intermediacy is commonly observed for interspecific hybrids, although a mosaic of parental traits is also seen, and in some cases F] hybrids are more similar to one of the parental species (e.g. Leary et al. 1983). It has also been shown that hybrids can exhibit intermediate behavioural traits, such as call site selection in tree frogs (Lamb 1987) or migration time and distance in warblers (Berthold & Querner 1981). Dol ly Varden and bull trout in Thutade Lake watershed are distinct enough for an F i to be intermediate in a morphological or behavioural trait without overlapping either parental distribution. A cross between F i individuals could, therefore, be seen as more likely than a backcross, given access to all three potential partners, because it would pair individuals of similar size and with similar preferences regarding spawning time and location. This type of cross has in fact been shown to have occurred three separate times, as the F n 104 hybrids were all fry taken from three different tributaries (Kemess Creek, Tributary 4, and Niven River). Two of the F n individuals have genotypes that are better explained by a mating between two Dolly Varden backcrosses than a cross between F i or F n hybrids (a 1997 fry was vvvbb and a 2+ age class individual was bvvvv, as per Appendix 6). These individuals were both taken in Kemess Creek at sites particular to Dol ly Varden of their age class: the fry was taken at a site 3 km upstream of the nearest bull trout fry in 1997, while the 2+ age class individual was taken in the tributary headwaters where older bull trout juveniles were not seen. That matings between Dol ly Varden backcrosses occur is even less surprising than those between F i hybrids. As backcrosses form 9.3% of the Dol ly Varden population in Kemess Creek, random mating alone would be expected to produce these crosses. Potential intermediacy in morphological or behavioural traits only serves to increase the chances. Throughout the watershed, F i hybrids also appear to backcross with both parental species. In fact, most of the hybrids in the watershed have genotypes which suggest that they are the descendants of backcrosses between an F i and a bull trout (n = 28) or between an F i and a Dol ly Varden (n = 51). These individuals are likely both immediate backcrosses (i.e. the F i is an immediate parent) and repeat backcrosses (i.e. the F i was an ancestor from one to several generations previously). The same morphological and reproductive characteristics that impede interspecific crosses wi l l also be expected to hinder backcrosses and repeat backcrosses, but with decreasing success. Differences in size and in behavioural traits between the F i and either parental species would be smaller than between bull trout and Dol ly Varden, 105 while the differences between backcrosses and their respective parental species would be smaller still. In terms of numerical availability of partners, a backcross or repeat backcross is certainly more likely than a cross between F i hybrids or between immediate backcrosses. One potential explanation for bi-directional backcrossing is the fact that char are repeat spawners (iteroparous) and continue to increase in size after sexual maturity is reached. Thus, a single F] individual could mature in or near the Dol ly Varden size range ( 1 2 - 1 7 cm, 21 cm max) and grow into the bull trout size range (40 - 90 cm). Further, as the F i would be smaller than bull trout for much of their lifespan, the sneaking mechanism would continue to be available for the bull trout backcross, allowing an F i male to sneak on a larger bull trout mating pair. Mitochondrial D N A cannot be used to distinguish between F i and bull trout, and so is of little use in examining bull trout backcrosses. This is unfortunate, as it would be quite interesting to distinguish F] sneaking from Fi /bul l trout pairing. Mitochondrial D N A can, however, be used in determining directionality of Dol ly Varden backcrosses. Dol ly Varden backcrosses contain both types of m t D N A , so no absolute conclusions can be made about the direction of the first generation backcross. However, the genotypic patterns within the backcrosses differ between the two m t D N A types. Six of the eleven (54.5%) Dol ly Varden backcrosses with bull trout m t D N A have more than one heterozygous locus, while only three of the twenty (15.0%) with Dol ly Varden m t D N A have multiple heterozygous loci. As the number of heterozygous loci can hint at how many generations have passed since the initial backcross, this difference in proportions implies that the first 106 backcross is more commonly Dol ly Varden male by F i female, but that later generation crosses (i.e. between a Dol ly Varden backcross and a Dol ly Varden) may be in either direction. The bi-directionality of the later generation cross is predictable, as Dol ly Varden backcross individuals would be expected to be morphologically and behaviourally similar to Dol ly Varden. Why a Dol ly Varden male would cross with a F i female is less obvious, as sneaking generally occurs on a pair, not on a lone female. The lack of potential partners for the F i female might, however, play a role in the cross (a small male is better than none at all?). The distribution of backcross fry in Kemess Creek is also more useful in looking at the spawning location of Dol ly Varden backcrosses than that of bull trout backcrosses. A s bull trout spawn throughout the tributary, the location of bull trout backcross genotype fry is not particularly informative. Dol ly Varden backcross genotype fry, however, are fairly tightly associated with Dol ly Varden spawning sites. For the initial backcross (Fi by Dol ly Varden) this association might not be as tight, but for repeat backcrosses it would certainly be expected given that both parents would be expected to have Dol ly Varden or Dol ly Varden-like spawning preferences, and this was the case (Figure 10). Asymmetrical Nuclear and Mitochondrial Introgression Asymmetry in the introgression of m t D N A is often associated with directional interspecific hybridization (Arnold 1997). As the initial cross between Dol ly Varden and bull trout is unidirectional, and as reverse backcrossing appears to be quite rare, any mitochondrial introgression that occurs must be asymmetrical: that of bull trout m t D N A into Dol ly Varden. A more thorough discussion of mitochondrial introgression, 107 particularly where introgression is permanent (i.e. has become part of the species' permanent genetic population structure) has been undertaken in Chapter 2. In the Thutade Lake watershed, it is impossible to tell whether 'permanent' mitochondrial introgression has occurred, as we cannot know how many generations have passed since the m t D N A alleles entered the Dol ly Varden population. As there were only 20 Dol ly Varden scored with bull trout mtDNA, out of 348 Dol ly Varden, bull trout m t D N A forms 5.7% of the 'Dol ly Varden' m t D N A pool. Unfortunately, it is difficult to say whether the lack of bull trout m t D N A in Dol ly Varden is due to selection against the m t D N A itself, or selection against the heterozygotic nuclear status of hybrids (and their m t D N A by association). Of the 22 Dol ly Varden heterozygous at a single nuclear locus, however (i.e. later generation backcrosses), only 5 had bull trout m t D N A , suggesting that selection against the m t D N A itself may be a factor. Although cohort analysis did not reveal any selection against hybrids, there did appear to be selection acting against nuclear loci, discernible due to differential rates of introgression. In particular, nuclear loci are introgressing to different degrees both with respect to each other and with respect to the two parental species. A l l four loci in bull trout and three of the four loci in Dol ly Varden introgress at the same rate (i.e. they have the same proportion of foreign alleles to parental alleles). Bu l l trout Growth Hormone 2, however, has introgressed into Dol ly Varden 2.7 times more than the other nuclear loci (Figure 9). This implies that selection is acting against foreign alleles, but is acting less strongly on the Growth Hormone 2 locus than the on other nuclear loci. It is also worth noting that bull trout Growth Hormone 2 introgresses into Dolly Varden independently of bull trout m t D N A . Of the 18 introgressed G H 2 alleles noted in 108 17 Dol ly Varden individuals (i.e. bull trout G H 2 alleles in an otherwise pure Dol ly Varden nuclear background), only 2 were associated with bull trout m t D N A . Of the 5 bull trout m t D N A alleles found in Dol ly Varden containing a single bull trout nuclear allele (i.e. Dol ly Varden backcrosses with 1 heterozygous nuclear locus), 2, 0, 1, and 2 were associated with G H 2 , ITS-1, M T B and F O K , respectively. Nuclear introgression for the other nuclear loci occurs at the same rate in both directions. Whether this indicates that Fi hybrids actually backcross equally in both directions is uncertain, however, due to unknown levels of selection acting on the two types of backcrosses, as well as to potential differences in the reproductive capacity/success of backcrosses (the reproductive capacity of bull trout is much greater than that of Dol ly Varden). In summary, Thutade Lake watershed Dol ly Varden are more affected by hybridization overall, due to higher levels of introgression of bull trout alleles (GH2 and m t D N A loci). The comparatively elevated level of introgression of G H 2 into Dol ly Varden is observed across the sympatric range (Chapter 3), while historical introgression of bull trout m t D N A has resulted in its fixation in southern populations of Do l ly Varden (Chapter 2). Why Do They Not Collapse? In the Thutade Lake watershed, F i genotype hybrids formed approximately 0.5% of the juvenile population assayed. These hybrids were viable and at least partially fertile. Although they do mate together occasionally (0.3% of juveniles have post-Fi genotypes), a hybrid swarm has not formed in the watershed. The primary reproductive 109 fate of F i hybrids appeared to be backcrossing with both parental species (8.0% of juveniles have backcross genotypes). Population genetic analyses, however, revealed that bull trout and Dol ly Varden form two distinct, non-randomly intermating populations within the Thutade Lake watershed. With the hybridizing and backcrossing, why are these two species not collapsing into a single, randomly mating char admixture? To begin, I wi l l eliminate two alternative explanations given in other instances of natural hybridization. First, this is not a case of very recent sympatry and we are not seeing the early stages of collapse. These two populations have likely been sympatric for upwards of 10,000 years in the headwaters of the Peace River (and other interior drainages; Chapter 3). Second, the parental species in the Thutade Lake hybrid zone are not being 'replenished' by gene flow from allopatric populations of the two parental species. While this might have been true for bull trout, which are present throughout the Peace River, Thutade Lake Dol ly Varden are physically isolated from the remainder of allopatric Dol ly Varden, which are present only in coastal drainages and the extreme headwaters of other interior drainages. There are three main types of barriers to natural hybridization: prezygotic barriers, gamete incompatibility, and postzygotic barriers (Arnold 1997). As natural hybridization between bull trout and Dol ly Varden does occur, any prezygotic barriers (discussed above) or gametic incompatibility are obviously insufficient for maintaining two distinct species. Postzygotic barriers, therefore, in the form of intrinsic and/or extrinsic selection against hybrids, must be playing a key role in balancing the gene flow between species. As I have already discussed, extrinsic selection against hybrids as juveniles in the Thutade Lake watershed is unlikely, as both species have near-identical ecological roles 110 in the tributaries (Hagen 2000). As adults, however, the two species adopt vastly different life histories, with large, migratory, piscivorous adfluvial bull trout and small, stream resident, invertebrate-feeding Dol ly Varden. It has been proposed that such alternate life histories allow coexistence between species that are otherwise very similar in ecology (Hagen 2000; Chapter 3). If extrinsic selection were to act against hybrids, therefore, it would most likely be at the adult stage. First generation hybrid fry from both reciprocal artificial crosses have been shown to be morphologically intermediate to parental fry (Haas & McPha i l 1991). One of the three juvenile and/or adult F i genotype hybrids from Thutade Lake watershed was morphologically identified as a hybrid due to its intermediate morphology. Unfortunately, the sample size for F i genotype hybrids was very small, allowing for few life history conclusions to be drawn. A 143 mm F] genotype hybrid was found in Kemess Creek. This length of fish is fairly common among resident Dol ly Varden, but was not often observed among juvenile bull trout (i.e. bull trout juveniles migrate downstream by the time they are this large), implying that the F i hybrid was adopting a life history similar to Dol ly Varden. A second F i genotype hybrid, 220 mm long, was captured in Thutade Lake. Its stomach was empty, so we were unable to determine whether it had made the diet switch to piscivory typical of bull trout, but it had certainly adopted the migratory life history of bull trout. These two F] hybrids were mis-identified as Dol ly Varden and bull trout, respectively, implying that they were morphologically more similar to the species whose life history they were assuming. Thus, there were indications that both the Dol ly Varden and the bull trout life history were represented 111 among F i hybrids. Unfortunately, the low sample size makes firm conclusions impossible to make. Backcross hybrids would still have some traces of morphological intermediacy but would be more similar to the parental species they are backcrossed with: among the juveniles identified morphologically, seven of the nine bull trout backcrosses were identified as bull trout and 25 of the 26 Dol ly Varden backcrosses were identified as Dol ly Varden. Behavioural and reproductive traits of backcrosses would also be expected to be more similar to those of the parental species involved in the backcross. Amongst the hybrid classes, therefore, selection would have two very different sets of traits to act against. Hybrids with bull trout-like characteristics would migrate to the lake and would be subject to competition for food and threats of piscivory. B u l l trout backcross and F i hybrids would be expected to have slower growth rates than pure bull trout (as seen in the F i hybrids which had growth rates intermediate to the two parental species; Haas 1988). These hybrids would therefore be at a disadvantage for two reasons. First, their smaller size could decrease their relative efficiency at piscivory. Second, bull trout are rapacious cannibals, switching to piscivory at sizes as small as 100 mm in the Skagit River (McPhail & Taylor 1995). During sample collection in the Thutade Lake watershed, at least a half dozen fry were eaten by older juveniles while in sample buckets and one 96 mm bull trout was observed with the tail of a fry protruding from its mouth. A hybrid fish, with its slower growth rate, would also be subject to the dangers of piscivory for longer than a young bull trout. Selection against F i and Dol ly Varden backcross hybrids in the stream could be due to competition for food to support their increased growth rates. Dol ly Varden are 112 capable of reaching much larger sizes than they commonly do in their resident sympatric populations. In the Thutade Lake watershed, an experimental transplant of Dol ly Varden from a tributary where the maximum size of Dol ly Varden was approximately 15 cm to a barren creek above a barrier to fish migration resulted in 22 cm fish three years later (Bustard, pers. comm.). Bustard suspects that this increase in maximum body size was directly related to food availability (pers. comm.), implying that competition for food limited the maximum size possible for Dol ly Varden prior to transplant. If hybrid individuals had some of the morphological characteristics of bull trout (jaw length suited for piscivory, increased growth rate), which could make them less adept at invertebrate feeding than adult Dol ly Varden, then competition could certainly have negative effects on hybrid fitness. O f the two situations, hybrids existing in the bull trout sphere (adfluvial piscivory) would certainly appear to be at more of a disadvantage than hybrids in the Dol ly Varden sphere. This fact might explain the relative lack of bull trout backcrosses amongst the adult bull trout sampled. A second source of selection against hybrids would also be against adults, but this time as reproductive selection. Hybrids with intermediate size, behaviour, spawning colouration, and sex characteristic would be expected to have lower reproductive success than parental individuals. There is also the question of the fecundities of the various classes of hybrids. B u l l trout and Dol ly Varden have different chromosome numbers (n = 78 and 82, respectively) so the fertility of F i hybrids, F n hybrids, and backcrosses could vary according to hybrid genotype and is likely lower than the fecundity of parental individuals. It is also worth noting that the only F n hybrids seen were fry. Although it is 113 dangerous to draw conclusions from such a small sample size (n = 3), it did suggest some form of intrinsic selection against certain hybrid genotypes (i.e. via "hybrid breakdown"). I propose, therefore, that selection against adult hybrids, both ecological and reproductive, is the most likely possibility for the maintenance of two distinct char species in Thutade Lake watershed in the face of interspecific gene flow. 114 Chapter 5: General Discussion 115 Implications for Dolly Varden and Bull Trout M y thesis has yielded strong evidence that Dol ly Varden and bull trout are distinct species, despite geographically widespread hybridization. This result supports the conclusions of previous morphological (Cavender 1978; Haas & McPha i l 1991) and ecological (Baxter et al. 1997; Hagen 2000) studies. Reciprocally monophyletic nuclear phylogenies showed quite clearly that Dol ly Varden and bull trout collected throughout the sympatric range are more closely related to allopatric populations of the same species than to sympatric individuals of the other species. Evidence for low levels of interspecific, introgressive hybridization only serves to strengthen this conclusion. Population genetic analysis revealed that sympatric char populations are not randomly mating hybrid swarms. Hybridization has not caused the disintegration of the genetic distinctions between species. Assortative mating (prezygotic isolation) appears to play a major role in the reproductive isolation of the two species, likely due to the mature body size differential (an ecologically based resource polymorphism) and to various differences in reproductive behaviour (time and location of spawning). Sneaking by male Dol ly Varden, a parasitic reproductive tactic common in salmdhids (Taborsky 1998), appears to create the 'leak' in the species barrier. Postzygotic isolation mechanisms also appear to be firmly in place, if only because something must be keeping the two species distinct despite introgressive hybridization. As F i hybrids and both directions of backcrosses appear to be both viable and fertile, I propose that this mechanism is primarily extrinsic selection against hybrids and their offspring. I propose that selection against hybrids would occur primarily at the 116 adult stage, in terms of both reproductive limitations (due to hybrid body size intermediacy) and ecological limitations (due to an inability of hybrids to successfully adopt either of the two alternate sympatric life histories). These postzygotic barriers are obviously sufficient in preventing the collapse of bull trout and Dol ly Varden into a local char hybrid swarm. They cannot, however, preclude permanent introgression of alleles across the permeable interspecific barrier. M y phylogenetic analysis provided the first detailed description of Dol ly Varden population structure within B . C . It revealed the presence of two distinct phylogenetic groups of Dol ly Varden, with different refugial histories. The ranges of these two groups overlap in the southern portion of B . C . Similar m t D N A phylogeographic variation has been seen in bull trout, providing evidence for a coastal and an interior refugial history (Taylor et al. 1999). This mitochondrial D N A variation provides evidence of a major source of genetic diversity in B . C . , that conservation/management planning should take into account, for conservation of the bioheritage of the species. The widespread local sympatry and interspecific hybridization of Dol ly Varden and bull trout raise some fairly interesting and important conservation issues. These questions have important consequences as bull trout are blue-listed in B . C . and have been shown to be susceptible to over-fishing and environmental degradation in the United States, where they are now listed as an endangered species. A s far as management in B . C . is concerned, three different management plans wi l l be required, as Dol ly Varden (anadromous, lacustrine, and/or stream resident) and bull trout (adfluvial and/or stream resident) in allopatry have different requirements than a sympatric Dol ly Varden/bull trout char population. In the area of range overlap, local 117 sympatry and hybridization are common, but both would be essentially invisible to anyone not extremely well versed in char identification. Even the biologists involved in various Thutade Lake projects (D Bustard, G Haas, & J Hagen), who know their char, made errors identifying individuals by species. Juvenile bull trout were misidentified as Dol ly Varden and both bull trout and Dol ly Varden were misidentified as hybrids, implying that natural variation or phenotypic plasticity within the species can produce individuals that are morphologically intermediate or have a mosaic of traits. Further, the identification of hybrids by morphology alone is near impossible. Management plans wi l l require a system-by-system distinction between locally allopatric and sympatric sites and the determination of whether hybridization is occurring the locally sympatric sites. As the coexistence of Dol ly Varden and bull trout in the face of hybridization appears to depend largely on extrinsic natural selection against hybrids as adults (ecological and/or reproductive), this means that there is likely a balance achieved between gene flow and selection. If the density or habitat of one of the two species were altered in a system, the balance could be shifted to the potential detriment of either species or to that of the char population as a whole. Hybrid Zones, the Evolution of Reproductive Isolation, and Ecological Speciation The Salvelinus hybrid zone is atypical in the hybrid zone literature for several reasons. It comprises a vast area of sympatry, as opposed to the customarily narrow strip of sympatry between two largely allopatric species. Within the zone of range overlap, the levels of hybridization are also atypically low, although such bimodal hybrid zones becoming more prominent in the literature (Jiggins & Mallet 2000). Perhaps most 118 importantly, the life history shift of Dol ly Varden in this interspecific hybrid zone mimics the intraspecific resource polymorphism and ecological speciation commonly observed in freshwater temperate fishes (discussed below; Skulason & Smith 1995; Schluter 1996; Orr & Smith 1998). Despite these differences, this hybrid zone contained many characteristics originally observed in narrow, high-rate hybrid zones. 1) Asymmetric introgression of m t D N A , which provided evidence for unidirectionality in the initial interspecific crosses. 2) A suspected behavioural root for the unidirectionality (i.e. sneaking), which also caused (or at least predominated in) interspecific hybridization. 3) Asymmetry in the introgression of nuclear alleles, which provided evidence for differential selection across loci. 4) Evidence for prezygotic reproductive isolation, due to assortative mating. 5) Evidence for postzygotic reproductive isolation, in the form of extrinsic ecological and reproductive selection against morphologically intermediate hybrids. In short, not only was a study of process in hybridization successful within a dispersed, dilute hybrid zone, but many traits commonly associated with narrow hybrid zones were also observed. The study of pattern in hybridization was also successful. The discordant mitochondrial and nuclear D N A phytogenies are the most typical, reliable evidence for historical introgressive hybridization between two species (Arnold 1997). This evidence for historic hybridization provided excellent background information for the study of current hybridization. This is particularly important in that it showed that secondary contact since the last glaciation (and the associated lack of reproductive isolation) is not the sole cause of hybridization. Hybridization and sympatry between Do l ly Varden and bull trout have a long history. 119 In evolutionary biology, hybrid zones are important because they are living laboratories that can be used for the study of reinforcement and the evolution of reproductive isolation. In this case, however, the traits traditionally important in prezygotic isolation have already evolved. Differences in mature body size and spawning time, location, colouration and behaviour are already present. These traits appear to be sufficient in preventing interspecific pairings, but do not prevent hybridization due to sneaking. If the avenue for hybridization is limited, as I have proposed, to the sneaking of male Dol ly Varden on bull trout spawning pairs, then complete prezygotic reproductive isolation wi l l be difficult to establish. The Dol ly Varden are simply employing an intraspecific parasitic reproductive trait, which, presumably, is relatively successful in allopatric populations of Dol ly Varden (which have more variability in adult spawning size). The fact that, in sympatry, bull trout and Dol ly Varden adopt alternate life histories, with the co-requisite difference in body size, means that any sneaking done by Dol ly Varden males in sympatry will lead to interspecific mating. The selective penalty for interspecific straying would be minimal at best, since sneaking 'on the side' is not likely to significantly lower a Dol ly Varden male's intraspecific mating success. This fact limits the strength of reinforcement in enhancing isolation. Further, since bull trout are already extremely aggressive against intruding males of any species during spawning, they would play virtually no role in the evolution of prezygotic isolation. They are essentially 'victims' of Dol ly Varden sneaking. It is interesting that the instigating factor in this hybridization (i.e. sneaking) is a common life history trait amongst salmonid species. Why then is hybridization so 120 corrimon among Salvelinus spp., but not within Oncorhynchusl The variation in hybridization between taxa has been well documented in the literature (e.g. Hubbs 1955; Arnold 1997), but rarely is any evidence put forth as to the cause of these differences. One fundamental difference between these salmonid genera is the semelparous nature of most salmon, while char are iteroparous and so continue to increase in size as mature adults and have more than one chance to reproduce. Is there less pressure on Salvelinus spp. to evolve reproductive isolation? Is the variation in mature body size important? A n exception that helps to prove the rule is hybridization between cutthroat (O. clarki) and rainbow trout (O. mykiss), the only two Oncorhynchus species that can be repeat spawners. Such conjectures are certainly interesting, but not easily testable. (Perhaps this is why there is so little mention in the literature of the causes of variation in levels of hybridization between related taxa??) One potential barrier to reproductive isolation that received little attention in my thesis is reproductive isolation due to gamete incompatibility. Many species have such well-evolved gamete recognition systems that it can almost completely counter the effects of random interspecific mating (e.g. Howard et al. 1998; Rieseberg et al. 1998). Could the potential benefit of reproductive isolation select for such a recognition system in bull trout eggs? Or, less likely, in Dol ly Varden sperm? Intraspecific sperm competition is certainly well documented in fishes, particularly those species that use various 'parasitic' reproductive tactics (Taborsky 1998). Unfortunately, preliminary efforts to measure interspecific sperm competition for this thesis had to be abandoned, due to difficulties in obtaining gametes from both species simultaneously. 121 Perhaps the most important facet of my thesis, certainly one of the most interesting, is the fact that this hybrid zone mimics the conditions of ecological speciation so commonly observed in temperate fishes (Skulason & Smith 1995; Schluter 1996). In allopatry, both Dol ly Varden and bull trout are typically large-bodied, piscivorous, migratory species. In sympatry, however, Dol ly Varden adopt a vastly different life history, becoming small-bodied, stream resident, invertebrate feeders, thus reducing interspecific competition. This resource partitioning mimics a situation of intraspecific ecological speciation. The fact that interspecific hybridization and introgression can and do occur further improves, i f not completes, the parallel (Skulason & Smith 1995). The genetic plasticity required to produce the small-bodied form is present in both species. Both Dol ly Varden and bull trout have small-bodied, non-migratory populations. Admittedly, these populations are typically associated with barriers to migration. This is in contrast to, for example, the resource polymorphism of O. nerka (sockeye and kokanee), in which the freshwater resident form has evolved multiple times while in sympatry with the anadromous form (Taylor et al. 1996). The fact that Dol ly Varden adopt the resident life history, rather than bull trout, is a predictable result of disruptive selection. B u l l trout are slightly larger and have a jaw shape more specialized for piscivory (Cavender 1978). If one of the two species is to be 'forced' into a resident life history, Dol ly Varden is certainly the more likely victim. M y thesis has provided valuable detail about Dol ly Varden / bull trout hybridization, both in pattern and process. This information is crucial in management and conservation of these two salmonid species in B . C . On a larger scale, my study has shown the importance of including 'messier' hybrid zones, certainly quite common in 122 nature, in the study of hybridization in evolution. M y thesis has also described a case of interspecific resource partitioning which parallels the intraspecific ecological speciation and trophic polymorphisms so common in temperate freshwater fish species. 123 References Abbott R J 1992 Plant invasions, interspecific hybridization, and the evolution of new plant taxa. Trends in Ecology and Evolution 7(12): 401-405. Albuquerque G S , C A Tauber, & M J Tauber 1996 Postmating reproductive isolation between Chrysopa quadripunctata and Chrysopa slossonae: mechanisms and geographic variation. Evolution 50(4): 1598-1606. Armstrong R H & JE Morrow 1980 The Dolly Varden char, Salvelinus malma. p. 90-140 In: Charrs: salmonid fishes of the genus Salvelinus. Edited by: E K Balon, Dr . W . Junk Pubis., The Hague, The Netherlands, 928 p. Arnold M L 1997 Natural Hybridization and Evolution. Oxford University Press, New York, 215 p. Arnold M L & S A Hodges 1995 Are natural hybrids fit or unfit relative to their parents? Trends in Ecology and Evolution 10(2): 67-71. Asmussen M A , J Arnold, & JC Avise 1987 Definition and properties of disequilibrium statistics for associations between nuclear and cytoplasmic genotypes. Genetics 115: 755-768. Avise JC & N C Saunders 1984 Hybridization and introgression among species of sunfish (Lepomis): analysis by mitochondrial D N A and allozyme markers. Genetics 108: 237-55. Avise JC & M J Van den Avyle 1984 Genetic analysis of reproduction of hybrid white bass X striped bass in the Savanna River. Trans. Am. Fish. Soc. 113: 563-570. Barton D M 1979 The dynamics of hybrid zones. Heredity 43:341-359. Barton N H & G M Hewitt 1985 Analysis of hybrid zones. Annual Review of Ecology and Systematics 16: 113-48. Baxter JS, E B Taylor, R H Devlin, J Hagen, & JD McPhai l 1997 Evidence for natural hybridization between Dolly Varden (Salvelinus malma) and bull trout (Salvelinus confluentus) in a northcentral British Columbia watershed. Can. J. Fish. Aquat. Sci. 54: 421-429. Becker II, W S Grant, R Kirby, & F T Robb 1988 Evolutionary divergence between sympatric species of southern African Hakes, Merluccius capensis and M. paradoxus II Restriction enzyme analysis of mitochondrial D N A . Heredity 61:21-30. 124 BehnkeRJ 1980 A systematic review of the genus Salvelinus. pp. 441-481. In Charrs: Salmonid Fishes of the Genus Salvelinus. Edited by E . K . Balon Junk Publishers, The Hague, 928 p. Bermingham E & JC Avise 1986 Molecular zoogeography of freshwater fishes in the southeastern United States. Genetics 113:939-965. Bernatchez L & JJ Dodson 1991 Phylogeographic structure in mitochondrial D N A of the lake whitefish (Coregonus clupeaformis) and its relation to Pleistocene glaciations Evolution. 45: 1016-1035. Bernatchez L , H Glemet, C C Wilson, & R G Danzmann 1995 Introgression and fixation of Arctic char (Salvelinus alpinus) mitochondrial genome in an allopatric population of brook trout (Salvelinus fontinalis). Can. J. Fish. Aquat. Sci. 52: 179-85. Berst A H , P E Ihssen, G R Spnagler, G B Ayles, & G W Martin 1980 The splake, a hybrid charr Salvelinus namaycush x S. fontinalis. p 841-887 In: Charrs: salmonid fishes of the genus Salvelinus. Edited by: E K Balon, Dr. W . Junk Pubis., The Hague, The Netherlands, 928 p. Bert T M & W S Arnold 1995 A n empirical test of predictions of two competing models for the maintenance and fate of hybrid zones: both models are supported in a hard-clam hybrid zone. Evolution 49(2): 276-289. Berthold P & U Querner 1981 Genetic basis of migratory behaviour in European warblers. Science 212(3): 77-79. Billingdon N , P D N Hebert, & R D Ward 1990 Allozyme and mitochondrial D N A variation among three species of Stizostedion (Percidae): phylogenetic and zoogeographical implications. CanJ. Fish. Aquat. Sci. 47:1093-1102. Bullini L 1994 Origin and evolution of animal hybrid species. Trends in Ecology and Evolution 9(11): 422-426. Bustard D 1996 Kemess South Project: 1995 fisheries studies final report. Consult. Rept. prep, by D . Bustard and Assoc. for E l Condor Resources, L td . A n d Royal Oak Mines, Inc., Vancouver, B . C . Bustard D 1997 Kemess South Project: 1996 fish monitoring studies. Consult. Rept. prep, by D . Bustard and Assoc. for Kemess Mines, Inc., Seattle, W A . Bustard D 1998 Kemess South Project: 1997 fish monitoring studies. Consult. Rept. prep, by D . Bustard and Assoc. for Kemess Mines, Inc., Seattle, W A . 125 Bustard D 1999 Kemess South Project: 1998 fish monitoring studies. Consult. Rept. prep, by D . Bustard and Assoc. for Kemess Mines, Inc. Campton D E 1987 Natural hybridization and introgression in fishes: Methods of detection and genetic interpretations, pp 161-192. In Population Genetics & Fishery Management. Edited by N . Ryman and F. Utter. University of Washington Press, 420 p. Cavender T M 1978 Taxonomy and distribuation of the bull trout, Salvelinus confluentus (Suckley), from the American Northwest. Calif Fish and Game 64(3): 139-74. Crane P A , L W Seeb, & JE Seeb 1994 Genetic Relationships among Salvelinus species inferred from allozyme data. Can. J. Fish. Aquat. Sci. 51 (Suppl. 1): 182-197. Cruzan N B & M L Arnold 1993 Ecological and genetic associations in an Iris hybrid zone. Evolution 47(5): 1432-1445. Cruzan N B & M L Arnold 1993 Assortative mating and natural selection in an Iris hybrid zone. Evolution 48(6): 1946-1958. DeCicco A L 1992 Long-distance movements of anadromous Dol ly Varden between Alaska and the U.S .S .R. Arctic 45(2): 120-123. DeMarais B D , T E Dowling, M E Douglas, W L Minckley, & P C Marsh 1992 Origin of Gila seminuda (Teleostei: Cyprinidae) throught introgressive hybridization: Implications for evolution and conservation. Proc. Nat. Acad. Sci. U S A 89: 2747-2751. Dowling T E , R E Broughton, & B D DeMarais 1997 Significant role for historical effects in the evolution of reproductive isolation: evidence from patterns of introgression between the Cyprinid fishes, Luxilus cornutus and Luxilus chrysocephalus. Evolution 51(5): 1574-1583. Echelle A A & PJ Connor 1989 Rapid, geographically extensive genetic introgression after secondary contact between two pupfish species (Cyprinodon, Cyprinodontidae). Evolution 43(4): 717-727. FelsensteinJ 1995 P H Y L I P (Phylogeny Inference Package) Version 3.57c Department of Genetics, University of Washington, Seattle, W A . Ferris S D , R D Sage, C - M Huang, JT Neilsen, U Ritte, & A C Wilson 1983 Flow of mitochondrial D N A across a species boundary. Proc. Nat. Acad. Sci. U S A 80: 2290-2294. Glemet H , P Blier, L Bernatchez 1998 Geographical extent of Arctic char (Salvelinus alpinus) m t D N A introgression in brook char populations (S. fontinalis) from eastern Quebec, Canada. Mol. Ecol. 7: 1655-1662. 126 Grewe P M , N Billington, & P D N Hebert 1990 Phylogenetic relationships among members of Salvelinus inferred from mitochondrial D N A divergence. Can. J. Fish. Aquat. Sci. 47: 984-991. Gross M R 1985 Disruptive selection for alternative life histories in salmon. Nature 313: 47-48. Gyllensten U & A C Wilson 1986 Mitochondrial D N A of salmonids: inter- and intraspecific variability detected with restriction enzymes, pp. 301-317 In: Population Genetics and Fisheries Management. Edited by N Ryman & F Utter. University of Washington Press, Seattle, W A , 420 p. Haas G R 1988 The systematics, zoogeography, and evolution of Dolly Varden and bull trout in British Columbia. M . Sc. Thesis, University of British Columbia, B . C . , 201 p. Haas G R & I D McPhai l 1991 Systematics and distribution of Dol ly Varden (Salvelinus malma) and bull trout (Salvelinus confluentus) in North America. Can. J. Fish. Aquat. Sci. 48:2191-2211. Hagen J 2000 Reproductive isolation between Dolly Varden (Salvelinus malma) and bull trout (S. confluentus) in sympatry: the role of ecological factors. M S c thesis, University of British Columbia, 122 p. Hamada M , N Takasaki, ID Reist, A L DeCicco, A Goto & N Okada 1998 Detection of the ongoing sorting of ancestrally polymorphic SINEs toward fixation or loss in populations of two species of charr during speciation. Genetics 150: 301-311. Hammar J, JB Dempson, & E Skdld 1989 Natural hybridization between Arctic char (Salvelinus alpinus) and lake char (S. namaycush): evidence from Northern Labrador. Nordic J. Freshw. Res. 65: 54-70. Hammar J, JB Dempson, & E Verspoor 1991 Natural hybridization between Arctic char (Salvelinus alpinus) and brook trout (S. fontinalis): evidence from northern Labrador. Can. J. Fish. Aquat. Sci. 48: 1437-1445. Hartl D L & A G Clark 1997 Principles of Population Genetics. 3 r d E d . Sinauer Associates, Sunderland M A , 542 p. Hewitt G M 1993 After the ice: Parallelus meets Erythropus in the Pyrenees, pp 140-164. In: Hybrid Zones and the Evolutionary Process. Edited by: R G Harrison. Oxford University Press, New York, 364 p. Hewitt G M 1996 Some genetic consequences of ice ages, and their role in divergence and speciation. Biol. J. Linnean Soc. 58: 247-76. 127 Hillis D M & JJ Bu l l 1993 A n empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology 42: 182-192. Howard D J , P G Gregory, J Chu, & M L Cain 1998 Conspecific sperm precedence is an effective barrier to closely related species. Evolution 52:511-516. Hubbs C L 1955 Hybridization between fish species in nature. Syst. Zool. 4: 1-20. . Jiggins C D & J Mallet 2000 Bimodal hybrid zones and speciation. Trends in Ecology and Evolution 15(6): 250-255. Kimura M 1980 A simple model for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Molecular Evolution 16: 111-120. Kitano S 1996 Size-related factors causing individual variation in seasonal reproductive success of fluvial male Dolly Varden (Salvelinus malma). Ecology of Freshwater Fish 5: 59-67. Kitano S, K Iguchi, S Ishikawa, & M Nishida 2000 Directional hybridization between native white-spotted charr Salvelinus leucomaenis and introduced brook trout S. fontinalis in a stream in central Japan. Submitted to Kitano S, K Maekawa, S Nakano, & K D Fausch 1994 Spawning behaviour of bull trout in the Upper Flathead Drainage, Montana, with special reference to hybridization with brook trout. Trans. Am. Fish. Soc. 123: 988-992. L a m b T 1987 Call site selection in a hybrid population of treefrogs. Animal Behav. 35: 1140-1144. Lamb T & JC Avise 1986 Directional introgression of mitochondrial D N A in a hybrid population of tree frogs: the influence of mating behavior. Proc. Nat. Acad. Sci. U S A 83: 2526-2530. Lamb T & JC Avise 1987 Morphological variability in genetically defined categories of Anurian hybrids. Evolution 41(1): 157-65. Latham 2000 M S c thesis in progress. Leary R F & F W Allendorf 1997 Genetic confirmation of sympatric bull trout and Dolly Varden in western Washington. Trans. Am. Fish. Soc. 126: 715-720. Leary R F , F W Allendorf, & K L Knudsen 1983 Consistently high meristic counts in natural hybrids between brook trout and bull trout. Syst. Zool. 32: 369-376. 128 Leggett J W 1980 Reproductive ecology and behaviour of Dolly Varden charr in British Columbia, p 721-737 In: Charrs: salmonid fishes of the genus Salvelinus. Edited by: E K Balon, Dr. W . Junk Pubis., The Hague, The Netherlands, 928 p. Lindsey C C & JD McPhai l 1986 Zoogeography of fishes of the Yukon and Mackenzie basins, pp 639-674. In: The Zoogeography of North American Freshwater Fishes. Edited by: C H Hocutt and E O Wiley, John Wiley and Sons, New York, 866 p. Louis E J & E R Dempster 1987 A n exact test for Hardy-Weinberg and multiple alleles. Biometrics 43: 800-802. MacCallum CJ , B Nurnberger, N H Barton, & J M Szymura 1998 Habitat preference in the Bombina hybrid zone in Croatia. Evolution 52(1): 227-239. Maekawa K , T Hino, S Nakano, & W W Smoker 1993 Mate preference in anadromous and landlocked Dolly Varden (Salvelinus malma) females in two Alaskan streams. Can. J. Fish. Aquat. Sci. 50: 2375-2379. Maekawa K , N Shigeru, & S Yamamoto 1994 Spawning behaviour and size-assortative mating of Japanese charr in an artificial lake-inlet stream system. Env. Biol. Fishes 39(2): 109-117. McEl roy D , P Moran, E Bermingham, & I Kornfield 1992 R E A P , A n integrated environment for the manipulation and phylogenetic analysis of restriction data. J. Heredity 83: 157-158. McGowan C & W S Davidson 1992 Unidirectional hybridization between brown trout (Salmo trutta) and Atlantic salmon (S. salar) in Newfoundland. Can. J. Fish. Aquat. Sci. 49: 1953-1958. M c K a y SJ, R H Devlin, & M J Smith 1996 The phylogeny of Pacific salmon and trout based on growth hormone type-2 (GH2) and mitochondrial N A D H dehydrogenase subunit 3 (ND3) D N A sequences. Can. J. Fish. Aquat. Sci. 53: 1165-1174. McPhai l JD 1961 A systematic study of the Salvelinus alpinus complex in North America. J. Fish. Res. Board. Can. 18: 793-816. McPhai l JD & C C Lindsey 1970 Freshwater Fishes of Northwestern Canada and Alaska. Bulletin of the Fisheries Research Board of Canada, no. 173, Ottawa. McPhai l JD & C C Lindsey 1986 Zoogeography of the freshwater fishes of Cascadia (the Columbia System and rivers north to the Stikine). pp 615-637. In: The Zoogeography of North American Freshwater Fishes. Edited by: C H Hocutt and E O Wiley, John Wiley and Sons, New York, 866 p. 129 McPhai l I D & C B Murray 1979 The early life history of Dolly Varden (Salvelinus malma) in the upper Arrow Lakes. Report of the Institute of Animal Resource Ecology and Department of Zoology (University of British Columbia) to B C Hydro and Ministry of Environment, Victoria, B . C . McPhai l I D & E B Taylor 1995 Final Report to Skagit Environmental Endowment Commision: Skagit Char Project (Project 94-1). June 1995. Meagher S & T E Dowling 1991 Hybridization between the Cyprinid fishes Luxilus albeolus, L. cornutus, and L. cerasinus with comments on the proposed hybrid origin of L. albeolus. Copeia 1991(4): 979-991. Meyer A , T D Kocher, P Basasibwaki, & A C Wilson 1990 Monophyletic origin of Lake Victoria cichlid fishes suggested by mitochondrial D N A sequences. Nature 347: 550-553. Moore WS 1977 A n evaluation of narrow hybrid zones in vertebrates. Quart. Rev. Biol. 52: 263-277. Mukai T, K Naruse, T Sato, A Shima, & M Morisawa 1997 Multiregional introgressions inferred from the mitochondrial D N A phylogeny of a hybridizing species complex of gobiid fishes, genus Tridentiger. Mol. Biol. Evol. 14(12): 1258-65. Nei M 1987 Molecular Evolutionary Genetics Columbia University Press, New York , 512 p. Nei M & F Tajima 1981 D N A polymorphism detectable by restriction endonucleases. Genetics 97: 145-163. Neilsen E E , M M Hansen, & K - L D Mensberg 1998 Improved primer sequences for the mitochondrial N D 1 , ND3/4 and ND5/6 segments in salmonid fishes: application to R F L P analysis of Atlantic salmon. J. Fish. Biol. 53: 216-220. Orr M R & T B Smith 1998 Ecology and speciation. Trends in Ecology and Evolution 13(12): 502-506. Phillips R B , L I Gidex, K M Westrich, & A L DeCicco 1999 Combined phylogenetic analysis of ribosomal ITS1 sequences and new chromosome data supports three subgroups of Dolly Varden char (Salvelinus malma) Can. J. Fish. Aquat. Sci. 56: 1504-1511. Phillips R B , S A Manley, & T J Daniels 1994 Systematics of the salmonid genus Salvelinus inferred from ribosomal D N A sequences Can. J. Fish. Aquat. Sci. 51 (Suppl. 1): 198-204. 130 Phillips R B , K A Pleyte, & M R Brown 1992 Salmonid phylogeny inferred from ribosomal D N A restriction maps Can. J. Fish. Aquat. Sci. 49: 2345-53. Phillips R B , S L Sajdak, & M J Domanico 1995 Relationships among charrs based on D N A sequences Nordic. J. Freshw. Res. 71: 378-391. Pleyte K A , S D Duncan, & R B Phillips 1992 Evolutionary relationships of the salmonid fish genus Salvelinus inferred from D N A sequences of the first internal transcribed spacer (ITS 1) of ribosomal D N A Mol. Phyl. Evol. 1(3): 223-230. Quattro J M , JC Avise, & R C Vrijenhoek 1991 Molecular evidence for multiple origins of hybridogenic fish clones (Poeciliidae: Poeciliopsis) Genetics 127: 391-398. Quattro J M , JC Avise, & R C Vrijenhoek 1992 Mode of origin and sources of genotypic diversity in triploid gynogenetic fish clones (Poeciliopsis: Poeciliidae) Genetics 130: 621-628. Rand D M & R G Harrison 1989 Ecological genetics of a mosaic hybrid zone: mitochondrial, nuclear, and reproductive differentiation of crickets by soil type Evolution 43(2): 432-449. Raymond M & F Rousset 1995 G E N E P O P (version 1.2): population genetics software for exact tests and ecumenicism. J. Heredity 86: 248-249. Redenbach Z & E B Taylor 1999 Zoogeographical implications of variation in mitochondrial D N A of Arctic grayling (Thymallus arcticus) Molecular Ecology 8: 23-35. Rice W R 1989 Analyzing tables of statistical tests Evolution 43(1): 223-225. Rieseberg L H , SJE Baird, & A M Desrochers 1998 Patterns of mating in wild sunflower hybrid zones Evolution 52:713-726. Rieseberg L H & C R Linder 1999 Hybrid classification: insights from genetic map-based studies of experimental hybrids. Ecology 80(2): 361-370. Rundle H D & D Schluter 1998 Reinforcement of stickleback mate preferences: Sympatry breeds contempt Evolution 52(1): 200-208. Schluter D 1996 Ecological speciation in postglacial fishes Phil. Trans. R. Soc. Lond. B 351: 807-814. Scribner K T & JC Avise 1994 Cytonuclear genetics of experimental fish hybrid zones inside Biosphere 2. Proc. Natl. Acad. Sci. U S A 91: 5066-5069. 131 Scribner K T & JC Avise 1994 Population cage experiments with a vertebrate: the temporal demography and cytonuclear genetics of hybridization in Gambusia fishes. Evolution 48(1): 155-171. Sexauer H A 1994 Life history aspects of bull trout, Salvelinus confluentus, in the eastern Cascades, Washington M.Sc . thesis, Central Washington University, Ellensburg, Wash. Sigurjonsdottir & K Gunnarsson 1989 Alternative mating tactics of Arctic charr, Salvelinus alpinus, in Thingvallavatn, Iceland Env. Biol. Fishes 26: 159-176. Skulason S & T B Smith 1995 Resource polymorphism in vertebrates. Trends in Ecology and Evolution 10(9): 366-370. Smith G R 1992 Introgression in fishes: significance for paleontology, cladisitcs, and evolutionary rates. Systematic Biology 41:41-57. Spolsky C & T Uzzel l 1984 Natural interspecies transfer of mitochondrial D N A in amphibians. Proc. Nat. Acad. Sci. U S A 81: 5802-5805. Spruell P, B E Rieman, K L Knudsen, F M Utter, & F W Allendorf 1999 Genetic population structures within streams: microsatellite analysis of bull trout populations. EcolFreshw. Fish 8: 114-131. Svedang H 1992 Observations on interbreeding between dwarf and normal arctic charr, Salvelinus alpinus, from Stora Rosjon, central Sweden. Env. Biol. Fishes 33: 293-298. Taborsky M 1998 Sperm competition in fish: 'bourgeois' males and parasitic spawning. Trends in Ecology and Evolution 13(6): 222-227. Taylor E B , C J Foote, & C C Wood 1996 Molecular genetic evidence for parallel life history evolution within a Pacific salmon (sockeye salmon and kokanee, Oncorhynchus nerka). Evolution 50: 401-416. Taylor E B , S Harvey, S Pollard, & J Volpe 1997 Postglacial genetic differentiation of reproductive ecotypes of kokanee Oncorhynchus nerka in Okanagan Lake, British Columbia. Molecular Ecology 6:213-216. Taylor E B , S Pollard, & D Louie 1999 Mitochondrial D N A variation in bull trout (Salvelinus confluentus) from northwestern North America: implications for zoogeography and conservation. Mol. Ecol. 8: 1155-1170. Taylor E B , S Pollard, Z Redenbach, A Costello, & C Pacas 2000 Nested analysis of genetic diversity in bull trout (Salvelinus confluentus): implications for speciation, biogeography, and conservation. Submitted to: Can. J. Fish. Aquat. Sci. 132 Weir B S 1990 Genetic Data Analysis. Sinauer Publ., Sunderland, M A , 377 p. Weir B S & C C Cockerham 1984 Estimating F-statistics for the analysis of population structure. Evolution 38: 1358-1370. Wilde G R & A A Echelle 1992 Genetic status of Pecos pupfish populations after establishment of a hybrid swarm involving an introduced congener. Trans. Am. Fish. Soc. 121(3): 277-86. Wilson C & L Bernatchez 1998 The ghost of hybrids past: fixation of arctic charr (Salvelinus alpinus) mitochondrial D N A in an introgressed population of lake trout (S. namaycush). Mol. Ecol. 7: 127-132. Wilson C & P D N Hebert 1993 Natural hybridization between Arctic char (Salvelinus alpinus) and lake trout (S. namaycush) in the Canadian Arctic. Can. J. Fish. Aquat. Sci. 50: 2652-2658. 133 Appendix 1: N D 1 m t D N A sequences. Haplotype D V - 1 is given in full. Base pairs 1-36 represent a partial sequence for the t R N A - G l n gene, base pairs 39-111 represent the t R N A -Ile gene, and base pairs 119-503 represent the beginning of the N A D H - 1 gene (975 bp in full length). For the other haplotypes, only variant sites are listed. Sample locations for given haplotypes are listed in Table 1. BT- I , B T - C , and S. fontinalis sequences are from Taylor etal. (1999). D V - 1 1 T T T T G A T C T C T T G A G G A T G G G T T C G A A T C C C T T C T C T C T A G A A T T G A G G G 5 0 5 1 G A C T T G A A C C C C T A T C A G C C A C G C T A T C A A G G T G G T C C T T A A G C A T T C A G 1 0 0 1 0 1 G C A C A A T T C C T T G G A A T T A A A G T T G G G G A G G G A G G C C G G C T A G T G C A A T A 1 5 0 1 5 1 G G A A G C G C G A G G T G T C A T A G G A C A A G T G C T A G G G T T A A T G G A A G G A A G C T 2 0 0 2 0 1 T T T T C A A A C T A A A T G C A T A A G T T G G T C G T A T C G A A A T C G T G G G T A G G A A G 2 5 0 2 5 1 C T C G T A C T C A C A A A A A C A C A A C G G A G A G T A G G G C A G C T T T C G T T A T C A G A 3 0 0 3 0 1 T T T A C G G C T G T T A A T T C A G G G A A A G C G G G G A T G T G G G A T G C G C C T A G A A A 3 5 0 3 5 1 T A G G A C G G C T G A A A G T G T A T T T A T T A G A A G A A T A T T A G C G T A T T C A G C C A 4 0 0 4 0 1 G G A A G A A G A G G G C G A A T G G T C C T C C A G C G T A T T C T A C G T T G A A T C C T G A A 4 5 0 4 5 1 A C T A A T T C T G A T T C T C C T T C T G T G A G G T C A A A A G T G C A C G G T T T G T T T C G 5 0 0 5 0 1 G C C A A A G T A G A A A T G T A T C A T A T G G C G G C A A G G G T T C A G G C T G G T A C T A A 5 5 0 5 5 1 T A A T C A G A T G C T T T C T T G A G 5 7 0 Diagnostic mutations are designated below the variant sites for the following clades: * distinguishes Dol ly Varden Clade N from bull trout and Dol ly Varden Clade S, distinguishes Interior bull trout from remainder ° distinguishes Dol ly Varden Clade S and Coastal bull trout from remainder 134 Appendix 1 (cont.) LO O ro LD O rH LO O O r - rH VO 00 VO LO VO ro VO CN LO LO LD o • ^ rH CN ro r - r -ro VD VO ro LO cn ro LO ro LD ro ^ CN ro CN r-ro CN ro rH ro O LO ro O ro ro O o CO CN cn CTl 0) CM cn rH 4J CM oo CO •rl CN CO LO w CM o CTl CN r - CN 4J CN i> O ti CM VO r> a CN VO rH •rl CN o U CN ro rH CN CN CO > CN rH VO CN rH CN H rH r- ro Q rH V0 CO S rH LO vo rH LD rH rH •si* ro rH • ^ rH rH CN VO rH CN o rH rH o CX) o VO vo LO CN LO CN ro CN CN rH CT\ rH LO rH ro r> o CD CD FH FH FH & H FH C J L H L H F H F H F H F H F H F H F H F H F H • • FH * FH • FH • • < ! < ; < ! i < < ! r < < L l < C < ! < ! < ! • < ! < ! * C J C J C J C J C J C J C J C J <" CJ C J u FH CD CD CD FH U C J CD o o C J < C J FH < O C J FH < FH O <3 • • C J • • U • • CD • • U • • CD • • u • • < • • u • • FH • • C J • • CD • • CD • • FH • • O • • < • • C J • • <c • • u • • FH • • CD • • FH < • FH • • CD < <C CD • • CD • • FH • • FH • • CD < c < i i < ; < i < i < i < c < i < f i < ! < i a CD CD u FH FH < • CD FH FH • FH • <C • FH C D C D C D C D C D C D C D C D C D C D C D F H F H F H F H F H F H F H F H F H F H F H • FH • CD * FH FH * CD CD • C J FH FH FH FH • < • • • • C J • CD C J < C J CD FH FH CD < C J C J FH CD CD CD • • FH < CD C J FH CD CD FH FH C J • • C J < FH • • < • • • • CD • • C J C J C J C J C J • C J • • CD C J < CD FH r H C N r O ^ f L O V O r - C O C h r H I I I I I I I I I I I I > > > > > > • > > > > > > P Q P Q Q P P Q P P Q P O rH CCJ , Q U (fl rfjmupwuuuH I I I I I I I > > > FH FH FH FH CJ H i FH in p p p m m m m m m ^ t o t Q • H CO 4J «« R w CO 0 (IJ R 4J • • - H CO (/) U - H 4J cn 0 a - H p 135 Appendix 2: Growth Hormone 2 intron C sequences. The full sequence is given for D V - 1 , which is identical the those in D V - 1 0 , D V - B , D V - D . For the other haplotypes, only variant sites are listed. S. namaycush sequence was obtained from M c K a y et al. (1996). 1 C C T G A C G T T G C C G T C G C C C C C C A G A T T C T G A T A G T A G T T C C C G T A G G G G G 5 0 5 1 G C A G A T G C T G A G A G T C A T T G T C A T C C A G G C T C A G T A C G C C G T C C T G G C T C 1 0 0 1 0 1 C C C T G T G G G A G A C A G A G A G A G A T G C A T G G T G G G G T G G A T G T G T G C A T A T C 1 5 0 1 5 1 A A T A T T T G T T T A T T A G G C C A T T C T A A A A G C A C A T A T A G C A A A A G A C A C A T 2 0 0 2 0 1 G C G T A A A T G T C C T T G A A T G G G C C A A T A T T C A G T C A T T G A A T G G G G T G T T A 2 5 0 2 5 1 C T G A T A T A G T C A C T A T T A G T G A T G A T C A T C A C A A T G T A G C A G A C G T T T A G 3 0 0 3 0 1 A A A G C A T A G T T G T A T T T C A C T T T C A T G T C G G T G G A A G A T A A T A A C C A C A C 3 5 0 3 5 1 T G T A A A C C C C A A T G T G C T G T C A T T A C T C A A A A C T T T T G A G G A A A C C G T T G 4 0 0 4 0 1 C A C T T A A T A T A C A G A A T C T G A C T A C A G T T A C T T A A A G T G A T T T T T A T A G T 4 5 0 4 5 1 C A T T A C T C A A A C C T A T A G A G T C A C C C T G T A G A G G 4 8 4 D V - 1 V a r i a n t S i t e s 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 4 4 1 1 3 6 8 8 3 5 6 6 8 7 7 7 7 7 8 8 8 8 9 9 9 9 9 9 0 4 4 5 9 9 4 8 3 2 9 1 2 5 8 4 5 57 5 4 5 6 7 8 3 6 8 9 1 2 5 6 8 9 6 4 7 6 2 5 3 3 D V - 1 * D V - 2 D V - 6 B T - I B T - C T C C A G T G G T A G A G A T C A A G A G A G G T T A A A A A A C T G G . . G . C A T C G A G A T A - G T T - - T A - - A G . . G . . . . A . . . G . C A T C G A G A T A - G T T - - T A - - A G . . G . . . . A . T G - A G . - . - G A G . G C G . K C S.namay * Includes D V - 1 0 , D V - B , D V - D 136 Appendix 3: Ribosomal D N A first internal transcribed spacer region ( r D N A ITS-1) sequences. The sequence of D V - 1 is given in full. For the other haplotypes, only variant sites are listed. S. m. krascheninninovi, S. m. malma, and S. m. lordi sequences were obtained from Phillips et al. (1999), S. alpinus, S. leucomaenis and S. fontinalis sequences were obtained from Pleyte etal. (1992). D V - 1 1 G G T T G C C A G C C G C C G G C A T G G G G C T G T G C T C C G A A A A C C A A A G T C T G C T G 5 0 5 1 T G G G G T G G G T A G G G T A T G G G G G C T C A C C C C C C C C G C C T C G C C C A T C T C T C 1 0 0 1 0 1 G G C G C G G G T G T C C T C G G T C T T A G C C C G G - T T C C C C G C T A T T C C T T T T G C C 1 5 0 1 5 1 T A A G G G T T G C A C C C G A C C G G C T C C A T C C C T T T T C C C C G T T A G C C A C G G C C 2 0 0 2 0 1 A C A T G G C G C A C C T A T G G G C A G G T G A G T C G G C C G C T A C C G A A G G G G A C T G G 2 5 0 2 5 1 G G G T G T C C G G T G A A C C G G G A C T T C C C G A A A C G T A A T C C C A T T T A T A A G C G 3 0 0 3 0 1 G C T T G A G T A T C G C C C A G T A T C C T C G C T C G G C A C C G G G A A C C C A G T C A A C C 3 5 0 3 5 1 G C T C T G C G C C C C G G C G C A G G C G G G G G T T T A A T G T C T C C C C C A G C C C T C A C 4 0 0 4 0 1 K G C - - G C T T C 4 1 0 V a r i a n t S i t e s 111111111112222222222222222222222222223333 33333333333333 3344444 44557 8890123555689001222233355677 8888888999990000223333 566668899900000 131583426695238107463156923 8690171234569234590127480128856798918913456 D V - 1 AGTGCCCCGG-CAATATGTGTGAGGCGCGGGCGCGTAATCTTATCGGCGCCGCAAGCGCGCCCCAKCGCG D V - 2 T S T . . . . A C . S . . . R G A . . . D V - 1 0 C G G K G . . . . D V - B . . . . G SC G . . . . D V - D YS G G . . . K S . m . k r a G . . T T . . . . A C G . . . . S . m . l o r G G . . . . S . m . m a l G . . T T . . . . A G . . . . B T - I . C C T S G . . - A G . . C . . . . C . A . M . . . T . . . G T C . A . G T G . R T G R . . R . . . . S . . R G . . . . B T - C . C C T G G . . - A G . . C T . . . G T C . A . G T G . . T G G . . . . S . a l p i n G C T CG G . . . . S . namay. . C . T G — . A . . T — . GC . . A . C . . . TC T S G T C . . G . T T - . C G A . . . S . l e u c o . CC . . G G T . - C . C A GTC . . . G T G . . T G . . T . . C . . . CG . A . - A C G . . . . S . f o n t i . . C . . G . T . . . . T - - . T C . C A C A T . . T T A . . . . - . . . . A T . T . . G . . C G . . . - . C G . - - . 137 Appendix 4: Genotypes by age class and by tributary. Genotypes include four nuclear loci (GH2, ITS-1, M T B , & F O K ) and one mitochondrial locus (ND5/6). These hybrid classes designate genotype only, and are not intended to reflect ancestry. Locality Year, N B T 2 B C B T F , 3 F 1 , 4 B C D V I G D V 6 D V 2 Age Class 5 1.5 Kemess 1997, fry 185 127 4 0 1/0 0/4 0 49 Creek 1998, fry 211 121 3 0 1/0 7/4 1 74 1998, 1+ 122 52 1 0 0 2/2 0 65 1998, >1+ 80 16 1 2 0/1 0/6 3 51 Total 598 316 9 2 2/1 9/16 4 239 South Pass 1997, fry 22 21 0 0 0 0 0 1 Creek 1998, fry 15 15 0 0 0 0 0 0 1998, 1+ 2 1 0 0 0 0/1 0 0 1998, >1+ 25 10 0 0 0 0/1 5 9 Total 64 47 0 0 0 0/2 5 10 Tributary 4 1997, fry 26 20 1 0 0 0 4 1 1998, fry 24 19 1 0 1/0 0 1 2 1998, 1+ 20 12 1 0 0 0/1 1 5 1998, >1+ 17 10 2 0 0 0 3 2 Total 87 61 5 0 1/0 0/1 9 10 Attichika 1997, fry 40 31 2 2 0 0 0 5 River 1998, fry 33 29 1 l 7 0 0 0 2 1998, 1+ 3 2 1 0 0 0 0 0 1998, >1+ 12 12 0 0 0 0 0 0 Total 88 74 4 3 0 0 0 7 Niven 1997, fry 43 40 1 0 1/0 0 0 1 River 1998, fry 40 29 6 0 0 1/0 2 2 1998, 1+ 10 6 3 0 0 0 0 1 1998, >1+ 1 0 0 0 0 0 0 1 Total 94 75 10 0 1/0 1/0 2 5 Attycelley 1997, fry 11 0 0 0 0 0 0 11 River 1998, fry 11 0 0 0 0 1/0 0 10 1998, 1+ 22 0 0 0 0 0/1 0 21 1998, >1+ 15 0 0 0 0 0 0 15 Total 59 0 0 0 0 1/1 0 57 Combined 1997, fry 327 239 8 2 2/0 0/4 4 68 Watershed 1998, fry 334 213 11 1 2/0 9/4 4 90 1998,1+ 179 73 6 0 0 2/5 1 92 1998, >1+ 150 48 3 2 0/1 0/7 11 78 Total Total 990 573 28 5 4/1 11/20 20 328 1 # / # indicates: (individuals with BT mtDNA) / (individuals with D V mtDNA) 2 'Pure' S. malma and 5. confluentus are designated as D V and BT, respectively 3 Heterozygous at all nuclear loci 4 Homozygous for alternate alleles at two or more nuclear loci 5 Backcrosses (at least one heterozygous nuclear locus in a 'pure' nuclear background) 6 Introgressed (bull trout mtDNA in a 'pure' Dolly Varden nuclear background) 7 This individual is heteroplasmic (contains both BT and D V mtDNA) 138 Appendix 5: B u l l trout allele frequencies by age class and by tributary, n = number of individuals. Nuclear loci (GH2, ITS-1, M T B , & F O K ) have 2n alleles, while haploid m t D N A have n alleles. As only two alleles are known for each locus in the charr of Thutade watershed, Dol ly Varden allele frequencies are 1 - (bull trout freq.). Tributary Year, n G H 2 ITS-1 M T B F O K m t D N A Age Class Kemess 1997, fry 185 0.719 0.703 0.700 0.703 0.714 Creek 1998, fry 211 0.607 0.590 0.597 0.600 0.630 1998, 1+ 122 0.443 0.434 0.438 0.434 0.549 1998, >1+ 80 0.263 0.238 0.225 0.225 0.275 Total 598 0.562 0.546 0.547 0.548 0.572 South Pass 1997, fry 22 0.954 0.954 0.954 0.954 0.954 Creek 1998, fry 15 1.000 1.000 1.000 1.000 1.000 1998, 1+ 2 0.750 0.500 0.500 0.500 0.500 1998, >1+ 25 0.420 0.400 0.420 0.420 0.600 Total 64 0.750 0.734 0.742 0.742 0.812 Tributary 4 1997, fry 26 0.808 0.808 0.788 0.788 0.962 1998, fry 24 0.875 0.833 0.852 0.833 0.917 1998, 1+ 20 0.650 0.650 0.650 0.650 0.700 1998, >1+ 17 0.676 0.676 0.706 0.676 0.882 Total 87 0.764 0.753 0.759 0.747 0.874 Attichika 1997, fry 40 0.850 0.850 0.838 0.825 0.875 River 1998, fry 33 0.924 0.924 0.924 0.909 0.924* 1998, 1+ 3 1.000 1.000 0.833 0.833 1.000 1998, >1+ 12 1.000 1.000 1.000 1.000 1.000 Total 88 0.903 0.903 0.892 0.881 0.920 Niven 1997, fry 43 0.965 0.977 0.954 0.965 0.977 River 1998, fry 40 0.812 0.850 0.875 0.838 0.950 1998, 1+ 10 0.800 0.900 0.800 0.850 0.900 1998, >1+ 1 0 0 0 0 0 Total 94 0.872 0.904 0.894 0.888 0.947 Attycelley 1997, fry 11 0 0 0 0 0 River 1998, fry 11 0 0 0 0.046 0.091 1998, 1+ 22 0.023 0 0 0 0 1998, >1+ 15 0 0 0 0 0 Total 59 0.008 0 0 0.008 0.017 Combined 1997, fry 327 0.766 0.758 0.751 0.752 0.780 Watershed 1998, fry 334 0.680 0.671 0.680 0.675 0.716 1998,1+ 179 0.447 0.441 0.436 0.436 0.469 1998, >1+ 150 0.367 0.350 0.350 0.347 0.427 Total 990 0.619 0.610 0.609 0.608 0.659 * 30 had B T m t D N A , 2 had D V m t D N A , and 1 individual was heteroplasmic 139 Appendix 6: Hybrid genotypes. The five loci are G H 2 , ITS-1, M T B , F O K , m t D N A , and the genotypes list the loci in the above order. For the four nuclear alleles, b indicates homozygous bull trout alleles, v indicates homozygous Dol ly Varden alleles, and H indicates heterozygous alleles. For the mitochondrial locus, and b and v indicate bull trout and Dol ly Varden mitochondrial D N A , respectively, while H indicates heteroplasmy. For example, vvvvb is has a 'pure' Dol ly Varden nuclear genotype with bull trout m t D N A . Locality Year, Age Class N Hybrid Genotypes Kemess 1997, fry 9 bHbHb' , bbHHb' x 2, Hvvvv x 4, b H H H b ' , vvvbb* Creek 1998, fry 16 Hbbbb', bHbbb', bbHbb', Hvvvv x 2, v v H v b \ v v H H v , bvHHb*, v v v H v \ H H H v b \ Hvvvb v x 2, H H v v b \ H v H H b \ v v v H b \ vvvvb 1998, 1+ 5 H H H b b ' , Hvvvv v x 2, v H H v b \ H v H v b v 1998, >1+ 13 H H H H b x 2, bbHHb' , bvvvv*, H v v v v v x 4, vvvvb x 3, v H H H v \ vHvvv v Total 43 South Pass 1997, fry 0 Creek 1998, fry 0 1998, 1+ 1 Hvvvv v 1998, >1+ 6 H v H H v \ vvvvb x 5 Total 7 Tributary 4 1997, fry 1998, fry 5 3 bbHHb' , vvvvb x 4 bHbbb', bHHvb*, vvvvb 1998, 1+ 3 bHbbb', v H v v v \ vvvvb 1998, >1+ 5 bHbbb', HbbHb ' , vvvvb x 3 Total 16 Attichika 1997, fry 4 bbbHb', bbHHb' , H H H H b x 2 River 1998, fry 2 bbbHb', H H H H H 1998, 1+ 1 B b H H b ' 1998, >1+ 0 Total 7 Niven 1997, fry 2 Hbbbb', bbvHb* River 1998, fry 9 HbbHb ' , bbHHb' , Hbbbb', HHbbb ' x 2, vvvvb x 2 v H H v b \ H H b H b ' 1998, 1+ 3 Hbbbb', HbHbb ' , bbHHb ' 1998, >1+ 0 Total 14 Attycelley River 1997, fry 1998, fry 0 1 vvvHb" 1998, 1+ 1 Hvvvv v 1998, >1+ 0 Total 2 Genotype classes: HHHH_ = Fu * = F„ , ' = BT backcross," = D V backcross, vvwb = introgressed bull trout mitochondrial D N A in Dolly Varden 140 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089603/manifest

Comment

Related Items