UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Biodiversity of Arctic char (Salvelinus alpinus) : sympatric morphs and hybridization with Dolly Varden… May-McNally, Shannan Leigh 2014

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

Item Metadata

Download

Media
24-ubc_2014_november_maymcnally_shannanleigh.pdf [ 2.73MB ]
Metadata
JSON: 24-1.0135591.json
JSON-LD: 24-1.0135591-ld.json
RDF/XML (Pretty): 24-1.0135591-rdf.xml
RDF/JSON: 24-1.0135591-rdf.json
Turtle: 24-1.0135591-turtle.txt
N-Triples: 24-1.0135591-rdf-ntriples.txt
Original Record: 24-1.0135591-source.json
Full Text
24-1.0135591-fulltext.txt
Citation
24-1.0135591.ris

Full Text

BIODIVERSITY OF ARCTIC CHAR (SALVELINUS ALPINUS):  SYMPATRIC MORPHS AND HYBRIDIZATION WITH DOLLY VARDEN (S. MALMA) IN SOUTHWESTERN ALASKA by  SHANNAN LEIGH MAY-MCNALLY B.Sc., University of Ottawa, 2010   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Zoology)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2014  © Shannan Leigh May-McNally, 2014 ii  Abstract  Resource polymorphism and natural hybridization are evolutionary phenomena that play an important role in the development of reproductive isolation during speciation. The Arctic char (Salmonidae: Salvelinus alpinus) exhibits substantial phenotypic and genetic diversity across its Holarctic range, making it an ideal species to examine the role of intraspecific trophic polymorphism in driving reproductive isolation and the interactions between hybridizing species in sympatry. In one southwestern post-glacial lake previously analyzed for resource polymorphism (Lower Tazimina Lake), I found evidence for two genetic groups of char and for significant differences in the distribution of microsatellite variability among at least two of the three previously described body-size morphotypes (‘large’, ‘medium’, and ‘small’–bodied char; maximum FST = 0.09). I also found significant associations between genetic type and gill raker counts among body-size morphs (r = -0.73, P < 0.001). These data represent the first record of genetically distinct sympatric Arctic char morphs in Alaska and provides further evidence that morphological differences associated with feeding and growth trajectories reflect niche diversification and may promote genetic divergence. Furthermore, I evaluated the level of natural hybridization and ecological segregation between two sympatric sister species of char, Arctic char and Dolly Varden (Salvelinus malma), from two southwestern Alaskan lakes. A total of N = 725 char were collected from 44 microhabitats across two lakes (Lake Aleknagik and Lake Nerka) and genotyped at 13 microsatellite loci. Using genetic admixture (Q)-values generated through Bayesian-based clustering, hybridization levels between Arctic char and Dolly Varden were found to be minimal; less than 0.5% of samples were classified as hybrids. Concurrent analyses with NEWHYBRIDS, however, supported the presence of late-generation hybrids in Lake Aleknagik with up to 7% of all samples representing F2 hybrids or backcrossed Dolly Varden. The existence of two discrete gene pools in sympatry (FST= 0.172–0.187 across the two lakes) which were generally spatially separate and low levels of contemporary hybridization between Arctic char and Dolly Varden helps to solidify their taxonomic status as iii  distinct biological species. Moreover, high levels of reproductive isolation between these sympatric species may be driven largely by strong pre-mating spatial segregation.  iv  Preface  The sample collection carried out in Chapter 3 was covered by Alaska Department of Fish and Game (ADF&G) permits SF2012-114 and SF2013-226.  Chapter 2. A version of this chapter has been published as [May-McNally, S.L., Quinn, T.P., Woods, P.J. and Taylor E. B. 2014/ Evidence for genetic distinction among sympatric ecotypes of Arctic char (Salvelinus alpinus) in south-western Alaskan lakes. Ecology of Freshwater Fish, available online, doi: 10.1111/eff.12169]. I was responsible for concept formation, genetic analyses, data synthesis, and manuscript composition. Tissue samples used in this study were collected by P.J. Woods (School of Aquatic and Fishery Sciences, University of Washington, Seattle, WA, USA). Sample collection in Lake Clark National Park was facilitated by D. Young (U.S. National Park Service). J.S. Moore and E.B. Taylor developed the genetic methods that I employed in this study. T.P. Quinn and P.J. Woods were involved in concept formation and refinement of the manuscript. E. B. Taylor was the supervisory author on this project and was involved throughout the project in statistical analysis and manuscript preparation.           v  Table of contents  Abstract ....................................................................................................................... ii Preface ...................................................................................................................... iv Table of contents ....................................................................................................... v List of tables ............................................................................................................. viii List of figures ............................................................................................................. ix Acknowledgements .................................................................................................... xi Dedication ................................................................................................................ xiii Chapter 1: General introduction ................................................................................. 1 1.1 Arctic char .................................................................................................... 3 1.2 Sympatric Arctic char morphs ...................................................................... 5 1.3 Natural hybridization in fishes ...................................................................... 7 1.4 Hybridization between Arctic char and Dolly Varden ................................. 12 1.5 Research objectives .................................................................................. 17 Chapter 2: Evidence for genetic distinction among sympatric ecotypes of Arctic char (Salvelinus alpinus) in southwestern Alaskan lakes ................................................. 18 2.1 Introduction ................................................................................................ 18 2.2 Materials and methods .............................................................................. 22 2.2.1 Fish collection ..................................................................................... 22 2.2.2 DNA extractions and microsatellite DNA analyses ............................. 24 2.2.3 Population structure within and among lakes ..................................... 25 2.2.4 Comparisons of genetic and phenotypic data ..................................... 26 2.3 Results ....................................................................................................... 28 2.3.1 Microsatellite variation among and within lakes .................................. 28 2.3.2 Population structure among lakes ...................................................... 29 2.3.3 Comparisons between genetic and phenotypic data within Lower Tazimina Lake .................................................................................................. 32 vi  2.4 Discussion ................................................................................................. 36 2.4.1 Microsatellite DNA variation across remote freshwater lakes ............. 36 2.4.2 Genetic differentiation among Lower Tazimina Lake morphs ............. 37 2.4.3 Biogeography of sympatric morphs .................................................... 39 Chapter 3: Low levels of hybridization between sympatric Arctic char (Salvelinus alpinus) and Dolly Varden char (Salvelinus malma) highlights their genetic and spatial distinctiveness .............................................................................................. 44 3.1 Introduction ................................................................................................ 44 3.2 Materials and methods .............................................................................. 48 3.2.1 Sample collection ............................................................................... 48 3.2.2 Microsatellite analysis ......................................................................... 51 3.2.3 Statistical analysis .............................................................................. 52 3.2.4 Population structure ............................................................................ 52 3.2.5 Levels of hybridization ........................................................................ 53 3.2.6 Spatial distribution across microhabitats ............................................ 54 3.3 Results ....................................................................................................... 55 3.3.1 Genetic analysis ................................................................................. 55 3.3.2 Population structure ............................................................................ 56 3.3.3 Hybridization levels and genetic classification .................................... 57 3.3.4 Spatial discreteness between species ................................................ 65 3.4 Discussion ................................................................................................. 72 3.4.1 Evidence for distinct gene pools between species ............................. 73 3.4.2 Hybridization levels............................................................................. 74 3.4.3 Spatial segregation across lake-stream microhabitats ....................... 78 Chapter 4: General conclusions ............................................................................... 82 4.1 Summary of findings .................................................................................. 82 4.2 Defining a species ..................................................................................... 83 4.3 Implications for conservation of char diversity ........................................... 85 4.4 Concluding thoughts and future directions................................................. 88 4.4.1 Genetic basis of divergence ............................................................... 89 vii  4.4.2 Basis of reproductive isolation: ecology and life history...................... 90 References .............................................................................................................. 92 Appendix A   Chapter 2 .......................................................................................... 111 Appendix B   Chapter 3 .......................................................................................... 116  viii  List of tables  Table 1.1:  Estimate of the percentage of species that hybridize with at least one other species............................................................................................................ 10 Table 2.1:  Pairwise FST (θ) estimated by variation across 11 microsatellite DNA loci in Arctic char from five post-glacial lakes in southwestern Alaska. .................... 30 Table 2.2:   Numbers of Arctic char in different admixture (Q)-value and growth curve categories from Lower Tazimina Lake, southwestern Alaska......................... 33 Table 3.1:  Pairwise FST (θ) estimated by variation across 13 microsatellite DNA loci in sympatric Arctic char and Dolly Varden from the Wood River Lake system, southwestern Alaska ................................................................................................ 59 Table 3.2:  Sample site information, sample sizes, and summary of genotyping results for the Wood River Lake system. ................................................................. 62 Table A.1: Genetic diversity values for Lower Tazimina Lake morphs and four populations of Arctic char  from southwestern Alaska. ........................................... 111 Table B.1:    Summary of Lake Aleknagik and Lake Nerka site names, assigned site ID numbers (2012) with corresponding sample size and the given proportions of samples genotyped as Arctic char, Dolly Varden or hybrid.. .................................. 116 Table B.2:     Summary of Lake Aleknagik and Lake Nerka site names and assigned site ID numbers (2013) with corresponding sample size and the given proportions of samples genotyped as Arctic char, Dolly Varden or hybrid. ................................... 117 Table B.3:     Primer information and original refernces for the 13 microsatellite markers used in this study. .................................................................................... 118 Table B.4:  Genetic diversity values for Arctic char and Dolly Varden from Lake Aleknagik and Lake Nerka for 2012-2013 pooled. ................................................. 120 ix  List of figures  Figure 1.1: Lower Tazimina Lake Arctic char morphs. ............................................ 6 Figure 1.2: Distribution of Arctic char and Dolly Varden char. .............................. 13 Figure 1.3: Morphological features and colouration of Alaskan Arctic char and Dolly Varden. ........................................................................................................... 15 Figure 2.1: Sample collection sites of Arctic char in southwestern Alaska. .......... 23 Figure 2.2: Bayesian clustering analysis of five populations of Arctic char from southwestern Alaska. ............................................................................................... 31 Figure 2.3: Pearson’s correlation between genetic admixture coefficients (Q) and total gill raker counts for Lower Tazimina Lake morphs of Arctic char. .................... 35 Figure 3.1:    The Wood River Lake system in southwestern Alaska, USA. ............ 49 Figure 3.2: Bayesian clustering  of sympatric Alaskan Arctic char and Dolly Varden  sampled from Lake Aleknagik and Lake Nerka. ....................................................... 58 Figure 3.3: Overall genetic classification of sympatric Arctic char and Dolly Varden using 13 microsatellite markers from Lake Aleknagik and Lake Nerka  ................... 60 Figure 3.4: Summary of NEWHYBRIDS results for sympatric Arctic char and Dolly Varden from Lake Aleknagik. ................................................................................... 64 Figure 3.5:   Proportion of genotypes defined by admixture (Q) values corresponding to Arctic char and Dolly Varden from lake and beach sites in Lake Aleknagik (2012). ..................................................................................................... 66 Figure 3.6:   Proportion of genotypes defined by admixture (Q) values corresponding to Arctic char and Dolly Varden from stream sites around the southern portion of Lake Aleknagik (2012). ............................................................. 67 Figure 3.7:     Proportion of genotypes defined by admixture (Q) values corresponding to Arctic char and Dolly Varden from stream sites around the southern portion of Lake Aleknagik (2013). ............................................................. 68 Figure 3.8: Proportion of genotypes defined by admixture (Q) values corresponding to Arctic char and Dolly Varden from stream and river sites around Lake Nerka (2012). .................................................................................................. 69 x  Figure 3.9:  Proportion of genotypes defined by admixture (Q) values corresponding to Arctic char and Dolly Varden  from a single stream site in Lake Nerka (2013). ........................................................................................................... 70 Figure 3.10:   Size frequency histogram displaying approximate age classes of Arctic char and Dolly Varden for Lake Aleknagik. .................................................... 71 Figure A.1:     Factorial correspondence analysis (FCA) of five populations of Arctic char from southwestern Alaska. ............................................................................. 114 Figure A.2:     Factorial correspondence analysis (FCA) of Lower Tazimina Lake morphs and Iliamna Lake Arctic char ..................................................................... 115 Figure B.1:     Factorial correspondence analysis (FCA) of Arctic char and Dolly Varden from the Wood River Lake system.. .......................................................... 124 Figure B.2:     Factorial correspondence analysis (FCA) of Arctic char and Dolly Varden from the Wood River Lake system including allopatric populations ........... 125 Figure B.3:    STRUCTURE plot generated from allopatric Arctic char and Dolly Varden in the Wood River Lake system and associated simulated hybrids. .......... 127 Figure B.4:     STRUCTURE plot generated from non-admixed Wood River Lake system Arctic char and Dolly Varden with associated simulated hybrids. .............. 126   xi  Acknowledgements  I would first and foremost like to thank my supervisor, Dr. Eric B. Taylor, for giving me the opportunity to work on a project which has perfectly combined my interest in molecular biology with ecology, and for allowing me to develop my own research topics. I greatly appreciated his help when establishing collaborations with researchers from the US and Japan and for the freedom to maintain them on my own. I thank him for all the guidance and support provided and for giving me the chance to work on additional side projects that helped to improve my lab techniques and writing skills. I have enjoyed learning from him and I truly admire his passion for research and fish. I would like to thank my committee members, Darren Irwin and Sally Aitken for all their helpful suggestions and advice.  Sincere thanks to Dr. Thomas P. Quinn at the University of Washington for giving me an amazing opportunity to conduct field work in Alaska and for the mentorship he provided for both Chapter 2 and 3. Thank you to the FRI staff, notably Chris Boatright and Jackie L. Carter for all of their assistance. Jackie, thank you for caring so much about my project and for always managing to fit a “charlie” sampling day into a busy camp schedule. Special thanks also goes out to all the researchers at the Lake Nerka field camp, especially Kale Bentley for his attention to detail when collecting genetic samples for me during 2012 and Dr. Daniel Schindler for facilitating sample collection from Lake Nerka in 2013. Big thanks to Craig Schwanke and Jason Dye (ADF&G, Dillingham) for providing spawning Arctic char and Dolly Varden samples from Lake Aleknagik. To the students and technicians at FRI: Jason Ching, Casey Clark, Sydney Clark, Curry Cunningham, Lindsay Hart, Rachel Hovel, Wes Larson, Bianca Santos, Brenden Smith, and everyone else who collected char for this project, thank you so much for all your help, your enthusiasm and for the fun times in the field. I could not have done this work without your help.   xii  Thank you to Dr. Pamela J. Woods for collecting the samples used in Chapter 2 and for her help with the development of the manuscript. Thanks also to D. Young (U.S. National Park Service) who facilitated field work in Lake Clark National Park, and UW staff and students and National Park Service staff for field help.  Khang Hua, thank you for being there with me every step of the way and for all your thoughtful advice when it came to drafts, presentations and troubleshooting. Your love and support got me through some tough times, and I still adamantly believe that you should also make the switch to ecology.   I thank my fellow labmates and volunteers in the Taylor lab for all their support and encouragement: JS Moore, Jon Mee, Matt Siegle, Monica Yau, Carla Crossman, Jenn Ruskey, Amanda Moreira, Rebecca Piercey and Chloe Mighton. Thank you to JS Moore for providing the multiplexes used in my genetic assays and for the mentorship when I first began my research. To the best 502 cohort ever and my dear friends and scientists in Zoology and Biology, it was wonderful working with you. Thanks for all the help and support over the years. Thank you to all of the Zoology Department staff and professors and our graduate secretary Alice Liou. Special thanks are owed to my parents for supporting my career in the sciences and for initially sparking my interest in biology. Mom, thank you for all the nature walks you took me on as a kid: ID’ing the birds, feeding the ducks and collecting wildflowers. Dad, thanks for taking me camping, fossil hunting, fishing and for teaching me to be inquisitive about the natural world. I’d also like to thank my brother Cody who (I think) is proud of his big sister and has always been there for me.  Support for the genetic analysis was provided by a Natural Sciences and Engineering Research Council of Canada operating and equipment grants awarded to E.B.T., and support for field work was provided by the H. Mason Keeler Endowment to T.P.Q. and two Northern Scientific Training Program (NSTP) grants to S.M.M. Financial assistance was provided by an NSERC Alexander Graham Bell Graduate Scholarship CGS-M and Anne Vallée Ecological Fund grant to S.M.M. and Department of Zoology Teaching Assistantship.  xiii  Dedication  To my family and all fish enthusiasts.  1  Chapter 1:                                                                                                        General introduction  Understanding the origins and maintenance of reproductively isolated populations is a central focus of speciation biology. Reproductive isolation can arise indirectly when divergent natural selection acts upon morphological, behavioural or physiological traits associated with differential resource use in a population or through direct reinforcement of pre-mating isolation (Schluter 1996). For instance, divergent selection has been shown to occur through resource polymorphisms in sympatric populations when an ancestral population expands to exploit novel resources (niche invasion) or through subdivision of resources from a broad niche into two narrower niches (niche specialization) (Schluter 2000; Knudsen et al. 2006). In such cases, reproductive isolation may arise indirectly because traits that are the focus of divergent selection also influence mate choice (e.g., body size –Foote & Larkin, 1988; Conte & Schluter, 2013). Alternatively, reinforcement can occur through direct selection against hybrids causing selection for divergent mate-selection traits (Schluter 1996; Servedio & Noor 2003; Mallet 2005).  Hybridization is defined as the successful interbreeding of individuals from two populations, or groups of populations, that can be distinguished on the basis of one or more heritable characters (Harrison 1990; Arnold 1997; Dowling & Secor 1997). Hybridization was once considered to be both a source of evolutionary “noise” and a rare phenomenon (Arnold 1997). This view, however, is becoming increasingly challenged as modern molecular techniques facilitate the identification of hybrid individuals (Seehausen 2004; Aboim et al. 2010). Although botanists have long recognized the creative role of hybridization in evolution, the evolutionary role of hybridization as a process that can promote or limit gene flow, reinforce or break down reproductive barriers and/or generate novel genotypes on which natural selection can act has only been more widely appreciated by zoologists (Arnold 1992; Mallet 2005).  2  Post-Pleistocene glacial populations of fishes have provided key study systems in research designed to better understand the roles of resource polymorphism and hybridization in speciation (Schluter & Rambaut 1996). The rapid evolution of temperate fishes in post-glacial lacustrine systems younger than < 15 000 years old provide unique opportunities to examine the first steps in divergence (Schluter & Rambaut 1996). Secondary contact and introgression of species or populations isolated during the last glaciation in refugia can also be a means of understanding the evolutionary consequences of hybridization and determining the role of pre- or post-mating reproductive barriers in constraining gene flow in sympatry (e.g., Lu & Bernatchez, 1998). Unfortunately, like many Arctic environments, post-glacial lacustrine systems are fragile and are becoming increasingly at risk from human perturbations which include climate change, environmental pollution from resource extraction and over-fishing (Reist et al. 2013).  Fishes from the Salmonidae family (66 Holarctic species of salmon, trout, char, grayling and whitefishes) are frequent inhabitants of post-glacial waters (Hendry & Stearns 2003; Nelson 2006). Many species of salmonids are capable of forming spatially and temporally discrete breeding units that can exhibit distinct genetic and phenotypic differentiation (Behnke 1972; Schluter & Rambaut 1996; Hendry & Stearns 2003; Taylor et al. 2008). These discrete breeding units are often adapted to different habitats but are not isolated by geographic barriers (Taylor 2004). Salmonids may also have the capacity to quickly evolve reproductive isolation through adaptive divergence (reviewed in Schluter, 1996; Taylor, 1999). For instance, some populations of sockeye salmon (Oncorhynchus nerka) released into a novel lake system were capable of founding several new, genetically-distinct populations differing in niche use, male body depth, female body size, and embryo survival rate after only 13 generations (Hendry 2001). This result suggests that ecological speciation in salmonids often closely follows the colonization of new environments containing vacant niches.  Char (salmonid fishes of the genus Salvelinus) show an especially wide range of variation in ecology, habitat use, life history, and phenotype. Char often differentiate into morphologically distinct forms in sympatry and hybridization amongst the 3  various species is common. Consequently, Salvelinus has been an important group for understanding factors related to ecological speciation and the divergence of fishes (Behnke 1972; Klemetsen 2010). Char are common species in many Holarctic regions and are abundant in areas throughout temperate areas of Canada and the USA. Despite being widespread and phenotypically diverse in Alaska, it is surprising how little is known about the genetic diversity and evolutionary history of char from Alaskan post-glacial lakes.  1.1 Arctic char  The Arctic char (Salvelinus alpinus (L., 1758), Pisces: Salmonidae) is the most northerly distributed freshwater salmonid and is one of five recognized North American, and one of approximately 22 Eurasian, species of char (Johnson 1980; Maitland 1995; Reist et al. 2013). Generally considered to be a single taxon, Arctic char can display substantial variability in phenotype and ecology, leading to the assignment of species-level status of multiple forms or the grouping of diverse forms into a species ‘complex’ (Snorrason et al. 1994; Maitland 1995; Jonsson & Jonsson 2001; Klemetsen 2013). The native distribution of Arctic char extends across freshwater and near-shore marine habitats in North America, northern Britain, Scotland, Iceland, Europe and Russia (Johnson 1980). The species’ southern limit extends to latitudes as low as 43-45º N in North America and Europe and corresponds closely to the maximum extent of the last glaciation (Weichsel-Wisconsinan) in the Holarctic (Klemetsen et al. 2003; Power et al. 2009). Perhaps limited by its physiological preference for low temperature water, Arctic char are especially common in many northern ecosystems, especially in isolated post-glacial lakes with low interspecific competition (Johnson 1980; Skúlason & Smith 1995; Klemetsen et al. 2003). The capacity for Arctic char to specialize in a variety of habitats across its range such as lakes (including profundal, littoral or pelagic regions), rivers, streams, and oceans and exhibit different life histories (e.g. anadromy or non-anadromy) has been well established (Skúlason et al. 1996; Bernatchez et al. 1998; Jonsson & Jonsson 2001; 4  Klemetsen 2010; Woods et al. 2013). In fact, Klemetsen (2013) compared variation across a number of northern fishes and declared the Arctic char to be the “most variable vertebrate on Earth”. A combination of plasticity and adaptive variation in life history and various phenotypic traits are two key reasons why Arctic char are so popular for evolutionary studies of trophic polymorphism and speciation (Power et al. 2009).  Although thoroughly researched throughout Eurasia (Snorrason et al. 1994; Jonsson & Jonsson 2001; Adams & Huntingford 2002b) and, increasingly, in Canada (Reist et al. 1995; Gomez-Uchida et al. 2008; Power et al. 2009), Arctic char have been largely understudied in Alaska (USA). In Alaska, Arctic char are abundant along the Bering Sea and Arctic coast from the northern slope to the southwestern portion of the state where they are common in freshwater lakes and streams (McPhail 1961; McPhail & Lindsey 1970). McPhail (1961) described two forms of Arctic char in Alaska including a “western” form characterized by lower gill raker and pyloric caeca counts (13-14, 27-31, respectively) found in rivers in northern and western Alaska, and an “eastern” form with higher gill raker and pyloric caeca counts (15-17, 36-53, respectively) found east of the Mackenzie River, but also in southwestern Alaska near Bristol Bay. The current distribution of the western and eastern forms is postulated to have arisen due to isolation of various populations during the late Pleistocene (McPhail 1961). Arctic char in southern areas of Alaska tend to be lake-resident fish that complete their entire life-cycle in freshwater (i.e., non-anadromous). Colouration is often variable, but in these regions the body is typically silver in non-spawning periods (see Fig.1.3) or orange-red when sexually mature and is marked year-round with large violet, pink or white spots (McPhail & Lindsey 1970). The species generally spawns in September and October on gravel bottom beaches in lakes and is a repeat spawner (i.e., Arctic char can spawn each year over multiple years) (McPhail & Lindsey 1970; Johnson 1980). The eastern form of Arctic char is, by contrast, often sea-run (anadromous) and can also vary phenotypically. Throughout their Arctic and sub-Arctic range, Arctic char are valuable subsistence and sport fish for local Aboriginal peoples (Johnson 1980).   5  1.2 Sympatric Arctic char morphs  Arctic char can rapidly diverge into discrete resource-based forms following colonization of young, depauparuate post-glacial lakes (Griffiths 1994; Skúlason et al. 1999; Jonsson & Jonsson 2001; Klemetsen et al. 2003). Sympatric Arctic char morphotypes, or morphs, have been described from lacustrine systems across the species’ Holarctic range and have received significant attention (especially in European lakes) as excellent study organisms for investigating the factors related to the ecological and genetic basis of resource polymorphism. Across relatively short evolutionary timespans, Arctic char have the capacity and tendency to diverge into alternative forms which exhibit variation in traits including: colouration, body morphology, growth rates, parasite load, and size and age at maturity (e.g., Gíslason et al. 1999). This variation is thought to be associated with differential use of trophic resources and habitats and is often associated with reproductive isolation between the forms (Malmquist et al. 1992; Power et al. 2009). The most common pathway of divergence in Arctic char appears to occur along the benthic (typically benthivorous feeders) and limnetic (planktivorous or piscivorous feeders) resource axis which highlights the potential role of habitat diversity and resource allocation during adaptive radiation and speciation in fishes (Gíslason et al. 1999). The magnitude of phenotypic divergence between these forms has been well studied; however, the connection between degree of genetic isolation and phenotypic or ecological divergence is often unknown.  Although many deep post-glacial Alaskan lakes contain landlocked Arctic char, it was only more recently that sympatric Arctic char morphs were discovered from one Alaskan system, Lower Tazimina Lake in southwestern Alaska, which was found to contain at least two morphs (Woods et al. 2013). The morphs, classified as ‘large’ and ‘small’ differed significantly in terms of colouration, age at maturity and body shape and size (Fig. 1.1).    6                      Figure 1.1: Lower Tazimina Lake sympatric Arctic char (Salvelinus alpinus) morphs.       A) Large growth form B) Small growth form; note the differences in colouration, morphology and size. Photo credit: Pamela J. Woods      A) B) 7  To differentiate the morphs and to detect signatures of resource polymorphism, Woods et al. (2013) tested for the presence of multiple Von Bertalanffy growth curves among samples using a Von Bertalanffy three-parameter non-linear growth model. Von Bertalanffy growth curves model individual growth rate and are mathematically derived from measures of asymptotic body length and curvature as a function of age (Woods 2011; Woods et al. 2012, 2013). The authors found evidence for multiple Von Bertalanffy curves supporting the presence of two, or possibly three, Arctic char growth forms in Lower Tazimina Lake with differing growth trajectories. Furthermore, the large small and growth forms exhibited differences in gill raker counts, gonado-somatic indices, stable isotope levels of δC13 and δN15 and food consumption in benthic and limnetic regions of the lake (Woods et al. 2013). Collectively, this work provided evidence for a resource polymorphism in Lower Lake Tazimina due to divergence along the epipelagic-littoral resource axis, yet there has been no study of the genetic relationships among forms. To further explore this phenomenon, in Chapter 2 of my thesis, I addressed the question of whether pronounced ecological divergence and niche specialization is associated with genetic divergence and perhaps reproductive isolation among these sympatric morphs in Lower Tazimina Lake.  1.3 Natural hybridization in fishes Arnold (1992, 1997) defined natural hybrids as individuals derived from matings between individuals from populations in nature that are distinguishable from each other by at least one heritable character. Natural hybrids can potentially backcross with their respective parental populations, leading to introgression (Arnold 1992; Dowling & Secor 1997; Arnold & Martin 2009). Genetic exchange between parental populations may increase phenotypic variation, potentially altering the ecology and behaviour of hybridizing populations (Arnold 1997).  The evolutionary importance of natural hybridization and introgression has been long been established in plants, and has only more recently been accepted as a fundamental and pervasive process in animals (Rieseberg et al. 1990; Jiggins & 8  Mallet 2000; Roques et al. 2001). The primary reason zoologists have had difficulty accepting that natural hybridization was not an evolutionary dead end has mostly been due to a sometimes overly zealous adherence to the Biological Species Concept (Mayr 1942). Although most biologists at the time believed in the existence of species as distinct biological units, evolutionists could not (and many still can not) come to a consensus on what constituted a species (Coyne & Orr 2004). Many species concepts were created to answer this issue, but none were as influential or controversial as Ernst Mayr’s Biological Species Concept (BSC) (Mayr 1942; Dowling & Secor 1997; Coyne & Orr 2004). The original BSC states that: “Species are groups of actually or potentially interbreeding natural populations, which are reproductively isolated from each other” (Dobzhansky 1937; Mayr 1942). Although the BSC is perhaps the best explanation we have of what constitutes a species, there are some inherent issues with the original definition, especially when applied to fishes (see Turner, 1999). The original BSC defines species on the basis of reproductive isolation caused by the development of barriers to hybridization (Arnold 1997). In the strictest sense, populations that hybridize are in direct violation of the BSC because they exhibit incomplete reproductive isolation. Despite increasing evidence that hybridization was a prevalent and a potentially important phenomenon, Mayr countered that hybrids were rare “mistakes” and evolutionarily unimportant, at least in animal taxa, because “the majority of such hybrids are totally sterile” (Mayr 1963). Furthermore, he stated: “Even those hybrids that produce normal gametes in one or both sexes are nevertheless unsuccessful in most cases and do not participate in reproduction. Finally, when they do happen to backcross to the parental species, they normally produce genotypes of inferior viability that are eliminated by natural selection” (Mayr 1963). Contrary to Mayr’s argument, it has been shown through modern biochemical and molecular techniques that natural hybridization occurs frequently and is evolutionarily important to many animals (Grant & Grant 1992, 2010; Arnold 1997; Jiggins & Mallet 2000; Mallet 2005). Further, approximately 25% of plant species and 10% of animal species are now known to hybridize (Table 1.1, Mallet 2005). Finally, hybrids are not always sterile 9  and may show equivalent or higher fitness compared to parental genotypes (Barton & Hewitt 1985; Arnold 1992).  Due to the various limitations of the BSC, many evolutionists now deviate from the strict definition and use a modified version (Arnold 1992, 1997). The modified version, which I will also adhere to, states that: “distinct species are characterized by substantial but not necessarily complete reproductive isolation” (Coyne & Orr 2004). Gene flow between populations is permitted with the modified BSC, yet the process of speciation (i.e., the formation of reproductive barriers) is still present (Coyne and Orr 2004). This definition also allows species that occasionally hybridize, but remain genetically dissimilar in sympatry, to be considered distinct biological species (Baxter et al. 1997).  Many taxonomically distinct species of fishes are known to hybridize (Hubbs 1955; Schwartz 1972, 1981; Thorgaard & Allendorf 1988; Warren & Burr 1993; Scribner et al. 2001; Taylor 2004). In fact, fishes have some of the highest rates of hybridization across all vertebrates, particularly among species that live in or near regions that were recently deglaciated (Campton 1987; Allendorf & Waples 1996; Scribner et al. 2001). Moreover, the capacity for fishes to produce viable and fertile hybrids has been firmly established (Hubbs 1955; Campton 1987; Verspoor & Hammar 1991). The high rate of hybridization recorded in fishes has been attributed to various causes such as: external fertilization, the often unequal abundance of parental species, competition for limited spawning sites, weak behavioural isolation during mating, decreased habitat complexity due to natural or anthropogenic impacts, frequent secondary contact between recently evolved species, and lower susceptibility to severe developmental incompatibilities following interspecific crossing (Campton 1987; Scribner et al. 2001). The many cases of natural hybridization have been primarily observed in freshwater species, but has now been illustrated in many marine fishes as well (Pyle et al. 1994; Roques et al. 2001; Nielsen et al. 2003; Riginos & Cunningham 2007). As introgressive hybridization can occur readily in both freshwater and marine species, fishes make an excellent model to investigate the role of natural hybridization in speciation.   10    Table 1.1:  Estimate of the percentage of species that hybridize with at least one other distinct species. Adapted from Mallet (2005).  a Estimate of the number of North American freshwater species. From Verspoor and Hammar (1991) b Sympatric pairs of subgenera    Taxa Number of   species sampled Number of species known to hybridize with another species % of species hybridizing with at least one other species Study UK vascular plants 539 135 25% Mallet 2005 European Butterflies (Lepidoptera: Rhopalocera) 379 47 12.4% Mallet 2005 Passion flower butterflies (Lepidoptera: Heliconiina) 73 19 26.0% (34.8% in the genus Heliconius) Mallet 2005 Swallowtail butterflies (Lepitoptera: Papilionidae; Papilio) 216 14-32 6.5-14.8% Mallet 2005 Drosphila (Diptera: Drosophilidae) ~ 1750 18 1.0% Mallet 2005 Birds (Aves) 9672 895 9.2% Grant and Grant 1992, Mallet 2005 North American Freshwater Fishes (Pisces) ~1061a >168 > 6.31% Scribner et al. 2001, Warren and Burr 1993 Arroyo chub (Pices: Cyprinidae; Gila orcutti) ~5562 442 7.9-9% Hubbs 1955 Freshwater Shiners (Campostoma: Notropis, Notropis cornuta) 608b 68 11% Hubbs 1955 European Mammals 200 12 6% Mallet 2005 11  In fishes and in other animals and in plants, hybridization with introgression leads to a wide array of evolutionary trajectories such as the formation of various kinds of hybrid zones, adaptive radiation, and speciation (Arnold 1997; Dowling & Secor 1997; Seehausen 2004; Aboim et al. 2010; Culumber et al. 2011). The evolutionary outcomes of natural hybridization are highly variable and can depend on other factors such as environmental features, the strength of selection, pre-zygotic (pre-mating), and intrinsic and extrinsic post-zygotic (post-mating) barriers (Mayr 1963; Barton & Hewitt 1985; Arnold 1992; Barton 2001; Latch et al. 2011). Pre-zygotic barriers can prevent or limit initial hybridization through gamete incompatibility and/or differences in mating behaviours leading to assortative mating and differences in the location and timing of reproduction (Howard et al. 1998; Rieseberg et al. 1998; Redenbach & Taylor 2003). Post-zygotic barriers can involve intrinsic genomic incompatibilities (endogenous selection) and/or extrinsic environment-dependent selection (exogenous selection) (Arnold 1997). Understanding the ecological and genetic basis of pre- and post-zygotic isolation may provide insight into the evolutionary forces which have led to reproductive isolation (e.g., Rogers & Bernatchez 2006).  One possible outcome of natural hybridization is the formation of hybrid zones (Barton & Hewitt 1985; Rieseberg et al. 1990). Hybrid zones are defined as geographic localities where individuals from genetically distinct populations interbreed and produce offspring with genetically mixed ancestry (Hewitt 1988; Arnold 1997; Baxter et al. 1997; Anderson & Thompson 2002; Nielsen et al. 2003). Hybrid zones can vary both temporally and spatially and can range anywhere from a few hundred meters to several kilometers in width, can have a mosaic-like structure, and may persist throughout evolutionary time or rapidly collapse (Barton & Hewitt 1985; Baxter et al. 1997; Rubidge & Taylor 2004).   From an evolutionary standpoint, hybrid zones are viewed as excellent opportunities to investigate the processes that drive speciation and adaptive divergence in natural populations (Taylor 2004; Gow et al. 2006; Albert et al. 2006). Hybrid zones can also be used to examine barriers to reproductive isolation, gene flow across species, the role of selection in speciation, and the formation of new 12  species with hybrid origin (Barton & Hewitt 1985; Hermansen et al. 2011; Brelsford 2011; Jacobsen & Omland 2011). The stability and fate of a hybrid zone is ultimately linked to the relative fitness of the novel hybrid phenotypes (Barton & Hewitt 1985; Dowling & Secor 1997; Burke & Arnold 2001; Barton 2001; Culumber et al. 2011).   Documenting the structure of hybrid zones can yield important insights about the pre-zygotic and post-zygotic processes that help generate and maintain hybrid zones in natural populations in a variety of taxa (Irwin et al. 2009; Latch et al. 2011; Hamilton et al. 2013). For instance, extrinsic post-zygotic isolation, in which hybrid fitness is determined by an interaction between hybrid genotypes and the environment, may be more common than previously hypothesized (Schluter 2009; Culumber et al. 2011). Thus, a fair amount of research on natural hybridization focuses on extrinsic (exogenous) post-zygotic isolation as it is central to ecologically based divergent selection and is an area where empirical tests are lacking (Moore et al. 2007; Gow et al. 2007; Culumber et al. 2011).   1.4 Hybridization between Arctic char and Dolly Varden  Dolly Varden char (Salvelinus malma), another member of the genus Salvelinus, share many phenotypic and genetic similarities with Arctic char. The distribution of Dolly Varden encompasses nearshore coastal marine and freshwater habitats north of the Olympic Peninsula in Washington State (USA), western Canada and Alaska, and extends across the Pacific into northern Japan and Korea and east in North America to the Mackenzie River (McPhail 1961; Reist et al. 2002; Dunham et al. 2008; Oleinik et al. 2011). Similar to Arctic char, Dolly Varden exist as a variety of forms including lacustrine, riverine, and anadromous char (Armstrong & Morrow 1980). The range of Dolly Varden overlaps with Arctic char in Kamchatka and Alaska, and due to their morphological similarities to one another, it is here that most of the confusion between the two species exists (Armstrong & Morrow, 1980, Fig. 1.2). In Alaska, Dolly Varden are ubiquitously distributed and sympatric with Arctic char in northern and southern regions near the Pacific coast (Fig. 1.2).   13     Figure 1.2: Approximate distribution of A) Holarctic Arctic char (Salvelinus alpinus), shown in purple and Dolly Varden char (Salvelinus malma), found around the North Pacific rim, shown in pink and B) Dolly Varden in Alaska (USA) including sympatric regions with Arctic char. Re-drawn and modified from Dunham et al. (2008), a COSEWIC (Committee on the Status of Endangered Wildlife in Canada) assessment report (2010) and Reist (2002, 2013).  A) B) 14  In these areas, Dolly Varden can inhabit fully marine habitats, brackish estuaries or rivers, and streams and lakes. Large anadromous or sea-run Dolly Varden return to lakes and rivers annually to spawn and overwinter while small-bodied (“dwarf”) stream-resident forms can remain in a single stream for their entire life-cycle (Armstrong & Morrow 1980).  Arctic char and Dolly Varden are very similar morphologically especially as juveniles, but they can usually be distinguished from one another on the basis of gill raker counts and the number of pyloric caeca (McPhail & Lindsey 1970). Gill rakers (bony protuberances on the gill arches that aid in feeding) and pyloric caeca (finger-shaped appendages that assist in digestion) are, however, highly variable characters that can be under selection depending on habitat and diet differences between populations or individuals within species (Jonsson et al. 1988; Skúlason et al. 1989; Helen 2001). In southwestern Alaska, Arctic char are typically characterized by 12-17 gill rakers on the lower limb of the first arch and 24-74 pyloric caeca, while Dolly Varden distributed north of the Alaska Peninsula generally have 11-14 gill rakers on the first arch and 19-47 pyloric caeca on average (McPhail 1961). Although these counts are normally discrete in allopatry, they can overlap when the species are found in sympatry. Aside from gill rakers and pyloric caeca, the species may display slight differences in morphology and colouration. Arctic char in southwest Alaska have silver bodies, pale pink or white spots, delicate round heads and narrow caudal peduncles with a definite fork in the tail (McPhail 1961; Johnson 1980). Conversely, Dolly Varden char have deeper and darker olive-green coloured bodies, “blocky” parr marks in younger life stages (resembling the pattern of military camouflage) or, in stream resident forms, blunter heads, thicker caudal peduncles and paddle-shaped tails (Armstrong & Morrow 1980). The presence of orange, red or yellow spotting has also been observed in mature small-bodied, stream-resident forms (McPhail 1961; Koizumi & Maekawa 2004). Sea-run Dolly Varden are silvery along their dorsal surface with pale pink or white spots that change to bright red or orange during spawning periods (Fig 1.3).   15               Figure 1.3: A) Arctic char (Salvelinus alpinus) from Happy Creek, Lake Aleknagik, AK; note the silver body with white spots, narrow caudal peduncle and defined fork in the tail;   B) Stream-resident Dolly Varden from Yako Creek, Lake Aleknagik; note the small colorful orange spots along a darker green body, thick caudal peduncle, paddle-shaped tail and small size at maturity. Mature stream-resident Dolly Varden retain parr marks which generally disappear in anadromous Dolly Varden upon their first migration seaward;               C) Anadromous Dolly Varden from the Egegik fishing district, AK; silver bodied with light spots, thick caudal peduncle and paddle-shaped tail.   5 cm 5 cm A) B)  5 cm C)  5 cm 16  Historically, the systematic relationship between Arctic char and Dolly Varden was the subject of much debate (McPhail 1961). Prior to the development of modern genetic techniques, Arctic char and Dolly Varden were both characterized by highly plastic and often overlapping morphological features (e.g. gill rakers and pyloric caeca) and Dolly Varden were considered a sub-species of Arctic char (see Discussion in McPhail 1961). In regions of sympatry, these species are very similar in terms of mitochondrial DNA, and even more recent phylogeographic work by Brunner et al. (2001) could not taxonomically resolve the two species.  Taylor et al. (2008) conducted a microsatellite DNA survey of sympatric samples from southwestern Alaska and provided evidence that Arctic char and Dolly Varden are highly distinct in sympatry and should be designated as valid biological species. Interestingly, Taylor et al. (2008) also provided evidence for natural hybridization between Dolly Varden and Arctic char which hinted at the lack of complete reproductive boundaries between the species and suggested that they have not yet fully diverged.  Hybridization between Arctic char and Dolly Varden is probable given the many accounts of natural hybridization between other, more divergent species of char (Helen 2001). For example, there is evidence of historical and contemporary natural hybridization between Arctic char and brook trout (S. fontinalis) in northern Canada (Hammar et al. 1991) (Bernatchez et al. 1995; Wilson et al. 1998). Furthermore, natural hybridization has been described between Arctic char and lake trout (S. namaycush) in the Canadian Arctic (Wilson & Hebert 1993), Dolly Varden and bull trout (S. confluentus) in British Columbia (Baxter et al. 1997; Taylor et al. 2001) and Dolly Varden and white-spotted char (S. leucomaenis) in Japan (Yamamoto et al. 2006). If genetic and phenotypic similarities shared by the species reflect ongoing introgressive hybridization between Arctic char and Dolly Varden, knowledge of the level and factors which promote or limit hybridization in Alaska could provide important clues about their divergence into distinct taxa.  17  1.5 Research objectives  The northern Pacific and Beringian regions of Alaska represent important sources of biodiversity and have been instrumental in understanding factors relating to the diversity, taxonomy and speciation of temperate faunas (Taylor et al. 2008; Geml et al. 2010). Unfortunately, many Arctic and sub-Arctic species are rapidly being jeopardized by human activities such as the dispersal of bio-pollutants, extraction of natural resources and climate change (Berg et al. 2010). Diverse forms of landlocked Arctic char morphs and potentially hybridizing populations of Arctic char and Dolly Varden provide an excellent opportunity to examine the genetic relationships between the various forms and the factors promoting or limiting divergence within and between these species in regions most susceptible to environmental change. The overall aim of my research project was to better understand the basis of variation in Salvelinus in northern environments, and in particular, the relatively understudied area of Alaska. Specifically, I wanted to establish if the ecological divergence seen in the Lower Tazimina Lake morphs reflected divergence at the genetic level and to document the extent of natural hybridization between Arctic char and Dolly Varden.  To achieve my aims, I conducted the following studies:  1.  In Chapter 2, I used microsatellite DNA markers to test whether morphological and ecological variability between body-size Arctic char morphs in Alaska was associated with genetic differentiation.  2.  In Chapter 3, l evaluated the distribution and level of interspecific hybridization between sympatric Arctic char and Dolly Varden across various microhabitats in two southwestern Alaskan lakes to determine the degree of genetic and ecological segregation between the two species. Here, I also explored the possibility that spatial distribution limits opportunities for hybridization.   18  Chapter 2:                                                                                                         Evidence for genetic distinction among sympatric ecotypes of Arctic char (Salvelinus alpinus) in southwestern Alaskan lakes   2.1 Introduction  Resource polymorphism is the means by which alternative phenotypes exploit distinct niches within a given ecosystem (Skúlason & Smith 1995; Arbour et al. 2011). Resource polymorphism may promote reproductive isolation if the traits associated with differential resource use influence mate choice directly, or if hybrids between differentially adapted trophic ecotypes suffer reduced survival in parental niches (e.g., Hatfield & Schluter 1999; Corrigan et al. 2011; see also McPhee et al. 2012). Morphological diversification is often complex and can occur through environmentally driven plasticity, or it can reflect underlying genetic differences. As resource polymorphism has been proposed as a key factor behind the diversification of vertebrates, understanding if resource polymorphism is associated with genetic variation is critical to our knowledge of population divergence and speciation (Wimberger 1994; Skúlason & Smith 1995; Smith & Skúlason 1996; Schluter 1998).  Genetic data that accompany ecological measures of differentiation are especially informative, but are often lacking (Klemetsen 2002). In particular, the relationship between the degree of genetic isolation and resource-driven phenotypic segregation is often poorly understood (Gíslason et al. 1999; Loh et al. 2012). Furthermore, resource polymorphisms may signal underlying genetic differentiation within taxa; knowledge of these patterns and processes can result in the development of more nuanced conservation strategies by documenting cryptic variation within taxa (Taylor 1999; Foster et al. 2003; Taylor et al. 2011).  The Arctic char (Salvelinus alpinus) is a circumpolar species and the most northerly distributed freshwater fish. The Arctic char has been very successful in colonizing post-glacial lakes and river systems, and it is often the only salmonid 19  species present in northern freshwater environments (Klemetsen et al. 2003; Reist et al. 2013). Part of this success may be because Arctic char can vary greatly in resource and habitat utilization, growth patterns, life history traits, and morphology, resulting in a number of specialized forms inhabiting a range of habitats (e.g., Power et al. 2009; Klemetsen 2010). For example, the size and shape of the jaws and number of gill rakers vary among populations of Arctic char (Skúlason & Smith 1995) and these traits are typically heritable and associated with differences in trophic ecology (Skúlason et al. 1993; Adams & Huntingford 2002a). Because of its highly polymorphic nature and ability to rapidly undergo trophic diversification following colonization of relatively young ecosystems, the Arctic char is ideal for examining the role of trophic variation in adaptive radiation. Understanding natural genetic and ecological differentiation across sympatric populations can also provide important insights about how species may adapt to particular habitats during the early stages of divergence.  Differentiated populations from sympatric lake systems have long been used as natural laboratories to study questions regarding the ecological and genetic basis of intraspecific divergence within several fish species (e.g., Frost 1965; Bentzen & McPhail 1984; Bernatchez et al. 1999; Langerhans et al. 2007). Post-glacial environments containing alternative forms that differ in resource use, or morphotypes, are especially informative as these young systems often have low biodiversity and high resource availability (see review by Taylor 1999). As multiple unoccupied niches may be exploited by colonizing fish, resource polymorphism arising from alternative life history patterns can be maintained (Berg et al. 2010). Understanding the genetic basis of sympatric systems can thus provide important clues relating to the diversification of alternative forms (Gíslason et al. 1999). Sympatric morphotypes, or morphs, of Arctic char have been documented from European countries such as Scotland, Sweden, Norway, Switzerland and Iceland, and to a lesser extent, North America and Russia (Jonsson & Jonsson 2001; Adams & Huntingford 2002b). A well-known example that showcases the remarkable potential for divergence within Arctic char is the four morphs present in Lake 20  Thingvallavatn, Iceland (Skúlason et al. 1989, 1996; Snorrason et al. 1994). Within the lake, two morphs are benthivorous specialists, but differ in size at maturity, one morph is piscivorous, and the fourth is a pelagic zooplanktivore. Alaska has one of the richest salmonid faunas in the Northern Hemisphere, including Arctic char which are found from the Kenai Peninsula in southwestern Alaska, around the Bering Sea drainages north and east to and including the North Slope (McPhail & Lindsey 1970). Despite this extensive distribution, relatively little is known about the extent of polymorphism in Alaskan Arctic char, investigation of which has been hampered by historical taxonomic confusion involving the relationship between Arctic char and Dolly Varden char (S. malma). Distinguishing the two species has proved challenging due to their similar morphology and life history, but their status as distinct species is supported by strong genetic differentiation in sympatry (e.g., Taylor et al. 2008). Woods et al. (2013) recently examined the extent of resource polymorphism in Arctic char from three isolated lakes in southwestern Alaska as a function of lake size, species diversity, and degree of isolation from a much larger fourth lake postulated to be the source population of Arctic char in the study area. The lakes were chosen because they exhibit many of the physical criteria consistent with the formation of alternative forms of Arctic char (e.g., long-term isolation and few available fish species). Indeed, in one of the three isolated lakes (Lower Tazimina Lake), two sympatric morphs of Arctic char were identified: ‘large’ and ‘small’, which were morphologically and ecologically distinct from each other (Woods et al. 2013). Large morphs had an estimated asymptotic size of > 50 cm and were gold or orange in colouration. Small morphs had an estimated asymptotic size of ≤ 20 cm and had silver, deeper bodies with light spots, and more gill rakers than large morphs. Large morphs consumed more benthic resources whereas small morphs fed mostly on zooplankton in the limnetic zone. Large morphs generally had steeper size-at-age growth curves, corresponding to higher growth rates and the small morphs had shallower growth curves and lower growth rates (Woods et al. 2013). It is unclear, however, to what degree the different morphs in Lower Tazimina Lake are genetically distinct from one another.  21  Interestingly, Woods et al. (2013) also suggested that a third group of fish was present whose growth rate was intermediate to those of the large and small morphs. These “medium” morphs were orange to silver in colouration, had an estimated asymptotic size about 33 cm, and foraged in the benthic regions of the lake like large morphs, but fed on slightly different prey items (i.e., relied more on terrestrial insects and less on snails) and had intermediate gill raker counts. The size-at-age growth curve for medium morphs closely resembled the growth curve of the large morph (Woods et al. 2013); however, it is unknown whether medium morphs are derived from admixture between large and small morphs or whether they represent their own distinct population within Lower Tazimina Lake. In order to better understand the degree of genetic differentiation between sympatric ecological and morphological forms of Arctic char from southwestern Alaska, I used microsatellite DNA markers to: 1) test whether the forms from Lower Tazimina Lake were genetically distinct from one another and; 2) place the genetic variation within Lower Tazimina Lake into the context of genetic differentiation among a series of four other post-glacial lakes in southwestern Alaska that also contain char, but do not appear to exhibit obvious resource polymorphism. The lakes were selected to represent different scales of geographical isolation; four, including Lower Tazimina Lake, are in the same drainage basin. The fifth lake, Lake Aleknagik is from an adjacent river system and is accessible from the sea by fishes. I sampled these lakes for char to understand the level to which genetic differences among morphs within a lake compare to divergence between lakes that have been isolated since deglaciation (e.g., Caribou Lakes and Summit Lake) and those that may experience some gene flow via anadromous char (Iliamna Lake and Lake Aleknagik), but that are located in separate watersheds.     22  2.2 Materials and methods  2.2.1  Fish collection  A total of 217 tissue samples (fin clips and dorsal muscle plugs) of Arctic char were donated to this study by Pamela J. Woods and Thomas P. Quinn (University of Washington, School of Aquatic and Fishery Sciences, Seattle, WA) and represented a subset of samples analyzed previously in Woods et al. (2013). These samples were collected from four lakes in the Kvichak River (Iliamna Lake) system in Bristol Bay, southwestern Alaska, and one lake (Lake Aleknagik) in the Nushagak River system of Bristol Bay (Fig. 2.1). Gene flow is likely limited between three of the lakes, Caribou Lakes, Lower Tazimina Lake, and Summit Lake, due to the presence of physical barriers that restrict the upstream migration of Arctic char and other fish species into the lakes. Summit Lake and Lower Tazimina Lake are above large barrier waterfalls, and migration to Caribou Lakes is limited by a series of rapids in the downstream Koksetna River. Caribou Lakes are at the highest elevation (550 m above sea level, ASL) whereas Summit and Lower Tazimina lakes are lower (152 and 194 m ASL, respectively). Caribou (1.2 km2, max depth of 5 m) and Summit lakes (0.6 km2, max depth of 20 m) are smaller, shallower and more remote than Lower Tazimina Lake, which is larger and deeper (520 km2, max depth of 60 m). Caribou Lakes, and Lower Tazimina and Summit lakes eventually drain into Iliamna Lake (14 m ASL, 2,622 km2, 301 m max depth), the largest lake in Alaska, which in turn empties into the southeastern portion of Bristol Bay via the Kvichak River. Iliamna Lake has a much more diverse fish community and is freely accessible to migratory fishes, in contrast to the other three lakes. Fish were also sampled from Lake Aleknagik (12 m ASL, 83 km2, max depth of 101 m), found in the adjacent Wood River Lake system, which is open to migration from the sea. Sampling was conducted from August─September 2008 to 2011. Fin clips from char were taken from Summit Lake (n = 59), Lower Tazimina Lake (n = 91), Caribou Lakes (n = 25), and Iliamna Lake (n = 42) and stored in 95% ethanol.    23        Figure 2.1: Collection sites of Arctic char (Salvelinus alpinus) in southwestern Alaska.          1 = Caribou Lakes, 2 = Lower Tazimina Lake, 3 = Iliamna Lake, 4 = Summit Lake. See inset for Lake Aleknagik.      Q 24  For Summit Lake, Lower Tazimina Lake and Caribou Lakes, three gill nets with dimensions of 57 m length x 1.6 m depth, 38.1 m x 1.8 m, and 30.5 m x 1.8 m were set as bottom nets close to and perpendicular to shore in shallow water (1-7 m deep). An additional sinking gill net and two floating gill nets were set at 17, 9 and 10 m, respectively, in Summit Lake. Mesh sizes ranged from 10 to 102 mm (see Woods et al. 2013 for full details). Iliamna Lake samples were obtained through seining and angling off near-shore habitats that were approximately 1-3 m in depth at the eastern end of the lake (Fig. 2.1). Exact depth information for individual fish was not recorded for catches originating from bottom set nets, i.e., in Summit Lake, Lower Tazimina Lake, and Caribou Lakes fish caught at depths between 1 and 17 m in depth could not be distinguished.  I collected additional samples of Arctic char from Lake Aleknagik in July─August, 2012 (n = 70) using a combination of gill nets and stick seines to place the level of divergence and genetic diversity seen within the Kvichak River system in the context of differences from another basin. Lake Aleknagik is located in the Wood River Lake system, about 85 km west of the Kvichak River system, and also flows into Bristol Bay, but via the Nushagak River. The Wood River Lake system has a similarly diverse fish community with easy access to the sea by anadromous fishes (Hartman & Burgner 1972).   2.2.2  DNA extractions and microsatellite DNA analyses  I extracted genomic DNA from a total of 287 fin clip or dorsal plug samples stored in 95% ethanol using the DNeasy DNA blood and tissue extraction kit (Qiagen Inc., Valencia, CA, USA) following kit protocols. Extracted DNA samples were stored at -20 °C for later use in multiplex polymerase chain reactions (PCR) using the Qiagen multiplex kit (Cat. No. 206145). The DNA was amplified in 10-µL PCR reactions at 95 °C for 15 min, 94 °C for 30 sec, 35 cycles of 1.5 min at an annealing temperature of 55 °C followed by 72 °C for 1 min, and 60 °C for 30 min.  I assayed microsatellite variation using primers labeled with infrared fluorophores and a 3730S 48-capillary DNA Analyzer with GS 500 LIZ or 600 LIZ internal size 25  standards (Applied Biosystems, Carlsbad, California, USA). Alleles were manually scored using the program GeneMapper (GeneMapper v.3.7, Applied BioSystems). I genotyped fish using 11 microsatellite loci isolated from other salmonid species such as Atlantic salmon, Salmo salar (SSOSL456; Slettan et al. 1997), bull trout, Salvelinus confluentus (Sco200, Sco215, Sco216, Sco220; DeHaan & Ardren 2005), Chinook salmon, Oncorhynchus tshawytscha (OtsG83b, OstG253b; Williamson et al. 2002), Dolly Varden, Salvelinus malma (Smm-17, Smm-22, Smm-24; Crane et al. 2004) and rainbow trout, Oncorhynchus mykiss (OMM1105; Rexroad et al. 2002).  To minimize genotyping scoring errors I used MICRO-CHECKER software (van Oosterhout et al. 2004) to identify instances where alleles failed to amplify owing to mutations in the primer binding sites or failure to detect large alleles (“null alleles” and “large allele dropout”, respectively). Genetic polymorphism was estimated from sample size (N), number of alleles per locus (A), allelic richness (AR), and observed (Ho) and expected (HE) heterozygosity using FSTAT ver 2.9.3 (Goudet 2001). The following tests were implemented using GENEPOP ver 4.2 (Raymond & Rousset 1995). Tests for departures from Hardy-Weinberg equilibrium (HWE) were performed for each locus-population combination using an exact test in which probability values were determined using a Markov chain method (P). Tests for genotypic linkage disequilibrium for all combinations of locus pairs within a population were also made using a Markov chain method with GENEPOP default values.   2.2.3  Population structure within and among lakes   To assess the level of population structure within and among lakes I used the Bayesian clustering analysis within STRUCTURE (Pritchard et al. 2000). STRUCTURE uses a Markov Chain Monte-Carlo (MCMC) to cluster individuals into K randomized and interbreeding groups that minimize departures from Hardy-Weinberg Equilibrium and linkage disequilibrium within the groups. I used the admixture model with correlated allele frequencies and a burn-in of 50,000 iterations proceeded by an additional 450,000 iterations, replicated five times to verify 26  consistency across runs. Because I had prior indications of distinct morphotypes in Lower Tazimina Lake (Woods et al. 2013) I analyzed the data in two ways. First, to determine the most likely number of populations (K) across all lakes, I ran simulations of K = 1 to K = 10 for all samples from all lakes in one analysis. Next I assessed K = 1 to K = 5 for each lake separately. For both kinds of analysis, I used STRUCTURE HARVESTER to process the results from multiple runs of STRUCTURE (Earl & vonHoldt 2012)  which were visualized using DISTRUCT (Rosenberg 2004). Pairwise genetic divergence across samples was expressed as FST and was quantified by calculating θ (Weir & Cockerman 1984). Pairwise values were then tested for significance using GENETIX ver 4.05.2 (Belkhir et al. 2001). A factorial correspondence analysis (FCA) was also conducted on allele frequencies in GENETIX. An FCA is a type of factor analysis that summarizes allele frequencies among all samples by finding the linear combination of variables that best describes variation between individual samples and is well-suited analysis for categorical data such as allele frequency counts. All statistical tests of differences between samples were corrected for multiple comparisons following Narum (2006).  2.2.4  Comparisons of genetic and phenotypic data  To assess the distribution of admixture values ((Q), the proportion of each fish’s genome derived from K genetic groups) estimated by STRUCTURE amongst the Lower Tazimina Lake morphs, individual fish were assigned to ‘large’, ‘medium’, or ‘small’ morph groups according to their growth curve assignment probability as found in Woods et al. (2013). To obtain growth curve assignment probabilities, Woods et al. (2013) first generated a Von Bertalanffy growth curve for each fish,  µ = L∞(1- e-ka)  27  using the parameters L∞(asymptotic length), k (curvature) and a (age estimated from sagittal otoliths) to represent length predictions (µ). Using a mixed model approach to access the likelihood of multiple Von Bertalanffy curves, the authors found evidence for the presence of two or three growth curves in Lower Tazimina Lake but could not ascertain the exact number of growth forms. To access the likelihood of three growth forms in Lower Tazimina Lake, Von Bertalanffy growth curves for each fish were fit to a three-curve model corresponding to a small-, medium-, and large- growth type (based on asymptotic length, L∞). An assignment probability value was generated for each fish and was used to assign fish to the component (growth curve, according to L∞) whose proportion of that individual’s summed likelihood was the largest. Woods et al. (2013) then used these assignment values to assign each fish to the large-, medium- or small- body size groups (n = 74). Using these pre-determined groupings, I then assessed the distribution of Q (defined with STRUCTURE) within each of these body size groups.  I first tested for differences in the mean values of Q among the three body size groups with a one-way ANOVA, with proportional Q values arcsine transformed, and post-hoc tests for samples with uneven variances using PAST (vers. 2.17b), a general spreadsheet-based statistical software package (Hammer et al. 2001).  The STRUCTURE analysis (see Results) resolved two genetic groups within Lower Tazimina Lake with the small-bodied fish tending to have admixture values < 0.2, large-bodied fish had admixture values > 0.8, and medium-sized fish tended to have intermediate admixture values. To assess whether or not the medium body size char may represent hybrids between the large and small-bodied fish I used all fish classified as large-bodied and small-bodied according to growth curve assignment and the program HYBRIDLAB 1.0 (Nielsen et al. 2006) to generate simulated multilocus F1 hybrid genotypes (n = 100) between these two groups of Arctic char. The upper and lower bounds of Q values for putative hybrids between large and small-bodied fish were found by running all Lower Tazimina Lake samples and the simulated hybrids in STRUCTURE with 50,000 replications during the pre-simulation burn-in, proceeded by an additional 450,000 iterations. Simulated hybrids 28  had Q values that were ≤ 0.82 and ≥ 0.19. I then tested for a significant association between morph (large, medium, or small-bodied char, assigned using the growth curve assignment probabilities, and the STRUCTURE- defined genetic groupings as described above (parental = > 0.82 or < 0.19; hybrid = ≤ 0.82 and ≥ 0.19) by using a 3x3 contingency table in PAST.  Woods et al. (2013) also reported gill raker counts taken on the first arch (ranged from 20 – 32) for each body size morph (n = 74). I assessed the association between morphology and genetic identity by testing for a Pearson’s correlation between individual gill raker count and the admixture proportions (Q) from the STRUCTURE analysis using the Hmisc package and the cor function in R (version 3.0.2, R Development Core Team, 2011).    2.3 Results 2.3.1 Microsatellite variation among and within lakes  MICRO-CHECKER found no evidence of a failure to resolve alleles owing to “large allele dropout” or otherwise non-amplifying (“null”) alleles in any of the samples. All loci were polymorphic, except for Sco215 in Summit Lake which was monomorphic (Appendix A Table. A.1). When examining samples within each lake (i.e., before partitioning by morph or by genetic sub-groups identified by STRUCTURE), eight of 54 tests for deviations from Hardy-Weinberg Equilibrium (four samples x 11 loci plus 1 sample x 10 loci) were significant at P ≤ 0.0171, and three of the eight deviations were in the Lower Tazimina Lake sample owing to deficits of heterozygotes (Appendix A Table. A.1). Of the tests of linkage disequilibrium per population, 1/45 for Summit Lake, 3/55 for Lower Tazimina Lake, 1/55 for Caribou Lakes, 2/55 for Lake Iliamna, and 3/55 for Aleknagik Lake were significant. No locus pairs consistently deviated from linkage equilibrium across lakes so I retained all loci in subsequent analyses. The greatest average allelic richness was found in Lower Tazimina Lake (17.7) and Lake Aleknagik (16.4), followed by Iliamna Lake (16.1, Appendix A Table. A.1). In Caribou Lakes and Summit Lake fish, allelic richness was lower (7.2 and 7.6, 29  respectively); however, a large proportion of those alleles were unique to those populations. The highest level of gene diversity was observed at the locus Sco220 and Smm24 (1.0) while the locus Sco216 showed the lowest gene diversity (0.32).  Across loci and populations, the highest average expected heterozygosity was seen in Lake Aleknagik fish (0.84) and the lowest heterozygosity was seen in Summit Lake fish (0.68).   2.3.2 Population structure among lakes  Pairwise FST across all five lakes ranged from 0.050 between Iliamna Lake and Lower Tazimina Lake to 0.283 between Summit Lake and Caribou Lakes (P < 0.001, Table 2.1). In general, FST was greatest between Caribou Lakes and Summit Lake and all other samples (minimum FST = 0.100, Table 2.1). The FCA showed a large degree of divergence among populations, especially between Summit Lake and the char from the other four lakes (Appendix A, Fig. A.1). The FCA also revealed that the Lower Tazimina Lake char were more similar genetically to Iliamna Lake fish when compared across all populations in the Kvichak River system (Appendix A, Fig. A.1).  Analysis by STRUCTURE, indicated that the most likely number of genetic populations was K = 6 among the five lakes, as evidenced by a mean log likelihood of ─13199.7 vs. ─13298.7 for K = 7 as the next most likely model (Fig. 2.2). In the K = 6 model, each lake formed a distinct genetic group, but two groups were resolved within Lower Tazimina Lake (Fig. 2.2). When the Lower Tazimina Lake samples were analyzed separately, the most likely number of clusters was K = 2 (mean log likelihood = ─3507.1 vs. ─3581.1 for K = 1 and ─3659.5 for K = 3). No distinct clusters were observed within any of the other lakes when they were examined together (Fig. 2.2), or when the lakes were analyzed individually; a K = 1 was returned in all cases (May-McNally et al. unpubl. data).   30       Table 2.1:  Pairwise FST (θ) estimated by variation across 11 microsatellite DNA loci in Arctic char (Salvelinus alpinus) from Caribou Lakes, Summit Lake, Lower Tazimina Lake large (L), medium (M), and small-bodied (S) morphs, Iliamna Lake, and Lake Aleknagik. All comparisons except the starred values are significantly greater than 0 (P ≤ 0.0185; after adjustment for multiple tests incorporating the false discovery rate procedure of Narum, 2006).    Caribou Lakes Summit Lake Lower Tazimina Lake-L Lower Tazimina Lake-M Lower Tazimina Lake-S Iliamna Lake   Summit Lake 0.283 - - - - - Lower Tazimina Lake-L 0.178 0.153 - - - - Lower Tazimina Lake-M 0.149 0.132 0.017* - - - Lower Tazimina Lake-S 0.162 0.117 0.092 0.037 - - Iliamna Lake 0.162 0.131 0.060 0.053 0.080 - Lake Aleknagik 0.100 0.205 0.090 0.062 0.092 0.085 Pairwise comparisons between Caribou, Summit, Iliamna Lake, Lake Aleknagik, and the pooled sample from Lower Tazimina Lake were 0.143, 0.108, 0.050, and 0.069, respectively (all P < 0.0185).     31              Figure 2.2: DISTRUCT plots for STRUCTURE runs of five populations of Arctic char (Salvelinus alpinus) from southwestern Alaska using 11 microsatellite markers: Caribou Lake (n = 25), Summit Lake (n = 59), Lower Tazimina Lake (n = 91), Iliamna Lake (n = 42), Lake Aleknagik (n = 70), and simulated hybrids (SH, n = 100).  Each fish is represented by a vertical bar that denotes membership fractions (Q) in K clusters. Lower Tazimina Lake morphs (lower panel, blue and orange clusters) were defined as large-, medium- or small- based directly on growth curve assignment data from Woods et al. (2013) and organized according to increasing Q value (in reference to the orange cluster) within each group (n = 74). Asterisks denote genetic samples that lack growth curve information and were included in the appropriate genetic groups according to their Q values (n =18). Horizontal solid lines in the SH plot represent the upper and lower boundaries for Q value defined hybrids.K= 2 Q K= 6 K= 2 Q Q 32   2.3.3 Comparisons between genetic and phenotypic data within Lower Tazimina Lake   When Lower Tazimina Lake char were grouped according to body size using (see Woods et al. 2013 for results), there were no deviations from Hardy-Weinberg Equilibrium (Appendix A Table. A.1) and there was only a single deviation from linkage equilibrium in the large morph sample. Further, at 9 of 11 loci, small morphs exhibited a higher AR per locus than both large and medium morphs as well as similar or higher AR values relative to the populations from Iliamna Lake and Lake Aleknagik (Appendix A Table. A.1).  There was a significant association between admixture values generated by STRUCTURE and body size morph; a one-way ANOVA followed by a post hoc test revealed significant pairwise differences among the mean Q values of each of the three morphs (F2,71 = 50.9, P < 0.001, all Tukey P < 0.001, mean Q = 0.10, 0.53, and 0.83 for small, medium and large-bodied morphs, respectively). Further, the mean Q value of the simulated hybrids (0.48) was not significantly distinct from that of the medium-sized morphs (Tukey P = 0.72), but was significantly distinct from the mean Q values of large and small-bodied morphs (both P < 0.001). Nine of 13 fish found in the range of Q values of simulated hybrids were of the medium-sized morph, while most large-bodied fish had Q values > 0.82 and most small-bodied fish had Q values < 0.19 (X2 = 29.6, P < 0.05, d.f.=  4, Table 2.2).          33           Table 2.2:  Contingency table analysis of Q value categories and growth curve assignment for small, medium and large morphs of Arctic char (Salvelinus alpinus) sampled from Lower Tazimina Lake, southwestern Alaska (n = 74)  Q value Small Medium  Large 0.0 – 0.18 28 5 0 0.19 – 0.82 2 9 2 0.83 – 1.0 0 9 19            34  In the FCA projection, large and small morphs showed roughly equivalent divergence from Iliamna Lake fish on the same FCA plot (Appendix A, Fig. A.2). Genetic distinction in allele frequencies was greatest between the large and small morphs (FST = 0.092, P < 0.001, Table 2.1). The medium and small morphs were also distinct from one another (FST = 0.037, P < 0.001), but there was only marginal genetic differentiation detected between the large and medium morphs (FST = 0.006, P = 0.09). Generally, Lower Tazimina Lake morphs were most divergent from Caribou and Summit lakes fish (minimum FST all > 0.108) and least divergent from char in Iliamna Lake (maximum FST = 0.080, Table 2.1). Pairwise FST averaged 0.048 among the three morphs in Lower Tazimina Lake and 0.050 between Lower Tazimina Lake morphs and fish from Iliamna Lake (Table 2.1). I found a significant negative correlation (r =-0.73, t = -8.67, df = 72, P < 0.001) between genetic admixture value (Q) and gill raker count across individual char (Fig. 2.3). Fish with low Q values had higher average gill raker counts than fish with higher Q values reflecting the differentiation, genetically and phenotypically, between the small and large-bodied morphs, respectively. Medium-sized morphs were characterized by Q values of between 0.19 and 0.82 and had an average gill raker count, of 23.8, compared to 27.9 and 22.8 for the small and large-bodied morphs, respectively (Fig. 2.3).          35        Figure 2.3: Pearson’s correlation between genetic admixture coefficients (Q) and total gill raker counts for Lower Tazimina Lake, southwestern Alaska, morphs of Arctic char (Salvelinus alpinus, n = 74). Fish were grouped as large, medium or small morphs according to their growth curve assignment probability as outlined in Woods et al. (2013). Horizontal bars indicate upper and lower boundaries for Q value defined hybrids from the STRUCTURE- defined genetic groupings (parental = > 0.82 or < 0.19; hybrid = ≤ 0.82 and ≥ 0.19) Fish images on y-axis represent relative body size and colouration differences (images from Woods et al. 2013).   r = -0.73, P < 0.001 36   2.4 Discussion  The Arctic char is a highly polymorphic species with an impressive ability to undergo rapid trophic diversification in isolated post-glacial lakes (reviewed by Klemetsen 2010). The Kvichak River system in southwestern Alaska has phenotypically diverse forms of Arctic char, notably at least two sympatric morphs in Lower Tazimina Lake (Woods et al. 2013). I have provided evidence that this morphological and ecological variability is accompanied by significant genetic differentiation in that at least two of the morphs are strongly divergent from one another and their genetic identity can be predicted to some degree by growth curves and total gill raker counts.   2.4.1 Microsatellite DNA variation across remote freshwater lakes  My microsatellite data revealed substantial variability in pairwise FST across Arctic char from five lakes and is consistent with variability found in other studies involving post-glacial lake systems with Arctic char (Bernatchez et al. 1998; Westgaard et al. 2004). The high values of pairwise FST involving Summit and Caribou lakes are likely attributable to their isolation from Iliamna Lake and Lower Tazimina Lake. Arctic char from both lakes exhibited numerous private alleles as well as dissimilar morphology and Von Bertalanffy growth curves compared to fish from Lower Tazimina Lake and Iliamna Lake (see Woods et al. 2013).   Arctic char from Summit Lake and Caribou Lakes may represent populations that were derived from the earliest colonizing fish from Kvichak River system. Summit Lake fish, in particular, showed low heterozygosity at multiple loci which may be due to differing demographic histories (e.g., numerous generations of inbreeding and genetic drift) in this very small lake. In addition, the low values for pairwise FST between Lower Tazimina Lake and Iliamna Lake when contrasted with FST values involving Summit Lake and Caribou Lakes may indicate a more recent colonization of Lower Tazimina Lake. Alternatively, because Lower Tazimina Lake and Iliamna Lake (520 and 2,622 km2, respectively) are many times larger than Summit Lake 37  and Caribou Lakes (0.6 and 1.2 km2, respectively, Woods et al. 2013), they probably have had much larger effective population sizes historically that would constrain neutral divergence between these two large lakes.  The relatively low FST values between Iliamna Lake and Lake Aleknagik suggests that gene flow has occurred recently between them or is ongoing, although Arctic char in both lakes are generally regarded as non-anadromous populations because they are present in the watersheds throughout the summer when anadromous char are typically in the sea and, at least in Iliamna Lake, diet comparisons inferred from stable isotopes do not support anadromous behaviour of Arctic char (Denton et al. 2010). Anadromy, however, could exist as an alternative, but less common life history pattern in these systems (e.g., McBride 1980); otolith microchemistry analysis of Arctic char from these lakes would provide a definitive test of this idea. Alternatively, the low FST between Iliamna Lake and Lake Aleknagik may reflect shared ancestral genetic polymorphism in the face of relatively high historical effective population sizes and recent postglacial founding of these populations.   2.4.2 Genetic differentiation among Lower Tazimina Lake morphs  The results of the STRUCTURE analysis in Lower Tazimina Lake were consistent with other studies that have shown that morphologically differentiated forms may also be genetically distinct, especially in systems where profundal and littoral morphs of Arctic char have been identified (e.g., Westgaard et al. 2004; Adams et al. 2007; Gomez-Uchida et al. 2008; Power et al. 2009; Conejeros et al. 2014). The Lower Tazimina Lake morphs shared most of their alleles with fish from Iliamna Lake, which suggests that they were founded by colonists from Iliamna Lake. Moreover, Lower Tazimina Lake morphs were genetically more similar to each other than to fish from one or more other lakes, suggesting a single colonization event followed by divergence within the lake rather than two or more separate invasions of allopatrically derived forms. It is, however, impossible to discount the possibility that the large and small morphs were derived allopatrically and have become genetically similar by post-colonization gene flow without a much more extensive survey of 38  Arctic char throughout Alaska. A similar conclusion appears applicable to the sympatric “pale” and “dark” morphs on Gander Lake (Gomez-Uchida et al. 2008), and Garduño-Paz et al. (2010) presented genetic evidence that scenarios of allopatric and sympatric divergence of Arctic char may be lake-specific in Scotland.  In most systems with sympatric Arctic char populations, the large-bodied (and usually pelagic) morph is thought to most resemble the original anadromous colonizing individuals because anadromous Arctic char are large-bodied and commonly piscivorous (e.g., Dempson et al. 2002). This is likely the case for Lower Tazimina Lake morphs: the large morph is at least partially piscivorous and is more similar in terms of allelic identity, size, and morphology to Iliamna Lake fish than is the small morph (Woods et al. 2013). Small morphs from Lower Tazimina Lake showed greater divergence from Iliamna Lake Arctic char and they also had higher allelic richness and more private alleles than the large morphs and Iliamna Lake fish. These observations suggest that following colonization of Lower Tazimina Lake, body size divergence was accompanied by two trophic shifts: one from piscivory toward greater benthic food consumption in large morphs, and another towards foraging on small-sized limnetic prey by small morphs (cf. Woods et al. 2013). While the small morph represents a very distinct gene pool, the large morph population displayed greater admixture and contained more fish that were genetically intermediate between the large and small morphs. Nevertheless, mean admixture values were significantly different amongst all three morphs, but those of the medium-sized morph and simulated hybrids between large and small-sized morphs were not significantly different from one another. Further, gill raker counts of medium-sized fish were intermediate to those of the large and small-sized morphs and gill raker count typically has a genetic component in fishes (e.g., Bentzen & McPhail 1984). These observations are consistent with the idea that the medium-sized morphs are hybrids between the other two morphs of char rather than a third, less isolated population (cf. Snorrason et al. 1994). Also, the FST value between large and medium morphs was not significant which suggests that if medium-sized char are hybrids, they backcross preferentially with large-sized char. Despite feeding together in the benthic regions of the lake, large and medium morphs consumed 39  different prey items (Woods et al. 2013). The large morph ate more snails whereas the medium morph ate more terrestrial insects. Given the detectable genetic divergence of medium-sized morphs from at least the small-sized morph, and polymorphism in prey choice and differential size at maturity, the medium sized morph could be in the process of diverging into a third distinct population (e.g., Smith et al. 2003; Seehausen 2004).   2.4.3 Biogeography of sympatric morphs While sympatric morphs of Arctic char have been described from European and Asian portions of their Holarctic range, accounts of sympatric morphs in North America are scant and restricted to the central Arctic or northeastern North America. The two morphologically distinct ‘dark’ and ‘pale’ sympatric forms in Gander Lake, Newfoundland showed strong genetic segregation (Power et al. 2005; Gomez-Uchida et al. 2008). Divergence between forms was hypothesized to have occurred when the lake became ice-free following the end of the last ice age (c. 12,000 years BP). Similarly, Lake Aigneau, in northern Québec, Canada contains two morphs that are largely segregated in either the pelagic or littoral zones of the lake, and with very little gene flow between forms (FST = 0.174, Power et al. 2009). Conversely, analysis of ‘large’ and ‘small’ sympatric and morphologically distinct morphs from Lake Hazen, Nunavut, in the Canadian High Arctic revealed no evidence of genetic distinction (Arbour et al. 2011). Given that the last glacial retreat only reached the Lake Hazen region c. 8,000 years BP and glacial ice cover of the lake was likely present up until 4,200-3,300 years BP, the lack of genetic differentiation may be attributable to a relatively recent colonization of the lake (Arbour et al. 2011).   While previously documented in other regions of North America, Lower Tazimina Lake morphs represent the first example of sympatric morphotypes of Arctic char found on the west coast of North America. Further, Lower Tazimina Lake morphs exhibit some genetic divergence from one another, but most medium-sized morphs also had intermediate Q values (0.19 - 0.82) and accounted for approximately 18% of all individuals genotyped. If the medium-sized morphs are indeed hybrids, 40  perhaps the char in Lower Tazimina Lake are currently at an evolutionary intermediate stage of divergence between geologically older lakes such as Gander Lake or Lake Aigneau and younger lakes like Lake Hazen. This hypothesis is consistent with the much stronger degree of divergence found between small-bodied and large-bodied morphs in Lake Aigneau (FST = 0.174, Power et al. 2009) compared to Lower Tazimina Lake (FST = 0.092). Lower Tazimina Lake was likely colonized by Arctic char via the Newhalen River sometime after the area became ice-free (~12 000 ─ 15 000 BP, Stilwell & Kaufman 1996; Ramstad et al. 2004), with divergence of char morphs perhaps occurring later than for the Gander Lake morphs, but earlier than for morphs in Lake Hazen. Collectively, Gander Lake and Lake Aigneau, Lower Tazimina Lake, and Lake Hazen suggest a gradation of evolutionary change, which may be dependent, in part, on the time of glacial retreat and colonization (cf. Garduño-Paz et al. 2010). In addition to differences in divergence time, variation in environmental features of lakes that result in different levels of ecological opportunity probably also influence the levels of phenotypic and genetic divergence in sympatric populations of fishes (e.g., Lu & Bernatchez 1999; Landry et al. 2007; Siwertsson et al. 2010; Ormond et al. 2011). For instance, size-at-age growth curves were an effective means of assigning morphs into discrete populations and identifying intermediate individuals in Lower Tazimina Lake (Woods et al. 2013) which also differ genetically from one another as shown in our study. Further, small morphs had more gill rakers than large or medium morphs and gill raker counts were strongly associated with genetic admixture (Q) values. Gill rakers, which assist in feeding on small prey items from the water, are generally more numerous in forms with largely limnetic diets in a variety of fishes (e.g., Bentzen & McPhail 1984; Gowell et al. 2012; Roesch et al. 2013). Indeed, Woods et al. (2013) showed that the diet of the small morph was characterized by consumption of more zooplankton and fewer carbon-based prey items derived from benthic resources (fewer snails, terrestrial and aquatic insects, and fishes) and occupied a lower trophic position than either the medium or large morph. By contrast, the diets of the medium and large morphs differed only slightly from one another. Morphological adaptations such as gill rakers, diets, and growth 41  rates are, therefore, associated with neutral molecular differentiation which suggests that genetic changes have accompanied phenotypic shifts as the morphs diverged in Lower Tazimina Lake. Given that resource polymorphism may be the driving factor for the isolation of large and small morphs in Lower Tazimina Lake, an intriguing hypothesis is that variation in physiological and morphological traits has become partitioned across at least two different morphs and enabled optimal exploitation of available trophic resources (cf., Ohlberger et al. 2008; Evans et al. 2012). Further work should investigate whether differences in genes relevant to metabolism are also correlated with genetic identity (e.g., Bernatchez et al. 2010).  Past introgression of Arctic char with its sister taxa Dolly Varden (S. malma) could also act to complicate the evolutionary history of the Lower Tazimina morphs. Shared mitochondrial DNA haplotypes and demographic inferences led Taylor et al. (2008) to suggest that historical gene flow from Dolly Varden to Arctic char had occurred approximately 1100-10 900 years ago. Dolly Varden alleles may then have been introduced into certain Arctic char individuals or populations before the colonization of Lower Tazimina Lake (or after, if more than one colonizing event occurred) and perhaps contributed to a selective advantage in certain microhabitats for those individuals (hybrids) possessing Dolly Varden alleles when compared to pure Arctic char. The divergence of the morphs may then be attributable to the selectively advantageous signatures of past hybridization events and not purely to sympatric speciation.  Woods et al. (2013) presented evidence for two growth phenotypes in Summit Lake (although a model invoking a single growth form was not substantially less supported). These authors, however, found no evidence of diet differences (from prey occurrence or stable isotope analyses) in Summit Lake char, in contrast to the situation in Lower Tazimina Lake. Similarly, my genetic analysis found no evidence of distinct genetic groups in the char from Summit Lake so the existence of distinct forms in Summit Lake remains equivocal. In addition, there is no evidence of divergent phenotypes or gene pools within any of the other lakes that I examined (see also Woods et al. 2013). It is possible, however, that more extensive sampling especially with respect to depth could uncover as yet unknown diversity, especially 42  in a large lake such as Iliamna Lake (cf., Finstad & Berg 2004). Alternatively, why some lakes might have divergent forms and others not is an uncertainty common to many instances of postglacial adaptive radiations in fishes and may result from historical or contemporary factors and their interactions (e.g., Bernatchez & Dodson 1990; Lu & Bernatchez 1999; Taylor & McPhail 2000; Landry et al. 2007; Siwertsson et al. 2010; Ormond et al. 2011). Kristjánsson et al. (2011) and Woods et al. (2012) explored characteristics of lakes that might influence morphological variation in Icelandic Arctic char, not including sympatric populations, and concluded that fish community composition was one of the most important factors (see also Ormond et al. 2011 for Gasterosteus aculeatus). Fish community composition may be relevant to the potential for sympatric populations and in our system diversity ranged from two species (Summit Lake) to at least 10 (Iliamna Lake), but is confounded with many other differences amongst lakes (Woods et al. 2013). A model that incorporated historical factors (e.g., time since colonization, opportunities for multiple colonization) and contemporary aspects of lake physiography and fish community structure across the geographic range of lakes with sympatric char would perhaps shed light on the factors critical to the origin and persistence of sympatric populations.   The North Pacific and Beringia are increasingly being recognized as important wellsprings of biodiversity and have been instrumental in understanding biogeography and speciation in temperate faunas (Cook & Auster 2005; Ilves & Taylor 2008; Taylor et al. 2008). Alaska is home to a diverse and well-established assemblage of Arctic char; however, quantifying the level of biodiversity on the western regions of North America has lagged behind that in the eastern Arctic and Europe owing to taxonomic confusion with Dolly Varden, remoteness of sites and difficulty in obtaining samples. Considering that many Arctic and sub-Arctic species are rapidly becoming at risk due to anthropogenic impacts such as climate change, dispersal of bio-pollutants and the harvesting of natural resources, research into these regions is becoming increasingly critical (Berg et al. 2010; Reist et al. 2013). Moreover, the study of unique intralacustrine populations of char and other fishes can provide clues about the role of adaptive divergence in speciation. My findings 43  and those of Woods et al. (2013) suggest that ecological aspects of individuals (gill raker count and associated feeding behaviour) are correlated with genetic identity and supports the idea that the availability of divergent habitats and/or ecological opportunities within lakes may help to drive the development of specialized phenotypic traits in Arctic char and more generally (see Smith & Skúlason 1996; Jonsson & Jonsson 2001). Further, the emergence of divergent phenotypes can contribute to reduced gene flow and thus promote the evolution of reproductive isolation between sympatric populations (e.g., Schluter & Rambaut 1996; Adams & Huntingford 2002a; Klemetsen 2010; Bernatchez et al. 2010).   44  Chapter 3:                                                                                           Low levels of hybridization between sympatric Arctic char (Salvelinus alpinus) and Dolly Varden char (Salvelinus malma) highlights their genetic and spatial distinctiveness  3.1 Introduction  Natural hybridization is a fundamental evolutionary process in the biology of plants and animals (Mayr 1963; Arnold 1992; DeMarais et al. 1992; Barton 2001). When different species or reproductively distinct populations interbreed, the level of hybridization, structure of hybrid zones, and viability of hybrids can indicate factors relating to the origin and maintenance of species differences as well as processes important in the evolution of reproductive isolation (Barton & Hewitt 1989; Arnold 1997). Moreover, as natural hybridization can occur during primary intergradation or through secondary contact, studies of hybridization can provide us with a means to evaluate the strength of isolation between evolutionarily young lineages (Schluter 1996; Jiggins & Mallet 2000). For example, contact or hybrid zones can act as a test to access the status of sympatric populations as distinct biological species (Avise 1994; Taylor et al. 2008) where to pass the test, sympatric species must remain genetically distinct from one another despite the potential for, or actual occurrence of, gene flow between them. Further, defining species boundaries from studies of contact zones can also be critical when designing management goals for morphologically cryptic or hybridized populations (Campton 1987; Allendorf et al. 2001; Bickford et al. 2007). Natural hybridization is frequent in plants and several vertebrate groups; in the latter it occurs perhaps most commonly in fishes (Schwartz 1972; Campton 1987; Bernatchez et al. 1995; Allendorf & Waples 1996; Arnold 1997; Rieseberg 1997; Scribner et al. 2001). Factors such as external fertilization (Hubbs 1955), niche overlap (van Herwerden & Doherty 2006) and competition for limited spawning sites 45  (Campton & Utter 1985) have probably contributed to elevated hybridization in fishes. Hybridization and introgression can lead to a wide array of phenomena including the formation of hybrid zones, adaptive radiation, and reinforcement of pre- and post-mating reproductive barriers during speciation (Schluter 1996; Arnold 1997; Dowling & Secor 1997; Seehausen 2004; Aboim et al. 2010). For salmonid fishes (Salmo, Oncorhynchus, Salvelinus and related genera) hybridization is a typical phenomenon among particular pairs of species (reviewed in Taylor 2004). Evidence suggests that some species of salmonids exhibit little or no intrinsic post-zygotic barriers to interbreeding and this may be one of many contributing factors to the high levels of natural and anthropogenic induced hybridization within this group (Hendry & Stearns 2003; Rubidge & Taylor 2004). The lack of post-zygotic barriers to hybridization between some salmonid species makes them well-suited to studies on pre-zygotic isolation such as environment or ecologically dependent reproductive isolation, an important issue in research on speciation (Redenbach & Taylor 2003; Taylor 2004; Rogers & Bernatchez 2006). Among salmonids, char (Salvelinus) have proved instrumental in our understanding of evolution and speciation in fishes and more generally due to their incredible diversity, polymorphism, and adaptability to different habitats (Snorrason et al. 1994; Gíslason et al. 1999; Brunner et al. 2001; Jonsson & Jonsson 2001; Klemetsen 2010; Reist et al. 2013). Of seven well-recognized char species, the Arctic char (Salvelinus alpinus) arguably shows the greatest adaptive potential and exhibits extensive variability in terms of morphology, colouration, ecology and genetic structure (Klemetsen 2013). Highly polymorphic, Arctic char can have anadromous or non-anadromous life histories and can be found in various marine, estuarine and freshwater habitats across their circumpolar range (Johnson 1980). Multiple (sometimes up to four) diverse morphological forms or morphs have been described from lacustrine systems in Europe, Russia and North America, especially in recently colonized post-glacial lakes with low fish taxonomic biodiversity (Skúlason et al. 1989; Klemetsen et al. 2003; Power et al. 2009; Klemetsen 2010; Woods et al. 2013; May-McNally et al. 2014). Depending on the system and time since colonization, morphs can show significant phenotypic plasticity and genetic 46  differentiation from other Arctic char populations or even other sympatric morphs which has led to the naming of multiple taxa within the Arctic char species ‘complex’ (Reist et al. 2013). Historical contention over the status of the Dolly Varden char (Salvelinus malma) as a species distinct from Arctic char has also contributed to the taxonomic confusion within this group. Once considered to be part of the Arctic char species ‘complex’, the Dolly Varden char is now the sister species to Arctic char (Lindsey 1956; Johnson 1980; Taylor et al. 2008; Crête-Lafrenière et al. 2012). Like Arctic char, Dolly Varden also exhibit a variety of life history patterns and ecological and genetic diversity (Baxter et al. 1997; Koizumi et al. 2006; Taylor et al. 2008). Recently diverged from one another (c. 215,000–458,000 years ago) (Taylor et al. 2008), Arctic char and Dolly Varden have largely allopatric distributions, but within North America, their geographical ranges overlap only in central and southwestern Alaska (McPhail 1961; Fig. 1.2).  In Alaska, Arctic char are abundant in nearshore habitats on the western and southern portion of the state around the Alaska Peninsula and along the northern slope adjacent to the Canadian Arctic populations (McPhail 1961). Alaskan Arctic char are usually lacustrine and non-anadromous while Dolly Varden char are typically anadromous and reside in streams as juveniles. Although consistent morphological differences exist between Arctic char and Dolly Varden throughout this zone of sympatry (e.g. pyloric caeca, gill raker counts), McPhail (1961) also identified populations that had ambiguous morphology and speculated that they have resulted from hybridization between the species. Phylogeographical mitochondrial (mtDNA)-based surveys by Brunner et al. (2001) support the idea of recent or ongoing hybridization between the species in Alaska. Five distinct mtDNA clades across the Holarctic range of Arctic char have been identified (Bering, Siberia, Atlantic, Acadia and Arctic) and were speculated to have resulted from isolation in separate Pleistocene glacial refugia. All samples identified morphologically as Dolly Varden from the western Pacific collectively grouped in the Bering lineage and shared haplotypes with southern Alaskan Arctic char. Paraphyly in mtDNA between Beringian Arctic char and Dolly Varden resulted in a lack of 47  resolution between the lineages and Brunner et al. (2001) suggested that the separate species status of Dolly Varden was “questionable”. Shared haplotypes between the species also suggests that gene flow between the species in these areas has occurred recently or is ongoing.   Using a combination of morphology and genetics (microsatellite and mtDNA), Taylor et al. (2008) further examined Arctic char and Dolly Varden from lakes in western and southwestern Alaska. For the majority of systems, they found that the sympatric taxa still shared mtDNA sequence similarity, but were highly genetically distinct at nine microsatellite loci, supporting their status as valid biological species. In one southwestern Alaskan lake (Lake Aleknagik) however, 7% of fish had ambiguous genetic identity and were putatively classified as hybrids. Moreover, McPhail (1961) described fish from Lake Aleknagik with phenotypes intermediate between Arctic char and Dolly Varden and speculated that they were hybrids. Detecting putative hybrids in Lake Aleknagik is not surprising given that Dolly Varden have also been found to hybridize with other char, such as bull trout (Salvelinus confluentus) (Taylor et al. 2001; Redenbach & Taylor 2003). Given the genetic and morphological similarities between Arctic char and Dolly Varden in Lake Aleknagik, it is possible that a potential mosaic hybrid zone could exist between the species and habitats contributing to hybridization could be more diverse than expected. A combination of relatively low sample size and lack of sample site diversity in the Taylor et al. (2008) study, however, limited their ability to accurately access the level of ecological and genetic segregation between the char species in the Aleknagik system. Moreover, as the mode of selection structuring areas of hybridization between Arctic char and Dolly Varden has not yet been defined, sampling a variety of different microhabitats (e.g. across streams, various depths in the lake) could help establish whether environment-dependent selection may be an important factor in the spatial distribution of hybridization and perhaps yield hints to the processes involved in the divergence of the two species. In this chapter, I report the results of a robust microsatellite survey between the two species in Lake Aleknagik, Alaska, as well as another connected lake in the same lake system to test several ideas. First, I wanted to assess the degree of 48  interspecific ecological segregation, in the form of species specific spatial distribution maps and age class allocations across multiple sampling sites inferred from genotyping results. Second, I used the microsatellite data to assess the degree of genetic divergence between species, hybridization levels, and assignment of individuals to different genotypic classes. The microdistribution and level of hybridization between the species was accessed across a wide variety of microhabitat types including across the length of streams, varying depths in the lake and from different beach sites in order to access whether habitat structure may influence hybrid distribution and prevalence. Determining the extent of genetic and ecological diversity will hopefully help us understand factors important in the evolution of reproductive isolation between Arctic char and Dolly Varden and have implications for the role of natural hybridization in speciation as well as our knowledge of the evolution of Arctic biodiversity.  3.2 Materials and methods  3.2.1  Sample collection  Lake Aleknagik (832 km2, 32 km in length) is located near the southwestern coast of Alaska, in the central portion of Bristol Bay and near Dillingham, Alaska (Fig. 3.1). Lake Aleknagik is the first lake in the Wood River Lake system, a series of five interconnected lakes (425 km2) that drain via the Wood River into the Nushagak River to the south and eventually into Bristol Bay. The lakes are oligotrophic and range from 6 to 201 km2 in length. Lake Aleknagik is home to a diverse native fish community including pygmy whitefish (Prosopium coulterii), threespine sticklebacks (Gasterosteus aculeatus), slimy and coastrange sculpins (Cottus cognatus and C. aleuticus), and salmonids (Salmonidae), with char and sockeye salmon (Oncorhynchus nerka) being the most numerically dominant salmonids.    49        Figure 3.1:    The Wood River Lake system in southwestern Alaska, USA. 1 = Wood River, 2 = Lake Aleknagik, 3 = Agulowak River, 4 = Lake Nerka (South arm), 5 = Lake Nerka (North arm)    1 2 3 4 5 50  Sympatric char were collected during the period of July to August 2012-2013 from 33 sites in Lake Aleknagik (total N collected = 752, Table 3.2) representing four lake, three beach, one river, and 25 stream sites by using a combination of sinking gillnets, beach seines, stick seines, minnow traps and angling (Appendix B Table. B.1). Lake, river, and lower, mid and upper streams sites were sampled in 2012 while only stream sites were sampled in 2013 in an attempt to increase the sample size of Dolly Varden. Sinking gillnets were set by boat daily for 4-12 h and suspended above the bottom with floats. Lake sites were chosen to represent various depths and regions around the lake and included a shallow (Site 7, ~5─15m deep), intermediate (Site 4, ~15─20m deep) and deep (Site 5, ~20─50m deep) site. At each of these sites, two gillnets were used in an attempt to sample different size classes of char in Lake Aleknagik, both 30 m long by 2 m high, one with 30 mm and the other with 20 mm stretched mesh. Beach sites (Sites 2, 3, 7) were selected from various regions around the lake and were sampled using a 32 m long x 5 m deep beach seine net tapered from 5 m deep in the centre 6 m wide section with 6 mm mesh to 13 mm mesh in the 2.5 m side panels and 1 m deep with 30 mm mesh at the edge panels.  During the day it was deployed to shore by boat in a semi-circle reaching approximately 10-15 m offshore. The majority of samples originating from streams were collected using an 8 foot long stick seine with a mesh size of approximately 5 mm x 5 mm. Minnow traps baited with salmon eggs were also used in Hansen Creek sites 17-19.  Fin clips constituted the removal of the adipose fin to avoid re-captures and were stored in 95% ethanol until later laboratory analysis. Fish classified as young of the year (YOY) were euthanized immediately in a dilute solution of MS-222 (tricaine methanesulfonate) and stored whole in 95% ethanol. A visual estimate of species identity (Arctic char, Dolly Varden or hybrid), fork length, a photograph and GPS coordinates were taken for each fish sampled and microhabitat features of the sampling site noted. Samples originating from Lake Nerka, the second lake in the Wood River Lake system and just upstream of Lake Aleknagik, were collected by Daniel Schindler and Kale Bentley (University of Washington, Seattle, WA) in 2012-2013 by stick seining and angling (total N collected = 329). Lake Nerka sites include 51  one river and 14 stream sites (Appendix B Table. B.2). Additional samples of spawning Arctic char and Dolly Varden from Lake Aleknagik and the Wood River were collected from September-October, 2012 by angling and donated to this study by C. Schwanke (Alaska Department of Fish and Games, ADF&G, Dillingham, AK). The spawning Arctic char were obtained from outside the mouth of Youth Creek, located on the far north side of Lake Aleknagik. The spawning anadromous Dolly Varden return to freshwater in autumn and were obtained from the Wood River which drains Lake Aleknagik into Bristol Bay to the south.   3.2.2  Microsatellite analysis  A total of 496 Lake Aleknagik and 229 Lake Nerka fish were included in the genetic analysis. Fin clip samples were digested and total genomic DNA was extracted using the DNeasy DNA blood and tissue extraction kit (Qiagen Inc., Valencia, CA, USA). Extracted DNA samples were stored at -20 °C for later use in multiplex polymerase chain reactions (PCR) using the Qiagen multiplex kit (Cat. No. 206145). DNA was amplified in 10-µL PCR reactions at 95 °C for 15 min, 94 °C for 30 sec, 35 cycles of 1.5 min at an annealing temperature of 55 °C followed by 72 °C for 1 min, and 60 °C for 30 min. I assayed microsatellite variation using primers labeled with infrared fluorophores and a 3730S 48-capillary DNA Analyzer with GS 500 LIZ or 600 LIZ internal size standards (Applied Biosystems, Carlsbad, California, USA). Alleles were manually scored using the program GeneMapper (GeneMapper v.3.7, Applied BioSystems). I amplified 13 microsatellite loci isolated from other salmonid species including: Atlantic salmon, Salmo salar (SSOSL456; Slettan et al., 1997), bull trout, Salvelinus confluentus (Sco200, Sco202, Sco215, Sco216, Sco220; DeHaan & Ardren, 2005), Chinook salmon, Oncorhynchus tshawytscha (OtsG83b, OstG253b; Williamson et al., 2002), Dolly Varden, Salvelinus malma (Smm-17, Smm-21, Smm-22, Smm-24 ; Crane et al., 2004) and rainbow trout, Oncorhynchus mykiss (OMM1105; Rexroad et al., 2002) (Appendix B Table. B.3). With the exception of Smm-21, all markers were polymorphic (non-diagnostic) for Arctic char and Dolly Varden.  52  The locus, Smm-21 was, however, monomorphic and diagnostic in Arctic char; all fish were homozygous for a 110 base pair allele. By contrast, Dolly Varden were polymorphic at Smm-21 with alleles that ranged in size from 120-132.  3.2.3  Statistical analysis I used MICRO-CHECKER (van Oosterhout et al. 2004) to check the microsatellite data for errors in scoring such as stuttering and null alleles which can compromise subsequent analyses. Next, I used the program FSTAT ver 2.9.3 (Goudet 2001) and ARLEQUIN ver 3.5 (Excoffier & Lischer 2010) to generate the basic descriptive statistics of sample size (N), number of alleles (NA), allelic richness (AR), observed (HO) and expected (HE) heterozygosity. Tests for departures from Hardy-Weinberg Equilibrium (HWE) for each locus-population combination were performed with GENEPOP ver 4.2 (Raymond & Rousset 1995) using an exact test in which probability values were determined using a Markov chain method (P) controlling for multiple tests.   3.2.4 Population structure  To test for population subdivision for all samples pooled within Lake Aleknagik and Lake Nerka, I ran simulations of K = 1 to K = 10, repeated five times in the Bayesian program STRUCTURE  for each sampling year (2012, 2013) separately to account for inter-annual differences and visualized the data using DISTRUCT (Rosenberg 2004). To infer support for the most probable number of subpopulations for each sampling year, K (Evanno et al. 2005) was calculated across multiple runs of STRUCTURE using STRUCTURE HARVESTER (Earl & vonHoldt 2012). To assign individuals as Arctic char or Dolly Varden, an ad hoc approximation of species identity was used where fish whose Q-value were ≥ 0.95 (i.e. at least 95% of the genome suggestive of Dolly Varden) were classified as Dolly Varden and fish whose Q-values were ≤ 0.05 were classified as Arctic char. Any fish that had Q-values ≥ 0.05 but ≤ 0.95 were noted as possible hybrids. 53   3.2.5 Levels of hybridization  To determine the level of hybridization and the genetic classification of hybrids present, I used the programs STRUCTURE ver. 2.3.4 (Pritchard et al. 2000) and NEWHYBRIDS vers.1.1 Beta 3 (Anderson & Thompson 2002). For both programs, I used genotypic data from all 13 microsatellite loci for each individual sampled and analyzed from Lake Aleknagik and Lake Nerka. I used STRUCTURE to estimate an individual’s probability of belonging to a one of two populations (K= 2, i.e., Arctic char and Dolly Varden – see Results) while minimizing departures from HWE and linkage disequilibrium within groups (Evanno et al. 2005). The STRUCTURE analyses were conducted using a burnin-period of 50,000 Markov Chain Monte-Carlo (MCMC) iterations proceeded by an additional 450,000 steps, replicated five times to verify consistency across runs (Taylor et al. 2008). I then used STRUCTURE to estimate the admixture proportion value (Q) and posterior probability intervals for each fish. I defined QDV as the proportion of the genome estimated to originate from Dolly Varden. Thus, to establish general cutoff values for the two species I defined pure Dolly Varden as individuals with QDV values > 0.95 and pure Arctic char as individuals with QDV values < 0.05. Next, to generate a range of QDV-values that would be more indicative of hybrid individuals, I used the program HYBRIDLAB (Nielsen et al. 2006) to create 100 simulated F1, F2, and backcrossed hybrids generated from the random selection of alleles from different loci in non-admixed allopatric populations of Arctic char and Dolly Varden. The allopatric populations were then designated as priors or “learning samples” by implementing the using the USEPOPINFO model combined with simulated hybrids and char from Lake Aleknagik and Lake Nerka through five replicated analyses in STRUCTURE (Pritchard et al. 2000). I took the average area between the upper and lower limits of possible QDV-values for the simulated hybrids across the five replicated analyses to create a zone of hybridity. Admixed individuals from the Wood River basin whose QDV-values fell inside this range were classified as putative hybrids and were used to generate the level of hybridization for each lake. 54  To account for low sample sizes for allopatric populations, a second zone of hybridity was generated using 50 Arctic char from Lake Aleknagik and Lake Nerka with QDV values closest to 0 (QDV < 0.05, pure Arctic char) and 50 Dolly Varden with QDV values closest to 1.0 (QDV ≥ 0.88, due to limited samples) using the same methods as above. Both zones of hybridity were then compared to the range generated by Taylor et al. (2008) for Arctic char and Dolly Varden in southwestern Alaska.  I next used NEWHYBRIDS separately on Lake Aleknagik and Lake Nerka samples to determine an individual’s probability of belonging to one of six genetic classes: pure Arctic char, pure Dolly Varden, F1 hybrid, F2 hybrid, backcross with Arctic char, backcross with Dolly Varden. The analysis by NEWHYBRIDS uses allelic differences between hybridizing species in a Bayesian model-based clustering framework with MCMC to estimate the posterior probability of an individual belonging to distinct hybrid classes (Anderson & Thompson 2002). The same allopatric populations used in the STRUCTURE analysis were also run in NEWHYBRIDS amongst the samples as non-admixed reference populations for a minimum of 450,000 MCMC steps. To independently assess the use of a K= 2 in STRUCTURE and two parental species in NEWHYBRIDS (i.e., Arctic char and Dolly Varden), a factorial correspondence analysis (FCA) in GENETIX (Belkhir et al. 2001) was conducted on allele frequencies of Lake Aleknagik and Lake Nerka char alone and with allopatric populations of Arctic char from the Canadian Arctic and Dolly Varden from the Egegik fishing district located south of Dillingham and Bristol Bay, Alaska at the outflow of Becharof Lake. Allopatric populations were assigned to species using a combination of morphology and diagnostic microsatellite markers.  3.2.6 Spatial distribution across microhabitats  Genotyped samples of Arctic char, Dolly Varden and hybrids were georeferenced in order to map the spatial distribution of the species across microhabitats.  55  Global Positioning System (GPS) coordinates of sampling sites were mapped using online resources (http://boulter.com/gps). Variation in spatial distribution was assessed by analyzing sampling sites and sampling years separately. To determine the distribution of age classes between lake and stream sites for each char species in Lake Aleknagik, a size-frequency histogram generated from fork length and frequency of fish present in each sampling site was used to identify different age classes from size-frequency distributions. Species assignment was based upon admixture (QDV)-values and fish that fell into the generated zone of hybridity (see section 3.3.3 for details) were excluded.   3.3 Results 3.3.1 Genetic analysis  A total of 725 Arctic char and Dolly Varden were genotyped at 44 sites across Lake Aleknagik and Lake Nerka. MICRO-CHECKER did not find evidence of alleles failing to resolve owing to scoring error or to large allele dropout in any of the samples. Evidence for a non-amplifying (“null”) allele was found in some samples for locus (Sco216) although removal of this locus did not significantly alter results. When samples were partitioned into Dolly Varden and Arctic char (see below), all loci were polymorphic except for Smm-21 in Arctic char, which was monomorphic and diagnostic for each species (Appendix B Table. B.4). Arctic char and Dolly Varden had approximately equal average allelic richness (17.8 vs 17.3) when pooled across the lakes and across sites within Lake Aleknagik. Dolly Varden showed the highest number of alleles for a single locus (39 at locus OtsG83b followed by 38 at locus Sco200). When examining samples pooled across sites within each lake and for Arctic char and Dolly Varden separately (separation of species being based on STRUCTURE results, seen below), twenty of fifty tests for deviations from Hardy Weinberg Equilibrium (HWE; two samples x 13 loci for Arctic char plus two samples x 12 loci for Dolly Varden) were significant at P ≤ 0.011. The majority of deviations from HWE originated within Dolly Varden with ten out of the total 20 deviations found in Lake Aleknagik Dolly Varden samples and six from Lake Nerka Dolly Varden. 56   3.3.2 Population structure Analysis by STRUCTURE indicated that the most likely number of genetic populations across both lake and stream sites for both lakes in 2012 was K= 2 (Fig. 3.2 A), as evidenced by a mean K of 1,675 vs. 1.1 for K = 3 as the next most likely model. When sampling effort in stream microhabitats was increased in 2013, a K= 3 model was best supported by the K method as evidenced by a mean K likelihood of 14.6 vs. 3.5 for K= 5 as the next most likely model.  For the initial ad hoc assignment of individuals to species, 565 fish (78% of samples) had QDV values ≤ 0.05 and were classified as “pure” Arctic char while another 148 fish (20%) had QDV values ≥ 0.95 and were classified as “pure” Dolly Varden. All individuals classified as Arctic char by this method were homozygous for the diagnostic Arctic char locus Smm-21 (allele 110). Likewise, all individuals classified as Dolly Varden by this method were polymorphic at this same locus (ranging from 120-132) and did not possess the diagnostic 110 allele. A total of 12 fish (2%) of samples fell in between the range of 0.05 ≥ QDV ≤ 0.95. Of the 12 fish in this middle range, five samples had the diagnostic alleles indicative of Arctic char and QDV values ranging from 0.072–0.216 while the other seven samples were polymorphic at the Smm-21 locus and had QDV values ranging from 0.493–0.939.  The results from the FCA analysis using known allopatric populations of Arctic char and Dolly Varden and the examination of the diagnostic locus Smm-21 indicated that the two populations identified both in Lake Aleknagik and Lake Nerka in 2012 were Arctic char and Dolly Varden (Appendix B Fig. B.1). The numerically dominant genetic cluster in Lake Aleknagik was taken to represent Arctic char as those individuals were homozygous at the diagnostic Arctic char locus (Smm-21) and grouped with known allopatric populations of Arctic char in the FCA projection (Appendix B Fig. B.2). The second genetic cluster was numerically much smaller and grouped with allopatric Dolly Varden populations in the same FCA projection. Substantial differentiation as measured by pairwise FST was seen between the two 57  species in Lake Aleknagik (FST= 0.172) and in Lake Nerka (FST= 0.187; P < 0.001, Table 3.1). Similarly, for 2013, the K analysis supported a single population of Arctic char, but a two population model for Dolly Varden, with the Yako Creek sample suggesting the existence of two populations of Dolly Varden in that system (Fig. 3.2 B). In all analyses, including when species were run separately, only a single population of Arctic char was ever supported either within or between the two lakes.   3.3.3 Hybridization levels and genetic classification  Analysis of the dataset with STRUCTURE at K= 2 suggested a general absence of hybridization between the two discrete genetic clusters representing Arctic char and Dolly Varden individuals, respectively (Fig. 3.3). The FCA projections further supported the existence of two distinct genetic clusters with few intermediate individuals (Appendix B Fig. B.1-B.2). Simulated hybrids were generated from allopatric populations of Arctic char and Dolly Varden from Lake Aleknagik and Lake Nerka char with admixture values indicative of pure species. These simulated hybrids had QDV values that were ≤ 0.779 and ≥ 0.445 (using allopatric learning samples, Appendix B Table. B.3) and QDV ≤ 0.788 and ≥ 0.237, respectively (Lake Aleknagik and Lake Nerka learning samples, Appendix B Table. B.4).           58  K= 3 K= 2 QDV QDV           Figure 3.2: Results of STRUCTURE analysis visualized with DISTRUCT for sympatric Alaskan Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) assayed at 13 microsatellite DNA loci and sampled from A) Lake Aleknagik proper and streams and Lake Nerka stream and river sites in 2012 for K= 2; B) Yako Creek in Lake Aleknagik with support for two populations of Dolly Varden, Lake Aleknagik stream and Lake Nerka stream sites in 2013 for K= 3. Colour codes are: green = Arctic char, red = Dolly Varden, blue = additional population of Dolly Varden from Yako Creek as defined by admixture (Q) values.     Lake Aleknagik streams (n = 99) Nerka streams (n = 33) Yako Creek, Lake Aleknagik (n = 66)  K= 3 Lake Aleknagik proper      (n = 177)  Lake Aleknagik streams        (n = 154)  Lake Nerka streams                            (n = 196)  A) B) 59       Table 3.1:  Pairwise FST (θ) estimated by variation across 13 microsatellite DNA loci in sympatric Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from Lake Aleknagik and Lake Nerka, Alaska and allopatric populations of Arctic char and Dolly Varden from the Canadian Arctic and Egegik, Alaska. Arctic char = AC, Dolly Varden = DV. Values accompanied by asterisks are not significantly > 0 (P ≤ 0.0151; after adjustment for multiple simultaneous tests incorporating the false discovery rate procedure of Narum, 2006).   DV (Aleknagik) DV (Nerka) AC (Aleknagik) AC (Nerka) AC  (Canadian Arctic) DV (Nerka) 0.009* - - - - AC (Aleknagik) 0.172 0.187 - - - AC (Nerka) 0.172 0.187 0.007* - - AC (Canadian Arctic) 0.169 0.179 0.124 0.128 - DV (Egegik) 0.022 0.0163 0.156 0.149 0.181         60        Figure 3.3: Overall genetic classification of sympatric Alaskan Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) using 13 microsatellite markers from A) Lake Aleknagik (n = 496) and B) Lake Nerka (n = 229). Assignment in STRUCTURE was based on the probability of belonging to two a priori genetic groups (K= 2). Individuals are arranged according to decreasing QDV -values (i.e., the proportion of red in each line) and both sampling years, summer 2012 and 2013, are shown. Colour codes are: red = S. malma, green = S. alpinus.       A) B) QDV QDV DV AC DV AC 61  Small sample sizes for allopatric populations (< 25) likely constrained the lower limit of hybridity as the diversity in genotypes was restricted. As the range generated from char in the Wood River Lake system was most comparable to the range generated by Taylor et al. (2008) (QDV between 0.785 and 0.232) for the same system it was selected for use in the subsequent analyses. This range of QDV values was generated by constraining K= 2 in STRUCTURE, and in doing so, I found that the admixture coefficient for “pure” Arctic char (QDV) averaged 0.0049 (range 0.128-0.002) while the resultant admixture value for “pure” Dolly Varden (QDV) averaged 0.99 (range 0.812─0.998). The simulated hybrids had an average admixture value (QH) of 0.44 (range 0.237─0.788). Using these boundaries, I established a zone of hybridity expressed relative to the proportion of the Dolly Varden genome, represented by QDV, and considered fish with QDV values that fell within this range to be putative hybrids, fish with QDV values ≥ 0.788 to be Dolly Varden and fish with QDV values ≤ 0.237 to be Arctic char. The majority of fish fell within the parental ranges of Arctic char and Dolly Varden and only two fish (0.40%) from Lake Aleknagik and one fish (0.45%) from Lake Nerka had admixture values within the zone of hybridity (full summary in Table 3.2). These values did not change when samples were analyzed separately based on shared sampling year and site (e.g. same stream or lake site), and because of small numbers of putative hybrids, any tests of microhabitat differences affecting hybridization prevalence and occurrence were no longer deemed relevant. The three putative hybrids were collected from one lake (Bear Bay, site 5) and one stream site (Happy Creek, site 19) in Lake Aleknagik and one stream site (upper Lynx Creek, site 30) in Lake Nerka during the sampling period of 2012.       62      Table 3.2:  Summary of sampling locations, date, total number of fish genotyped (N), and proportion of samples assigned as Arctic char (Salvelinus alpinus) (AC), Dolly Varden (S. malma) (DV) and putative hybrids (HYB) in the Wood River Lake system, southwest Alaska. Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788.   Latitude and Longitude Date N Total AC Total DV Total HYB Lake Aleknagik  59.343301°N-158.806686°W July-August 2012 332 303 27 2  July-August 2013 164 77 87 0 Lake Nerka 59.559679°N -159.023666°W June-August 2012 196 173 22 1  July-August 2013 33 17 16 0 Wood River 59.262936°N -158.573313°W October 2012 8 0 8 0        63  Bayesian assignment of individuals into parental or hybrid (i.e., first- or later generation hybrids or backcrosses) genotypes with NEWHYBRIDS indicated a higher proportion of hybrids than estimated by STRUCTURE for Lake Aleknagik (7% vs. 0.4% by STRUCTURE) but was comparable for Lake Nerka (0% vs. 0.45% by STRUCTURE). A total of 76% of the Lake Aleknagik samples were classified as pure Arctic char, 17% as Dolly Varden, 0% as F1, 2% as F2, 0% as backcrosses with Arctic char and 5% as backcrosses with Dolly Varden using NEWHYBRIDS (Fig. 3.4). The fish assigned as backcrosses with Dolly Varden were scattered across stream habitats in Lake Aleknagik with most of the individuals originating from Yako Creek (2013 sites), Happy Creek (both years), Silversalmon Creek and Hansen Creek (2012). The F2 individuals were found in two stream sites (Happy Creek, site 19 and upper Bear Creek, site 10). Most (6/7) samples of Dolly Varden with spawning colouration collected from the Wood River were assigned to the F2 hybrid class. With the exception of one fish from Happy Creek, all F2 individuals possessed the Smm-21 alleles diagnostic for Dolly Varden. For Lake Nerka, all individuals were assigned to parental classes, with a total of 82.5% of samples classified as pure Arctic char and 17.5% as Dolly Varden.             64       Figure 3.4: Summary of NEWHYBRIDS results for sympatric Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from Lake Aleknagik into six genetic classes. AC = Arctic char, DV = Dolly Varden, Bx = backcross.         65  3.3.4 Spatial discreteness between species  The frequency and distribution of Arctic char (i.e., QDV < 0.05) in lake and stream habitats was widespread when compared to Dolly Varden. Arctic char were the only char species found in lake habitats at the time of sampling (Fig. 3.5), but could also be found in most streams both in Lake Aleknagik and Lake Nerka (Fig. 3.6─3.9). In streams, Arctic char typically occupied sites near the mouths of streams, but they could also be present in upper stream sites, located several kilometers upstream from the lake. The distribution of Dolly Varden was generally found in the middle to upper reaches of most streams although some streams were exclusively Dolly Varden (e.g., Site 44, Silversalmon Creek). In October, during the spawning period for both species, spawning Arctic char were collected off beaches at the outflow of large creeks on the northernmost region of the lake while mature Dolly Varden were found spawning in the Wood River to the south (C. Schwanke, ADF&G, personal communication). Stream resident Dolly Varden were also observed spawning in early-mid September at the outflow of Lynx Lake, located above Lynx Creek in Lake Nerka (D. Schindler, University of Washington, Seattle, personal communication). The distribution of the different age classes across stream and lake habitats differed for Arctic char and Dolly Varden in Lake Aleknagik (Fig. 3.10). Young-of-the-year (YOY) to mature ages classes of Dolly Varden were seen in stream habitats, but were not collected during the sampling period from any lake sites. Age 3+ Dolly Varden appeared to reach a maximum size of approximately 270 mm in fork length and displayed vibrant spotting patterns. Young-of-the-year to age 3+ or 4+ Arctic char were also present in the majority of stream microhabitats similar to Dolly Varden, but were also common at lake sites. The numerically dominant char species in the lake, Arctic char, was present at beach sites as YOY and again as 2+ age classes in deeper microhabitats in the lake.     66      Figure 3.5:    Proportion of genotypes defined by admixture (QDV) values corresponding to Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from within the lake proper and on beaches sites in Lake Aleknagik during summer 2012 (black = S. alpinus, white = S. malma). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788. A single putative hybrid was found at the starred site (Bear Bay, site 5).    * 67      Figure 3.6: Proportion of genotypes defined by admixture (QDV) values corresponding to Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from stream sites around the southern portion of Lake Aleknagik during summer 2012 (black = S. alpinus, white = S. malma). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788. A single putative hybrid was found at the starred site (Happy Creek, site 19).     * 68      Figure 3.7: Proportion of genotypes defined by admixture (QDV) values corresponding to Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from stream sites around the southern portion of Lake Aleknagik during summer 2013 (black = S. alpinus, white = S. malma). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788.    69      Figure 3.8:  Proportion of genotypes defined by admixture (QDV) values corresponding to Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from stream and river sites around Lake Nerka during summer 2012 (black = S. alpinus, white = S. malma). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788. A single putative hybrid was found at the starred site (Lynx Creek 4, site 30).    * 70      Figure 3.9: Proportion of genotypes defined by admixture (QDV) values corresponding to Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from stream sites around Lake Nerka during summer 2013 (black = S. alpinus, white = S. malma). Arctic char were defined as having QDV values of ≤ 0.237 and Dolly Varden were defined as having QDV ≥ 0.788 while hybrids, defined as having QDV values between 0.237 —0.788, were not found.     71     Figure 3.10: Size frequency histogram displaying approximate age classes with estimated age classes outlined by coloured boxes for Lake Aleknagik (n = 494) A) Dolly Varden (Salvelinus malma) in stream habitats with the exception of starred individuals collected from the Wood River in October B) Arctic char (S. alpinus) in stream habitats C) Arctic char in lake habitats. Age 0 represents young of the year fish and ′+' represents all older age classes combined. Dolly Varden were not present in the lake during the July-August sampling period. Arctic char were defined as having QDV values of ≤ 0.237 and Dolly Varden were defined as having QDV ≥ 0.788 while hybrids with QDV values between 0.237 —0.788 were excluded. A) B) C) * 72  3.4 Discussion  Quantification of the degree of natural hybridization between lineages allows us to evaluate the strength of reproductive and genetic barriers between lineages and define species boundaries and can further our understanding of processes that help maintain genetic distinctiveness in the presence of gene flow (Arnold 1992; Jiggins & Mallet 2000; Hagen & Taylor 2001). In many situations, parental species can maintain their genetic integrity in the face of gene flow (Hammar et al. 1991); however, the formation of hybrid zones with a variety of evolutionary outcomes (e.g. species collapse, reinforcement of reproductive isolation) can also occur (Barton 2001; Latch et al. 2011). Arctic char and Dolly Varden hybrids have been identified across their sympatric range from lakes in southwestern Alaska (McPhail 1961; Taylor et al. 2008), but the lack of a robust genetic survey across multiple lake and stream microhabitats limited the ability to properly address the hybridization issue between the species. As Arctic char and Dolly Varden are considered sister species and hybridization occurs frequently in char, I initially predicted that hybrids would be more prevalent in sympatric systems and that a hybrid zone between the species may exist in the Wood River basin due to intermediacy in normally discrete morphological counts (McPhail 1961; Taylor et al. 2008). In this chapter I have addressed the limitations of the previous studies in two southwestern Alaskan lakes and find evidence supporting much lower levels of hybridization between the two species than previously shown when using similar genetic methods to Taylor et al. (2008). Understanding the factors which underlie low levels of natural hybridization may help us to resolve important steps in the divergence of Arctic char and Dolly Varden as models in speciation.        73   3.4.1 Evidence for distinct gene pools between species  The microsatellite data revealed significant differentiation in pairwise FST between sympatric Arctic char and Dolly Varden samples from the Wood River Lake system and the degree of differentiation in sympatry was equivalent to, or higher than, divergence seen between either species and the and allopatric populations (i.e., between Dolly Varden in sympatry and allopatric Arctic char and vice versa). While Arctic char and Dolly Varden clearly separated into two distinct genetic clusters at K= 2, a model supporting two subpopulations of Dolly Varden was the most supported genetic structure in the 2013 samples from Lake Aleknagik (i.e., in Yako Creek). This result suggests that Dolly Varden may not disperse far from their presumably natal streams, but more sampling is required to test this idea rigorously. Spatial population genetic structure has also been observed in stream-resident Dolly Varden populations from tributaries in central Hokkaido, Japan and has been associated with the presence of physical barriers (Koizumi & Maekawa 2004; Koizumi 2011). Yako Creek in particular had high habitat complexity with abundance of fallen woody debris and boulders (Pess 2009) which could act as physical barriers isolating pockets of individuals. This has important implications for hybridization as Dolly Varden may be less likely to move from stream habitats to the spawning grounds of Arctic char on the lake beaches (see below). By contrast, across the two interconnected lakes, Arctic char populations had a higher degree of genetic homogeneity as evidenced by a population cluster of one; this suggests that there is extensive gene flow among all Arctic char from different areas both among stream and lake habitats.      74  3.4.2 Hybridization levels  My results strongly suggest that contemporary hybridization between Arctic char and Dolly Varden is very infrequent in the Wood River Lake system. First, without using any prior designation of species or providing any information collected from different microhabitats (e.g., same site in a particular stream), analyses in STRUCTURE supported the existence of two highly distinct genetic clusters identified as Arctic char and Dolly Varden and there are at least two known diagnostic microsatellite markers (Sfo18 is also diagnostic between the species – see Taylor et al. 2008). These distinct groups were further supported in the FCA projection as Arctic char and Dolly Varden from the Wood River Lake system clustered closely with their respective allopatric populations and separately from one another. Moreover, the FCA projection identified few, if any, obvious hybrid samples that fell into intermediate regions between the two clusters.  Using comparable methods to Taylor et al. (2008), my genetic survey and STRUCTURE analysis indicated that the level of hybridization between Arctic char and Dolly Varden in the Wood River Lake system is minimal (0.4%) and much lower than that reported by Taylor et al. (2008, 7%). Subsequent analyses with NEWHYBRIDS supported the presence of later-generation hybrids (up to 7% in Lake Aleknagik); however, the proportion of F1 and F2 generation hybrids in the Wood River Lake system was still minimal (2% in Lake Aleknagik and 0% in Lake Nerka). The lower level of hybridization reported in my study could be a result of increased sample size (N = 720 vs.105), additional microsatellite DNA markers used (13 loci vs. nine), a greater range of microhabitats sampled (20 discrete streams and various lake sites vs. six sites total) and more intensive top-to-bottom sub-sampling of streams in my study which should have facilitated a more representative sample of genetic diversity in Wood River Lake system. The difference observed in the level of hybridization between the two studies could also result from the five year gap between studies as temporal fluctuations in environmental conditions across years can influence hybridization levels or the fitness of hybrids (Arnold 1997; Chenuil et al. 2000).  75  Low hybridization levels between the species in the Wood River basin may be comparable to other nearby systems containing sympatric populations of Arctic char and Dolly Varden. For example, in the nearby Iliamna Lake system, located approximately 85 km east of the Kvichak River, Taylor et al. (2008) found no evidence of F1 hybrids between the species, but classified approximately 10 out of 65 individuals (15%) through admixture (Q) values as post-F1 hybrids. Similar to the findings of this study, the high number of putative later-generation hybrids can perhaps be explained by low sampling site diversity and sample size. Only two discrete habitats were sampled from Iliamna Lake, one a series of spring-fed ponds (Pedro Ponds) comprised exclusively of small-bodied Dolly Varden and a beach site from Iliamna Lake proper (Taylor et al. 2008). Given that Iliamna Lake is over 2,600 km2 and only a small proportion of the genetic diversity of char was previously analyzed (Taylor et al. 2008), I suggest that a more robust genetic survey across the wide range of microhabitats present within the Iliamna Lake system would produce similar results, i.e. low hybridization levels; to that seen in the Wood River Lake system.  In contrast to other sympatric char systems where hybridization is often a regular occurrence e.g., Arctic char x lake char (Hammar et al. 1989), Arctic char x brook trout (Hammar et al. 1991), Dolly Varden x bull trout (with 0-25% of individuals presumed to be of hybrid origin, Baxter et al. 1997, Redenbach & Taylor 2003), the level of hybridization between Arctic char and Dolly Varden in the Wood River Lake  system is unexpectedly low, especially considering their sister species status. While other sympatric char systems provide evidence of spatial overlap of species during breeding (e.g. bull trout spawning in upstream habitats where Dolly Varden predominate, Bustard & Royea 1995, but see Hagen & Taylor 2001), Arctic char and Dolly Varden appear to be longitudinally segregated along streams and presumably spawn in different habitats at slightly different times of the year. Given that no direct observations of hybridization between Arctic char and Dolly Varden currently exist (Reist & Sawatzky 2010), low levels of interspecific competition for spawning sites and pre-mating spatial segregation may be important aspects of pre-mating isolation between the two species; however, further work is needed to formally establish the 76  basis and degree of pre-mating reproductive isolation between the species in sympatry.  Despite assigning approximately 7% of samples as hybrids in Lake Aleknagik, Taylor et al. (2008) found striking evidence for a complete absence of heterozygous individuals at a diagnostic marker (Sfo18). Given that species-specific alleles in heterozygous fish have been observed in hybridized stock populations of brook trout and Arctic char (Gross et al. 2004), Taylor et al. (2008) speculated that the likely cause of this apparent lack of F1 individuals was rarity in hybridization events because the majority of hybrids had admixture values near the boundaries in the zone of hybridity and no hybrid had a QDV value of 0.5 which is suggestive of an F1 hybrid. They could not, however, rule out the possibility that this locus was under selection or closely linked to another locus which may bias their estimates of hybridization. To circumvent these unknowns, I used another putative diagnostic locus (Smm-21) which is polymorphic in Dolly Varden, but fixed in Arctic char (May-McNally, S.L. unpublished). The same absence of heterozygotes seen with Sfo18 was observed with Smm-21; therefore, it is even more unlikely that selection or linkage at diagnostic markers is underestimating current levels of hybridization. Moreover, a random sample from 2012-2013 analyzed at Sfo18 (n = 75) showed the same discreteness in species-specific alleles and lack of heterozygotes as seen in Taylor et al. (2008).  The STRUCTURE analyses did not find evidence of F1 hybrid individuals at any age class and little evidence for post-F1 hybrids in the two lakes (~0.4 % across the two lakes). The lack of F1 hybrids is further supported through analyses in NEWHYBRIDS. In general, hybrid zones between Dolly Varden and bull trout also show a relative paucity of F1 hybrids relative to other genotypic classes (e.g. Redenbach & Taylor 2003). Conversely, NEWHYBRIDS suggested that post-F1 hybrids may constitute up to 7% of the Lake Aleknagik population and that backcrossed hybrids are biased towards introgression with Dolly Varden. As NEWHYBRIDS has been shown to be more sensitive at detecting the correct proportion of hybrid classes present in a sample than STRUCTURE it could be detecting second or third generation hybrids that were classified as parental species 77  when using only STRUCTURE (Vähä & Primmer 2006). As the credible regions and error margins associated with each sample are accounted for differently in each program, many of the potential late-generation hybrids with borderline QDV-values nearing parental levels could have easily been misclassified by the STRUCTURE analysis, but detected by NEWHYBRIDS. Although the programs are not directly comparable (i.e., a 0.4% level of hybridization in STRUCTURE ≠ 7% level of hybridization in NEWHYBRIDS), the presence of advanced generation hybrids as suggested by my NEWHYBRIDS analyses could perhaps explain the overlap in morphology between species (e.g. gill rakers, pyloric caeca) observed by McPhail (1961) and Taylor et al. (2008) in Lake Aleknagik.  Taken together, it seems plausible that F1 hybrids are generated very rarely in this system, but signatures of their influence may be detected through backcrossing, as suggested by the presence of post-F1 individuals. This situation perhaps best fits the conceptual framework of Arnold’s (1997) “Evolutionary Novelty” model which posits that evolutionary stable lineages can arise when environment-dependent and environment-independent selection act concurrently on hybrid genotypes and when hybrids can exhibit equivalent or higher fitness than parental genotypes. The model was derived based on observations that under natural conditions F1 individuals are rarely formed in ecosystems without well-defined ecotones or disturbed sites that often act to promote sympatry and inter-breeding between parental species that are otherwise restricted and adapted to alternative habitats (Anderson 1948; Moore 1977). In other words, the presence of well-established habitat differences (lakes versus streams) has perhaps promoted the evolution of distinct habitat use and life history in Dolly Varden and Arctic char that limits the opportunities for hybridization especially in the absence of any obvious ecotones or disturbed habitats.  The relative rarity of Dolly Varden in the Wood River Lake system may also contribute to the low observed levels of hybridization, i.e., the level of hybridization is to some extent dependent on the relative density of parental species (see Wirtz 1999). Across the habitats surveyed, Dolly Varden were numerically under-represented in comparison to Arctic char as seen by the ~76% of samples genotyped from Lake Aleknagik that were classified as Arctic char. In Lake Nerka, 78  Arctic char were also the dominant species in the streams and rivers. The lower abundance of Dolly Varden in this system probably limits interspecific encounters during reproductive periods and thus may have an impact on the dynamics of hybridization and directionality of introgression (Hubbs 1955; Mayr 1963; Burgess et al. 2005; Lepais et al. 2009). Furthermore, through studies of extrinsic selection on hybrid zones, it has observed that phenotypically and physiologically intermediate hybrid individuals often occupy intermediate habitats across ecotones or clines (Arnold 1997; Fritsche & Kaltz 2000; Culumber et al. 2011). Although the few post-F1 generation hybrids that were observed were scattered throughout the Lake Aleknagik system, they could show highest fitness in currently unknown, and presumably less common, intermediate habitats (e.g., stream outflow sites near the lake). If high densities of hybrid offspring were present in Lake Aleknagik, the carrying capacity of these intermediate habitats would perhaps be exceeded and extrinsic selection could act against hybrid individuals in parental environments. When the density of hybrids is low and intermediate habitat sites are available, perhaps selection against hybrids is weaker and they can better persist throughout their lifecycle. If hybrid survival is high when hybrids are rare, an evolutionary stable situation that promotes low levels of gene flow between two otherwise discrete species may develop and be potentially long-standing (Arnold 1997).   3.4.3 Spatial segregation across lake-stream microhabitats Arctic char are the predominant and more widely distributed char species across the majority of microhabitats within Lake Aleknagik and Lake Nerka. During the summer sampling period, Arctic char were the only species collected from the lake, with younger (age 0) age classes inhabiting nearshore beach habitats and older individuals collected from a variety of depths. My spatial distribution mapping, however, also supports the presence of Arctic char in some stream habitats. The species is generally abundant from the mouth to middle regions of streams, but was also occasionally present at the top-most reaches of streams located several kilometers from the mouth.  79  From age class distributions, it appears that Arctic char migrate into stream habitats from the lake at a young life stage, demonstrated by a dramatic absence of age 1 Arctic char from any lake microhabitat and a high abundance of this year class near the lower reaches of streams. Although age 1 Arctic char might have been underrepresented in the gill nets and beach seine collections, it seems more plausible that they move into the small streams to exploit available invertebrate resources while still at an appropriate size for the shallow depths. Once certain size thresholds are met, estimated to be around ages 2+ or 3+ due to the sharp decline in stream habitat abundance at these age classes, Arctic char may migrate back into the lake (or deep rivers) where they remain at least until the time when large numbers of spawning sockeye salmon (Oncorhynchus nerka) enter streams. Large Arctic char were observed swimming back into the mouths of larger streams, perhaps as a transient habitat shift, in July-August to capitalize on pulses of sockeye salmon eggs that become available when the salmon begin spawning.   At the time of sampling, Dolly Varden were only found in stream microhabitats and were in highest abundance in the upper reaches of streams, with the exception of a few Dolly Varden dominated streams (e.g., Silversalmon creek). A complete range of age classes, from young-of-the-year (YOY) fish to those exhibiting spotting colours characteristic of maturing fish were also present in stream microhabitats. When plentiful, Dolly Varden could be found feeding near the mouths of streams on sockeye salmon eggs, a behaviour which may relate to the daily upstream-downstream migration seen with juvenile coho salmon (Oncorhynchus kisutch) in Lake Aleknagik (Armstrong & Schindler 2013). Small YOY Dolly Varden were also observed in upper reaches of certain streams such as Yako Creek which suggests that adult spawning areas extend well upstream. Anadromous Dolly Varden were excluded from this study due to their absence in the system during summer periods, but are presumed to spawn in the Wood River or near the stream mouths of Lake Aleknagik’s creeks such as Yako and Happy creeks (C. Schwanke, ADF&G, personal communication). Thus, it seems possible that both young anadromous and small-bodied non-anadromous Dolly Varden co-exist in similar stream microhabitats until they adopt their alternative life histories. 80  Arctic char and Dolly Varden overlap spatially in streams, but my results indicate that this sympatry occurs primarily with Dolly Varden and immature Arctic char or in periods when sockeye salmon eggs are plentiful. Previous work in the Karluk River system, located just southeast of my study area, further supports my findings that Dolly Varden prefer stream habitats while Arctic char favour the lake (Delacy & Morton 1943). Consequently, I suggest that well-established niche segregation limits the potential for hybridization events to occur during spawning periods in Dolly Varden and Arctic char and in sympatric char more generally. The spatial distribution of smaller Dolly Varden in upper stream reaches and larger Arctic char in the remaining available niche space mirrors the distribution observed with other sympatric char such as Dolly Varden and bull trout (Salvelinus confluentus). In northwestern British Columbia where they occur in sympatry, Dolly Varden are apparently limited to stream habitats while bull trout are adfluvial; they spatially overlap with Dolly Varden in many tributaries as juveniles in their first year or two of life, but migrate to lakes as older juveniles to feed before returning to streams to spawn as adults (Hagen & Taylor 2001). Although bull trout and Dolly Varden do not appear to be adapted to alternative trophic or habitat resources while sympatric in streams (Hagen & Taylor 2001), the extent of such differentiation between Arctic char and Dolly Varden is unknown, but could contribute to resource driven ecological segregation in these fish. Preliminary diet analyses reveal that Arctic char have higher numbers of salmon eggs, a calorie rich food source, in their stomach than do similar sized Dolly Varden feeding in the same microhabitat (May-McNally, S.L., unpublished data) suggesting that Arctic char may be able to outcompete Dolly Varden in certain situations and environments. This finding could help explain the current distributions and the predominance of Arctic char in the Wood River Lake system, but further work (e.g. detailed diet analysis, stable isotopes) is needed to examine the degree of trophic niche overlap and resource partitioning between the species.    Overall, this chapter of my thesis has demonstrated that Dolly Varden and Arctic are highly distinct genetically in the Wood River Lake system. Despite their sympatry, hybridization between the two species is very rare and a high degree of 81  spatial segregation occurs between them at various life stages suggesting that pre-mating barriers to interbreeding are important in this system.   82  Chapter 4:                                                                                                  General conclusions  4.1 Summary of findings  My thesis produced two main findings: 1)  I provided strong support for three genetically distinct sympatric Arctic char (Salvelinus alpinus) morphs (‘large’, ‘medium’, and ‘small’–bodied char) from a southwestern Alaskan lake showing substantial FST (up to 0.09) and differences in admixture proportions across forms. Significant associations between genetic admixture proportions and gill raker counts among body-size morphs (r = -0.73, P < 0.001) and growth trajectories further support the idea that resource polymorphism may promote genetic divergence in Holarctic populations of Arctic char.  2)  My work also found that the level of natural hybridization between Arctic char and Dolly Varden (Salvelinus malma) across multiple microhabitats in two southwestern Alaskan lakes is much lower than estimated from earlier genetic surveys. The percentage of recent hybrid genotypes in the population was found to be less than 0.5% across lakes through STRUCTURE analyses, but up to 7% of the samples may represent late-generation backcrosses as suggested through analyses by NEWHYBRIDS. I also found that Dolly Varden char are concentrated in stream habitats while Arctic char dominated lake habitats and downstream portions of streams. Pre-mating spatial segregation in discrete habitats (lake vs. streams) may contribute to the low levels of contemporary hybridization that I have documented.    83  4.2  Defining a species The diversity of Arctic char ecophenotypes and morphs has led to taxonomic uncertainty within the group and has been summarized as the ‘char problem’ (Nordeng 1983; Klemetsen 2010). Some authors have suggested that many of these forms are distinct species, as seen with the up to 32 taxa recognized in Europe (Kottelat & Freyhof 2007), while others promote the idea of a single species ‘complex’ which encompasses all the possible variants of Arctic char (McPhail 1961; Reist et al. 2013). The morphs of Lower Tazimina Lake showed substantial ecological differences; they feed at different trophic levels and differ in colouration and size at maturity (Woods et al. 2013). Genetic differentiation also exists among the morphs despite evidence for some gene flow among them and was correlated with morphological trait differences (e.g., gill rakers) associated with divergence in trophic ecology (May-McNally et al. 2014) Although there is no evidence of behavioural reproductive isolation between large and small forms and there does seem to be some gene flow between them, they do exhibit substantial ecological and genetic divergence (Woods et al. 2013; May-McNally et al. 2014). The question of when these ecological and genetic differences “cross the line” to become species-level differences is important topic in evolutionary biology. This issue is even more apparent in some other systems where Arctic char morphs show complete reproductive isolation in sympatry (e.g., Westgaard et al. 2004). Reist (2013) argued that Arctic char is probably best described as a species ‘complex’ given that the various forms reflect contemporary-scale differentiation caused by local habitat differences, morphological and diet driven trophic polymorphism and not deeper evolutionary-scale differentiation driven by large-scale geomorphological events such as glaciation. This characterization seems to apply to the Lower Tazimina Lake and Lake Hazen (Nunavut) morphs which showed a lack of complete reproductive isolation, but exhibited morphological differences and trophic polymorphisms. Further, trophic differentiation in Arctic char may be possible in relatively short evolutionary time-frames, as suggested by translocation experiments of lacustrine populations in Maine where Arctic char 84  shifted from their original food source to a novel diet across only ~ 25 years or about 5-6 generations (Michaud et al. 2008). In addition to changes in diet, the Arctic char accumulated significant phenotypic differences suggested to have arisen as a result of this trophic shift (Michaud et al. 2008). Rapid phenotypic change accompanying trophic specialization is not unique to char and has also been observed in Galapagos finches (Geospiza spp.) which can develop substantial phenotypic variability and divergence in resource use within a few generations depending on food availability (Grant & Grant 2006). Given the estimated recent time of deglaciation in the Lower Lake Tazimina area (~12 000 ─ 15 000 BP) (Stilwell & Kaufman 1996; Ramstad et al. 2004) and lack of complete reproductive isolation, the Lower Tazimina Lake morphs are probably best associated with local or contemporary-scale differentiation and as variants within the Arctic char species ‘complex’. This argument, however, becomes more complicated in systems with morphs such as the ‘dark’ and ‘pale’ sympatric forms of Arctic char in Gander Lake, Newfoundland, and the two morphs in Lake Aigneau, Québec, which represent populations that have diverged to a much deeper level (Reist et al. 1995; Gomez-Uchida et al. 2008; Power et al. 2009; Arbour et al. 2011; May-McNally et al. 2014).  Understanding the nature of species and their interrelationships is central to evolutionary biology, but is complicated in char and many other taxa by hybridization and introgression even between what are thought to be relatively divergent species. The status of Arctic char and Dolly Varden as distinct biological species has only recently become better supported with genetic data (Taylor et al. 2008). My study in the Wood River Lake system of Alaska provided a more robust ‘acid test’ of the biological species status of Arctic char and Dolly Varden in sympatry. The data presented in Chapter 3, in addition to the work of Taylor et al. (2008), supports the substantial genetic distinctiveness of these taxa at microsatellite DNA loci and established that the species only rarely hybridize. The general lack of hybridization between Dolly Varden and Arctic char in the Wood River Lake system is consistent with the observation that the species appear to spawn in different areas of the Aleknagik Lake watershed (Arctic char on beaches, Dolly Varden in rivers and streams) even though they clearly overlap in stream habitats as juveniles (Chapter 85  3). The occasional gene flow between Dolly Varden and Arctic char is likely facilitated by infrequent species encounters during spawning, perhaps to low occurrences of smaller-bodied ‘sneaker’ males as seen with sympatric bull trout and Dolly Varden in northwestern British Columbia (Hagen & Taylor 2001). Opportunistic sneaker behaviour can be a common evolutionary stable strategy across many species of salmonids (Hendry & Stearns 2003). The formation of what I inferred to be some post-F1 individuals also suggests that F1 hybrids are viable and that no post-zygotic intrinsic genomic incompatibilities exist between Arctic char and Dolly Varden (Jiggins & Mallet 2000). Following strict adherence to the original biological species concept (BSC, Mayr 1942), incomplete reproductive isolation and genomic congruence would suggest that Arctic char and Dolly Varden represent a single taxon. Recent work, however, suggests that natural hybridization can be transient and result in the reinforcement of pre-mating isolating behaviours (Arnold 1997; Gow et al. 2006; Hausdorf 2011). Substantial, but not necessarily complete, reproductive isolation between sympatric lineages may signify that the majority of pre- and post-mating barriers are already in place. Given environmental stability, it is expected that reinforcement through hybridization would lead to the finalization of mating barriers (Servedio & Noor 2003; Mallet 2005). It therefore seems most appropriate for taxonomic classification to follow a modified BSC as proposed by Coyne and Orr (2004) and designate evolutionary-scale lineages that border on the edge of complete reproductive isolation, such as Dolly Varden and Arctic char, as valid species (Hausdorf 2011).    4.3  Implications for conservation of char diversity Climate change, in combination with direct human influence on ecosystems, e.g., habitat destruction, pollution, introduced species, and over-fishing, may result in novel selection pressures on many northern fishes potentially resulting in a loss of species diversity (Maitland 1995; Hoffmann & Sgro 2011; Reist et al. 2013). Arctic char and Dolly Varden may be particularly at risk from climate change due to the combination of a highly northern range (especially for Arctic char), long lifespans of 86  up to 40 years for Arctic char and 15 years for Dolly Varden, and old ages at maturation (Hendry & Stearns 2003; Berg et al. 2010). Under most models of climatic change, increased average water temperatures and longer seasonal periods are expected in the coming decades (Minns & Moore 1992; Reist et al. 2006; Hoffmann & Sgro 2011). Cold-adapted fishes with low resistance to warm temperatures such as Arctic char may not be capable of tolerating thermal stress in conjunction with increasing anthropogenic perturbation from northern development (Johnson 1980; Baroudy & Elliott 1994; Reist et al. 2013). Further, Arctic char populations balancing multiple stressors may be even more susceptible to the spread of parasites and diseases (Reist et al. 2006). As Arctic char are a common subsistence species for Aboriginal peoples in many northern regions including Alaska and Dolly Varden can be harvested year-round for subsistence or through commercial fisheries, an increase in parasite and disease load may have a negative impact on fish quality for consumption (Armstrong & Morrow 1980; Johnson 1980; Gordeeva et al. 2010).  Climate change may be especially critical in fragile Arctic and sub-Arctic lacustrine systems containing landlocked or sympatric morphotypes of Arctic char. These typically species-poor ecosystems may be especially impacted by the reduction or loss of species incapable of handing or dispersing from the changing thermal environment (Reist et al. 2013). Reduction in relative species abundance and increased seasonal length of may impact food web dynamics and composition of other species and prey abundance at different tropic levels (Reist et al. 2006, 2013). The Arctic char morphs present in Lower Tazimina Lake consumed discrete diets (Woods et al. 2013) and resource competition between the large and small forms is likely to be impacted if conditions do not remain stable. Additionally, Arctic char may be spatially limited to deeper, cooler regions of the lake (predicted to be as deep as >30 m) if average water temperatures continue to rise (Power et al. 2012). Currently in Lower Tazimina Lake, the large and small morphs segregate between the limnetic and benthic regions of the lake as predicted through stable isotope differences (Woods et al. 2013). If the uppermost regions of the lake (in particular the shallow limnetic zone) increase in temperature the small morph may be 87  displaced and potentially lost. As these morphs represent a unique component of the genetic diversity of Arctic char present in Alaska, protection from additional human-induced stressors that might exacerbate the effects of climate change and monitoring of these populations in response to increased warming may be essential for their survival. Environmental changes could also result in increased gene flow amongst morphs and change the genetic structure of the complex in Lower Tazimina Lake (Taylor et al. 2006; Vonlanthen et al. 2012). For sympatric populations of Arctic char and Dolly Varden in Alaska, climate change may result in alterations to the frequency and therefore the pattern of hybridization (Hoffmann & Sgro 2011). Disturbance of stable climate regimes could cause the balance between the species to shift and may result in changes in ecological selection against F1 hybrids (Gilman & Behm 2011). Given that disturbed environments are often more favourable for hybridization, the survival and viability of F1 hybrids may be facilitated through climate change (Seehausen 2006; Seehausen et al. 2008). The extreme situation would be the breakdown of pre-mating isolating barriers between the two species, potentially leading to the formation of a hybrid swarm and substantial genetic admixture between the species i.e., ‘speciation in reverse’ as observed with sympatric pairs of threespine sticklebacks (Gasterosteus aculeatus) of Enos Lake on Vancouver Island (Taylor et al. 2006) and Lake Victoria cichlids (Seehausen et al. 1997, 2008). Although the stickleback and cichlid divergences are probably less than 15,000 years old while Dolly Varden and Arctic char probably diverged from one another in pre-Wisconsinan glaciation times (i.e., ~ 1 million years ago, Grewe et al., 1990), McPhail (1961) proposed that pre-mating isolation is also central to the genetic discreteness of Dolly Varden and Arctic char in sympatry. Some breakdown of spatial mating isolation between Arctic char and Dolly Varden may be possible if warming of once thermally suitable spawning grounds leads to a reduction of available spawning sites in the lake, increasing the odds of interspecific interactions during mating periods if the streams and rivers become more thermally suitable for Arctic char. In addition to spatial shifts in spawning sites, warmer and more sporadic lake and ocean (in terms of anadromous Dolly Varden) temperatures could affect the timing and success of spawning events (Mieszkowska 88  et al. 2006; Sundby & Nakken 2008). Lake Aleknagik Arctic char and Dolly Varden spawn at slightly different, but nevertheless overlapping, times in the autumn (September-October for Arctic char and October-November for Dolly Varden, C. Schwanke, ADF&G, personal communication) and changes in climate could result in a higher degree of overlap and thus increase the potential for hybridization. To conserve the most char genetic diversity, continual monitoring of the Arctic char and Dolly Varden populations of the Wood River Lake system would be required to assess the degree of impact from climate warming. Reducing additional human induced habitat destruction may also help to mitigate any effects of climate change by limiting additional stressors that may act alone or synergistically to promote a species reversal for sympatric populations (Hubbs 1955; Taylor et al. 2006; Vonlanthen et al. 2012).  4.4  Concluding thoughts and future directions  My research builds on previous work surrounding sympatric Arctic char morphs and sympatric populations of Arctic char and Dolly Varden from lakes in southwestern Alaska and has firmly established that the two char are strikingly genetically distinct in sympatry. I have also provided better information on the spatial segregation of the two char that may contribute to their high degree of inferred reproductive isolation. Taken together, my work contributes to a better understanding of the evolutionary history of Arctic char and it fits broadly into a continuum of evolutionary scenarios from the origin of resource polymorphism and divergence of sympatric forms to species-level interactions. Further, as Arctic char are an important food and recreational source for many people in the north, understanding the evolutionary processes underlying adaptive divergence may facilitate predictions about how they will respond to habitat loss or climate change; knowledge that can result in more nuanced and detailed conservation strategies.   89  4.4.1 Genetic basis of divergence  Using the recent advancements in genome-level technology, the Lower Tazimina Lake morphs may represent a system in which to further study genetic changes surrounding the early stages of intraspecific sympatric ecological divergence. For the sympatric Arctic char morphs, it remains unknown whether differences in particular gene families also exist between the large, medium and small forms─ knowledge of which could provide important clues about the type of genetic change which may have occurred during the first stages of adaptive divergence. Genetic differences between intraspecific morphs associated with the earliest stages of divergence could be identified given that their genomes have not yet been entirely confounded by further genetic differences associated with complete divergence which can remove so-called genomic “islands of speciation” (Bernatchez et al. 2010). The Lower Tazimina Lake morphs therefore represent ideal organisms in which to conduct a transcriptome analysis targeting functional and regulatory genes that differ between the two or potentially three growth forms (Renaut et al. 2010). By taking samples of various tissues from juvenile and mature large, medium and small-bodied morphs in Lower Tazimina Lake, the extracts of total RNA could be reverse-transcribed to create cDNA libraries, which in turn, could be used to probe a 16,006 cDNA microarray originally designed for the Atlantic salmon (Salmo salar) ─ a microarray proven to be a successful substitute for other Salmonidae species without a fully sequenced genome such as lake whitefish (e.g., Coregonus clupeaformis,  Bernatchez et al. 2010). Using the information from the microchip array one could test for differences in gene expression (up-regulation or parallel expression) between large, medium and small morphs and contrast age and tissue type. This could identify genes or gene families important in the early stages of divergence and whether they are associated with known ecological differences between alternative Arctic char growth forms.   90  4.4.2 Basis of reproductive isolation: ecology and life history By conducting a robust genetic survey of sympatric populations of Arctic char and Dolly Varden to better estimate the degree of admixture, my data clearly support the hypothesis that hybridization between the species is rare (given by the complete lack of F1 hybrids) in the Wood River Lake system and that they remain highly distinct across various alternative microhabitats. This result supports previous, less extensive work by Taylor et al. (2008) and the separate biological species designation of Arctic char and Dolly Varden. These data indicate that there is simply no basis for the contention that the separate species status of Dolly Varden is “questionable” based on the sharing of some mtDNA haplotypes as suggested by Brunner et al. (2001).  With the addition of genome-wide genetic assays, a more detailed analysis of the level of hybridization, admixture and population structure would be possible. Although the species show some degree of spatial segregation, fine-scale spatial mapping of individuals through GIS and/or PIT tagging could be used to access the degree of niche and resource overlap. As the movement and diet of older Arctic char (> 300 mm) in Lake Aleknagik has been previously studied (McBride 1980) and it was found that the species largely capitalizes on sockeye salmon smolts, it would be informative to analyze diets across both lake and stream habitats where the two species co-exist and at younger life stages when the gape size of their mouths are too small to consume smolts. Interestingly, a preliminary diet analysis of the species in streams has suggested that Arctic char and Dolly Varden consume different amounts of sockeye salmon eggs (with Arctic char appearing to consume more) which could be related to differences in gape sizes or competitive interactions as seen other studies of char (Baxter et al. 1997; Hagen & Taylor 2001). Competitive exclusion may help to explain the current spatial distribution of char in Lake Aleknagik, where Dolly Varden are generally restricted to upstream areas of streams and Arctic char to lakes, and why the prevalence of Dolly Varden is lower than Arctic char in certain stream habitats during the summer months.  91  My results, in combination with previous work by McPhail (1961), suggest that Arctic char and Dolly Varden of the Wood River Lake system possess alternative life history patterns. In Alaska, Arctic char are presumed to be non-anadromous, lake-resident fish (Johnson 1980; Denton et al. 2010), and Dolly Varden are presumed to show both anadromous and non-anadromous life histories in Lake Aleknagik (C. Schwanke, ADF&G, personal communication), but these characterizations have not yet been explicitly tested in Wood River Lake system char using otolith chemistry analyses (Limburg 1995; Kraus & Secor 2004; Lowe et al. 2010).  Using genetic methods, my results agree with and extend previous work suggesting that the evolution of alternative sympatric Arctic char ecophenotypes can be associated with, and may indeed promote, reproductive isolation as a component of ecological speciation. Additionally, I found contemporary levels of natural hybridization between sympatric Arctic char and Dolly Varden to be minimal, likely reflecting the evolution of strong pre-mating barriers to reproduction. In conclusion, my thesis highlights the complex evolutionary history and genetic diversity of Arctic char in Alaska - an understudied portion of its range.        92  References Aboim MA, Mavárez J, Bernatchez L, Coelho MM (2010) Introgressive hybridization between two Iberian endemic cyprinid fish: a comparison between two independent hybrid zones. Journal of Evolutionary Biology, 23, 817–28. Adams CE, Fraser D, Wilson AJ et al. (2007) Patterns of phenotypic and genetic variability show hidden diversity in Scottish Arctic charr. In: Ecology of Freshwater Fish , pp. 78–86. Adams CE, Huntingford FA (2002a) The functional significance of inherited differences in feeding morphology in a sympatric polymorphic population of Arctic charr. Evolutionary Ecology, 16, 15–25. Adams CE, Huntingford FA (2002b) Inherited differences in head allometry in polymorphic Arctic charr from Loch Rannoch, Scotland. Journal of Fish Biology, 60, 515–520. Albert V, Jónsson B, Bernatchez L (2006) Natural hybrids in Atlantic eels (Anguilla anguilla, A. rostrata): evidence for successful reproduction and fluctuating abundance in space and time. Molecular Ecology, 15, 1903–16. Allendorf FW, Leary RF, Spruell P, Wenburg JK (2001) The problems with hybrids: setting conservation guidelines. Trends in Ecology & Evolution, 16, 613–622. Allendorf FW, Waples RS (1996) Conservation and genetics of salmonid fishes. In: Avise J.C. and Hamrick J.L. (eds.), Conservation Genetics: Case Histories from Nature. Chapman and Hall, New York, USA. Anderson EC (1948) Hybridization of the habitat. Evolution, 2, 1–9. Anderson EC, Thompson EA (2002) A Model-Based Method for Identifying Species Hybrids Using Multilocus Genetic Data. Genetics, 160, 1217–29. Arbour JH, Hardie DC, Hutchings JA (2011) Morphometric and genetic analyses of two sympatric morphs of Arctic char (Salvelinus alpinus) in the Canadian High Arctic. Canadian Journal of Zoology, 89, 19–30. Armstrong RH, Morrow JE (1980) The Dolly Varden charr, Salvelinus malma. In: Charrs: Salmonid fishes of the genus Salvelinus, Vol. 1 (ed. Balon E). Armstrong JB, Schindler DE (2013) Going with the Flow: Spatial Distributions of Juvenile Coho Salmon Track an Annually Shifting Mosaic of Water Temperature. Ecosystems, 16, 1429–1441. 93  Arnold ML (1992) Natural Hybridization as an Evolutionary Process. Annual Review of Ecology and Systematics, 23, 237–61. Arnold ML (1997) Natural Hybridization and Evolution. Oxford University Press, Cary, NC, USA. Arnold ML, Martin NH (2009) Adaptation by introgression. Journal of Biology, 8, 82. Avise JC (1994) Molecular Markers, Natural History and Evolution. Oxford University Press, New York. Baroudy E, Elliott JM (1994) The critical thermal limits for juveline Arctic charr Salvelinus alpinus. Journal of Fish Biology, 45, 1041–1053. Barton NH (2001) The role of hybridization in evolution. Molecular Ecology, 10, 551–68. Barton NH, Hewitt GM (1985) Analysis of hybrid zones. Annual Review of Ecology and Systematics, 16, 113–148. Barton NH, Hewitt GM (1989) Adaptation, speciation, and hybrid zones. Nature, 341, 497–503. Baxter JS, Taylor EB, Devlin RH, Hagen J, McPhail JD (1997) Evidence for natural hybridization between Dolly Varden (Salvelinus malma) and bull trout (Salvelinus confluentus) in a northcentral British Columbia watershed. Canadian Journal of Fisheries and Aquatic Sciences, 54, 421–29. Behnke RJ (1972) The systematics of salmonid fishes of recently glaciated lakes. Journal of the Fisheries Board of Canada, 29, 639–671. Belkhir K, Borsa P, Chikhi N, Raufaste N, Bonhomme F (2001) GENETIX 4.02, Logiciel Sous Windows TM Pour la Genetique des Populations. Laboratoire Genome, Populations, Interactions, CNRS UMR 5000, Universite de Montpellier II, Montpellier, France. Bentzen P, McPhail JD (1984) Ecology and evolution of sympatric sticklebacks (Gasterosteus): specialization for alternative trophic niches in the Enos Lake species pair. Canadian Journal of Zoology, 62, 2280–86. Berg OK, Finstad AG, Olsen PH, Arnekleiv JV, Nilssen K (2010) Dwarfs and cannibals in the Arctic: production of Arctic char (Salvelinus alpinus (L.)) at two trophic levels. Hydrobiologia, 652, 337–47. 94  Bernatchez L, Chouinard A, Lu G (1999) Integrating molecular genetics and ecology in studies of adaptive radiation: whitefish, Coregonus sp., as a case study. Biological Journal of the Linnean Society, 68, 173–194. Bernatchez L, Dempson JB, Martin S, Dempson BJ (1998) Microsatellite gene diversity analysis in anadromous arctic char, Salvelinus alpinus, from Labrador, Canada. Canadian Journal of Fisheries and Aquatic Sciences, 55, 1264–72. Bernatchez L, Dodson JJ (1990) Allopatric Origin of Sympatric Populations of Lake Whitefish (Coregonus clupeaformis) as Revealed by Mitochondrial-DNA Restriction Analysis. Evolution, 44, 1263–1271. Bernatchez L, Glemet H, Wilson CC, Danzmann RG (1995) lntrogression and fixation of Arctic char (Salvelinus alpinus) mitochondrial genome in an allopatric population of brook trout (Salvelinus fontinalis). Canadian Journal of Fisheries and Aquatic Sciences, 52, 179–185. Bernatchez L, Renaut S, Whiteley AR et al. (2010) On the origin of species: insights from the ecological genomics of lake whitefish. Philosophical transactions of the Royal Society of London. Series B, Biological Sciences, 365, 1783–800. Bickford D, Lohman DJ, Sodhi NS et al. (2007) Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution, 22, 148–55. Brelsford A (2011) Hybrid speciation in birds: allopatry more important than ecology? Molecular Ecology, 20, 3705–7. Brunner PC, Douglas MR, Osinov A et al. (2001) Holarctic Phylogeography of Arctic Charr (Salvelinus alpinus L .) Inferred from Mitochondrial DNA Sequences. Evolution, 55, 573–586. Burgess KS, Morgan M, Deverno L, Husband BC (2005) Asymmetrical introgression between two Morus species (M. alba, M. rubra) that differ in abundance. Molecular Ecology, 14, 3471–83. Burke JM, Arnold ML (2001) Genetics and the Fitness of Hybrids. Annual Review of Genetics, 35, 31–52. Bustard D, Royea D (1995) Kemess South Gold–Copper Project 1994 Fisheries Studies. Report of Dave Bustard and Associates and Hallam Knight Piésold Limited to El Condor Resources Limited and St. Phillips Resources Incorporated. Vancouver, B.C. Campton DE (1987) Natural hybridization and introgression in fishes: Methods of detection and genetic interpretations. In: Ryman N. and Utter F. (eds.), 95  Population Genetics and Fishery Management. University of Washington Press, Seattle, WA, USA. Campton DE, Utter F (1985) Natural Hybridization between Steelhead Trout (Salmo gairdneri) and Coastal Cutthroat Trout (Salmo clarki clarki) in Two Puget Sound Streams. Canadian Journal of Fisheries and Aquatic Sciences, 42, 110–119. Chenuil A, Crespin L, Pouyaud L, Patrick B (2000) Movements of adult fish in a hybrid zone revealed by microsatellite genetic analysis and capture-recapture data. Freshwater Biology, 43, 121–131. Conejeros P, Phan A, Power M et al. (2014) Differentiation of Sympatric Arctic Char Morphotypes Using Major Histocompatibility Class II Genes. Transactions of the American Fisheries Society, 143, 586–594. Conte GL, Schluter D (2013) Experimental confirmation that body size determines mate preference via phenotype matching in a stickleback species pair. Evolution, 67, 1477–1484. Cook RR, Auster PJ (2005) Use of Simulated Annealing for Identifying Essential Fish Habitat in a Multispecies Context. Conservation Biology, 19, 876–886. Corrigan LJ, Lucas MC, Winfield IJ, Hoelzel a R (2011) Environmental factors associated with genetic and phenotypic divergence among sympatric populations of Arctic charr (Salvelinus alpinus). Journal of Evolutionary Biology, 24, 1906–17. COSEWIC (2010) Assessment and Status Report on the Dolly Varden (Salvelinus malma malma) Western Arctic population in Canada. Ottawa, ON. Coyne JA, Orr HA (2004) Speciation. Sinauer Associates, Sunderland, Massachusetts. Crane PA, Lewis CJ, Kretschmer EJ et al. (2004) Characterization and inheritance of seven microsatellite loci from Dolly Varden, Salvelinus malma, and cross-species amplification in Arctic char, S. alpinus. Conservation Genetics, 5, 737–741. Crête-Lafrenière A, Weir LK, Bernatchez L (2012) Framing the Salmonidae family phylogenetic portrait: a more complete picture from increased taxon sampling. PLoS One, 7, e46662. Culumber ZW, Fisher HS, Tobler M et al. (2011) Replicated hybrid zones of Xiphophorus swordtails along an elevational gradient. Molecular Ecology, 20, 342–56. 96  Dehaan PW, Ardren WR (2005) Characterization of 20 highly variable tetranucleotide microsatellite loci for bull trout (Salvelinus confluentus) and cross-amplification in other Salvelinus species. Molecular Ecology Notes, 5, 582–585. Delacy AC, Morton WM (1943) Taxonomy and habits of the charrs, Salvelinus malma and Salvelinus alpinus, of the Karluk River drainage. Transactions of the American Fisheries Society, 72, 79–91. DeMarais BD, Dowling TE, Douglas ME, Minckley WL, Marsh PC (1992) Origin of Gila seminuda (Teleostei: Cyprinidae) through introgressive hybridization: implications for evolution and conservation. Proceedings of the National Academy of Sciences, U.S.A., 89, 2747–51. Dempson J, Shears M, Bloom M (2002) Spatial and temporal variability in the diet of anadromous Arctic charr, Salvelinus alpinus, in northern Labrador. ( eds. E In: Magnan, P., Audet, C., Glémet, H., Legault, M., Rodríguez, M., & Taylor, E.B., Ed,). Denton KP, Rich HB, Moore JW, Quinn TP (2010) The utilization of a Pacific salmon Oncorhynchus nerka subsidy by three populations of charr Salvelinus spp. Journal of Fish Biology, 77, 1006–23. Dobzhansky T (1937) Genetics and the origin of species. Columbia University Press, New York. Dowling TE, Secor CL (1997) The role of hybridization and introgression in the diversification of animals. Annual Review of Ecology and Systematics, 28, 593–619. Dunham J, Baxter C, Fausch K et al. (2008) Evolution, Ecology, and Conservation of Dolly Varden , White-spotted Char, and Bull Trout. Fisheries, 33, 537–50. Earl DA, VonHoldt BM (2012) STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources, 4, 359–361. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611–20. Evans ML, Præbel K, Peruzzi S, Bernatchez L (2012) Parallelism in the oxygen transport system of the lake whitefish: the role of physiological divergence in ecological speciation. Molecular Ecology, 21, 4038–4050. 97  Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10, 564–567. Finstad AG, Berg OK (2004) Bimodal population size distributions and biased gillnet sampling. Canadian Journal of Fisheries and Aquatic Sciences, 61, 2151–2157. Foote CJ, Larkin PA (1988) The role of male choice in the assortative mating of anadromous and non-anadromous sockeye salmon (Oncorhynchus nerka). Behaviour, 43–62. Foster SA, Baker JA, Bell MA (2003) The Case for Conserving Threespine Stickleback Populations. Fisheries, 28, 10–18. Fritsche F, Kaltz O (2000) Is the Prunella (Lamiaceae) hybrid zone structured by an environmental gradient? Evidence from a reciprocal transplant experiment. American Journal of Botany, 87, 995–1003. Frost WE (1965) Breeding habits of Windermere charr, Salvelinus willughbii (Günther), and their bearing on speciation of these fish. Proceedings of the Royal Society of London. Series B., 163, 232–284. Garduño-Paz M V., Demetriou M, Adams CE (2010) Variation in scale shape among alternative sympatric phenotypes of Arctic charr Salvelinus alpinus from two lakes in Scotland. Journal of Fish Biology, 76, 1491–1497. Geml J, Kauff F, Laursen GA, Taylor DL (2010) Genetic studies point to Beringia as a biodiversity hotspot for high-latitude fungi. Alaska Park Science, 8, 37–41. Gilman RT, Behm JE (2011) Hybridization, species collapse, and species reemergence after disturbance to premating mechanisms of reproductive isolation. Evolution, 65, 2592–605. Gíslason D, Ferguson MM, Skúlason S, Snorrason SS (1999) Rapid and coupled phenotypic and genetic divergence in Icelandic Arctic char (Salvelinus alpinus). Canadian Journal of Fisheries and Aquatic Sciences, 56, 2229–34. Gomez-Uchida D, Dunphy KP, O’Connell MF, Ruzzante DE (2008) Genetic divergence between sympatric Arctic charr Salvelinus alpinus morphs in Gander Lake, Newfoundland: roles of migration, mutation and unequal effective population sizes. Journal of Fish Biology, 73, 2040–57. Gordeeva N V., Chukova EI, Oleinik AG (2010) Microsatellite genetic variation of Asian populations of Dolly Varden char. Hydrobiologia, 650, 133–144. 98  Goudet J (2001) FSTAT version 2.9.3.1. Updated from Goudet, J. 1995. Journal of Heredity, 86, 485–486. Gow JL, Peichel CL, Taylor EB (2006) Contrasting hybridization rates between sympatric three-spined sticklebacks highlight the fragility of reproductive barriers between evolutionarily young species. Molecular Ecology, 15, 739–52. Gow JL, Peichel CL, Taylor EB (2007) Ecological selection against hybrids in natural populations of sympatric threespine sticklebacks. Journal of Evolutionary Biology, 20, 2173–80. Gowell CP, Quinn TP, Taylor EB (2012) Coexistence and origin of trophic ecotypes of pygmy whitefish, Prosopium coulterii, in a south-western Alaskan lake. Journal of Evolutionary Biology, 25, 2432–48. Grant PR, Grant BR (1992) Hybridization of bird species. Science (New York, N.Y.), 256, 193–7. Grant PR, Grant BR (2006) Evolution of character displacement in Darwin’s finches. Science, 313, 224–26. Grant PR, Grant BR (2010) Conspecific versus heterospecific gene exchange between populations of Darwin’s finches. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 365, 1065–76. Grewe PM, Billington N, Hebert PDN (1990) Phylogenetic Relationships Among Members of Salvelinus Inferred from Mitochondrial DNA Divergence. Canadian Journal of Fisheries and Aquatic Sciences, 47, 984–991. Griffiths D (1994) The size structure of lacustrine Arctic charr (Pisces: Salmonidae) populations. Biological Journal of the Linnean Society, 51, 337–57. Gross R, Gum B, Reiter R, Kühn R (2004) Genetic Introgression Between Arctic Charr (Salvelinus alpinus) and Brook Trout (Salvelinus fontinalis) in Bavarian Hatchery Stocks Inferred from Nuclear and Mitochondrial DNA Markers. Aquaculture International, 12, 19–32. Hagen J, Taylor EB (2001) Resource partitioning as a factor limiting gene flow in hybridizing populations of Dolly Varden char (Salvelinus malma) and bull trout (Salvelinus confluentus). Canadian Journal of Fisheries and Aquatic Sciences, 58, 2037–2047. Hamilton JA, Lexer C, Aitken SN (2013) Genomic and phenotypic architecture of a spruce hybrid zone (Picea sitchensis× P. glauca). Molecular Ecology, 22, 827–841. 99  Hammar J, Dempson JB, Skold E (1989) Natural hybridization between Arctic char (Salvelinus alpinus) and lake char (S. namaycush): evidence from northern Labrador. Nord. J. Freshwater Res., 65, 54–70. Hammar J, Dempson B, Verspoor E (1991) Hybridization between Arctic Char (Salvelinus alpinus) and Brook Trout (S . fontinalis): Evidence from Northern Labrador. Canadian Journal of Fisheries and Aquatic Sciences, 48, 1437–1445. Hammer Ø, Harper DAT, Ryan PD (2001) PAST : Paleontological statistics software package for education and data analysis. Palaentol. Electronica, 4, 1–9. Harrison RG (1990) Hybrid zones: windows on evolutionary process. Oxford Surveys in Evolutionary Biology, 7, 69–128. Hartman WL, Burgner RL (1972) Limnology and Fish Ecology of Sockeye Salmon Nursery Lakes of the World. Journal of the Fisheries Research Board of Canada, 29, 699–715. Hatfield T, Schluter D (1999) Ecological Speciation in Sticklebacks: Environment-Dependent Hybrid Fitness. Evolution, 53, 866–873. Hausdorf B (2011) Progress Toward A General Species Concept. Evolution, 65, 923–31. Helen E (2001) Phylogenetic Analysis of the Arctic Char-Dolly Varden Species Complex. PhD Thesis, University of Wisconsin-Milwaukee, WI, USA. Hendry AP (2001) Adaptive divergence and the evolution of reproductive isolation in the wild: an empirical demonstration using introduced sockeye salmon. Genetica, 112-113, 515–34. Hendry AP, Stearns SC (2003) Evolution Illuminated: Salmon and Their Relatives. Oxford University Press, Cary, NC, USA. Hermansen JS, Saether SA, Elgvin TO et al. (2011) Hybrid speciation in sparrows I: phenotypic intermediacy, genetic admixture and barriers to gene flow. Molecular Ecology, 20, 3812–3822. Van Herwerden L, Doherty PJ (2006) Contrasting genetic structures across two hybrid zones of a tropical reef fish, Acanthochromis polyacanthus (Bleeker 1855). Journal of Evolutionary Biology, 19, 239–52. Hewitt GM (1988) Hybrid Zones- Natural Laboratories for Evolutionary Studies. Trends in Ecology & Evolution, 3, 158–167. 100  Hoffmann AA, Sgro CM (2011) Climate change and evolutionary adaptation. Nature, 470, 479–485. Howard DJ, Gregory PG, Chu J, Cain ML (1998) Conspecific Sperm Precedence is an Effective Barrier to Hybridization Between Closely Related Species. Evolution, 52, 511–16. Hubbs CL (1955) Hybridization Between Fish Species in Nature. Society of Systematic Biologists, 4, 1–20. Ilves KL, Taylor EB (2008) Evolutionary and biogeographical patterns within the smelt genus Hypomesus in the North Pacific Ocean. Journal of Biogeography, 35, 48–64. Irwin DE, Brelsford A, Toews, D P, MacDonald C, Phinney M (2009) Extensive hybridization in a contact zone between MacGillivray’s warblers O. porornistolmiei and mourning warblers O. philadelphia detected using molecular and morphological analyses. Journal of Avian Biology, 40, 539–552. Jacobsen F, Omland KE (2011) Species tree inference in a recent radiation of orioles (Genus Icterus): multiple markers and methods reveal cytonuclear discordance in the northern oriole group. Molecular Phylogenetics and Evolution, 61, 460–9. Jiggins CD, Mallet J (2000) Bimodal hybrid zones and speciation. Trends in Ecology & Evolution, 15, 250–55. Johnson L (1980) The Arctic charr: Salvelinus alpinus. In Charrs: salmonid fishes of the genus Salvelinus, ed. E.K. Balon. The Hague: Sr W. Junk Publishers. Jonsson B, Jonsson N (2001) Polymorphism and speciation in Arctic charr. Journal of Fish Biology, 58, 605–638. Jonsson B, Skulason S, Snorrason SS et al. (1988) Life history variation of polymorphic arctic charr (Salvelinus alpinus) in Thingvallavatn, Iceland. Canadian Journal of Fisheries and Aquatic Sciences, 45, 1537–1547. Klemetsen A (2002) Evidence for genetic differences in the offspring of two sympatric morphs of Arctic charr. Journal of Fish Biology, 60, 933–950. Klemetsen A (2010) The charr problem revisited: exceptional phenotypic plasticity promotes ecological speciation in postglacial lakes. Freshwater Reviews, 3, 49–74. Klemetsen A (2013) The Most Variable Vertebrate on Earth. Journal of Ichthyology, 53, 781–791. 101  Klemetsen A, Amundsen P-A, Dempson J, Jónsson B (2003) Atlantic salmon Salmo salar L., brown trout Salmo trutta L. and Arctic charr Salvelinus alpinus (L .): a review of aspects of their life histories. Ecology of Freshwater Fish, 12, 1–59. Knudsen R, Klemetsen A, Amundsen P-A, Hermansen B (2006) Incipient speciation through niche expansion: an example from the Arctic charr in a subarctic lake. Proceedings of the Royal Society of Britian, 273, 2291–2298. Koizumi I (2011) Integration of ecology, demography and genetics to reveal population structure and persistence: a mini review and case study of stream-dwelling Dolly Varden. Ecology of Freshwater Fish, 20, 352–363. Koizumi I, Maekawa K (2004) Metapopulation structure of stream-dwelling Dolly Varden charr inferred from patterns of occurrence in the Sorachi River basin, Hokkaido, Japan. Freshwater Biology, 49, 973–981. Koizumi I, Yamamoto S, Maekawa K (2006) Female-biased migration of stream-dwelling Dolly Varden in the Shiisorapuchi River, Hokkaido, Japan. Journal of Fish Biology, 68, 1513–1529. Kottelat M, Freyhof J (2007) Handbook of European freshwater fishes. Cornol, Switzerland: Kottelat and Berlin, Germany: Freyhof. Kraus RT, Secor DH (2004) Incorporation of strontium into otoliths of an estuarine fish. Journal of Experimental Marine Biology and Ecology, 302, 85–106. Kristjánsson BK, Malmquist HJ, Ingimarsson F et al. (2011) Relationships between lake ecology and morphological characters in Icelandic Arctic charr, Salvelinus alpinus. Biological Journal of the Linnean Society, 103, 761–771. Landry L, Vincent WF, Bernatchez L (2007) Parallel evolution of lake whitefish dwarf ecotypes in association with limnological features of their adaptive landscape. Journal of Evolutionary Biology, 20, 971–984. Langerhans RB, Gifford ME, Joseph EO (2007) Ecological speciation in Gambusia fishes. Evolution, 61, 2056–2074. Latch EK, Kierepka EM, Heffelfinger JR, Rhodes OE (2011) Hybrid swarm between divergent lineages of mule deer (Odocoileus hemionus). Molecular Ecology, 20, 5265–79. Lepais O, Petit RJ, Guichoux E et al. (2009) Species relative abundance and direction of introgression in oaks. Molecular Ecology, 18, 2228–42. 102  Limburg KE (1995) Otolith strontium traces environmental history of subyearling American shad Alosa sapidissima. Marine Ecology Progress Series, 119, 25–35. Lindsey CC (1956) Distribution and Taxonomy of Fishes in the Mackenzie Drainage of British Columbia. J. Fish. Res. BD. Canada, 13, 759–789. Loh Y-HE, Bezault E, Muenzel FM et al. (2012) Origins of shared genetic variation in African cichlids. Molecular Biology and Evolution, 30, 906–17. Lowe MR, DeVries DR, Wright RA, Ludsin SA, Fryer BJ (2010) Otolith Microchemistry Reveals Substantial Use of Freshwater by Southern Flounder in the Northern Gulf of Mexico. Estuaries and Coasts, 34, 630–639. Lu G, Bernatchez L (1998) Experimental evidence for reduced hybrid viability between dwarf and normal ecotypes of lake whitefish (Coregonus clupeaformis Mitchill). Proceedings of the Royal Society of London. Series B: Biological Sciences, 265, 1025–30. Lu G, Bernatchez L (1999) Correlated Trophic Specialization and Genetic Divergence in Sympatric Lake Whitefish Ecotypes (Coregonus clupeaformis): Support for the Ecological Speciation Hypothesis. Evolution, 53, 1491–1505. Maitland PS (1995) World status and conservation of the arctic charr Salvelinus alpinus (L.). Nordic Journal of Freshwater Research, 71, 113–127. Mallet J (2005) Hybridization as an invasion of the genome. Trends in Ecology & Evolution, 20, 229–37. Malmquist HJ, Snorrason SS, Skúlason S et al. (1992) Diet differentiation in polymorphic Arctic charr in Thingvallavatn, Iceland. Journal of Animal Ecology, 61, 21–35. May-McNally SL, Quinn TP, Woods PJ, Taylor EB (2014) Evidence for genetic distinction among sympatric ecotypes of Arctic char (Salvelinus alpinus) in south-western Alaskan lakes. Ecology of Freshwater Fish. Mayr E (1942) Systematics and the Origin of Species. Columbia University Press, New York. Mayr E (1963) Animal species and evolution. Belknap Press, Cambridge, MA. McBride DN (1980) Homing of Arctic Char, Salvelinus alpinus (Linnaeus) to feeding and spawning sites in the Wood River Lake system, Alaska. Juneau, State of Alaska Department of Fish and Game. 103  McPhail JD (1961) A Systematic Study of the Salvelinus alpinus Complex in North America. J. Fish. Res. BD. Canada, 18, 793–815. McPhail JD, Lindsey CC (1970) Freshwater Fishes of Northwestern Canada and Alaska (JC Stevenson, Ed,). Fisheries Research Board of Canada, Ottawa, Canada. McPhee M V, Noakes DLG, Allendorf FW (2012) Developmental rate: A unifying mechanism for sympatric divergence in postglacial fishes? Current Zoology, 58, 21–34. Michaud WK, Power M, Kinnison MT (2008) Trophically mediated divergence of Arctic charr (Salvelinus alpinus L.) populations in contemporary time. Evolutionary Ecology Research, 10, 1051–66. Mieszkowska N, Kendall MA, Hawkins SJ et al. (2006) Changes in the range of some common rocky shore species in Britain – a response to climate change? Hydrobiologia, 555, 241–251. Minns CK, Moore JE (1992) Predicting the impact of climate change on the spatial pattern of freshwater fish yield capability in eastern Canadian lakes. Climatic Change, 22, 327–346. Moore WS (1977) An Evaluation of Narrow Hybrid Zones in Vertebrates. The Quarterly Review of Biology, 52, 263–277. Moore J-S, Gow JL, Taylor EB, Hendry AP (2007) Quantifying the constraining influence of gene flow on adaptive divergence in the lake-stream threespine stickleback system. Evolution, 61, 2015–26. Narum SRSR (2006) Beyond Bonferroni: Less conservative analyses for conservation genetics. Conservation Genetics, 7, 783–787. Nelson JS (2006) Fishes of the World, 4th edn. John Wiley and Sons, New York. Nielsen EE, Bach LA, Kotlicki P (2006) HYBRIDLAB (version 1.0): a program for generating simulated hybrids from population samples. Molecular Ecology Notes, 6, 971–73. Nielsen EE, Hansen MM, Ruzzante DE, Meldrup D, Gronkjaer P (2003) Evidence of a hybrid-zone in Atlantic cod (Gadus morhua) in the Baltic and the Danish Belt Sea revealed by individual admixture analysis. Molecular Ecology, 12, 1497–1508. 104  Nordeng H (1983) Solution to the “char problem” based on Arctic char (Salvelinus alpinus) in Norway. Canadian Journal of Fisheries and Aquatic Sciences, 40, 372–87. Ohlberger J, Staaks G, Petzoldt T, Mehner T, Hölker F (2008) Physiological specialization by thermal adaptation drives ecological divergence in a sympatric fish species pair. Evolutionary Ecology Research, 10, 1173–1185. Oleinik AG, Skurikhina LA, Brykov VA (2011) Population Genetic Structure of Northern Dolly Varden Char Salvelinus malma malma in Asia and North America. Russian Journal of Genetics, 47, 1456–1467. Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) Micro-Checker: Software for Identifying and Correcting Genotyping Errors in Microsatellite Data. Molecular Ecology Notes, 4, 535–538. Ormond CI, Rosenfeld JS, Taylor EB (2011) Environmental determinants of threespine stickleback species pair evolution and persistence. Canadian Journal of Fisheries and Aquatic Sciences, 68, 1983–1997. Pess GR (2009) Patterns and Processes of Salmon Colonization. University of Washington, Seattle (USA). Power M, O’Connell MF, Dempson BJ (2005) Ecological segregation within and among Arctic char morphotypes in Gander Lake, Newfoundland. Environmental Biology of Fishes, 73, 263–274. Power M, O’Connell MF, Dempson B (2012) Determining the consistency of thermal habitat segregation within and among Arctic charr morphotypes in Gander Lake, Newfoundland. Ecology of Freshwater Fish, 21, 245–254. Power M, Power G, Reist JD, Bajno R (2009) Ecological and genetic differentiation among the Arctic charr of Lake Aigneau, Northern Québec. Ecology of Freshwater Fish, 18, 445–460. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics, 155, 945–59. Pyle RL, Randall JE, Museum BIB (1994) A review of hybridization in marine angelfishes (Perciformes : Pomacanthidae ). Environmental Biology of Fishes, 41, 127–145. R Development Core Team (2011) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 105  Ramstad KM, Woody CA, Sage GK, Allendorf FW (2004) Founding events influence genetic population structure of sockeye salmon (Oncorhynchus nerka) in Lake Clark , Alaska. Molecular Ecology, 13, 277–290. Raymond M, Rousset F (1995) GENEPOP (version 1.2), population genetics software for exact tests and ecumenicism. J. Hered., 86, 248–249. Redenbach Z, Taylor EB (2003) Evidence for bimodal hybrid zones between two species of char (Pisces: Salvelinus) in northwestern North America. Journal of Evolutionary Biology, 16, 1135–1148. Reist JD, Gyselman E, Babaluk JA, Johnson JD, Wissink R (1995) Evidence for two morphotypes of Arctic char (Salvelinus alpinus (L.)) from Lake Hazen, Ellesmere Island, Northwest Territories, Canada. Nordic Journal of Freshwater Research, 71, 396–410. Reist JD, Low G, Johnson JD, McDowell D (2002) Range extension of bull trout, Salvelinus confluentus, to the central Northwest Territories, with notes on identification and distribution of Dolly Varden, Salvelinus malma, in the western Canadian Arctic. Arctic, 55, 70–76. Reist JD, Power M, Dempson B (2013) Arctic charr (Salvelinus alpinus): a case study of the importance of understanding biodiversity and taxonomic issues in northern fishes. Biodiversity, 14, 45–56. Reist JD, Sawatzky CD (2010) Diversity and Distribution of Chars, Genus Salvelinus, in Northwestern North America in the Context of Northern Dolly Varden (Salvelinus malma malma (Walbaum 1792)). Reist JD, Wrona FJ, Prowse TD et al. (2006) Effect of Climate Change and UV Radiation on Fisheries for Arctic Freshwater and Anadrmous Species. Royal Swedish Academy of Science, 35, 402–10. Renaut S, Nolte AW, Bernatchez L (2010) Mining transcriptome sequences towards identifying adaptive single nucleotide polymorphisms in lake whitefish species pairs (Coregonus spp. Salmonidae). Molecular Ecology, 19, 115–31. Rexroad CE, Coleman RL, Hershberger WK, Killefer J (2002) Rapid communication: Thirty-eight polymorphic microsatellite markers for mapping in rainbow trout. Journal of Animal Science, 80, 541–542. Rieseberg LH (1997) Hybrid Origins of Plant Species. Annual Review of Ecology and Systematics, 28, 359–389. Rieseberg LH, Baird SJE, Desrochers AM (1998) Patterns of Mating in Wild Sunflower Hybrid Zones. Evolution, 52, 713–726. 106  Rieseberg LH, Carter R, Zona S (1990) Molecular Tests of the Hypothesized Hybrid Origin of Two Diploid Helianthus species (Asteraceae). Evolution, 44, 1498–1511. Riginos C, Cunningham CW (2007) Hybridization in postglacial marine habitats. Molecular Ecology, 16, 3971–2. Roesch C, Lundsgaard-Hansen B, Vonlanthen P, Taverna A, Seehausen O (2013) Experimental evidence for trait utility of gill raker number in adaptive radiation of a north temperate fish. Journal of Evolutionary Biology, 26, 1578–1587. Rogers SM, Bernatchez L (2006) The genetic basis of intrinsic and extrinsic post-zygotic reproductive isolation jointly promoting speciation in the lake whitefish species complex (Coregonus clupeaformis). Journal of Evolutionary Biology, 19, 1979–94. Roques S, Sévigny JM, Bernatchez L (2001) Evidence for broadscale introgressive hybridization between two redfish (genus Sebastes) in the North-west Atlantic: a rare marine example. Molecular Ecology, 10, 149–65. Rosenberg NA (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes, 4, 137–138. Rubidge EM, Taylor EB (2004) Hybrid zone structure and the potential role of selection in hybridizing populations of native westslope cutthroat trout (Oncorhynchus clarki lewisi) and introduced rainbow trout (O. mykiss). Molecular Ecology, 13, 3735–49. Schluter D (1996) Ecological Causes of Adaptive Radiation. The American Naturalist, 148, S40–S64. Schluter D (1998) Ecological causes of speciation. In Endless forms: species and speciation. Edited by D.J. Howard and S.H. Berlocher., Oxford University Press, Oxford, U.K. Schluter D (2000) The ecology of adaptive radiation. Oxford, UK: Oxford University Press. Schluter D (2009) Evidence for ecological speciation and its alternative. Science, 323, 737–41. Schluter D, Rambaut A (1996) Ecological Speciation in Postglacial Fishes. Philosophical Transactions of the Royal Society B: Biological Sciences, 351, 807–14. 107  Schwartz FJ (1972) World Literature to Fish Hybrids, with an Analysis by Family, Species, and Hybrid. , 1–328. Schwartz FJ (1981) World Literature to Fish Hybrids, with an Analysis by Family, Species, and Hybrid: Suppl. 1. NOAA. Scribner KT, Page KS, Bartron ML (2001) Hybridization in freshwater fishes: a review of case studies and cytonuclear methods of biological inference. Reviews in Fish Biology and Fisheries, 10, 293–323. Seehausen O (2004) Hybridization and adaptive radiation. Trends in Ecology & Evolution, 19, 198–207. Seehausen O (2006) Conservation: Losing Biodiversity by Reverse Speciation. Current Biology, 16, R334–337. Seehausen O, Alphen JJM van, Witte F (1997) Cichlid Fish Diversity Threatened by Eutrophication That Curbs Sexual Selection. Science, 277, 1808–1811. Seehausen O, Takimoto G, Roy D, Jokela J (2008) Speciation reversal and biodiversity dynamics with hybridization in changing environments. Molecular Ecology, 17, 30–44. Servedio MR, Noor MAF (2003) The Role of Reinforcement in Speciation: Theory and Data. Annual Review of Ecology and Systematics, 34, 339–364. Siwertsson A, Knudsen R, Kahilainen KK, Præbel K, Primicerio R (2010) Sympatric diversification as influenced by ecological opportunity and historical contingency in a young species lineage of whitefish. Evolutionary Ecology Research, 12, 929–947. Skúlason S, Noakess DLG, Snorrason SS (1989) Ontogeny of trophic morphology in four sympatric morphs of arctic charr Salvelinus alpinus in Thingvallavatn , Iceland. Biological Journal of the Linnean Society, 38, 281–301. Skúlason S, Smith TB (1995) Resource polymorphisms in vertebrates. Trends in Ecology & Evolution, 10, 366–70. Skúlason S, Snorrason SS, Jonasson B (1999) Sympatric morphs, populations and speciation in freshwater fish with emphasis on Arctic char. In evolution of biological diversity. Edited by A.E. Magurran and R. May. Oxford University Press, Oxford, U.K. Skúlason S, Snorrason SS, Noakes DLG, Ferguson MM (1996) Genetic basis of life history variations among sympatric morphs of Arctic char, Salvelinus alpinus. Canadian Journal of Fisheries and Aquatic Sciences, 53, 1807–13. 108  Skúlason S, Snorrason SS, Ota D, Noakes DLG (1993) Genetically based differences in foraging behaviour among sympatric morphs of arctic charr (Pisces: Salmonidae). Animal Behaviour, 45, 1179–1192. Slettan A, Olsaker I, Lie O (1997) Segregation studies and linkage analysis of Atlantic salmon microsatellites using haploid genetics. Heredity, 78, 620–627. Smith PF, Konings A, Kornfield I (2003) Hybrid origin of a cichlid population in Lake Malawi: Implications for genetic variation and species diversity. Molecular Ecology, 12, 2497–2504. Smith TB, Skúlason S (1996) Evolutionary Significance of Resource Polymorphism in Fishes, Amphibians, and Birds. Annual Review of Ecology and Systematics, 27, 111–133. Snorrason SS, Skúlason S, Jónsson B et al. (1994) Trophic specialization in Arctic charr Salvelinus alpinus (Pisces; Salmonidae): morphological divergence and ontogenetic niche shifts. Biological Journal of the Linnean Society, 52, 1–18. Stilwell KB, Kaufman DS (1996) Late Wisconsin Glacial History of the Northern Alaska Peninsula, Southwestern Alaska, U.S.A. Arctic and Alpine Research, 28, 475–487. Sundby S, Nakken O (2008) Spatial shifts in spawning habitats of Arcto-Norwegian cod related to multidecadal climate oscillations and climate change. ICES Journal of Marine Science, 65, 953–962. Taylor EB (1999) Species pairs of north temperate freshwater fishes: evolution, taxonomy, and conservation. Reviews in Fish Biology and Fisheries, 9, 299–324. Taylor EB (2004) Evolution in mixed company. Evolutionary inferences from studies of natural hybridization in Salmonidae. Pp. 232-263. In A.P. Hendry and S. Stearns, eds. Evolution Illuminated. Salmon and their Relatives. Oxford University Press, Oxford. Taylor EB, Boughman JW, Groenenboom M et al. (2006) Speciation in reverse: morphological and genetic evidence of the collapse of a three-spined stickleback (Gasterosteus aculeatus) species pair. Molecular Ecology, 15, 343–55. Taylor EB, Gow JL, Witt J, Zemlak R (2011) Connectivity among populations of pygmy whitefish (Prosopium coulterii) in northwestern North America inferred from microsatellite DNA analyses. Canadian Journal of Zoology, 266, 255–266. 109  Taylor EB, Lowery E, Lilliestråle A, Elz A, Quinn TP (2008) Genetic analysis of sympatric char populations in western Alaska: Arctic char (Salvelinus alpinus) and Dolly Varden (Salvelinus malma) are not two sides of the same coin. Journal of Evolutionary Biology, 21, 1609–25. Taylor EB, McPhail JD (2000) Historical contingency and ecological determinism interact to prime speciation in sticklebacks, Gasterosteus. Proceedings of the Royal Society of London. Series B., 267, 2375–84. Taylor EB, Redenbach Z, Costello AB, Pollard SJ, Pacas CJ (2001) Nested analysis of genetic diversity in northwestern North American char, Dolly Varden (Salvelinus malma) and bull trout (Salvelinus confluentus). Canadian Journal of Fisheries and Aquatic Sciences, 58, 406–420. Thorgaard GH, Allendorf FW (1988) Developmental genetics in fishes. In: Malacinski G.M. (ed.), Developmental Genetics of Animals and Plants. MacMillan, New York. Turner GF (1999) What is a fish species? Reviews in Fish Biology and Fisheries, 9, 281–297. Vähä J-P, Primmer CR (2006) Efficiency of model-based Bayesian methods for detecting hybrid individuals under different hybridization scenarios and with different numbers of loci. Molecular Ecology, 15, 63–72. Verspoor E, Hammar J (1991) Introgressive hybridization in fishes: the biochemical evidence. Journal of Fish Biology, 39, 309–334. Vonlanthen P, Bittner D, Hudson AG et al. (2012) Eutrophication causes speciation reversal in whitefish adaptive radiations. Nature, 482, 357–362. Warren BML, Burr BM (1993) Status of Freshwater Fishes of the United States : Overview of an Imperiled Fauna. Fisheries, 19, 6–18. Weir BS, Cockerman CC (1984) Estimating F-Statistics for the Analysis of Population Structure. Evolution, 38, 1358–1370. Westgaard JI, Klemetsen A, Knudsen R (2004) Genetic differences between two sympatric morphs of Arctic charr confirmed by microsatellite DNA. Journal of Fish Biology, 65, 1185–91. Williamson, Kevin S, Cordes JF, May B (2002) Characterization of microsatellite loci in chinook salmon (Oncorhynchus tshawytscha) and cross-species amplification in other salmonids. Molecular Ecology Notes, 2, 17–19. 110  Wilson CC, Bernatchez L, Vachon P (1998) The ghost of hybrids past: fixation of arctic charr (Salvelinus alpinus) mitochondrial DNA in an introgressed population of lake trout (S . namaycush ). Molecular Ecology, 7, 127–132. Wilson CC, Hebert PDN (1993) Natural Hybridization between Arctic Char (Salvelinus alpinus) and Lake Trout (S. namaycush) in the Canadian Arctic. Canadian Journal of Fisheries and Aquatic Sciences, 50, 2652–2658. Wimberger PH (1994) Trophic polymorphisms, plasticity, and speciation in vertebrates. In: Theory and application of fish feeding ecology , pp. 19–43. Wirtz P (1999) Mother species-father species: unidirectional hybridization in animals with female choice. Animal Behaviour, 58, 1–12. Woods PJ (2011) Ecological diversity in the polymorphic fish Arctic charr (Salvelinus alpinus). PhD Thesis, University of Washington, Seattle, WA, USA and University of Iceland, Reykjavik, Iceland. Available at http://fish.washington.edu/people/tquinn/ pubs/Woods_. Woods PJ, Skúlason S, Snorrason SS et al. (2012) Intraspecific diversity in Arctic charr, Salvelinus alpinus, in Iceland: II. Which environmental factors influence resource polymorphism in lakes? Evolutionary Ecology Research, 14, 993–1013. Woods PJ, Young D, Skúlason S, Snorrason SS, Quinn TP (2013) Resource polymorphism and diversity of Arctic charr Salvelinus alpinus in a series of isolated lakes. Journal of Fish Biology, 82, 569–87. Yamamoto S, Kitano S, Maekawa K, Koizumi I, Morita K (2006) Introgressive hybridization between Dolly Varden (Salvelinus malma) and white-spotted charr (Salvelinus leucomaenis) on Hokkaido Island, Japan. Journal of Fish Biology, 68, 68–85. 111  Appendix A   Chapter 2   Table A.1: Genetic diversity values for Lower Tazimina Lake morphs and four populations of Arctic char (Salvelinus alpinus) from southwestern Alaska: sample codes; N, number of samples; A, number of alleles per locus; AR, Allelic richness; HO, observed heterozygosity; HE, expected heterozygosity; P (HW), probability of departure from HWE (p≤ 0.0185).    Sco 220 Sco 200 Smm 22 Sco 215 Ssos 1456 Otsg 253b Smm 24 Omm 1105 Otsg 83b Smm 17 Sco 216 Lower Tazimina Lake           N 89 89 89 89 90 90 90 90 90 90 89 A 30 20 20 3 8 16 22 16 22 11 27 AR 30 20 20 3 7.9 16 21.9 16 22 11 27 Ho 0.966 0.899 0.888 0.438 0.344 0.811 0.856 0.756 0.956 0.633 0.910 HE 0.947 0.890 0.920 0.493 0.597 0.881 0.925 0.873 0.923 0.746 0.900 FIS -0.021 -0.01 0.035 0.112 0.424 0.080 0.076 0.135 -0.035 0.152 -0.012 P(HW) 0.242 0.363 0.170 0.248 0* 0.0075* 0.042 0.028 0.084 0* 0.988 Large Morph                     N 31 31 31 31 31 31 31 31 31 31 31 A 20 11 11 2 3 11 11 11 11 6 12 AR 20 11 11 2 3 11 11 11 11 6 12 Ho 1 0.935 0.903 0.323 0.258 0.742 0.742 0.742 0.935 0.645 0.935 HE 0.932 0.867 0.881 0.444 0.286 0.855 0.864 0.845 0.866 0.691 0.875 FIS -0.074 -0.081 -0.026 0.277 0.099 0.134 0.143 0.124 -0.082 0.087 -0.071 P(HW) 0.750 0.425 0.639 0.209 0.584 0.213 0.347 0.389 0.618 0.189 0.708             112   Sco 220 Sco 200 Smm 22 Sco 215 Ssos 1456 Otsg 253b Smm 24 Omm 1105 Otsg 83b Smm 17 Sco 216 Intermediate Morph          N 18 18 18 18 19 19 19 19 19 19 19 A 14 12 13 2 4 10 15 11 14 7 14 AR 14 12 13 2 3.95 9.89 14.63 10.84 13.68 6.89 13.58 Ho 0.944 0.944 0.944 0.667 0.526 0.789 0.842 0.842 0.947 0.579 0.895 HE 0.921 0.903 0.924 0.514 0.585 0.879 0.923 0.896 0.905 0.686 0.863 FIS -0.027 -0.047 -0.023 -0.308 0.102 0.104 0.09 0.062 -0.049 0.166 -0.034 P(HW) 0.172 0.702 0.637 0.343 0.675 0.193 0.662 0.447 0.865 0.0756 0.532 Small Morph           N 40 40 40 40 40 40 40 40 40 40 39 A 30 14 19 3 7 11 21 10 21 9 24 AR 29.7 13.9 18.9 3 6.9 11 20.7 10 20.9 9 24 Ho 0.950 0.850 0.850 0.425 0.325 0.875 0.950 0.725 0.975 0.650 0.897 HE 0.963 0.879 0.932 0.512 0.350 0.843 0.917 0.834 0.943 0.685 0.930 FIS 0.014 0.034 0.089 0.172 0.071 -0.0380 -0.037 0.133 -0.034 0.144 0.035 P(HW) 0.841 0.314 0.075 0.265 0.164 0.415 0.468 0.253 0.670 0* 0.459 Caribou Lakes           N 25 25 25 25 24 25 25 25 25 25 24 A 10 13 12 2 3 9 9 2 9 4 6 AR 9.96 12.72 11.88 2 3 8.84 8.92 2 9 4 6 Ho 0.920 0.800 0.760 0.400 0.583 0.720 0.760 0.080 0.800 0.480 0.167 HE 0.87 0.84 0.908 0.372 0.582 0.745 0.83 0.075 0.887 0.552 0.782 FIS -0.057 0.048 0.163 -0.076 0.039 0.03 0.091 -0.021 0.099 0.127 0.783 P(HW) 0.0981 0.666 0.0706 1 0.706 0.603 0.315 1 0.606 0.385 0* Summit Lake           N 59 59 59 59 59 59 59 59 59 57 59 113   Sco 220 Sco 200 Smm 22 Sco 215 Ssos 1456 Otsg 253b Smm 24 Omm 1105 Otsg 83b Smm 17 Sco 216 A 10 7 14 1 2 6 8 8 6 7 15 AR 10 7 14 1 2 5.97 7.97 8 5.97 7 14.90 Ho 0.881 0.627 0.898 NA 0.373 0.678 0.644 0.729 0.780 0.351 0.831 HE 0.875 0.763 0.882 NA 0.472 0.712 0.738 0.799 0.757 0.441 0.864 FIS -0.008 0.180 -0.019 NA 0.211 0.048 0.128 0.089 -0.031 0.374 0.023 P(HW) 0.0812 0.0219 0.364 — 0.166 0.120 0.093 0.078 0.888 0* 0.0586 Iliamna Lake           N 42 42 42 42 42 42 42 42 41 41 41 A 22 27 23 5 5 17 19 7 18 12 23 AR 21.81 26.83 22.83 4.98 4.98 16.95 18.90 6.98 18 12 23 Ho 0.857 0.905 0.857 0.381 0.548 0.952 1 0.857 0.854 0.732 0.829 HE 0.940 0.957 0.939 0.434 0.654 0.919 0.917 0.792 0.917 0.780 0.931 FIS 0.090 0.056 0.088 0.124 0.165 -0.037 -0.092 -0.083 0.070 0.063 0.110 P(HW) 0.011* 0.0300 0.0806 0.018* 0.086 0.290 0.997 0.829 0.570 0.186 0.106 Lake Aleknagik           N 70 70 70 70 70 70 70 70 70 70 70 A 24 19 23 3 5 18 18 9 17 11 33 AR 24 19 23 3 5 18 18 9 17 11 33 Ho 0.929 0.886 0.971 0.271 0.557 0.910 0.886 0.800 0.886 0.729 0.929 HE 0.946 0.928 0.927 0.351 0.651 0.922 0.902 0.721 0.895 0.841 0.962 FIS 0.019 0.046 -0.047 0.227 0.094 0.008 0.018 -0.11 0.010 0.134 0.034 P(HW) 0.420 0.076 0.275 0.109 0.316 0.085 0.143 0.699 0.279 0.0048* 0.0344 114         Figure A.1: Factorial correspondence analysis (FCA) of five populations of Arctic char (Salvelinus alpinus) from southwestern Alaska. Lower Tazimina Lake morphs have been separated into large, intermediate and small populations according to the assigned genetic groupings based on Q value. Populations are denoted as follows: Yellow = Caribou Lakes; Blue = Summit Lake; White = Iliamna Lake; Grey = Lake Aleknagik; Dark Blue = Small morphs; Teal = Intermediate morphs; Pink = Large morphs       115         Figure A.2: Factorial correspondence analysis (FCA) of Lower Tazimina Lake, southwestern Alaska, morphs and Iliamna Lake Arctic char (Salvelinus alpinus). Populations are denoted by circles as follows: Yellow = Iliamna Lake; Blue and White = Large and Intermediate morphs; Green = Small morphs    116  Appendix B   Chapter 3  Table B.1:    Lake Aleknagik and Lake Nerka site names and assigned site ID numbers (July- August, 2012) with corresponding sample size, sampling locale, total number of samples collected, total samples genetically analyzed, and the given proportions of samples successfully genotyped as Arctic char (Salvelinus alpinus) (AC), Dolly Varden (S. malma) (DV) and hybrid (HYB). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788.  Site ID Site Name Locale Total N N Analyzed AC/DV/HYB 1 Youth Creek Aleknagik 24 24 24/0/0 2 Beach Seine 2N Aleknagik 98 40 40/0/0 3 Beach Seine 2S Aleknagik 20 16 16/0/0 4 Island off Hansen Creek Aleknagik 35 18 18/0/0 5 Bear Bay Aleknagik 37 30 29/0/1 6 Beach Seine 7S Aleknagik 30 24 24/0/0 7 Whitefish Bay Aleknagik 31 20 20/0/0 8 Wood River Wood River 8 8 0/8/0 9 Lower Bear Creek Aleknagik 57 15 15/0/0 10 Upper Bear Creek Aleknagik 3 3 1/2/0 11 Lower Whitefish Creek Aleknagik 73 31 31/0/0 12 Upper Whitefish Creek Aleknagik 5 5 5/0/0 13 Yako Creek Aleknagik 51 30 29/1/0 14 Lower Eagle Creek Aleknagik 35 15 15/0/0 15 Mid Eagle Creek Aleknagik 4 4 2/2/0 16 Upper Eagle Creek Aleknagik 12 12 11/1/0 17 Mid Hansen Creek Aleknagik 30 17 13/4/0 18 Upper Hansen Creek Aleknagik 6 4 3/1/0 19 Happy Creek Aleknagik 31 22 14/7/1 20 Fenno Creek Nerka 14 14 14/0/0 21 Little Togiak River Nerka 18 18 18/0/0 22 Teal Creek Nerka 20 20 18/2/0 23 Hidden Lake 1 (lower) Nerka 54 10 10/0/0 24 Hidden Lake 2 (lower-mid) Nerka 15 15 15/0/0 25 Hidden Lake 3 (mid-upper) Nerka 34 15 15/0/0 26 Hidden Lake 4 (upper) Nerka 20 18 18/0/0 27 Lynx Creek 1 (lower) Nerka 19 11 10/1/0 28 Lynx Creek 2 (lower-mid) Nerka 3 3 1/2/0 29 Lynx Creek 3 (mid-upper) Nerka 6 6 6/0/0 30 Lynx Creek 4(upper) Nerka 93 61 43/17/1 117    Table B.2:     Lake Aleknagik and Lake Nerka site names and assigned site ID numbers (July-August, 2013) with corresponding sample size, sampling locale, total number of samples collected, total samples genetically analyzed, and the given proportions of samples successfully genotyped as Arctic char (Salvelinus alpinus) (AC), Dolly Varden (S. malma) (DV) and hybrid (HYB). Arctic char were defined as having QDV values of ≤ 0.237, Dolly Varden were defined as having QDV ≥ 0.788, and hybrids were defined as having QDV values between 0.237 —0.788.  Site ID Site Name Locale Total N N Analyzed AC/DV/HYB 31 Mid Yako Creek Aleknagik 29 29 1/28/0 32 Upper Yako Creek Aleknagik 40 37 0/37/0 33 Lower Mission Creek Aleknagik 4 4 3/1/0 34 Upper Mission Creek Aleknagik 8 7 4/3/0 35 Lower Eagle Creek Aleknagik 22 21 20/1/0 36 Mid Eagle Creek Aleknagik 4 3 3/0/0 37 Upper Eagle Creek Aleknagik 2 2 2/0/0 38 Mid Hansen Creek Aleknagik 4 4 4/0/0 39 Happy Creek Aleknagik 49 49 39/10/0 40 Lynx Creek 1 (lower) Nerka 6 6 10/1/0 41 Lynx Creek 2 (lower-mid) Nerka 14 14 1/2/0 42 Lynx Creek 3 (mid-upper) Nerka 6 11 6/0/0 43 Lynx Creek 4(upper) Nerka 2 2 1/1/0 44 Silversalmon Creek Wood River 6 6 0/6/0       118   Table B.3:     Repeat motif, optimal annealing temperature (°C), species of origin, GenBank accession number, characteristics of marker in Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) and original reference for the 13 microsatellite markers used in this study.   Locus Repeat motif °C Species of Origin GenBank No. Diagnostic/  Polymorphic Reference OMM1105 (AGAC)23 (GATA)16 62 Oncorhynchus mykiss AF352768 Polymorphic Rexroad et al., 2002 OtsG83b (TGTC)7-Na51-(TATC)34 52 Oncorhynchus tshawytscha AF393189 Polymorphic Williamson et al., 2002 OtsG253b (GACA)10 (GATA)14 52 Oncorhynchus tshawytscha AF393193 Polymorphic Williamson et al., 2002 Sco200 (ATAG)15 56 Salvelinus confluentus AY88869 Polymorphic DeHaan & Ardren, 2005 Sco202 CTAT(10) 60 Salvelinus confluentus AY88871 Polymorphic DeHaan & Ardren, 2005 Sco215 (GAAA)6(GA)6(GGGA)1 (GA)13 60 Salvelinus confluentus AY88884 Polymorphic DeHaan & Ardren, 2005 Sco216 (CAGT)18(CAGG) 10C1(AGAT)7 60 Salvelinus confluentus AY88885 Polymorphic DeHaan & Ardren, 2005 Sco220 (ATAG)5(ATAC)2 (ATAG)15     (ATTG)1(ATAG)17 60 Salvelinus confluentus AY88889 Polymorphic DeHaan & Ardren, 2005 119  Locus Repeat motif °C Species of Origin GenBank No. Diagnostic/  Polymorphic Reference Smm-21 (TC)4TTTC(TC)21 56 Salvelinus malma AY327128 Diagnostic for ACb Crane et al., 2004 Smm-22 (TAgA)19 55 Salvelinus malma AY327129 Polymorphic Crane et al., 2004 Smm-24 (TATC)33 56 Salvelinus malma AY327130 Polymorphic Crane et al., 2004 SSOSL456 (AC)12AG(AC)10 58 Salmo salar Z69645 Polymorphic Slettan et al., 1997 a Any nucleotide b  Arctic char           120  Table B.4: Genetic diversity values at 13 microsatelllite loci for Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from two southwestern Alaskan Lakes with two sampling years (2012-2013) pooled. Sample codes: DV, Dolly Varden; AC, Arctic char; N, number of samples; A, number of alleles per locus; AR, Allelic richness; HO, observed heterozygosity; HE, expected heterozygosity; FIS, inbreeding coefficient of an individual relative to the subpopulation; P (HWE), probability of departure from Hardy-Weinburg equilibrium (p≤ 0.011).  Locus  DV (Aleknagik) DV (Nerka) AC (Aleknagik) AC (Nerka) Sco220 N 109 36 383 179  A 32 23 34 29  AR 3.7 5.2 33.8 28.8  HO 0.771 0.667 0.943 0.922  HE 0.939 0.944 0.95 0.937  FIS 0.18 0.297 0.008 0.016  P(HWE) 0* 0* 0.0327 0.119 Sco200 N 112 35 383 181  A 38 27 28 23  AR 3.7 5.3 27.3 22.7  HO 0.9375 0.943 0.898 0.873  HE 0.951 0.948 0.921 0.906  FIS 0.015 0.006 0.025 0.037  P(HWE) 0* 0.0126* 0.0297 0.108 Smm22 N 112 37 383 182  A 18 19 32 23  AR 3.4 5 31.4 22.9  HO 0.884 0.919 0.922 0.885  HE 0.886 0.924 0.916 0.934  FIS 0.002 0.06 -0.006 0.053  P(HWE) 0.0089* 0.305 0.175 0.0607 121  Locus  DV (Aleknagik) DV (Nerka) AC (Aleknagik) AC (Nerka) Sco215 N 2 3 381 171  A 3 3 5 6  AR 3 3 5 6  HO 0.5 0 0.276 0.222  HE 0.833 0.8 0.319 0.287  FIS 0.5 1 0.14 0.225  P(HWE) 0.333 0.0701 0.0374 0.0013* Ssosl456 N 111 37 381 183  A 8 4 7 6  AR 2.5 2.8 6.9 5.9  HO 0.784 0.703 0.617 0.623  HE 0.682 0.649 0.653 0.662  FIS -0.15 -0.084 0.055 0.059  P(HWE) 0.0855 0.921 0.0214 0.389 Otsg253b N 112 37 382 183  A 20 17 23 17  AR 3.3 4.7 22.6 17  HO 0.893 0.811 0.88 0.847  HE 0.861 0.891 0.907 0.901  FIS -0.038 0.091 0.031 0.06  P(HWE) 0* 0.001* 0.102 0* Smm24 N 101 36 348 183  A 25 21 21 21  AR 3.5 5.2 21 20.8  HO 0.871 0.944 0.905 0.907  HE 0.907 0.942 0.902 0.908       122  Locus  DV (Aleknagik) DV (Nerka) AC (Aleknagik) AC (Nerka)  FIS 0.040 -0.0030 -0.0030 -0.0050  P(HWE) 0* 0.0218 0.658 0.987 Omm1105 N 108 37 383 182  A 18 15 11 12  AR 3.3 4.8 10.9 11.9  HO 0.694 0.833 0.773 0.747  HE 0.861 0.906 0.752 0.803  FIS 0.194 0.081 -0.027 0.069  P(HWE) 0* 0.242 0.311 0.369 Otsg83b N 110 37 382 183  A 39 30 28 25  AR 3.7 5.45 27.5 24.6  HO 0.927 0.865 0.893 0.869  HE 0.951 0.958 0.911 0.917  FIS 0.026 0.101 0.02 0.052  P(HWE) 0* 0.0144* 0.113 0.177 Smm17 N 99 35 381 183  A 15 15 14 15  AR 3.3 4.5 13.7 14.9  HO 0.485 0.6 0.822 0.754  HE 0.857 0.88 0.856 0.851  FIS 0.436 0.321 0.04 0.114  P(HWE) 0* 0* 0.711 0.0882 Sco216 N 102 36 382 183  A 37 21 38 36  AR 3.7 5.26 37.6 35.8  HO 0.36 0.417 0.901 0.891 123  Locus  DV (Aleknagik) DV (Nerka) AC (Aleknagik) AC (Nerka)  HE 0.943 0.946 0.958 0.962  FIS 0.616 0.563 0.06 0.075  P(HWE) 0.0043* 0* 0.028 0.0074* Smm21 N 108 35 NA NA  A 6 4 1 1  AR 2.19 2.3 1 1  HO 0.574 0.429 NA NA  HE 0.572 0.444 NA NA  FIS -0.003 0.035 NA NA  P(HWE) 0.0114* 0.261 NA NA Sco202 N 107 35 374 180  A 7 3 17 14  AR 1.56 1.7 16.9 14  HO 0.28 0.314 0.797 0.806  HE 0.275 0.273 0.822 0.838  FIS -0.019 -0.154 0.03 0.038  P(HWE) 0.157 1 0.439 0.0014*      124        Figure B.1: Factorial correspondence analysis (FCA) based on variation at 13 microsatellite DNA loci in Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from the Wood River Lake system in southwestern Alaska. Dolly Varden are denoted by yellow squares in Lake Aleknagik and blue in Lake Nerka and Arctic char by grey squares in Lake Aleknagik and white in Lake Nerka.       125        Figure B.2: Factorial correspondence analysis (FCA) based on variation at 13 microsatellite DNA loci in Arctic char (Salvelinus alpinus) and Dolly Varden (S. malma) from the Wood River Lake system in southwestern Alaska compared to Arctic char from Resolute Bay, Nunavut in the Canadian Arctic and Dolly Varden from the Egegik fishing district on the southwestern coast of Alaska. Allopatric Arctic char are represented by white squares, allopatric Dolly Varden by grey squares and sympatric Arctic char and Dolly Varden from the Wood River Lake System by blue and yellow squares respectively.       126  QDV QDV > 0.779  QDV < 0.445                            AC                     DV                                      SH   Figure B.3: STRUCTURE plot generated from allopatric Arctic char (AC, n = 21) from the Canadian Arctic, allopatric Dolly Varden (DV, n = 29) from the Egegik fishing district in southern Alaska and associated simulated hybrids (SH, n = 50). Each fish is represented by a vertical bar that denotes admixture fractions (Q) for K =2 genetic clusters (red and green portions of each bar). Horizontal bars indicate the upper and lower in the SH plot represent the upper and lower boundaries for Q-value defined hybrids.          127  QDV QDV > 0.788    QDV < 0.237                                  DV                        AC                                        SH    Figure B.4:  STRUCTURE plot generated from presumed non-admixed Arctic char (AC, n = 50) and Dolly Varden (DV, n = 50) from the Wood River Lake system in southwestern Alaska and associated simulated hybrids (SH, n = 150). Each fish is represented by a vertical bar that denotes admixture fractions (Q) for K =2 genetic clusters (red and green portions of each bar). Horizontal bars indicate the upper and lower in the SH plot represent the upper and lower boundaries for QDV-value defined hybrids.   

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.24.1-0135591/manifest

Comment

Related Items