UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Physiological, transcriptomic and genomic mechanisms of thermal adaptation in Oncorhynchus mykiss Chen, Zhongqi 2017

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

Item Metadata

Download

Media
24-ubc_2017_february_zhongqi_chen.pdf [ 5.15MB ]
Metadata
JSON: 24-1.0340726.json
JSON-LD: 24-1.0340726-ld.json
RDF/XML (Pretty): 24-1.0340726-rdf.xml
RDF/JSON: 24-1.0340726-rdf.json
Turtle: 24-1.0340726-turtle.txt
N-Triples: 24-1.0340726-rdf-ntriples.txt
Original Record: 24-1.0340726-source.json
Full Text
24-1.0340726-fulltext.txt
Citation
24-1.0340726.ris

Full Text

PHYSIOLOGICAL, TRANSCRIPTOMIC AND GENOMIC MECHANISMS OF THERMAL ADAPTATION IN ONCORHYNCHUS MYKISS by  Zhongqi Chen  B.Sc., Xuzhou Normal University, 2007 M.Sc., Northwest A&F University, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Zoology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  January 2017  © Zhongqi Chen, 2017 ii  Abstract Given the rate and magnitude of the ongoing global warming, there is some urgency to understand the underlying mechanism of thermal adaptation to evaluate and predict the ecological consequences. This thesis used Oncorhynchus mykiss from different thermal regimes to examine thermal adaptation at physiological, transcriptomic and genomic levels. Thermal tolerance was examined in three redband trout (O. mykiss gairdneri) populations from warm desert and cool montane climates (Idaho, USA), as well as in a domesticated rainbow trout (O. mykiss) strain raised in a thermally challenging environment for over 19 generations (Western Australia, Australia). Acclimated to 15°C, the desert redband trout had the highest critical thermal maximum (CTMAX; 29.7°C) and maintained an almost constant absolute aerobic scope (AAS) across a broader range of test temperatures (12-24°C) than seen in other strains, but had the lowest peak AAS, suggesting a tradeoff between thermal performance and tolerance. Western Australian rainbow trout had the highest AAS, even when tested at 21°C, which may be a result of hatchery selection for thermal performance. Although the rate transition temperatures for maximum heart rate (Arrhenius breakpoint and arrhythmia temperature for fH,max) were similar among all populations, fH,max was the highest in the desert redband trout population at all temperatures. Cardiac RNA sequencing revealed different patterns of gene regulation among redband trout populations during acute warming. Many genes had different mRNA abundances between populations due to constitutive and induced expression, and the number of differentially expressed genes among populations was positively correlated to the genetic distance, suggesting intraspecific cellular regulatory strategies in response to acute warming.  iii  Population and quantitative genetic studies identified potential genomic markers for thermal adaptation. A total of twenty-one loci were putatively under positive thermal selection (“outliers”). In addition, genotypes of some outlier loci had significantly different CTMAX. Genome-wide association study identified twelve loci that were significantly associated with individual CTMAX. Altogether, results in my dissertation demonstrated the capacity of thermal adaptation in O. mykiss populations at multiple organismal levels. This data lays a foundation to improve our understanding on the potential impact of global warming on wild aquatic populations. iv  Preface This thesis is a collaborative work by the University of British Columbia, Fisheries and Oceans Canada, Department of Fisheries (Western Australia, Australia) and Columbia River Inter-Tribal Fish Commission (USA). All experiment procedures were approved by the University of British Columbia Committee on Animal Care in accordance with the Canadian Council on Animal Care (A10-0335) and by the University of Idaho (IACUC protocol 2013-80). Redband trout fry collection from natural streams was approved by Idaho Department of Fish and Game (Permit F-13-06-13). A version of chapter 2 has been published. Chen, Z., Snow, M., Lawrence, C., Church, A., Narum, S., Devlin, R. and Farrell, A. (2015). Selection for upper thermal tolerance in rainbow trout (Oncorhynchus mykiss Walbaum). J. Exp. Biol. 218, 803–812. Anthony Farrell and I conceived and designed the experiments with input from Shawn Narum and Robert Devlin; I performed experiments with assistance from Anthony Church. I analyzed results with input from Tony Farrell. I also drafted the manuscript. Mike Snow and Craig Lawrence organized the fish breeding. Craig Lawrence and Anthony Church performed the breeding and rearing of fish. Mike Snow conduced the hemoglobin and hematocrit analyses.  Chapter 3 is based on works conducted in Hagerman Fish Culture Experimental Station in Hagerman, Idaho, USA. I designed the experiments with input from Anthony Farrell, Robert Devlin, and Shawn Narum. I assisted the fish catching from natural streams, which was mainly performed by Shawn Narum, Ben Hecht and Nick Hoffman. I carried out the experiments and most result analyses. Amanda Matala performed the Illumina sequencing. Ben Hecht carried out the bioinformatics analyses and drafted the “Genotyping” v  section in “Materials & Methods”. I wrote the rest manuscript. Anthony Farrell, Robert Devlin and Shawn Narum made key contributions to the subsequent edits. Temperature data loggers were placed and retrieved by Shawn Narum. Chapter 4 is also based on works conducted in Hagerman Fish Culture Experimental Station in Hagerman, Idaho, USA. I designed the experiments with input from Anthony Farrell, Robert Devlin, and Shawn Narum. I performed physiological experiments and prepared RNA sequencing library. Amanda Matala conducted the Illumina sequencing. Shawn Narum conducted the differential gene expression analyses and I analyzed the rest results. I drafted the writing. Anthony Farrell, Robert Devlin, and Shawn Narum provided revisions to this manuscript. vi  Table of Contents Abstract ................................................................................................................................... ii Preface .................................................................................................................................... iv Table of Contents ................................................................................................................... vi List of Tables ........................................................................................................................ xiii List of Figures ........................................................................................................................ xv List of Abbreviations .......................................................................................................... xvii Acknowledgements ................................................................................................................ xx Chapter 1: Introduction .......................................................................................................... 1 1.1 Characterization of thermal tolerance ....................................................................... 1 1.1.1 Critical temperatures ............................................................................................. 2 1.1.2 Aerobic scope and thermal performance .............................................................. 3 1.2 Cardiac response to temperature and its potential to limit aerobic scope................. 8 1.2.1 Role of heart rate in limiting aerobic scope .......................................................... 9 1.2.2 Cardiac arrhythmia at high temperatures ............................................................ 11 1.2.3 Thermal performance curve for maximum heart rate ......................................... 12 1.3 Cellular response to temperature and its potential to limit thermal tolerance ........ 15 1.3.1 Cellular stress response....................................................................................... 15 1.3.2 Cardiac myocyte responses................................................................................. 16 1.3.3 Using transcriptome analysis to study cellular stress ......................................... 19 1.3.4 Transcriptomic response under different thermal regimes ................................. 20 1.4 Genomic basis of thermal adaptation ..................................................................... 20 1.5 High-throughput sequencing .................................................................................. 23 vii  1.6 Thermal adaptation of Oncorhynchus mykiss ......................................................... 24 1.6.1 Known thermal requirements of O. mykiss......................................................... 25 1.6.2 Adaptation of O. mykiss populations to warm climates ..................................... 26 1.7 Thesis objectives and chapters................................................................................ 28 Chapter 2: Selection for upper thermal tolerance in rainbow trout (Oncorhynchus mykiss) ..................................................................................................................................... 33 2.1 Introduction............................................................................................................. 34 2.2 Materials and methods ............................................................................................ 37 2.2.1 Fish culture and rearing conditions..................................................................... 37 2.2.2 Critical thermal maximum .................................................................................. 38 2.2.3 Routine and maximum metabolic rate ................................................................ 39 2.2.4 Maximum heart rate ............................................................................................ 41 2.2.5 Hemoglobin and hematocrit analyses ................................................................. 43 2.2.6 Statistical analyses .............................................................................................. 43 2.3 Results .................................................................................................................... 44 2.3.1 Critical thermal maximum (CTMAX) ................................................................... 44 2.3.2 Metabolic rates and absolute aerobic scope........................................................ 45 2.3.3 Maximum heart rate (fH,max) ................................................................................ 46 2.3.4 Ventricle mass, hemoglobin (Hb) and hematocrit (Hct) .................................... 47 2.4 Discussion ............................................................................................................... 47 2.4.1 Critical thermal maximum (CTMAX) ................................................................... 48 2.4.2 Metabolic rates and absolute aerobic scope........................................................ 50 2.4.3 Maximum heart rate (fH,max) ................................................................................ 52 viii  2.4.4 Ventricle mass, hemoglobin (Hb) and hematocrit (Hct) .................................... 53 2.4.5 Conclusion .......................................................................................................... 53 Chapter 3: Intraspecific genomic variation and associations related to thermal performance in redband trout (Oncorhynchus mykiss gairdneri) ..................................... 61 3.1 Introduction............................................................................................................. 62 3.2 Materials and methods ............................................................................................ 66 3.2.1 Fish culture and rearing conditions..................................................................... 66 3.2.2 Phenotyping ........................................................................................................ 68 3.2.2.1 Critical thermal maximum (CTMAX) ........................................................... 68 3.2.2.2 Routine and maximum metabolic rate ........................................................ 68 3.2.3 Genotyping ......................................................................................................... 70 3.2.3.1 DNA extraction ........................................................................................... 70 3.2.3.2 Doubled haploid samples ............................................................................ 70 3.2.3.3 RAD library preparation and sequencing ................................................... 71 3.2.3.4 De Novo SNP discovery and genotyping.................................................... 71 3.2.3.5 Annotation and mapping of RAD tag sequences ........................................ 73 3.2.4 Population genetics ............................................................................................. 74 3.2.4.1 Genetic diversity (heterozygosity) .............................................................. 74 3.2.4.2 Population differentiation (FST) .................................................................. 74 3.2.4.3 Outlier test .................................................................................................. 74 3.2.5 Genetic association ............................................................................................. 75 3.2.6 Statistics .............................................................................................................. 76 3.3 Results .................................................................................................................... 77 ix  3.3.1 Habitat temperature ............................................................................................ 77 3.3.2 Critical thermal maximum (CTMAX) ................................................................... 78 3.3.3 Aerobic scope ..................................................................................................... 79 3.3.4 Genetic diversity ................................................................................................. 80 3.3.5 Loci under positive adaptive selection ............................................................... 81 3.3.6 Association analysis............................................................................................ 82 3.3.7 Annotation of candidate loci ............................................................................... 83 3.4 Discussion ............................................................................................................... 83 3.4.1 Intra-specific differences in redband trout .......................................................... 84 3.4.2 Population genetics ............................................................................................. 89 3.4.3 Genome-wide association study ......................................................................... 93 3.4.4 Conclusion .......................................................................................................... 95 Chapter 4: Cardiac transcriptomic response to acute warming in redband trout populations (Oncorhynchus mykiss gairdneri) ................................................................... 112 4.1 Introduction........................................................................................................... 113 4.2 Materials and methods .......................................................................................... 117 4.2.1 Fish culture and rearing condition .................................................................... 117 4.2.2 Maximum heart rate .......................................................................................... 117 4.2.3 RNA Sequencing .............................................................................................. 118 4.2.3.1 Total RNA extraction ............................................................................... 118 4.2.3.2 RNA library preparation ........................................................................... 118 4.2.3.3 Illumina sequencing .................................................................................. 119 4.2.3.4 Alignment to reference O. mykiss mRNA ................................................ 119 x  4.2.4 Differentially expressed genes analysis ............................................................ 120 4.2.5 Annotation and gene ontology enrichment analysis ......................................... 120 4.2.6 Association transcriptomics .............................................................................. 121 4.2.7 Statistics ............................................................................................................ 121 4.3 Results .................................................................................................................. 122 4.3.1 Maximum heart rate (ƒH,max) ............................................................................. 122 4.3.2 RNA sequencing ............................................................................................... 123 4.3.3 Effect of temperature on cardiac gene expression ............................................ 124 4.3.3.1 Redband trout as a species ........................................................................ 124 4.3.3.2 Intraspecific regulation of gene expression .............................................. 125 4.3.4 Differentially expressed (DE) genes between desert and montane populations127 4.3.5 Association between gene expression and arrhythmia temperature (TAR) ....... 128 4.4 Discussion ............................................................................................................. 128 4.4.1 Maximum heart rate (ƒH,max) in response to warming ...................................... 129 4.4.2 Effect of temperature on cardiac gene expression ............................................ 131 4.4.2.1 Important pathways in response to warming ............................................ 132 4.4.2.2 Intraspecific pattern of gene expression ................................................... 135 4.4.3 Gene expression among populations ................................................................ 137 4.4.4 Association between gene expression and arrhythmia temperature (TAR) ....... 139 4.4.5 Conclusion ........................................................................................................ 140 Chapter 5: Discussion and conclusions .............................................................................. 158 5.1 Study of thermal adaptation to local environments .............................................. 159 5.2 Thermal adaptation of CTMAX .............................................................................. 161 xi  5.3 Thermal adaptation of Fry Curve ......................................................................... 163 5.4 Role of fH.max in limiting AAS .............................................................................. 165 5.5 Cellular response to acute warming ...................................................................... 167 5.6 Genetic changes in thermal adaptation ................................................................. 170 5.7 Conservation and management implications ........................................................ 173 5.8 Future directions ................................................................................................... 174 References ............................................................................................................................. 183 Appendices ........................................................................................................................... 211 Appendix A Additional tables and figures for Chapter 1. ................................................ 211 A.1 Critical thermal maximum (CTMAX) for nine sockeye salmon populations ..... 211 A.2 The CTMAX for various populations of O. mykiss ............................................. 212 Appendix B Additional tables and figures for Chapter 3. ................................................ 213 B.1 Correlation between CTMAX and body mass ..................................................... 213 B.2 Body mass corrected RMR, MMR, AAS and FAS of redband trout. .............. 214 B.3 Quadratic fitting analysis for AAS in redband trout populations. .................... 215 B.4 Parameters in quadratic fitting analysis for aerobic scope ............................... 215 Appendix C Additional tables and figures for Chapter 4. ................................................ 216 C.1 Alignment of RNA sequencing reads to a reference mRNA of rainbow trout . 216 C.2 Fold change of genes during warming for each redband trout population ....... 218 C.3 Gene ontology distribution of significantly regulated genes ............................ 219 C.4 Summary of significantly regulated genes in selected pathways. .................... 220 C.5 Fold change of heat shock proteins mRNA in redband trout populations ........ 222 C.6 KEGG pathway analysis for metabolism pathways ......................................... 225 xii  C.7 Differential gene expression between redband trout populations ..................... 226 C.8 Genes that are associated with cardiac arrhythmia temperature. ...................... 227  xiii  List of Tables Table 2.1 Body size of the PFRC rainbow trout O. mykiss used in each measurement. ........ 55 Table 2.2  Topt, Tpej, Tcrit of absolute aerobic scope (AAS) in three family groups of PFRC rainbow trout. .......................................................................................................................... 55 Table 2.3 Rate transition temperatures of maximum heart rate in three family groups of PFRC rainbow trout. ............................................................................................................... 56 Table 2.4 Wet ventricle mass, hematocrit (Hct) and hemoglobin concentration (Hb) for PFRC rainbow trout. ............................................................................................................... 56 Table 3.1 Summary of summer water temperatures logged for Little Jacks Creek, Keithley Creek and Fawn Creek in Idaho, USA. .................................................................................. 96 Table 3.2 Total time of stream temperatures above the test temperatures in AAS measurements for three redband trout O. mykiss gairdneri habitats from July 10 to September 10. ........................................................................................................................................... 96 Table 3.3 Body size of redband trout in CTMAX and AAS measurements. ............................ 96 Table 3.4 Topt amd Tpej for AAS in redband trout populations. .............................................. 97 Table 3.5 Average expected (HE) heterozygosity of redband trout populations. Different superscripted letters indicate significant differences. ............................................................. 97 Table 3.6 Pairwise FST distance of redband trout populations. .............................................. 97 Table 3.7 Summary of outlier loci and the nearby genes (within 15 kb). .............................. 98 Table 3.8 Markers significantly associated with critical thermal maximum in GWAS analysis and the nearby genes within 15 kb range in reference genome. ........................................... 100 Table 3.9 Annotation for genes within the 15 kb flanking region of significant markers from outlier tests and GWAS. ....................................................................................................... 101 xiv  Table 4.1 Number of heart samples harvested during acute warming for RNA sequencing. .............................................................................................................................................. 142 Table 4.2 Maximum heart rate (ƒH,max) at 15ºC, 20ºC and the temperature with peak ƒH,max (TPEAK) for each redband trout population. ........................................................................... 142 Table 4.3 Thermal indices for maximum heart rate (ƒH,max) in redband trout. ..................... 143 Table 4.4  Number of significantly regulated genes during acute warming for each population (FDR < 0.05). ........................................................................................................................ 143 Table 4.5 Gene Ontology (GO) enrichment analysis for the significantly up-regulated genes. .............................................................................................................................................. 144 Table 4.6 Gene Ontology (GO) enrichment analysis for the down-regulated DE genes. GO terms are categorized into biological process (BP), cellular component (CC) and molecular function (MF)........................................................................................................................ 146 Table 4.7 Number of differentially expressed genes between redband trout populations.... 147 Table 4.8 Differential gene expression between desert and montane populations. .............. 148  xv  List of Figures Figure 1.1 Parameters used to characterize thermal tolerance. .............................................. 31 Figure 1.2 Cellular response to temperature and the Fry Curve. ............................................ 32 Figure 2.1 Temperature of water passing through Pemberton Freshwater Research Centre (PFRC) ponds from January to May, 2014. ............................................................................ 57 Figure 2.2 CTMAX of five family groups of PFRC rainbow trout. .......................................... 57 Figure 2.3 Routine (RMR) and maximum (MMR) metabolic rate, and absolute aerobic scope (AAS) for three family groups of PFRC rainbow trout. ......................................................... 58 Figure 2.4 Maximum heart rate (fH,max) in response to temperature increase in three family groups of rainbow trout. ......................................................................................................... 59 Figure 3.1 Water temperatures of Little Jacks Creek, Keithley Creek and Fawn Creek in the Snake River tributary of southern Idaho. .............................................................................. 104 Figure 3.2 CTMAX of redband trout populations. .................................................................. 105 Figure 3.3 Effect of temperature on metabolic rate and aerobic scope in redband trout populations. ........................................................................................................................... 106 Figure 3.4 Outlier loci analyses in redband trout populations using Lositan and BayeScan. .............................................................................................................................................. 107 Figure 3.5 Association between CTMAX and genotypes of the candidate outlier loci. ......... 108 Figure 3.6 Association between AAS and genotypes of the candidate outlier loci. ............. 109 Figure 3.7 Principal component analyses of genetic differentiation in redband trout using neutral loci (A) and outlier loci (B). ..................................................................................... 110 Figure 3.8 Summary of genome-wide association study results for CTMAX. ....................... 111 Figure 4.1 Correction of maximum heart rate for body mass using four scaling exponents.150 xvi  Figure 4.2 Percentage of fish showing arrhythmia during acute warming for each redband trout population. .................................................................................................................... 151 Figure 4.3 Number of significantly regulated transcripts during acute warming in redband trout. ...................................................................................................................................... 152 Figure 4.4 Genes that are most significantly up- and down- regulated during acute warming. .............................................................................................................................................. 152 Figure 4.5 Pathways of significantly regulated genes in cardiac myocyte functions. .......... 153 Figure 4.6 Pathway of significantly regulated genes in glycolysis and citric acid cycle. .... 154 Figure 4.7 Number of significantly regulated transcripts for each population during acute warming. ............................................................................................................................... 155 Figure 4.8 Number of enriched gene ontology terms in enrichment analysis. ..................... 155 Figure 4.9 Correlation between the number of overall differentially expressed (DE) genes and FST (A) as well as thermal regimes (B). ......................................................................... 156 Figure 4.10 Gene expression that significantly associated with cardiac arrhythmia temperature. .......................................................................................................................... 157 Figure 5.1 Optimum and upper critical temperatures of O. mykiss populations. ................. 179 Figure 5.2 RMR, MMR and AAS of O. mykiss populations. ............................................... 180 Figure 5.3 Fry Curves of several fish species. ...................................................................... 181 Figure 5.4 Maximum heart rate of O. mykiss populations. ................................................... 182  xvii  List of Abbreviations AAS absolute aerobic scope (calculated as MMR−RMR) BH-FDR Benjamini - Hochberg adjusted p - value BY-FDR Benjamini - Yekutieli adjusted p - value CHR cellular homeostasis response CRITFC Columbia River Inter-Tribal Fish Commission CSR cellular stress response CTMAX critical thermal maximum DE differential expression of genes DNA deoxyribonucleic acid ECG electrocardiogram F redband trout from Fawn Creek, Idaho, USA FAS factorial aerobic scope (calculated as MMR/RMR) FDR false discovery rate fH heart rate fH,max maximum heart rate FST fixation index gDNA genomic DNA GO gene ontology GWAS genome-wide association study Hb haemoglobin Hct haematocrit xviii  HE heterozygosity HSP heat shock protein HSR heat shock response H-W Hardy–Weinberg K redband trout from Keithley Creek, Idaho, USA K×LJ Hybrid of Keithley × Little Jacks Redband Trout KEGG Kyoto Encyclopedia of Genes and Genomes LJ redband trout from Little Jacks Creek, Idaho, USA Mb body mass MMR maximum metabolic rate ?̇?𝑂2 oxygen consumption rate (mg O2 kg−1 h−1) MS222 tricaine methanesulfonate NaHCO3 sodium bicarbonate OCLTT oxygen- and  capacity- limited thermal tolerance [O2]t0 oxygen concentration at time t0 [O2]t1 oxygen concentration at time t1 PFRC Pemberton Freshwater Research Centre PSV paralogous sequence variants PCA principal component analysis QTL quantitative trait locus RAD restriction site associated DNA RMR routine metabolic rate xix    RNA ribonucleic acid rRNA ribosome RNA RVM relative wet ventricle mass to body mass SMR standard metabolic rate SNP single nucleotide polymorphism TAB Arrhenius breakpoint temperature Tcrit critical temperature Topt optimum temperature TPEAK temperature where peak fH,max is reached Tpej pejus temperature TQB Q10 breakpoint temperature UILT upper incipient lethal temperature UTT upper thermal tolerance V volume Vs stroke volume of the heart   xx  Acknowledgements First and foremost, I would like to sincerely thank my thesis supervisors, Drs. Tony Farrell and Bob Devlin, for their patience, support, mentorship and friendship. Tony has put in countless efforts to help me expand my comfortable circle (sometimes with pushes) in many aspects, which are all crucial to my success. As a global collaborator himself having research projects in all continents, Tony also supported my visit to five labs in three countries, where I had the opportunity to communicate, share and collaborate.  I thank Bob Devlin for providing me tremendous freedom and support, even when my research organisms and questions were somewhat beyond his primary interests. Bob has taught me lots of philosophies in scientific research, e.g. "always step back and ask biological questions, without which it is just chemistry or comparison of strings”. Through several years of working with Tony and Bob at UBC, I have become not only a better scientist, but also a better person. I also sincerely appreciate my dissertation committee: Drs. Scott Hinch, Rick Taylor, and Shawn Narum, for their encouragements, questions and comments. Scott gave me an opportunity to work in his lab on my first experiment at UBC, where I gained plenty of training and hands-on experience with salmon. I would like to thank Rick for his knowledge in fish biology and input in some critical evolution concepts. I would like to express thankfulness to Shawn, who guided me into the field of evolution genomics. Shawn generously provided me full access to his state-of-the-art genetic lab at the Columbia River Inter-Tribal Fish Commission (CRITFC) in Hagerman, Idaho, where I was given adequate freedom and trust to develop my skills in high throughput sequencing. I greatly appreciate his support and am proud to have worked with him. I would like to thank the members of Farrell Lab, past and present (Linda Hanson, Matt Casselman, Christine Verhille, Chris Wilson, Georgina Cox, Helen Drost, Joy Wu, Beth Ferreira, Yangfan Zhang, Hamid Safi, Sabine Laguë, Rachel Sutcliffe, Matt Gilbert, James Marchant, Mandy Lo and Adam Goulding). I would like to particularly thank Dr. Katja Anttila, my “wet-lab supervisor” and a wonderful friend. Katja helped me in the development of my thesis and guided me through my earlier years of Ph.D. study.  I appreciate Chris, Adam and Mandy for their efforts in improving the quality of my presentation and thesis. I thank Chris, Adam, Ryan Shartau and Nicolas Muñoz for their assistance in the 1,200 km field trips from British Columbia to Idaho. I also thank all other faculties and students of the comparative physiology group at the Department of Zoology, UBC. I own many thanks to Dr. Ronald Hardy at the Fish Culture Experimental Station (Hagerman, Idaho) for his support and encouragement in my project, sharing of personal stories and discussion of fish biology, which broadened my horizons. I am indebted to the xxi  entire genetic lab of CRITFC: with special mention to Amanda Matala, Nick Hoffmann, Vanessa Morman, Travis Jacobson, Megan Moore, Lori Maxwell and Stephanie Harmon, Jeff Stephenson and Nate Campbell. This phenomenal group of people demonstrated incredible technical expertise, knowledge and willingness to help. I am very grateful to Ben Hecht for his professional statistic advices and technical training in bioinformatics, without which I would not have been able to present the genetic data in Chapter 3.  I also thank other fellow graduate students (Alejandro Villasante and Andreas Brezas) and visiting scholars (Dr. Cristian Araneda, Christina Der, Caroline Nebo and Qingchao Wang) for the intellectual discussion and the celebration of good times in the outlying but bustling dormitory. I would like to thank Drs. Michael Snow and Craig Lawrence at the Department of Fisheries, Western Australia, for their support and commitment in my thesis. I thank Tony Church, Terry Cabassi and Chris Church at the Pemberton Freshwater Research Center, for their diligent work in fish rearing, for their support in my experiment, and for making me feel at home. Additionally, my trip to Australian would have been several months longer if without the help from Neil Rutherford, who did an amazing job to build the fish rearing facility and prepared everything I requested for my experiment.  I would also like to acknowledge the funding from China Scholarship Council, Department of Zoology (UBC), Graduate and Postdoctoral Studies (UBC), Faculty of Science (UBC), the Canadian Society of Zoologists, Natural Sciences and Engineering Research Council (NSERC) grants to Dr. Tony Farrell and a Canadian Regulatory System for Biotechnology grant to Dr. Bob Devlin.   Last, but not least, this dissertation would not be possible without the support from my family. I thank my parents: Chaomin Chen and Meihong Lu, for their faith in me and steadfast support throughout my student career. I would like to express my deepest gratitude to my wonderful wife, Lijuan Hu, for her unconditional love, for her sensible indulgence and insistence, for the encouragement and support, and for being my best friend.  1  Chapter 1: Introduction 1.1 Characterization of thermal tolerance Body temperature of most fishes closely matches the ambient environment (Clausen, 1934), even when temperature changes acutely (e.g. 1°C min-1) (Lutterschmidt and Hutchison, 1997a). This temperature conformance is due to the efficient exchange of heat as well as gases at the gills and the lack of an insulation mechanism to act against the high heat capacity of water. Therefore, biochemical processes, physiological performances and geographical distributions of ectothermic fish are directly affected and sometimes limited by water temperature (Fry, 1947). Thermal tolerance is bounded by the maximum and minimum critical temperatures, which bracket the optimum and sub-optimum thermal ranges, where fish spend most of their lives (Fry, 1947; Hofmann and Todgham, 2010). Thermal performance curves (also called thermal reaction norm) are often used to provide an overall picture of functional performances across the entire thermal window (Schulte et al., 2011). Thermal effects on fish performance was first conceptualized about 70 years ago by Dr. Fred Fry using incipient lethal levels of temperature and thermal performance curves for aerobic scope (Fry, 1947). The Fry aerobic scope curve (often called a “Fry Curve”) was recently adopted as a framework for the oxygen- and capacity- limited thermal tolerance (OCLTT) hypothesis (Pörtner, 2002; Pörtner and Farrell, 2008), which emphasizes the role of oxygen limitation and systemic capacity limitation in cardiorespiratory functions during heat stress. To explore the mechanisms of thermal adaptation, this thesis measures critical temperatures to examine the thermal limits for survival and uses OCLTT to evaluate the thermal indices for performance in populations 2  from diverse climates within a single species, the rainbow trout (Oncorhynchus mykiss). In doing so, I acknowledge that thermal adaptation, which results in altered phenotypes, is complex and needs comprehensive examination at all biological levels. Also, I only focus on the upper thermal tolerance to provide relevance to the global warming scenarios. 1.1.1 Critical temperatures Every fish species has a lower and an upper critical temperature that define its thermal limits for survival. Under acute warming situations, fish can only survive upper critical temperatures for a limited time, and thus largely rely on behavioral thermoregulation by seeking cooler micro-environments (Block et al., 1984; Nielsen et al., 1994).  Critical temperatures for aquatic ectotherms are different among species (interspecific) and among populations within a species (intraspecific), which is a result of thermal adaptation. Interspecifically, upper critical temperature varies from as low as 13.3°C in the Antarctic icefishes (Chionodraco rastrospinosus) (Beers and Sidell, 2011) to >42.5°C in the California pupfish (Cyprinodon salinus) (Stuenkel and Hillyard, 1981). Intraspecific differences are less profound but still commonly exist (Beitinger et al., 2000). Upper critical temperatures have been widely measured using either thermal static (upper incipient lethal temperature, UILT) or thermal dynamic (critical thermal maximum, CTMAX) methodologies (Fry, 1947; Becker and Genoway, 1979; Bennett and Beitinger, 1997). Both UILT and CTMAX have been applied to measure upper critical temperatures in fish species (Beitinger and Bennett, 2000; Beitinger et al., 2000), but they differ substantially when it comes to the time and magnitude of heat exposure.  The UILT measures the upper temperature where 50% mortality can be observed for a given acclimation temperature (Armour, 1991; Beitinger and Bennett, 2000). The ultimate UILT represents the maximum 3  acclimation temperature where 50% of fish can survive for a long period (e.g. > one week in experimental situations).   Different from the static UILT method, CTMAX is a means of quantifying the upper critical temperature by raising water temperature at a constant rate (usually 0.3°C min-1) until test organisms display a physical disorganization response that indicates the loss of ability to escape from the adverse thermal condition, which promptly leads to fish mortality (Becker and Genoway, 1979; Beitinger and Bennett, 2000). Thus, UILT measures the critical temperature for long-term survival, while CTMAX is for short-term survival (within minutes) (Fry, 1947).  Compared to UILT, CTMAX is less time-consuming to measure and, more importantly, can be estimated for individual fish. Hence, CTMAX is now the most widely used methodology to quantify upper critical temperature in ectothermic fishes (Lutterschmidt and Hutchison, 1997b). The CTMAX is typically above the mean maximal habitat temperatures in most species and is rarely experienced by animals throughout their life cycle. Despite that, occasional occurrences of extreme temperatures do cause immediate mortality (Huntsman, 1942; Huntsman, 1946; Bailey, 1955; Brooker et al., 1977; Mundahl, 1990; Durham et al., 2006) and impose strong and direct selection pressures. As a result, critical temperatures have presented an intraspecific pattern among populations from habitats with different thermal maxima (Carline and Machung, 2001; Fangue et al., 2006; Kelley et al., 2011; also see Appendix A.1 for a preliminary study in sockeye salmon, Oncorhynchus nerka). 1.1.2 Aerobic scope and thermal performance Within the thermal tolerance limits, temperature is an important factor that influences the performance of activities such as locomotion, predator avoidance, food digestion and assimilation. The capacity and thermal limits of performance can be evaluated by the thermal 4  dependence of aerobic scope (Fry, 1947; Farrell, 2009), because oxidative metabolism is the only long-term option to supply adenosine triphosphate (ATP) for work. Absolute aerobic scope (AAS) is calculated as the difference between standard metabolic rate (SMR) and maximum metabolic rate (MMR), but using oxygen uptake ( ?̇?𝑂2) from the water by fish as an index of metabolic rate (Figure 1.1 A). Factorial aerobic scope (FAS) is a different but widely used expression of aerobic scope and is calculated as MMR/SMR. In this thesis, and in most other studies, the expression of “aerobic scope” refers to AAS except when specified for FAS. In this thesis I replace SMR with routine metabolic rate (RMR) because minor activity may exist during respirometry measurements. RMR is slightly higher (< 15%) than SMR (Chabot et al., 2016), which therefore slightly underestimates AAS and FAS. Thus, RMR is defined here as the energy required to maintain a basic post-absorptive state in the absence of reproduction, plus some minor cost for growth and activity. The RMR may also have greater ecological relevance than SMR because fish in natural conditions are rarely in SMR state. The MMR is defined as the maximum energy expenditure that can be produced by aerobic metabolism and is usually measured during prolonged swimming tests (Fry and Hart, 1948; Brett, 1964; Clark et al., 2011; Eliason et al., 2011) or post-exhaustion (Reidy et al., 1995; Casselman et al., 2012; Clark et al., 2013; Ferreira et al., 2014; Gräns et al., 2014). While O2 uptake rate can be easily measured in respirometry experiments, one big challenge is to obtain the corresponding physiological status during measurements. To control the stress-related activities during acute measurements for RMR, water needs to be gradually heated from the holding temperature to test temperature, which gives fish time to adjust. Despite the effort, chances still exist for an overvalued RMR and an undervalued MMR, which in combination results in an 5  underestimated AAS and FAS. Nevertheless, it should not affect the intraspecific and interspecific comparisons because mismatch between the measured and true values should be proportionally constant across measurements. The effect of temperature on AAS can be described by a Fry Curve (i.e. an acute thermal performance curve for AAS in this thesis), which has three distinct phases: (1) a rising phase near the critical thermal minimum (CTMIN); (2) a plateau and peak phase around an optimum temperature (Topt); and (3) a declining phase near CTMAX (Figure 1.1 B) (Fry, 1947; Fry, 1948; Eliason et al., 2011; Farrell, 2016). By examining the position, height and width of each phase in the Fry Curve, thermal performance of a fish can be assessed (Pörtner, 2002; Pörtner and Farrell, 2008).  First, the plateau and peak phase of AAS between lower and upper “pejus” (Latin word meaning “getting worse”) temperatures (Tpej) represents the optimum thermal window. The broader an optimum thermal window is, the greater ability a fish has to maintain maximum functional performance across a wide range of temperatures. Eurythermal fish typically have a broader optimum thermal window than stenothermal fish. The Tpej for AAS is set at some arbitrary level below the peak value, e.g. 90% of peak AAS (Eliason et al., 2011; Farrell, 2016), but this level could be set higher or lower.  In my thesis, I also adopt 90% as the threshold for Tpej. Second, the rising and declining phases between the Tpej and critical temperatures are considered the pejus zone or sub-optimum zone. The declining phase is often steeper and narrower than the rising phase, which leads to a left skewed Fry Curve. Third, the Fry Curve is bounded by lower and upper Tcrit, where AAS is zero (i.e. no sustained activity is possible) (Figure 1.1 B), which means survival is only temporary. Due to the difficulty of mathematically modeling Fry Curves, some authors use an arbitrary 10% of peak AAS to estimate Tcrit (e.g. Ferreira et al., 2014). Regardless, upper 6  Tcrit and CTMAX are close in value despite being different measurement endpoints. CTMAX is usually above Tcrit because fish resort to anaerobic metabolism before losing their ability to maintain an upright position (the endpoint in CTMAX measurements). Compared to Tcrit and CTMAX, optimum thermal window and pejus zones are of more functional interest because they are critical to perform fitness-related functions (e.g. growth and swimming) and are often the most encountered thermal ranges for fishes in natural conditions. However, there are concerns when using OCLTT to predict thermal performances for eurythermal fish species (Norin et al., 2014) and benthic fish species (Gräns et al., 2014).  For example, in eurythermal species, AAS has been found either to be relatively thermal insensitive across wide range of temperatures (10-33°C) in killifish (Fundulus heteroclitus) (Healy and Schulte, 2012) or to reach the maximum at a temperature (38°C) that is just 3°C below its Tcrit in barramundi (Lates calcarifer) (Norin et al., 2014). Despite that, OCLTT hypothesis seems valid in salmonids with abundant evidences from swimming performance tests (Farrell et al., 2008). For example, in pink salmon (Oncorhynchus gorbuscha), maximum attainable swimming speed is impaired at temperatures beyond Topt for AAS (Clark et al., 2011). Application of these aerobic scope data to estimate the ecological performance of fishes remains a challenge and is in need of much more work.  Nevertheless, in sockeye salmon populations from the Fraser River (British Columbia, Canada), AAS and critical swimming velocity have similarly shaped thermal performance curves with similar Topt values in Weaver Creek (14.5°C and 15.2°C, respectively) and Gates Creek (16.6°C and 16.2°C, respectively) populations (Lee et al., 2003b). Further studies with a larger sample size strengthened the discovery of population differences within sockeye salmon by comparing eight populations (including the data for Weaver and Gates Creek populations) 7  and provided considerable mechanistic support (Eliason et al., 2011; Eliason et al., 2013). From an ecological perspective, the population-specific Fry Curves and other aspects of their cardiac physiology seem to be tailored for the local environmental challenges during upstream swimming to perform the once-in-a-life-time spawning event (Eliason et al., 2011). The tailoring of a Fry Curve to the local thermal environment is likely an evolutionary strategy and exists across species (Farrell, 2009). For instance, five tropical reef fish species of Australia have Topt values for AAS (29-31°C) close to the current habitat temperatures (Rummer et al., 2014). In two other reef fish species, aerobic scope significantly decreased by ~50% when the water temperature was increased from 29°C to 31°C (over 1-2 days), which means the current habitat temperature is quite close to their upper thermal limit (Nilsson et al., 2009). Besides swimming performances, thermal limits for growth can also be related to the Fry Curve, especially for maximum growth rate (Brett, 1976). For instance, sockeye salmon has a common optimum temperature (15°C) both for growth rate (Brett et al., 1969) and aerobic scope (Brett, 1964, 1976). Similarly, in rainbow trout, the optimum temperature for aerobic scope is between 15-20°C (Dickson and Kramer, 1971; Anttila et al., 2013), which is equivalent to the optimum thermal range for growth rate (Hokanson et al., 1977; Myrick and Cech, 2005). Fry Curves have been mostly used to describe the thermal performances at optimal or sub-optimal temperatures largely because of the difficulties of holding fish at lethal critical temperatures. Nonetheless, Eliason et al., (2011) was able to measure RMR and MMR for one sockeye salmon population (Quesnel River) at high temperatures and obtained a better estimate of Tcrit (25.8°C), which is interestingly the same as the CTMAX of a 8  geographically nearby (thus possibly genetically similar) population (25.5°C) from Horsefly River (Appendix A.1). 1.2 Cardiac response to temperature and its potential to limit aerobic scope The declining phase of a Fry Curve is caused by the peaking and decline in MMR, which fails to maintain AAS above RMR. The underlying mechanism of a restricted MMR has been associated with the capacity limitation of the circulatory system in delivering O2 to tissues for ATP production (Pörtner and Knust, 2007; Pörtner and Farrell, 2008; Clark et al., 2008). The critical observation behind this association is the progressive accumulation of plasma lactate as animals are warmed to temperatures beyond Topt and towards Tcrit (Meka and McCormick, 2005; Clark et al., 2008; Jeffries et al., 2012), and certainly as fish approach their critical temperatures (Steinhausen et al., 2008; Iftikar and Hickey, 2013). Release of lactate from cells is an indicator of anaerobic metabolism, which is recruited when O2 supply is insufficient to fulfill the ATP demand.  The linkage between O2 uptake and cardiac function can be explored by inspecting the Fick Equation for O2 consumption: M ̇ O2 = fH∙ VS ∙ ([CaO2] - [CvO2]) ?̇?𝑂2, rate of O2 consumption (mg min-1); fH, heart rate (beats min-1); VS, stroke volume (ml beat-1); CaO2, arterial O2 content (mg O2 ml-1); CvO2, venous O2 content (mg O2 ml-1). 9  1.2.1 Role of heart rate in limiting aerobic scope To increase or maintain AAS during acute warming conditions, the increase in maximum M ̇ O2 must be equivalent to or exceed the exponential increase in routine M ̇ O2 through either cardiac output (fH∙VS) or tissue O2 extraction (CaO2 - CvO2), or some combination of both.   The O2 content of blood is determined primarily by a fish’s hemoglobin concentration and the partial pressure of O2 in the blood. In other words, each component in Fick Equation has the potential to interrupt the increase in maximum M ̇ O2 and hence limits AAS at temperatures above Topt. Among all the components, fH has been proposed to be the main dictator in both RMR and MMR, and thus AAS for three reasons.   First, during acute warming, fish invariably increase fH with no appreciable change in VS under both resting and swimming state (Gollock et al., 2006; Sandblom and Axelsson, 2007; Clark et al., 2008; Steinhausen et al., 2008). In rainbow trout, fH changes with temperature at Q10 > 2 (Aho and Vornanen, 2001; Verhille and Farrell, 2012), meaning fH can double with a 10°C temperature increase (Figure 1.1 C), due to the intrinsic response to warming and extrinsic autonomic regulation (Farrell and Jones, 1992; Sandblom and Axelsson, 2011; Ekström et al., 2014). Stroke volume, however, is thermally unresponsive in both the resting and swimming fish, although swimming fish have a higher VS and fH than resting fish (Steinhausen et al., 2008). Second, fH has a capacity limitation. Peak fH is reached at temperatures (TPEAK) near CTMAX in resting fishes (i.e. in RMR status) (Brett, 1971; Steinhausen et al., 2008), but at lower temperatures in swimming fish (i.e. in MMR status) (Steinhausen et al., 2008; Clark et al., 2011; Eliason et al., 2013). The different TPEAK is largely because swimming per se stimulates a higher fH (Altimiras and Larsen, 2000; Steinhausen et al., 2008; Eliason et al., 10  2013) to increase cardiac output, which forces the swimming fish to have a higher fH than the resting fish at common temperatures. Third, if there is a limitation in the cardiac activity, fish must resort to changes in tissue O2 extraction, according to the Fick equation, to increase M ̇ O2. Indeed, tissue O2 extraction increases with warming and is higher in swimming fish than resting fish (Steinhausen et al., 2008; Eliason et al., 2013). However, the difference in O2 extraction between swimming and resting fish during warming is either unchanged (Eliason et al., 2013) or constantly increases to the pejus zone (Steinhausen et al., 2008), which makes tissue O2 extraction a possible contributor for the ascending phase of Fry Curve but less likely to be the limiting factor for the peaking and declining phase. The contribution of O2 extraction to ?̇?𝑂2 is also limited because it largely rely on the decrease in CvO2. The CaO2 and PaO2 remain constant during acute warming (Steinhausen et al., 2008; Eliason et al., 2013) owning to the increased ventilation frequency and efficient counter-current gas exchange at gills. In addition, neither swimming nor resting fish release splenic store of red blood cells during acute warming to increase the hemoglobin concentration in blood (Eliason et al., 2013). Even if the stored red blood cells are released, O2 carrying capacity would only increase by around 20~30% (Black et al., 1966; Pearson and Stevens, 1991; Sandblom and Axelsson, 2007), which offers minor help as compared to the magnitude of increase in O2 demand. Furthermore, tissue O2 extraction is ~30% in resting and ~70% in swimming fish (Randall and Daxboeck, 1982), which leaves limited potential for further O2 extraction when fish are swimming even if there is no diffusion limitation.  Altogether, the ascending phase of Fry Curve is primarily driven by an increase in fH that increases cardiac output. A capacity limitation in fH, which further limits cardiac output 11  and M ̇ O2 in swimming fish (MMR), may set the peaking of AAS at the optimum thermal window.  1.2.2 Cardiac arrhythmia at high temperatures Near Tcrit, an additional problem with fH occurs. Heartbeats become dysrhythmic or bradycardiac, which significantly decreases cardiac output (Clark et al., 2008). As the studies of thermal effect on fH grow, there is increased interest on physiological processes that trigger the cardiac failure at supra-optimal temperatures (Vornanen, 2016). Common expressions of the dysrhythmia are missing ventricular but not atrial contractions (or ventricular bradycardia) as seen from a missing QRS complex in the electrocardiogram (ECG), atrium tachycardia (more P waves), irregular ventricular contraction (inconstant R-R interval) and a complete cessation of heart beat (missing cardiac events) (Casselman et al., 2012; Verhille et al., 2013; Ferreira et al., 2014; Muñoz et al., 2014; Sidhu et al., 2014; Anttila et al., 2014b; Badr et al., 2016).  Because cardiac contraction is the result of a series of electric events, the cause of dysrhythmia must be related to the underlying electrical signal generation and conduction, which has recently been hypothesized as temperature-dependent depression of electrical excitation (TDEE) (Vornanen, 2016). Regular heartbeats rely on action potentials (APs) to coordinate atrial and ventricular contractions, while APs depend on a series of transmembrane ion currents. Heart rate is set by the APs of pacemaker cells that are located at the junction between the sinus venosus and atrium, known as SA node. Pacemaker cells have an unstable resting membrane potential, which slowly depolarizes predominantly due to a leaky inward sodium (Na +) current through hyperpolarization-activated cyclic nucleotide-gated channels (Mangoni and Nargeot, 2008; Lakatta et al., 2010; Wilson and Farrell, 2013; 12  Vornanen, 2016) and other ion currents, such as the Na+/Ca2+ exchanger (NCX) produced inwards Na+ current and the T-type Ca2+ channels produced inwards Ca2+ current. When membrane voltage reaches the threshold potential, sarcolemma L-type Ca2+ channels (also called dihydropyridine receptors, DHPR) and sarcoplasmic reticulum ryanodine receptors (RYR)activates the influx of Ca2+, which causes the upstroke of the action potential. Then, K+ channels open and repolarize the cell membrane potential back to the resting state. Consequently, two important currents in action potential are the sodium current (INa), which is responsible for the rapid depolarization in AP Phase 0, and the inward rectifier K+ current (IK1), which repolarizes the AP in Phase 3. It seems that INa is less thermal resistant than IK1 (Vornanen et al., 2014), which breaks the balance between depolarization and repolarization during acute warming. TDEE suggests that different thermal sensitivity of INa and IK1 affects the development of AP, which causes the dysrhythmia events.  For example, a reduced inward INa decreases the amplitude of depolarization, which may fail to trigger the contraction of atrium and ventricle. Similarly, the combined effect of a reduced INa and elevated IK1 increased the relative threshold potential, which could slowly become problematic in developing a conductable AP and ventricular contraction. Thus, the phenomenon of atrial tachycardia and ventricular bradycardia at high temperatures could be due to the deleterious thermal effect on pacemaker APs.  1.2.3 Thermal performance curve for maximum heart rate Due to the aforementioned predominate role of fH in ?̇?𝑂2 and limited contribution of other variables in Fick Equation, rate transition temperatures for AAS should be comparable to those for fH scope (maximum fH (fH,max)-routine fH). Indeed, AAS and fH scope had Topt at similar temperatures (Steinhausen et al., 2008; Casselman et al., 2012; Eliason et al., 2013).  13  Furthermore, the plateau in a Fry Curve is largely because of the failure to increase MMR to match with the exponential increase of RMR, suggesting there is a rate limitation or transition in MMR. Therefore, Casselman et al., (2012) explored the relationship between aerobic scope and fH,max, using pharmacological stimulation by atropine and isoproterenol, and found that the Arrhenius breakpoint temperature (TAB) of fH,max was the same as the Topt for aerobic scope, again suggesting the association between fH and AAS. The thermal performance curve for fH,max has been studied in many other fish species such as in eurythermal goldfish fish (Carassius auratus) (Ferreira et al., 2014) to replicate the earlier work of Dr. Fred Fry on this species, tropical Danio species (Sidhu et al., 2014), other temperate salmonid species (Verhille et al., 2013; Anttila et al., 2014b; Muñoz et al., 2015), and Arctic cod (Boreogadus saida) (Drost et al., 2014; Drost et al., 2016). All these measurements used the same protocol and thus generated comparable results. In general, TAB for fH,max showed good agreement with the already known physiological optimum temperatures for AAS. For example, TAB was below 4°C for Arctic cod (Drost et al., 2014; Drost et al., 2016), between 14-19°C in temperate salmonids  (Verhille et al., 2013; Anttila et al., 2014b; Muñoz et al., 2015),  and over 20°C in goldfish and Danios species (Ferreira et al., 2014; Sidhu et al., 2014). Since the cardiac arrhythmia could affect cardiac output and aerobic scope, it was not surprising that the arrhythmia temperature (TAR) was close to or slightly below either Tcrit for AAS or CTMAX. Indeed, TAR in goldfish was close to the previously measured CTMAX across three acclimation temperatures, however, TAR was much lower than the Tcrit (Ferreira et al., 2014), which is probably because Tcrit was extrapolated using statistical models instead of from direct measurements. Interestingly, TAR always occur at no more than 4°C below CTMAX (Muñoz et al., 2014; Sidhu et al., 2014), which suggests 14  the possible usage of fH,max as a surrogate for CTMAX in some species.  The precision of thermal indices generated from fH,max was also impressive. In three Danio species, species with a higher CTMAX always had a higher TAB and TAR, despite the CTMAX difference is within 3°C (Sidhu et al., 2014). Given that AAS measurements are time-consuming and the components in Fick equation are often measured in very few swimming fishes due to the technical difficulties of making necessary simultaneous measurements, fH,max has been suggested as a useful proxy to examine the thermal performance in fish (Casselman et al., 2012; Farrell, 2016).  In this thesis, I measured the thermal performance curves for both AAS and fH,max, to further explore their interdependence and potential usage in studying thermal adaptation within O. mykiss. Rate transition temperature of fH,max indicates a switch or limitation in the intra-cellular processes, as suggested by the different “activation energy” (Ea) for fH,max in the Arrhenius equation, the plateau and dysrhythmia of fH,max (Figure 1.1 D). The underlying mechanism is still not clear, but could be related to the quantitative and qualitative properties of the molecular system that control cardiac functions. For example, rate-limiting enzymes, which catalyze the slowest steps in a biochemical pathways, could be quantitatively not enough to ensure a faster reaction. Similarly, cross-membrane ion gradients may be disturbed due to the limitation in the number of voltage-gated ion channels. Mitochondrial failure has also been linked to heart failure (Iftikar and Hickey, 2013; Iftikar et al., 2014; Rodnick et al., 2014). In addition, cellular membrane changes its phases during warming from an ordered state to a less ordered state (hyperfluid), which affects the cellular function (McMurchie et al., 1973; Cossins and Prosser, 1978). One way to identify these putative key “weak links” is to study the cellular responses that help defending the adverse thermal effect. 15  1.3 Cellular response to temperature and its potential to limit thermal tolerance Response to stressors is an intrinsic characteristic of all cells when exposed to environmental stimuli that affect homeostasis. Cellular responses can be broadly classified into two types (Figure 1.2), the stress response (CSR) and the homeostasis response (CHR) (Kültz, 2005). CSR is a highly conserved response across taxa and mainly consists of heat shock response (HSR) and oxidative stress response. The main cellular functions affected by CSR include growth control, apoptosis (Fulda et al., 2010), and energy metabolism. Unlike CSR, CHR largely depends on the type of stressor (temperature, salinity, etc.) and is tissue specific. Temperature induced CHR typically invokes exponential changes in rate functions of biochemical reactions and hence changes in energy demand. 1.3.1 Cellular stress response Temperatures above Topt increases the production of reactive oxygen species (ROS) (Pörtner, 2002; Kassahn et al., 2007a). ROS is a group of reactive molecules and free radicals derived from molecular oxygen (Martínez-Álvarez et al., 2005; Tomanek, 2015). Most ROS are generated as by-products from the electron transport chain in mitochondria. While always present in cells, ROS are maintained at relatively low concentrations through a balance between production and elimination. In aquatic animals, the increase of ROS during warming is mostly caused by intensified metabolism (Vinagre et al., 2012; Hemmer-Brepson et al., 2014). The antioxidant enzymes system include superoxide dismutase and catalase (Wdziczak et al., 1982; Bagnyukova et al., 2007), as well as glutathione and glutathione-dependent enzymes (Hayes and McLellan, 1999). The cellular transcription/translation regulation during antioxidant stress involves several pathways, among which the well-studied are Keap1-Nrf2 and Hif-1α signaling pathways (Lushchak, 2011).  16  The HSR has become the most studied cellular stress response (Kregel, 2002) since its first discovery in Drosophila (Ritossa, 1962). Later studies showed that HSR is ubiquitous in most species and responsive to different types of stressors (Feder and Hofmann, 1999; Kregel, 2002). The trigger of HSR is the aggregation of damaged or misfolded macromolecules, such as DNA/RNA, proteins and membranes, which removes the repressive effect of heat shock proteins (HSPs) on heat shock transcription factors (HSFs). The HSFs then migrate to the nucleus and triggers the HSR by binding to the regulatory region of HSP genes  (Morimoto et al., 1992). HSR produces a family of highly conserved HSPs that function as molecular chaperons to facilitate macromolecules assembly, correct folding, translocation and degradation (Iwama and Thomas, 1998; Feder and Hofmann, 1999; Iwama et al., 1999). In fish, the onset and magnitude of HSR has been found to be dependent on tissue types (Currie et al., 2000), acclimation temperatures (Dietz and Somero, 1992; Logan and Somero, 2011; Fangue et al., 2011) and genetic origins (Fangue et al., 2006). 1.3.2 Cardiac myocyte responses The CHR of fish heart in response to temperature fluctuations involves changes in the morphology, electrical property and contractility. Salmonids have a four-chambered heart that sequentially moves blood through the sinus venosus, atrium, ventricle and bulbus arteriosus into the ventral aorta. Among the four chambers, the ventricle contains the most cardiac muscle and thus generates the highest blood pressure. The ventricle muscle is organized into two layers, an inner spongy layer and an outer compact layer. The inner spongy layer has a trabecular structure with large surface area, which has the advantage of not only maximizing O2 extraction from poor oxygenated venous blood but also helps developing tension. The outer compact layer is thinner but stiffer, and helps maintain the 17  integrity of heart. The relative ventricular mass to body mass (RVM) and the proportion of compact to spongy layer are species specific and both can change with acclimation temperature. Warm acclimation decreases the RVM in rainbow trout (Farrell et al., 1988b; Graham and Farrell, 1989; Taylor et al., 1996; Klaiman et al., 2011), which has been linked to less myofibrils (Clark and Rodnick, 1998; Vornanen et al., 2005; Keen et al., 2015) and connective tissue (Klaiman et al., 2011; Keen et al., 2015).  Furthermore, warm acclimation also decreases the proportion of the compact layer relative to the spongy layer (Klaiman et al., 2011), which suggests that less stiffness is needed as a result of the decreased blood viscosity (Clark and Rodnick, 1999).  Apparently, the remodeling of cardiac anatomy is a slow process and can only occur over a time scale of weeks. However, transcriptomic studies show that the regulation of cell growth and apoptosis are actually initiated immediately (minutes) after the thermal stress (Logan and Somero, 2011; Tomalty et al., 2015). Heart rate is also very thermally sensitive through the intrinsic and extrinsic controls. First, temperature itself directly affects the pacemaker action potential (Harper et al., 1995), through the thermodynamics of molecular kinetics. Second, heart rate can be affected by extrinsic controls, including the parasympathetic cholinergic vagal innervation (inhibitory), the sympathetic adrenergic nervous control (stimulatory) and circulating catecholamine released by adrenal medulla (stimulatory). In salmonids, heart rate increases with acute warming at a Q10 around two (Farrell, 1996; Farrell, 2007; Farrell, 2009). Elevated heart rate increases the work of cardiac myocytes and the turnover of channel proteins. But hearts have the ability to compensate for the adverse thermal effect by reducing fH after exposure to a warmer environment for a longer period (Farrell, 1996; Aho and Vornanen, 2001; Ekström et al., 2016). The underlying molecular mechanism is a slow depolarization of the resting 18  membrane potential. Since fH seems to have a limit (Farrell, 1996), the compensation of fH after warm acclimation brings another advantage. Compared to cold acclimated fish, warm acclimated fish have a decreased fH at the same temperature and thus delays the occurrence of peak fH to higher temperatures, which could benefit thermal tolerance (Aho and Vornanen, 2001). In cardiac myocytes, excitation–contraction coupling (E-C coupling) plays a critical role and describes the physiological process of converting electrical stimuli (action potential) to mechanical responses (myocardial contraction), which involves a number of ion transporting proteins (Roden et al., 2002). The trigger of E-C coupling is the increase of intracellular Ca2+, which is transported by voltage-gated Ca2+ channels (i.e. dihydropyridine receptor or DHPR) during sarcolemma depolarization (Vornanen, 1997; Hove-Madsen et al., 1998). Then, intracellular Ca2+ triggers more Ca2+ release from the sarcoplasmic reticulum through RYR (Hove-Madsen, 1992), which is an important process in myocytes called the calcium induced calcium release (CICR). The CICR causes spontaneous increase in cytoplasmic Ca2+ (Ca2+ sparks) and activates muscle contraction through actin-myosin cross-bridges. To resume the relaxed state, intracellular Ca2+ is mostly removed by sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) to the sarcoplasmic reticulum lumen (Korajoki and Vornanen, 2012) and by NCX to the extracellular space (Birkedal and Shiels, 2007). The rate at which SERCA removes Ca2+ can be controlled by a regulatory protein, phospholamban (PLB) (MacLennan and Kranias, 2003; Korajoki and Vornanen, 2012). Sympathetic control of E-C coupling is mediated by the cAMP-PKA signaling pathway (Lohse et al., 2003; Lissandron and Zaccolo, 2006).  The CICR is also a temperature sensitive process. In rainbow trout, peak influx of Ca2+ increased seven-fold when 19  temperature was acutely increased from 7 to 21°C (Shiels et al., 2003a). Elevated temperatures also increase the Ca2+ sensitivity of contractile elements, such as the myosin ATPase (Churcott et al., 1994) and troponin C. On the other hand, elevated temperatures often cause acidosis, which decreases the Ca2+ sensitivity of myofilaments and impairs cardiac function in fishes (Farrell et al., 1983; Farrell et al., 1984; Farrell and Milligan, 1986; Farrell et al., 1988a). 1.3.3 Using transcriptome analysis to study cellular stress Transcriptome analyses can reveal molecular hallmarks of cellular response and thus has been extensively used in thermal stress studies. Transcriptomic approaches are useful in studying cellular stress responses for a number of reasons: (1) gene expression is an essential step in the regulation on proteins synthesis; (2) mRNA changes rapidly and has a higher turnover rate than proteins.  For instance, HSR under thermal stress occurs within minutes and often in a large magnitude (Ju et al., 2002; Narum et al., 2013b), while changes at protein level is a relatively slower process (Tomanek and Somero, 2000); (3) transcriptomic analysis is relatively feasible and cost-efficient by using microarray or RNA sequencing; (4) there are quantitative correlations (either constitutive or delayed response) between mRNA and protein abundance (Buckley, 2006; Newman et al., 2006); and (5) multiple cellular responses can be studied simultaneously. In rainbow trout, cellular responses to temperature have been studied using microarray (Vornanen et al., 2005) and RNA sequencing (Tan et al., 2012; Narum and Campbell, 2015; Tan et al., 2016). However, transcriptome studies have limitations for a number of reasons (Evans, 2015). One of the most questioned is the exact correlation and timing of regulation between gene expression and protein abundances. 20  1.3.4 Transcriptomic response under different thermal regimes Most fish in the wild live in thermally dynamic environments with temperature changes at acute, diurnal and seasonal timescales. Cellular responses also differ when organisms are exposed to different thermal regimes (e.g. acute or chronic, steady or fluctuating) (Podrabsky and Somero, 2004; Logan and Buckley, 2015). Acute and chronic exposures to temperature fluctuations elicit different transcriptomic responses depending on the degree of acclimation or compensation effect. For example, in redband trout, Hsps gene expression was significantly upregulated after acute exposure, but declined to near basal levels after prolonged exposure (Narum et al., 2013b). Thus, transcriptomic responses should be studied under corresponding thermal regimes to provide functional significance. However, a response during acute exposure may be the signal to initiate a long-term response. Similarly, the response under extended exposure may also be acutely inducible. Among the literature that looked at the response of gene expression to temperature changes, thermal exposure time ranged from less than an hour (Tan et al., 2016) to over three months (Castilho et al., 2009). The majority have used long-term thermal exposure (>one week) (e.g. (Vornanen et al., 2005; Fangue et al., 2006; Rissanen et al., 2006; Hassinen et al., 2008; Castilho et al., 2009; Korajoki and Vornanen, 2012; Jayasundara et al., 2013), while only some used exposure < 24 h (e.g. (Buckley, 2006; Buckley and Somero, 2009; Healy et al., 2010; Tan et al., 2016) and rarely <2 h (Tan et al., 2016). 1.4 Genomic basis of thermal adaptation Genetic diversity in natural populations are caused by two types of genetic variations (Holderegger et al., 2006). One type is neutral genetic variation and the other type is adaptive genetic variation. The difference between them is that neutral genetic variation has no effect 21  (i.e. functionally neutral) on the phenotypes or performances that increase the chance of survival or reproduction in an altered environment (e.g. temperature). Thus, neutral genetic variation is mostly affected by demographic factors, such as gene flow and genetic drift, while adaptive genetic variation is affected by selection pressures from local environments. Determining the genetic basis of adaptive phenotypes is critical to understanding the process of local adaptation.  For a trait to be adaptable, it needs to be functional, heritable and able to increase fitness. First, thermal tolerance is no doubt a functional trait because of its role in limiting performance and survival. For example, a higher optimum temperature for aerobic scope in sockeye salmon has ensured the up-stream swimming performance during spawning migration in the warm waters of the Fraser River, British Columbia (Eliason et al., 2011). Similarly, a higher critical temperature tolerance ensured survival at higher temperatures. Second, while heritability studies of temperature tolerance are limited, all of them suggest a genetic basis. Using acute death as the endpoint, CTMAX had a narrow-sense heritability (h2) of 0.32 in the mosquito fish (Gambusia holbrooki) (Meffe et al., 1995), and 0.2 in the livebearing fish (Heterandria formosa) (Doyle et al., 2011). In rainbow trout, one generation of selection for CTMAX achieved a 48% improvement (realized heritability) (Ihssen, 1986). In a recent study in chinook salmon (Oncorhynchus tshawytscha), additive genetic effect has been found in several cardiac functions, such a peak fH,max (13%) and TAB (9%) (Muñoz et al., 2015). Lastly, the contribution of temperature tolerance to fitness is often quite direct. For instance, failure to complete the once-in-a-lifetime spawning migration will lead to a failure to reproduce, which leads to zero fitness. Likewise, a fish that cannot tolerate extreme summer temperatures will either have an immediate reduction in fitness (high temperature 22  induced mortality) or chronic reduction in fitness (e.g. impaired gonad development and fecundity). Two approaches are often used to identify the genetic basis of evolutionary adaptation (Nadeau and Jiggins, 2010). One is forward genetics, which assumes no prior knowledge of genes involved and uses quantitative genetic approaches to scan the genome for regions that are associated with the phenotypic variation. Quantitative genetic approaches, e.g. quantitative trait locus (QTL) mapping and genome-wide association study (GWAS), have been widely used to identify molecular markers for traits of interest in agriculture and human diseases. In thermal tolerance studies in salmonids, the most studied trait is upper thermal tolerance (UTT) (Araneda et al., 2008). In rainbow trout, QTLs have been identified using backcrosses derived from a thermal-selected line in Ontario, Canada (Ihssen, 1986). In total, six QTLs have been identified in rainbow trout (Jackson et al., 1998; Danzmann et al., 1999; Perry et al., 2001; Perry et al., 2005). GWAS for mortality at high temperatures has also been studied in redband trout and identified some heat shock protein genes (hsp47, hsf2 and hsc71) as candidates (Narum et al., 2013b). The other approach is reverse genetics using population genetics and candidate gene approaches, which study how genes that were differentiated between populations affect phenotypic variations. Unlike forward genetics, reverse genetics does not require the prior knowledge of the phenotypic trait. Through adaptation, genes that contribute to the adaptive traits change in allele frequency as a result of natural selection. Continuous selection leads to the fixation of genes and genetic differentiation between populations.  The fixation index (FST) is a widely used method to detect the signal of selection by identifying markers with higher genetic distance than the neutral expectations, also called “outlier” loci. By mapping 23  the “outlier” loci to genome, genes that are presumably linked to “outlier” loci are candidates for further functional verifications. The limitation of the outlier test is the false positives that can be caused by other factors such as random drift and demographic factors (founder effect and bottlenecks).  A variety of outlier detection tools based on FST have been developed (Narum and Hess, 2011) and widely applied into population genetics.  1.5 High-throughput sequencing The development of next generation sequencing (NGS) technology has greatly improved the ability to identify dense single-nucleotide polymorphism (SNP) markers across the genome, which include both neutral and adaptive genetic variation and provide more opportunities to study evolution  (Stapley et al., 2010; Davey et al., 2011; Narum et al., 2013a; McMahon et al., 2014). The most abundant genetic markers compared to other genetic markers such as microsatellites, restriction fragment length polymorphisms (RFLPs) and amplified fragment-length polymorphisms (AFLPs) are SNPs. Although the ideal situation would be to identify all SNP variations across the entire genome in a large number of individuals, it is not only infeasible but also unnecessary for many studies (Baird et al., 2008; Emerson et al., 2010; Hohenlohe et al., 2010). In addition, the genome of rainbow trout and many other species has increased in complexity because of the whole genome duplication events and the ongoing re-diploidization process (Berthelot et al., 2014). Therefore, reduced genome representation approaches have developed (Baird et al., 2008; Elshire et al., 2011; Wang et al., 2012) and can discover large numbers of genome-wide SNPs in a rapid and cost-efficient fashion, thus has been widely used in evolutionary genetic studies (Narum et al., 2013a).  24  One popular method is restriction-site associated DNA (RAD) sequencing (Davey and Blaxter, 2010; Davey et al., 2013). This method uses a restriction enzyme (e.g. sbf1) to cut DNA into fragments, which are then ligated with an adapter containing molecular identifiers assigned for each individual sample. The ligated sequences (or RAD-tags) represent a genome greatly reduced in complexity and are sequenced using the Illumina sequencing platform (Hecht et al., 2013). Sequencing coverage of genome can be changed by using a combination of restriction enzymes so that genomic DNA is cut into fragments of different sizes. Furthermore, RAD sequencing can be readily analyzed without the requirement of an existing sequenced genome, which makes the technique particularly applicable to non-model organisms (Baxter et al., 2011). Therefore, RAD sequencing is a rapid, cost-efficient and flexible platform for simultaneous discovery of tens of thousands of genome-scale SNPs (Miller et al., 2007; Baird et al., 2008). 1.6 Thermal adaptation of Oncorhynchus mykiss Salmonid species, including Oncorhynchus spp., are good candidates for studying adaptation (Taylor, 1991; Adkison, 1995; Behnke, 2002; Quinn, 2005; Waples et al., 2008; Fraser et al., 2011) because of their wide distribution throughout the world and successful adaptation to harsh environments. Rainbow trout is a temperate fish species and is native to most coastal drainages in the northern Pacific from the Bering Sea to southern California and to southern Kamchatka. The diversification of rainbow trout is, in large part, a result of the postglacial re-colonization and rapid expansion from the coast to inland regions since the end of the most recent Pleistocene glaciation beginning about 18,000 years before present, plus contemporary changes in local environment and landscape (Waples et al., 2008; Currens et al., 2009; Tamkee et al., 2010). Due to their popularity in aquaculture and sport fishing, as 25  well as presumably their adaptability to new environments, rainbow trout has also been introduced to all continents except Antarctica (MacCrimmon, 1971; Brown, 1983) including high elevation streams in some tropical countries such as Sri Lanka (Jinadasa et al., 2005), Papua New Guinea (Glucksman et al., 1976) and Kenya (Ngugi and Green, 2007). In much of North America, O. mykiss has been recognized as comprising several subspecies: a coastal form (O. mykiss irideus) and several inland forms (e.g., O. mykiss gairdneri, O. mykiss newberrii, O. mykiss stonei) using the Cascades and Sierra Nevada as natural barriers (Allendorf and Utter, 1979; Behnke, 2002; Currens et al., 2009). Phenotypically, inland rainbow trout are often collectively called “redband trout” because of the red stripe along the lateral line.   In my thesis, I use rainbow trout populations that have presumably adapted to the extreme warm environments in Idaho, USA (O. mykiss gairdneri) and Western Australia (O. mykiss irideus) to study the mechanism of thermal adaptation. 1.6.1 Known thermal requirements of O. mykiss Most of the existing knowledge of temperature tolerance in rainbow trout are generated from the coastal form. Generally, rainbow trout have optimum temperatures below 20°C and upper critical temperatures above 24°C (McCullough et al., 2001; Richter and Kolmes, 2005). UILT ranged from 24.3-26.2°C (Hokanson et al., 1977; Kaya, 1978; Bear et al., 2007).   CTMAX has also been widely measured and ranges from 26.9-31.8°C (Appendix A.2). Overall, although both UILT and CTMAX are affected by acclimation temperature as well as other abiotic and biotic factors, the ultimate value for UILT and CTMAX is seemingly around 26°C and 31°C, respectively. The preferred temperatures of rainbow trout are between 15.0-18.4°C (McCauley and Pond, 1971; Myrick et al., 2004; McMahon et al., 2006), while the optimum growth 26  temperatures overlap mostly with the preferred temperatures with a range between 13.1-19.0°C (Hokanson et al., 1977; Myrick and Cech, 2000a; Myrick and Cech, 2005; Bear et al., 2007). Intraspecific differences in the optimum thermal window for AAS differed significantly, particularly the upper Tpej. Peak AAS was maintained across a broad range of temperatures (17.8-24.6°C) in California (US) rainbow trout (Verhille et al., 2016), but was narrower (16.5-20.5°C) in British Columbia (Canada) rainbow trout (Anttila et al., 2013). 1.6.2 Adaptation of O. mykiss populations to warm climates Columbia River redband trout (O. mykiss gairdneri) is a major subgroup of the inland forms of rainbow trout. They inhabit the Fraser River and Columbia River systems on the east side of the Cascades (Behnke, 1992; Currens et al., 2009).  Habitats of redband trout have been diverse, including harsh deserts with shallow creeks and hot summers, as well as cool montane regions with mild summers, making them a good candidate to study thermal adaptation. Redband trout that currently live in desert streams probably routinely experience the highest temperatures in Oncorhynchus. For instance, diel summer water temperatures fluctuated over 10°C and can exceed 29°C in redband trout dwelling streams in southern Oregon and Idaho (Zoellick, 1999; Rodnick et al., 2004). In 1972, trout biologist Dr. Robert J. Behnke had an angling trip in Chino Creek, Owyhee Desert (Nevada), and caught a redband trout in “fine condition and excellent fighting” at the water temperature of 28.3°C (Behnke, 1979).  Impressively, redband trout were also observed feeding at 27°C, which is very close to their lethal temperatures (Sonski, 1982; Sonski, 1983). This is a surprising observation because specific dynamic action requires substantial AAS.  These observations of exceptional performances at high temperatures, where other populations would succumb, suggest the thermal adaptation in redband trout. However, Rodnick et al., (2004) showed that 27  CTMAX was not different among three redband trout populations using the fish caught from distinct climates (peak summer temperature differed >12°C) in the field (mean CTMAX = 29.4°C) and therefore questioned the usage of CTMAX as a sole indicator of thermal adaptation. A common garden rearing condition will provide a more accurate evaluation of CTMAX by controlling the thermal history (Myrick and Cech, 2000b; Myrick and Cech, 2005). Despite the impressive capability of enduring high water temperatures, research on the thermal physiology in redband trout is limited (Gamperl et al., 2002; Rodnick et al., 2004). Compared to the pace of natural selection, artificially selecting individuals for desired performance is a faster process using the same genetic rule. Therefore, artificial selection has been an important approach to help understand phenotypic variation and adaptation (Fuller et al., 2005). The application of artificial selection has been mostly successful in agriculture sciences and in model organism researches. Purposely breeding fish for thermal tolerance is rather rare (Ihssen, 1986; Ineno et al., 2005) as compared to the stress studies in plants (Hall, 1992). During the introduction of rainbow trout across the globe, human-caused changes in thermal environments may have, to some extent, caused selection on thermal tolerance if it is an adaptable trait. A good example is the introduction of rainbow trout to Western Australia, where summer temperatures are high and arid conditions prevail. The translocation of rainbow trout to Oceania can date back to the early 1890s, from their native habitats in California, USA, to New Zealand and eastern Australia (Molony, 2001; Morrissy et al., 2002; Ward et al., 2003). The current rainbow trout stock in Western Australia was introduced from eastern Australia in 1936 (eggs, n = 50,000), 1941 (fry, n = 50,000) and 28  1972 (fry, n = 20,000) (Morrissy et al., 2002). To control the risk of spreading disease during the transportation of live fish, the introduction has been completely ceased since 1972. The maintenance of rainbow trout stock today in Western Australia has largely relied on the broodstock at the Pemberton Freshwater Research Centre (PFRC). The PFRC rainbow trout have been self-sustaining since 1975 (i.e. over 19 generations as the fish were bred every two years). During the hatchery culture, PFRC rainbow trout have been living in shallow concrete ponds with water supply from a head reservoir, where water level often decreases in summer. Several bouts of severe mortality (> 90%) have occurred due to the exceptionally hot summers (Morrissy, 1973; Molony et al., 2004).  Therefore, temperature may have acted as a selecting factor by eliminating fish that cannot tolerate high temperatures. Indeed, the time to mortality at 27°C of PFRC rainbow trout is approximately twice as long as that of a self-sustaining wild-type from a deep reservoir (Molony et al., 2004), which was stocked by PFRC before 1961 and thus the resident rainbow trout resemble the original gene pool. 1.7 Thesis objectives and chapters The overall objective of my dissertation was to examine the mechanism of thermal adaptation in Oncorhynchus mykiss. I choose populations that are presumably adapted to warm environments in Idaho (USA) and Western Australia (Australia) as research organisms to address specific scientific questions, which appear in Chapters 2, 3 and 4. Specifically, I comprehensively characterized thermal tolerance by measuring CTMAX, AAS and fH,max in each study system. To identify genetic markers that are under positive adaptive selection, I utilized RAD-sequencing and discovered genome-wide SNP markers for divergent loci discovery and phenotypic association studies. To investigate the cellular response to acute warming, I conducted RNA sequencing on the cardiac transcriptome and identified candidate 29  genes associated with thermal adaptation. My overall objective has been partitioned into three main sub-objectives and the work is presented in Chapters 2-4: Objective one: Quantify the thermal tolerance of warm adapted O. mykiss population To characterize thermal tolerance, I used integrated approaches to examine the performance at the levels of whole animal (CTMAX), physiological processes (RMR, MMR and AAS) and organ (fH.max). All measurements were made during acute warming (< 7h). Due to the importance of the circulatory system and fH in thermal tolerance, I measured fH.max to compare cardiac performances.  The hypothesis was that adaptation to different thermal regimes is accompanied by adjustments in cardiorespiratory functions to achieve a more suitable thermal performance in local environment.  To test this hypothesis, in Chapter 2, I measured CTMAX, aerobic scope and fH.max for an artificially selected rainbow trout strain in Western Australia. In the physiology section of Chapters 3 and 4, I measured CTMAX, aerobic scope and fH.max for redband trout populations from desert and montane climates.  I predicted that rainbow trout from warm climate have a greater aerobic scope and fH.max, broader optimum thermal window and higher CTMAX than those from cool or cold climate. Objective two: Genomic mapping of thermal tolerance and adaptation Thermal tolerance is a heritable and quantitative trait with multiple genes contribute to the phenotype. Previous QTL analyses have identified microsatellite loci that were significantly associated with critical temperatures in rainbow trout. However, the genetic architecture for thermal tolerance is still largely unclear. Similarly, knowledge in the genetic components underlying natural events of warm adaptation is scarce. Previous studies have 30  been largely limited by the low density of genetic markers. The development of sequencing technologies, e.g. RAD-sequencing, enables the discovery of dense genomic SNP markers, which can be used to better pinpoint relevant genes for thermal tolerance and adaptation.  The hypothesis that intraspecific variation in thermal tolerance is associated with allelic variation in genomic markers was examined. Also, locally warm adapted populations will have different allele frequency in genes that are under positive selection for high thermal tolerance.   In Chapter 3, I identified SNPs across the genome using RAD-sequencing to test the hypothesis in redband trout populations.  I predicted that intraspecific differences of thermal tolerance in rainbow trout at the individual and population level are correlated with the variations in SNP markers.  Objective three: Gene expression response to temperature tolerance The heart is an important life-support organ. In thermal tolerance, the heart has been proposed to be the limiting factor according to the OCLTT hypothesis.  However, cellular mechanism of cardiac responses to temperature is still largely unclear. Microarray and quantitative PCR have been widely used to study the cellular response to temperature. Here, I use RNA sequencing as an unbiased and large-scale approach to identify genes that may be involved in stress response.   The hypothesis was that expression of functional important genes are thermally inducible and the expression pattern is population specific.   In Chapter 4, I used RNA sequencing to quantify gene expression during acute warming in redband trout, contrasting populations from desert and montane streams. I predicted that acute warming will cause genes to respond to thermal stress differently between locally adapted populations.     31    Figure 1.1 Parameters used to characterize thermal tolerance. (A) Routine metabolic rate (RMR) and maximum metabolic rate (MMR); (B) Fry Curve of absolute aerobic scope (AAS) between lower critical temperature (Tcrit) and higher Tcrit. The optimum temperature (Topt) is when AAS is maximum. Optimum thermal window is bound by two pejus (“getting worse”) temperatures (Tpej); (C) Maximum heart rate (fH,max) in response to temperature. Arrhenius breakpoint temperature (TAB), calculated from the discontinuous slope (-Ea/R) of Arrhenius plot (D), indicates the rate change of exponential increase in fH,max.  The fH,max reaches peak value at TPEAK and is followed by the presence of cardiac arrhythmia (TAR). 32    Maximum performance is achieved at the optimum temperature (Topt) without thermal stress. Performances decrease at temperatures below the lower pejus temperature (Tpej) or above the higher Tpej. Intracellular homeostasis environment became unbalanced and thus needs to be adjusted (cellular homeostasis response, or CHR). For example, to increase cardiac output at temperatures above Tpej, pacemaker cells can generate more action potentials and stimulate more frequent myocyte contractions. Mitochondria function is regulated to meet the elevated demand for ATP. Protein synthesis processes are regulated to produce more functioning molecules, such as rate-limiting enzyme for biochemical reactions. At critical temperatures (Tcrit), AAS becomes limited and causes the accumulation of reactive oxygen species. Also, high temperatures disturb the correct folding, transporting and even cause protein denaturation. Therefore, cellular stress response (CSR) is triggered, while CHR still persists with additional tasks, such as to repair the damaged macromolecules to restore homeostasis. This diagram is modified from Pörtner et al., (2007), Kassahn et al., (2009) and Pörtner, (2010).  Heat shock response Oxidative stress response critical pejus optimum pejus critical CHR CSR Mitochondrial function Cardiac action potential Excitation-contraction coupling Transcription and translation  Topt Tpej Tcrit Tcrit Tpej Temperature Absolute aerobic scope Figure 1.2 Cellular response to temperature and the Fry Curve. 33  Chapter 2: Selection for upper thermal tolerance in rainbow trout (Oncorhynchus mykiss)1 Synopsis Using a combination of approaches, this chapter comprehensively examines the thermal tolerance of a rainbow trout population (Oncorhynchus mykiss) that has been raised in a warm climate for over 19 generations in southern Western Australia. This chapter evaluates the oxygen- and capacity- limited thermal tolerance (OCLTT) hypothesis by measuring the absolute aerobic scope (AAS) and maximum heart rate (fH,max). Furthermore, this chapter validates the feasibility of using thermal indices derived from different approaches to estimate the optimum temperature, optimum thermal window and critical temperatures. Results suggest that thermal tolerance in the Western Australia rainbow trout is correlated with oxygen and capacity limitations. Maximum AAS values are centered at 15.8±0.3°C (Topt) for 15°C-acclimated fish. Optimum thermal window for AAS is between 11.8 ± 0.4°C (low Tpej) and 19.9±0.3°C (high Tpej). Upper Tpej for AAS is close to the Arrhenius breakpoint temperature (TAB) for fH,max (20.5 ± 0.2°C). AAS decreases beyond the optimum thermal window and reaches zero (extrapolated to Tcrit using the Quadratic function) at temperatures close to CTMAX (29.0±0.02°C).  These results collectively suggest a high thermal performance for a domesticated rainbow trout population.                                                   1 A version of this chapter has been published: Chen, Z., Snow, M., Lawrence, C., Church, A., Narum, S., Devlin, R. and Farrell, A. (2015). Selection for upper thermal tolerance in rainbow trout (Oncorhynchus mykiss Walbaum). J. Exp. Biol. 218, 803–812. 34  2.1 Introduction Thermal adaptation of many fish species in the Northern Hemisphere has been the result of the range expansion since the last glaciation period (Bernatchez and Wilson, 1998), which is no more evident than within the genus Oncorhynchus. This genus is generally regarded as a post-glaciation invader and is naturally distributed between subarctic to central Mexico in North America. Nowadays, Oncorhynchus is globally distributed after being introduced to many countries due to their high values in fisheries and aquaculture. Despite having a preference for temperatures well below 20°C, members of Oncorhynchus are remarkably adaptable. Indeed, recent studies suggest that there may have been local intraspecific adaptation to different thermal regimes that are experienced by adult sockeye salmon (Oncorhynchus nerka) during the spawning migration (Lee et al., 2003a; Eliason et al., 2011). Similarly, the redband trout (Oncorhynchus mykiss gairdneri) appears to have uniquely adjusted its physiology to tolerate high water temperatures, which can reach 29°C in desert streams (Rodnick et al., 2004), a sign of thermal adaptation that has been supported by genetic evidence (Narum et al., 2010). In a more rapid fashion, artificially transplanting animals to a new environment may also lead to selection by accumulating favorable heritable adaptations to create strains that are better suited to the introduced habitat. Here, rainbow trout (Oncorhynchus mykiss) introduced to a warm environment in Western Australia were used to study their thermal tolerance and help understand the adaptive capacity of this species to elevated temperatures. Since early 1890s, rainbow trout has been translocated to New Zealand and Australia from their native habitat in California, USA (Ward et al., 2003). Inevitably, these introductions exposed this species to many new environmental challenges, which led to 35  either extirpation or possibly natural selection of the surviving population. Indeed, the establishment of rainbow trout in Western Australia has proven to be difficult (Molony, 2001) with an extremely hot climate being the critical limiting factors (Morrissy, 1973). Stocking locations in Western Australia also have to be carefully selected to balance the economic and recreational fishing outcomes while limiting negative impacts on native fish species. Despite the environmental and regulative challenges, a rainbow trout strain has been successfully maintained at the Pemberton Freshwater Research Centre (PFRC) in Pemberton (Western Australia) for several decades and is being used to annually stock the reservoirs and creeks in southern Western Australia. During the hatchery culture, PFRC rainbow trout not only battled the averagely higher water temperatures, but also experienced exceptionally hot summers, which caused several severe mortality events (> 90%) (Morrissy, 1973; Molony et al., 2004). Therefore, the PFRC hatchery strain of rainbow trout (termed the H-strain) is a particularly good candidate to study thermal adaptation given the strength, form and duration of selection. Phenotypic evidences exist to support the hypothesized thermal selection in PFRC H-strain rainbow trout. Compared with a self-sustaining wild-type population in the deep Serpentine reservoir (S-strain) in Perth (Western Australia), which was stocked by PFRC prior to the 1960s, the PFRC H-strain has been shown to tolerate the extreme temperature of 27°C for twice as long as the wild-type S-strain (Molony et al., 2004).  More broadly, an earlier study showed that PFRC rainbow trout survived for a longer time at high temperatures (25-29°C) than those from New South Wales and Victoria, where the PFRC rainbow trout were developed and first introduced (Morrissy, 1973). These results suggest the possibility that poorly adapted individuals have been eliminated during hatchery rearing because water 36  temperature cannot be operationally maintained below 25°C (Figure 2.1), whereas similar selection pressure was avoided by wild fish in the reservoir that had access to cool waters at depth (< 20°C). Therefore, to better elucidate the underlying physiological changes in the PFRC H-strain rainbow trout that render them more heat tolerant, this chapter uses comprehensive approaches to assess their upper thermal tolerance. Specifically, the ability of PFRC H-strain rainbow trout to tolerate acute warming was characterized by measuring CTMAX, AAS and the response of fH,max to acute warming. These experiments test the hypotheses that: 1) PFRC H-strain rainbow trout can maintain AAS and fH,max at high temperatures; and 2) phenotypic variation has reduced among individuals due to the intensive selection. The approaches used here stem from the pioneering physiological studies of thermal tolerance in fishes by Fry, (1947) and the more recent hypothesis of “oxygen- and capacity- limited thermal tolerance (OCLTT)” (Pörtner and Knust, 2007; Pörtner and Farrell, 2008), as well as the recent finding that rate transition temperatures for heart rate (fH) reveal much about the upper thermal tolerance that governs the biogeographical distribution of a species (Tepolt and Somero, 2014). Indeed, OCLTT in fishes has been widely studied from warm water species such as coral reef fishes (Nilsson et al., 2009; Rummer et al., 2014) and Danio (Sidhu et al., 2014), through several temperate salmonid species (Casselman et al., 2012; Chen et al., 2013; Verhille et al., 2013; Anttila et al., 2014b; Muñoz et al., 2015) and goldfish (Carassius auratus) (Ferreira et al., 2014), to polar species such as Antarctic nototheniid fish (Pagothenia borchgrevinki) (Franklin et al., 2007) and Arctic cod (Boreogadus saida) (Drost et al., 2014; Drost et al., 2016). Limitations in the OCLTT have been questioned in general (Clark et al., 2013) and rebutted (Farrell, 2013; Pörtner and 37  Giomi, 2013). More specifically, the questioning was for a fish species (Norin et al., 2014), an amphibian species (Overgaard et al., 2012) and a tropical shrimp (Ern et al., 2014). Yet, AAS, fH,max and CTMAX have been satisfactorily compared through the studies in goldfish (Ferreira et al., 2014), Danio species (Sidhu et al., 2014) and salmonids (Casselman et al., 2012; Chen et al., 2013). Therefore, the present study examines the optimum temperature for AAS (Topt) and the extent of AAS diminishes at supra-optimal temperatures up to 25°C. In addition, the rate transition temperatures when fH,max fails to keep up with a steady Q10 during acute warming are characterized because in salmonids increasing fH is the primary response to deliver more O2 to tissues (Steinhausen et al., 2008; Farrell, 2009; Eliason et al., 2011; Casselman et al., 2012; Eliason et al., 2013). 2.2 Materials and methods All experiment procedures were approved by the University of British Columbia Committee on Animal Care in accordance with the Canadian Council on Animal Care (A10-0335). 2.2.1 Fish culture and rearing conditions Rainbow trout at PFRC originated from the San Francisco Bay area - Sonoma Creek and/or Russian Creek (Morrissy et al., 2002). After the introduction to Western Australia was ceased in the 1970s, rainbow trout at PFRC have been self-sustaining for nearly half a century. For my study, the brood stock had been reared in shallow and circular concrete ponds (0.7 m deep, 7.5 m diameter). These ponds were aerated by fresh flow-through water from a reservoir. Water temperature of the ponds reflects that of the reservoir, which increases in summer as a result of the insolation and a greatly reduced water level. All fish were bred on June 10th and 11th, 2013 using 25 males and 25 females that were dry stripped. 38  Milts from five males were pooled and fertilized separately with eggs from five females to generate a group with five half-sibling families. In total, I made five such family groups, each of which contained five half-sibling families. Fertilized eggs were then incubated in boxes made from rigid fly wire (length × width × height: 120 mm × 50 mm × 35 mm). Prior to hatching, eggs were removed from the incubator and placed into 25 hatching/larval rearing boxes (340 mm × 270 mm × 200 mm) made from perforated (1.0 mm diameter holes) stainless steel sheet. These boxes were suspended (half submerged) into separate 200 L cones, with a common water supply from a recirculating system. Animals were raised in thermally controlled water (15.0 ± 0.4°C) with 12 hours light and 12 hours dark cycles. The system received an input of fresh creek water at the rate of ~6 L min−1. Excess food and wastes were cleaned daily. Two biological filters were also used to remove suspended particles. Water nitrite and ammonia levels were checked twice a week using aquarium water quality test kit (Tetratest, Tetra, Germany). After hatching, 200 fish were released from each hatching box into the rearing cone. Fish were fed twice a day with commercial freshwater trout diet (Skretting, Cambridge, Tasmania, Australia), but not on the day before an experiment, which resulted in a 48-hour fast period for the experimental fish. Physiological experiments (described in followed sections) were conducted between November 2013 and January 2014. Different fish were used in each experiment.  2.2.2 Critical thermal maximum Each CTMAX measurement used a batch of 15 fish from each half-sibling family (a total of 75 fish from each family group, Table 2.1). Fish were quickly dip-netted from their rearing tank into a 40 L plastic container with aerated and temperature-controlled circulating water by a heating/cooling unit (F32-MD, Julabo Labortechnik GmbH, Seelbach, BW, 39  Germany). Influx water flow was sufficient to maintain thermal homogeneity in the container throughout the experiment. Fish were given one hour to recover from minor handling stress at their acclimation temperature (15°C). Water temperature was then increased at 0.3°C min−1 to 22°C and at 0.1°C min−1 thereafter (Beitinger et al., 2000) until CTMAX was reached, which is defined as the temperature when an individual fish is unable to maintain an upright posture (a sign of loss of equilibrium) for 10 s. Subsequently, fish was immediately removed from the test chamber and put into a separate and labeled bucket at 15°C to recover. After the experiment, fish were anesthetized in 80 mg L−1 MS222 buffered with 160 mg L−1 NaHCO3 and body size was measured. Water temperature was measured using a digital thermometer (precision: 0.1°C; Total-Range Thermometer, Fisher Scientific, Nepean, ON, Canada). Post-testing fish mortality < 3.2%. All thermometers were calibrated in ice water bath (0°C) and 50°C water bath. Two mercury thermometers were used to validate the water bath temperatures. 2.2.3 Routine and maximum metabolic rate Routine metabolic rate (RMR) and maximum metabolic rate (MMR) were measured to calculate AAS and FAS. Metabolic rate was only measured in three family groups that showed the greatest collective variability in CTMAX. Six fish were randomly selected from each family group and placed into separate floating boxes made from rigid plastic mesh (diameter = 15 cm; height: 20 cm) for two weeks before experiment. Metabolic rate measurements were completed in 18 days, with daily measurements of six fish (two from each of the three family groups) (Table 2.1). As a result, fish growth during the entire course of the metabolic rate measurements was not a significant factor (see below). RMR and MMR 40  were measured on individual fish (n = 6 per group) at each of the six test temperatures (8, 12, 15, 18, 21, and 25°C) as follows. After a 48-hour fasting period, fish were individually transferred from the holding tank to one of the six custom-made, intermittent-flow respirometers (~500 ml). Water temperature (15°C) in the respirometers was controlled with an in-line, recirculating chillers (F32-MD, Julabo Labortechnik GmbH, Seelbach, BW, Germany). Fish were left overnight for a minimum of 12 hours to adjust to the respirometer environment. Water temperature was increased at the rate of 2°C h−1 to the desired test temperature. Then the fish were held at the test temperature for one hour before RMR measurement, which was simply close the influx water to the respirometer and record the depletion of O2 in water with a fiber optic oxygen meter (Firesting O2, PyroScience GmbH, Aachen, NRW, Germany). Dissolved oxygen was allowed to decrease by no more than 25% (i.e. maintain ≥ 75% air saturation), which generally took less than 10 min. After the RMR measurement, one fish at a time was removed from the respirometer and placed in a 50 cm diameter circular tank that contains aerated water at the test temperature, where the fish was exercised to exhaustion over a roughly 5-min period by combining hand chases and gentle tail pinches until a sign of unresponsive to handling (Norin and Clark, 2016). During the exhaustion protocol, fish was occasionally exposed to air for no more than a total of 20 s to help stimulate the MMR (Clark et al., 2013). Then, fish was immediately returned to the respirometer (within one min) and had the oxygen uptake measured for three min. MMR was determined by taking the maximum rate of oxygen removal for a minimum of 1.5 min during this recording period, which typically occurred immediately after the fish was returned to the respirometer. In both RMR and MMR calculations, the water O2 concentration-time relationship was essentially 41  linear with R2>0.98. After the MMR measurement, fish were weighed and returned to the rearing tank. Each fish was given a minimum of three days to recover before the measurement at the next test temperature. Before metabolic rate experiments, fish weighed an average of 31.8 ± 0.9 g (N = 18 in total). Body mass did not differ significantly among family groups. Body mass increased by an average of 7.9% during the eighteen-day testing despite food restriction, handling and experiment tests. RMR and MMR (mass specific) was calculated as: ?̇?𝑂2 = ([O2]𝑡0 − [O2]𝑡1) ∙  V / (T ∙  Mb) ?̇?𝑜2: O2 consumption rate (mg O2 kg−1 h−1); [O2]t0: O2 concentration at time t0 (mg O2 L−1); [O2]t1: O2 concentration at time t1 (mg O2 L−1); V: respirometer volume minus the volume of experimental animal (L). The value of V (L) was approximated as the same as the fish mass (kg) by assuming the similar density between water and fish; T = t1 – t0 (hour); Mb: body mass of experimental animal (kg). 2.2.4 Maximum heart rate Maximum heart rate (fH,max) was measured in individually anaesthetized fish using the same protocol that was originally developed in coho salmon (Oncorhynchus kisutch) (Casselman et al., 2012) and since then was widely used in a variety of fish species (see Introduction, Section 1.2.3).  Briefly, each fish was anaesthetized (80 mg L−1 MS222 buffered with 160 mg L−1 NaHCO3; Sigma-Aldrich, St. Louis, MO, USA) for around six minutes to reach stage III 42  sedation (total loss of reactivity but with slowed opercular movements) and placed (submerged in water) in the electrocardiogram (ECG) measuring system, where the gills were continuously supplied with aerated and temperature-controlled water containing a maintenance anaesthetic concentration (65 mg L–1 of MS222 and 130 mg L–1 NaHCO3). Dissolved O2 was typically >90% air saturation, as measured by a galvanic oxygen sensor (Oxyguard, Farum, Denmark). Water temperature of the entire system was controlled with a heating/chilling unit (F32-MD, Julabo). To capture the ECG signal, two silver electrodes (30 gauge; 5 cm in length) were placed to touch the ventral side of the body (positive one in the middle of the pectoral fins, and the negative one 4 cm away from positive electrode toward the pelvic fins). The electrodes were connected to an Animal Bio Amp (AD Instruments Inc., Bella Vista, NSW, Australia) that amplified (1000×) and filtered (60 Hz line filter; low-pass: 30-50 Hz; high-pass 0.1-0.3 kHz) the ECG signal. The conditioned ECG signal was digitalized using a Powerlab 8/35 data acquisition system and was analyzed using Labchart software version 7.0 (AD Instruments Inc., Bella Vista, NSW, Australia). Anaesthetized fish was stabilized at the initial test temperature (15°C) for 1 hour. After the 1-hour stabilization, ƒH,max was achieved with pharmacological interventions. An intraperitoneal injection of atropine sulfate (2.7 mg kg−1, Sigma-Aldrich, St. Louis, MO, USA) blocked inhibitory vagal tonus to the heart and an isoproterenol injection (9 μg kg−1, Sigma-Aldrich, St. Louis, MO, USA) stimulated cardiac adrenergic β-receptors. Each injection was given 15 min to take effect, these injections resulted in an elevated and stable ƒH,max. Then, water temperature was changed at the rate of 1°C increment over a four-min period and maintained at the new temperature for an additional two min, which was a sufficient equilibration time for ƒH,max to reach a new stable level. Thus, the 43  overall warming rate was 1°C every six min, which continued until the heartbeat developed an arrhythmia, usually a missed ventricular depolarization (can be visualized as a missed QRS complex in the ECG). The ƒH,max was calculated by counting the R-waves in ECG over the final 30 sec of a temperature increment. Once an arrhythmia was noticed, the ventricle was immediately separated from fish and weighed. 2.2.5 Hemoglobin and hematocrit analyses Subsequent to the completion of above tests, a subsample of fish was removed from the remaining stock and acclimated at 15°C. Five individuals from each group were randomly selected for the determination of individual haematocrit (Hct) and haemoglobin (Hb) levels. Fish were anesthetized (80 mg L−1 MS222 buffered with 160 mg L−1 NaHCO3; Sigma-Aldrich, St. Louis, Missouri, USA), weighed and blood samples immediately taken using heparinized vacutainers. Hb was measured on a handheld hemoglobin analyzer (Hemacue 201+, Ängelholm, Sweden) using 0.1 ml of blood, and Hct was measured using microhematocrit capillary tubes centrifuged at 10,000 g for five minutes. Results of family group-4 is not present because of the blood clotting during measurements. 2.2.6 Statistical analyses All data analyses were performed using SigmaPlot version 11.0 (Systat Software Inc., San Jose, CA, USA) except where specified. Values are presented as mean ± s.e.m. Statistical significances for all analyses were set at α = 0.05. Analysis of covariance (SPSS version 21.0, IBM Corp., Armonk, NY, USA) was used to test for differences in CTMAX among family groups and account for the effect of body size. RMR, MMR and AAS were determined for each individual fish at each of the six test temperatures. Differences in RMR, MMR and AAS between family groups were tested using a two-way ANOVA with Holm-44  Sidak post-hoc test. In addition, regression analysis was used to generate best-fit curves (polynomial quadratic function for RMR, and Gaussian 3-parameter function for AAS and MMR) for each of the three family groups. The equations of the fitted curves for AAS were used to generate the rate transition temperatures to generate Fry Curves: 1) optimum temperature (Topt) is the temperature; 2) pejus temperature (Tpej) is when AAS was reduced to 90% of peak AAS; 3) critical temperature (Tcrit) is the temperature when AAS was 0% of peak AAS. These rate transition temperatures were compared among family groups using a one-way ANOVA with Holm-Sidak post-hoc test. In fH,max analyses, the rate transition temperatures (TAB – the first Arrhenius breakpoint temperature; TQB – the breakpoint temperature for Q10 of fH,max; TPEAK – the temperature at which fH,max reached its maximum absolute value; TAR – the arrhythmic temperature) were determined from the responses of individual fish and are reported as mean values. According to Yeager and Ultsch, (1989), the analyses for TAB were performed by identifying the breakpoint within the best-fit two-segment regression line in an Arrhenius plot of fH,max. TQB was arbitrarily set as the temperature beyond which the incremental Q10 for fH,max fell below 1.8. For each individual, incremental Q10 values for fH,max during incremental warming were plotted against temperature. 2.3 Results 2.3.1 Critical thermal maximum (CTMAX) Average CTMAX of all individuals measured was 29.0 ± 0.02°C. Individual variance in CTMAX was quantitatively small, with CTMAX values (N=375) for 375 fish from 25 families ranged from 28.1 to 29.7°C. Minor CTMAX differences could be statistically distinguished among some Family Groups: Family Group-1 had the highest CTMAX (29.3 ± 45  0.02°C, p < 0.01), while Family Group-2 and 5 had the lowest CTMAX (28.9 ± 0.03°C and 28.8 ± 0.03°C, p < 0.01) (Figure 2.2).  2.3.2 Metabolic rates and absolute aerobic scope In Family Group-1, 2 and 5, RMR increased by about six-fold with acute warming from 8°C to 25°C (Figure 2.3 A). RMR values for Family Group-1 and 5 were significantly different (p = 0.049). Family Group-1 had the lowest Q10 for RMR (2.9) and the highest CTMAX, whereas Family Group-5 had the highest Q10 for RMR (3.1) and the lowest CTMAX. No significant difference was found among the three Family Groups in MMR at any test temperatures (p > 0.13). Before reaching its maximum value, MMR increased with a similar average Q10 of 1.8 for all three Family Groups. According to the Gaussian function, MMR had its maximum value at 20.0°C for Family Group-1 (753.0 mg O2 kg−1 h−1), 19.2°C for Family Group-2 (747.2 mg O2 kg−1 h−1) and 20.6°C for Family Group-5 (775.8 mg O2 kg−1 h−1). Although the regression equations for AAS in Figure 2.3 B were based on average values for each family group, regression analysis was also conducted for each individual fish and was used to calculate the rate transition temperatures, which are present in Table 2.2. According to Gaussian functions, AAS reached a maximum value around 16°C in all three Family Groups (550.0 ± 25.6 mg O2 kg−1 h−1 at 16.5 ± 0.6°C for group 1, 559.7 ± 22.0 mg O2 kg−1 h−1 at 15.3 ± 0.2°C for group 2 and 541.2 ± 24.3 mg O2 kg−1 h−1 at 15.8 ± 0.7°C for group 5). Neither the maximum AAS values nor the rate transition temperatures were significantly different among family groups. Pejus temperatures were arbitrarily assigned at 90% of the maximum AAS (Eliason et al., 2011). Lower Tpej values were clustered between 11.5 ± 1.0°C and 12.4 ± 0.4°C, while upper Tpej values were clustered between 19.0 ± 0.4°C 46  and 20.5 ± 0.8°C. Upper Tcrit values were between 31.6 ± 1.7 and 36.0 ± 1.6°C. Factorial aerobic scope (FAS = MMR/RMR) was the greatest at 8°C in all family groups (5.9 – 6.4) and decreased with warming, but remained an impressive 1.4 – 1.8 even at the highest test temperature (25°C). At 25°C, AAS was 29 - 43% of the peak AAS at Topt. On average, Family Group-1 had the highest FAS. 2.3.3 Maximum heart rate (fH,max) The thermal performance curve for fH,max is left-skewed and bell-shaped. In all three Family Groups, fH,max increased with incremental warming. The further warming resulted cardiac arrhythmia (Figure 2.4 A), which occurred during the rising phase in most fish (22 fish), the plateau phase (7 fish) and the decline phase (7 fish) in some fish. The peak value for fH,max occurred between 23.5 ± 0.4 and 24.0 ± 0.4°C (TPEAK) among families (Table 2.3), which were above the upper Tpej determined from the Fry Curve. The cardiac arrhythmia temperature (TAR) averaged around 25°C for all three Family Groups. The lowest individual TAR in all three Family Groups was 23°C and no fish maintained a regular heart beat beyond 27°C (Figure 2.4 A). At 25°C, which was the highest test temperature used to measure AAS, 50% of the tested fish displayed cardiac arrhythmias. Family Group-2 maintained the lowest fH,max across all test temperatures. Rate transition temperatures for fH,max were compared among the three Family Groups (Table 2.3), but no significant difference was discovered. The incremental Q10 for fH,max decreased progressively with acute warming (Figure 2.4 B). TQB values ranged from 18.8 ± 0.3 to 19.0 ± 0.2°C and they were statistically indistinguishable from TAB values, which ranged from 20.3 ± 0.3 to 20.7 ± 0.4°C among Family Groups. 47  2.3.4 Ventricle mass, hemoglobin (Hb) and hematocrit (Hct) Fish from the fH,max measurements had a mean body mass of 37.2 ± 1.2 g (Table 2.4) and a mean wet ventricle mass of 30.6 ± 1.2 mg. There were no significant differences among Family Groups. The mean relative wet ventricle mass (RVM) was between 0.081 – 0.084%. Hematocrit (Hct) level and hemoglobin (Hb) content were also not significantly different among Family Groups. Hct level ranged from 32.9 ± 1.1 to 36.9 ± 2.6%, while Hb content was around 10.1 g dL−1. 2.4 Discussion For the first time, the present study comprehensively examined the upper thermal tolerance of an introduced fish species in Western Australia (PFRC H-strain of rainbow trout) that has undergone thermal selection for over 19 generations of hatchery culture. Results supported the two hypotheses that I set out to test: 1) PFRC H-strain rainbow trout are able to maintain aerobic scope and fH,max at high temperatures; and 2) phenotypic variation has been reduced due to intensive thermal selection. The mean fH,max was sustained without showing arrhythmia until 25°C, at which temperature AAS was still 40% greater than RMR. Limited phenotypic variation was particularly evident for the rate transition temperatures in both AAS and fH,max as indicated by the subtle differences among independently tested family groups. In CTMAX, differences among family group that did emerge as significant were quantitatively small. The occurrence and intensity of local adaptation rely on many factors: the scale and magnitude of selection; the standing genetic variation in population; the extent of gene flow; and the heritability of adaptive traits. Because these factors differ widely between species and populations, so too does the magnitude of local adaptation (Leimu and Fischer, 2008). 48  Artificially translocating fish to a new environment could potentially accelerate and intensify adaptation.  Indeed, both phenotype and genetic results from precedent studies conform to the prediction of adaptive changes in introduced animals to warm climate. Ward et al., (2003) examined the genetic structure of the Western Australia rainbow trout, comparing Serpentine wild type S-strain and PFRC hatchery H-strain with North American rainbow trout. The Australian PFRC fish were genetically more homozygous than their North American counterparts. Furthermore, the study of gene expression in the rainbow trout introduced to Japan revealed a higher gene expression of heat shock proteins after the strain was raised at high temperatures (20 – 24°C, occasionally 30°C) in summer for more than fourteen successive generations (Ojima et al., 2012; Tan et al., 2012). Similarly, transplanted sticklebacks (Gasterosteus aculeatus) became more cold tolerant in just three generations (Barrett et al., 2011). However, phenotypic plasticity can also cause exceptional variability on thermal performance (Currie et al., 1998) and must be investigated as a complement to the evolutionary adaptation for thermal tolerance (Narum et al., 2013b). In this study, I had kept fish at 15°C in a common garden environment since hatching stage to minimize the acclimation effect (plasticity).  2.4.1 Critical thermal maximum (CTMAX) CTMAX is a widely accepted measure of acute thermal tolerance and is defined as "the thermal point at which locomotion becomes disorganized and the animal loses its ability to escape from conditions that will promptly lead to its death" (Lutterschmidt and Hutchison, 1997a). Thus, CTMAX can be used to infer the ability of animals to tolerate rapid regional or diurnal temperature increase in the wild. For example, CTMAX has been found to be coincident with the extreme habitat temperatures in a number of fish species (Stuenkel and 49  Hillyard, 1981; Beitinger et al., 2000; Beers and Sidell, 2011). In redband trout adapted to desert climate in Oregon, USA (Rodnick et al., 2004), where summer water temperature can oscillate from 18 to 30°C daily, CTMAX (29.7 ± 0.3°C) was close to the upper daily water temperature recorded for those streams. The present study found that CTMAX value of individual PFRC rainbow trout was between 28.1 and 29.7°C, which is very close to the highest historic water temperatures passing through PFRC (Molony et al., 2004). In the summer of 2014, a relatively cool year in Western Australia, the highest water temperature was recorded as 26.1°C for 3 hours (Figure 2.1) and this extreme did not produce any thermal-related fish mortality. Instead, more surprisingly, active feeding behavior was observed.  The reason that PFRC trout survived historic extreme water temperatures that clearly exceed their CTMAX (> 30°C) is because cooling towers were used to chill the water for each pond as best as possible. When CTMAX values were compared among rainbow trout populations from different locations and habitats, I found large variations among studies (26.9 - 31.8°C, Appendix A.2). However, acclimation temperature played a significant role in this variability. When acclimated at ~10°C, CTMAX was within a smaller range between 27.5 and 29°C among indigenous strains from British Columbia, Canada (Scott et al., 2015), California (Myrick and Cech, 2000a; Myrick and Cech, 2005) and recently established strains in Pennsylvania, USA (Carline and Machung, 2001). When acclimated at 15°C, CTMAX varied from 27.7°C in a North Carolina population (Galbreath et al., 2006) to 29.7°C in a Japanese rainbow trout strain, which was originated from California, USA and has been intensely selected for upper temperature tolerance since 1966 (Ineno et al., 2005). By contrast, there are larger differences among species, which ranged from 13.3°C in Antarctic icefishes (Chionodraco 50  rastrospinosus) (Beers and Sidell, 2011) to >42.5°C in California pupfish (Cyprinodon salinus) (Stuenkel and Hillyard, 1981). Despite the mean CTMAX among family groups were <0.5°C, statistically significant differences did exist at the family level in the present study, as well as in many previous studies (Carline and Machung, 2001; Myrick and Cech, 2005). The underlying reason for the limited intraspecific variance of CTMAX is unclear, but may arise from the modification of key biochemical pathways during adaptation (Hochachka and Somero, 2002). With the rapid development of high-throughput next generation sequencing, it may be possible to address this question in the near future. All the same, artificially selecting a particular strain of rainbow trout (and perhaps other species) to perform at significantly higher temperatures may be difficult using CTMAX as the only metric to make such a distinction, because CTMAX is a measure of thermal tolerance limits rather than performance. 2.4.2 Metabolic rates and absolute aerobic scope It is proposed that thermal performance of an aquatic ectotherm is limited by the capacity to deliver O2 to its tissues (Pörtner and Knust, 2007; Pörtner and Farrell, 2008). Thus, Fry Curve, which is a function of temperature and was originally conceived for fishes by Fry, (1947), graphically describes the thermal performance of a fish by illustrating the capacity to increase its tissue O2 delivery above and beyond the routine needs (Farrell, 2009). According to the Fry Curve, an elevated metabolic rate without a corresponding increase in MMR will reduce AAS, which consequently will cause the decrease of upper critical temperatures. In an earlier research in Italian rainbow trout (3 - 10 g), AAS at 20°C averaged at 355.2 mg O2 kg−1 h−1 (Wieser, 1985), which is lower than the value in the present study (474.3 mg O2 kg−1 h−1), suggesting the PFRC H-line trout may have a greater capacity for 51  performance at warmer temperatures (>20°C). However, naturally warm adapted populations of redband trout may have evolved an even greater AAS compared with the artificially selected stock in the present study. For example, RMR of redband trout at 24°C (200 ± 13 mg O2 kg−0.83 h−1; Rodnick et al., 2004) was lower than the values generated here, yet MMR was the same in both studies. As a result, AAS of redband trout at 24°C (533 ± 22 mg O2 kg−0.882 h−1) is higher than the rainbow trout from Western Australia, which suggests that redband trout may be better adapted to warm temperatures even though they have similar CTMAX (29.4°C versus 29.0°C, Figure 2.2; Appendix A.2). While Gaussian curves fitted the MMR and AAS data well, extrapolation of these lines beyond the actual data points must be undertaken with great caution. For example, here, the extrapolated values for upper Tcrit exceeded the CTMAX values, which is a highly unlikely situation. Instead, this anomaly reflects the difficulty in measuring metabolic rates at test temperatures > 25°C, where AAS falls precipitously with increasing temperature. Thus, I included a quadratic fit (dash lines) to the data (Figure 2.3 B) as an alternate extrapolation for Tcrit. Gaussian functions (solid lines in Figure 2.3 B), however, provided a better-fit regression than quadratic functions and thus are used to predict Topt for AAS (from 15.3 to 16.5°C). Results showed that 90% of maximum AAS could be maintained up to 19.0 – 20.5°C depending on the family, and FAS was an impressive 2.7 to 3.1 at these temperatures. By comparing the upper Tpej with other congeneric samonids species, PFRC H-strain was at the upper end range of the values for sockeye salmon populations (16.4 – 20.7°C; Eliason et al., 2011). Where the PFRC H-strain truly stands out was its potential ability to maintain a large aerobic scope at temperatures close to CTMAX. At 25°C, RMR could still be increased by 40% to 80% depending on the family groups. Given that salmonids can double their 52  metabolic rate after consume a large meal (Farrell et al., 2001; Fu et al., 2005; Eliason and Farrell, 2014), I conclude that the PFRC H-strain at 25°C would have some (but not full) capacity for both digestion and activity. This conclusion is consistent with empirical observations of fish swimming and eating in the holding aquaria on a hot summer day (e.g. 26.1°C in the present study). Even so, the temperature margin between these activities and death is small. A previous research found that no PFRC can survive 27°C for over 20 hours, but the mortality rate varied across time (Molony et al., 2004). The group variability in aerobic scope and RMR observed here might help explain the variable mortality observed earlier. Further studies will be needed to explain how aerobic scope can be maintained at such high temperatures in this strain.  2.4.3 Maximum heart rate (fH,max) The heart plays a critical role in delivering O2 to tissues. The Fick equation states that O2 uptake is the product of cardiac output and tissue O2 extraction from the blood. Moreover, cardiac performance becomes even more critical during acute warming because it is predominately the heart rate, much more so than either cardiac stroke volume or tissue O2 extraction, that increases as metabolic rate increases (Farrell, 2009). In salmonids, data suggest that the inability of fH,max to continue to increase exponentially with temperature may be the initial trigger for the leveling of aerobic scope as Topt is approached (Farrell, 2009; Casselman et al., 2012; Eliason et al., 2013). Here, as in earlier studies (Casselman et al., 2012; Chen et al., 2013; Sidhu et al., 2014), I forced the heart rate of fish to reach the fH,max status using pharmacological approaches and examined the response to acute warming. The proximity of TAB (Table 2.3) and upper Tpej (Table 2.2) contrasts with the previous findings that TAB was closer to either Topt (Casselman et al., 2012) or Tpej (Ferreira et al., 2014). 53  Another important association is that cardiac arrhythmias began when Tpej was exceeded and aerobic scope was in decline. The reason is that if the fH,max cannot increase further, then the only means to increase maximum aerobic scope (or maintain if fH,max is in decline) would be to increase cardiac stroke volume and/or tissue O2 extraction, which are less efficient. Indeed, the present study found that cardiac arrhythmias developed well beyond the upper Tpej, but always at temperatures prior to CTMAX. 2.4.4 Ventricle mass, hemoglobin (Hb) and hematocrit (Hct) The Hct level was higher than the mean value in rainbow trout (Gallaugher and Farrell, 1998), while Hb concentration was similar to the previous findings in redband trout (Rodnick et al., 2004). However, the relative ventricle mass (RVM) of PFRC rainbow trout was smaller (7.7 – 33.1%) than RVM in redband trout (Rodnick et al., 2004), which may be a hatchery effect. If the difference in cardiac mass translates to a different cardiac output or arterial blood pressure, it may help explain why redband trout appears to be better adapted to warm temperatures. 2.4.5 Conclusion On the whole, despite the fact that Western Australia is a marginal habitat for rainbow trout (high temperatures, low dissolved O2 content and loss of aquatic habitat as creeks and reservoirs dry up), the stocking program carried out annually by PFRC is able to maintain a viable recreational fishery based on the PFRC H-strain of rainbow trout in stocked habitats in Western Australia. The present study provided physiological evidences to support the previous evidences that the PFRC H-strain has undergone selection and demonstrated the ability to survive warm climate. Future comparative physiological and genomic studies could help to elucidate the mechanism of thermal adaptation at both the inter- and intra-specific 54  level. Because of the adequate performances at both growth (Molony et al., 2004) and temperature tolerance, PFRC rainbow trout may represent a promising strain that is suited for aquaculture in a warming climate.55  Table 2.1 Body size of the PFRC rainbow trout O. mykiss used in each measurement. Family CTMAX measurements  Aerobic scope   fH,max measurements Group N Mass (g) Length (cm)  N Mass (g)  N Mass (g) Length (cm) 1 75 27.7 ± 0.6c 12.6 ± 0.08c  6 33.6 ± 0.7  12 35.8 ± 2.4 14.0 ± 0.3 2 75 31.0 ± 0.6ab 13.1 ± 0.08ab  6 33.0 ± 0.5  12 39.6 ± 1.9 14.3 ± 0.2 3 75 29.5 ± 0.6bc 12.8 ± 0.09bc        4 75 32.8 ± 0.5a 13.4 ± 0.06a        5 75 30.1 ± 0.5b 13.0 ± 0.08b  6 34.2 ± 0.8  12 35.4 ± 1.9 13.9 ± 0.3 Mean 375 32.0 ± 0.3  13.0 ±0.04  18 33.6 ± 0.4  36 37.2 ± 1.2 14.1 ± 0.2 Note: different superscript letters indicate the significant differences among groups (p < 0.05). All values are mean ± s.e.m   Table 2.2  Topt, Tpej, Tcrit of absolute aerobic scope (AAS) in three family groups of PFRC rainbow trout.  Fry Curve Estimates of optimum temperatures for AAS (Topt), pejus temperatures (Tpej), critical temperatures (Tcrit) values and their corresponding AAS as derived from the Fry Curves for three family groups of H-strain rainbow trout O. mykiss.  Family group 1  Family group 2  Family group 5  Family Mean  Temperature (°C) AAS (mg O2 h-1 kg-1)  Temperature (°C) AAS (mg O2 h-1 kg-1)  Temperature (°C) AAS (mg O2 h-1 kg-1)  Temperature (°C) AAS (mg O2 h-1 kg-1) Lower Tcrit -2.4 ± 0.7 55.0 ± 2.6  -1.0 ± 1.4 56.0 ± 2.2  -4.3 ± 2.5 54.1 ± 2.4  -2.6 ± 1.0 55.0 ± 1.3 Lower Tpej 12.4 ± 0.4 495.0 ± 23.1  11.6 ± 0.2 503.7 ± 19.8  11.5 ± 1.0 487.0 ± 21.9  11.8 ± 0.4 495.2 ± 11.8 Topt 16.5 ± 0.6 550.0 ± 25.6  15.3 ± 0.2 559.7 ± 22.0  15.8 ± 0.7 541.2 ± 24.3  15.8 ± 0.3 550.3 ± 13.2 Upper Tpej 20.5 ± 0.8 495.0 ± 23.1  19.0 ± 0.4 503.7 ± 19.8  20.2 ± 0.5 487.0 ± 21.9  19.9 ± 0.3 495.2 ± 11.8 Upper Tcrit 35.3 ± 1.6 55.0 ± 2.6  31.6 ± 1.7 56.0 ± 2.2  36.0 ± 1.6 54.1 ± 2.4  34.3 ± 1.0 55.0 ± 1.3 Note: Estimates were generated from individual fish data using either a Gaussian 3-parameter function to generate the Fry Curve (similar to the fitting for the mean value in Figure 2.3 B). Using these equations, Topt was defined as the temperature with maximum AAS, upper and lower Tpej were defined as the maximum and minimum temperatures at which aerobic remained above 90% of the maximum AAS, and the upper and lower Tcrit were defined as the maximum and minimum temperatures at which AAS was at least 10% of the maximum aerobic scope at Topt. All values are mean ± s.e.m.   56  Table 2.3 Rate transition temperatures of maximum heart rate in three family groups of PFRC rainbow trout. Arrhenius breakpoint temperature (TAB), Q10 breakpoint temperature (TQB), temperature of highest fH,max (TPEAK), cardiac arrhythmia temperature (TAR) values and their corresponding heart rate for rainbow trout family groups.  Family group 1  Family group 2  Family group 5  Family Mean  Temperature (°C) fH,max (beats min−1)  Temperature (°C) fH,max (beats min−1)  Temperature (°C) fH,max (beats min−1)  Temperature (°C) fH,max (beats min−1) TAB 20.7 ± 0.4b 152.6 ± 4.9a  20.3 ± 0.3b 151.0 ± 3.4a  20.6 ± 0.1b 148.6 ± 1.5a  20.5 ± 0.2 150.8 ± 2.1 TQB 18.9 ± 0.2b 137.0 ± 2.6b  19.0 ± 0.2b 141.6 ± 2.0b  18.8 ± 0.3b 135.2 ± 2.8b  18.9 ± 0.1 137.8 ± 1.5 TPEAK 24.0 ± 0.4a 173.7 ± 4.6b  23.5 ± 0.4a 173.1 ± 3.5b  23.7 ± 0.3a 168.0 ± 2.7b  23.7 ± 0.2 171.6 ± 2.2 TAR 25.7 ± 0.4a   24.7 ± 0.4a   24.9 ± 0.4a   25.1 ± 0.3  Note: TAB represents the transition temperature beyond which the exponential changes of fH,max slows. TQB represents the transition temperature where the incremental Q10 for fH,max decreases and remains below 1.8. The TMAX is the highest absolute value achieved for fH,max. TAR is the temperature where cardiac arrhythmia first develops. Different superscript letters indicates the significant differences within each group. All values are mean ± s.e.m   Table 2.4 Wet ventricle mass, hematocrit (Hct) and hemoglobin concentration (Hb) for PFRC rainbow trout. Family Group Ventricle size   Blood analysis N Mass (mg) RVM (%)  N Hct (%) Hb (g dL−1) 1 12 29.1 ± 2.3 0.081 ± 0.003  5 33.8 ± 3.7 11.1 ± 0.1 2 12 32.1 ± 1.6 0.081 ± 0.002  3 35.3 ± 2.1 10.0 ± 0.2 3     5 36.9 ± 2.6 9.8 ± 0.5 5 12 30.6 ± 2.4 0.084 ± 0.003  4 32.9 ± 1.1 9.9 ± 0.3 Mean 36 30.6 ± 1.2 0.082 ± 0.002  17 37.4 ± 1.3 10.1 ± 0.2 Note: RMV, Relative wet ventricular mass to body mass. All values are mean ± s.e.m.  57    Figure 2.1 Temperature of water passing through Pemberton Freshwater Research Centre (PFRC) ponds from January to May, 2014.     Figure 2.2 CTMAX of five family groups of PFRC rainbow trout.  Each family group contained five sub-families that shared the same sires. The individual mean values (s.e.m.) for the half-sibling families (white symbols; n= 15 per half-sibling family) and the mean value for the family groups (red symbol; n= 75 per group) are jittered on the x-axis from each other for clarity. Letters above the symbols denote a statistically significant difference in means among family groups (p < 0.05). 58    Figure 2.3 Routine (RMR) and maximum (MMR) metabolic rate, and absolute aerobic scope (AAS) for three family groups of PFRC rainbow trout. A: RMR and MMR (mean ± s.e.m.; N = 6 for each Family Group). RMR was fitted with polynomial quadratic function (Group 1, red, y=74.7+(−6.8)*x+0.6*x2, R2 = 0.99, p < 0.01; Group 2, blue, y=149.7+(−18.9)*x+1.2*x2, R2 = 0.99, p < 0.01; Group 5, black open, y=59.2+(−6.3)*x+0.9*x2, R2 = 0.99, p < 0.01). MMR was fitted with Gaussian 3-parameter function (Group 1, red, y=753.0*e^(−0.5*((x–20.0)/11.23)^2), R2 = 0.99, p < 0.01; Group 2, blue, y=747.2*e^(−0.5*((x–19.2)/11.1)^2), R2 = 0.93, p = 0.019; Group 5, black, y=775.8*e^(−0.5*((x–20.6)/11.3)^2), R2 = 0.98, p < 0.01). Gray area represents the ± 95% confidence interval of the aforementioned fittings for the overall means in three groups.  B: AAS (absolute difference between MMR and RMR). Solid lines: Gaussian 3-parameter function was fitted for each group to generate Fry Curve (Group 1, red, y=548.1*e^(−0.5*((x – 16.0)/8.3)^2), R2 = 0.98, p < 0.01; Group 2, blue, y=558.5*e^(−0.5*((x – 15.2)/7.6)^2), R2 = 0.84, p = 0.062; Group 5, black, y=537.1*e^(−0.5*((x – 16.0)/8.6)^2), R2 = 0.95, p = 0.012). The broken horizontal lines represent 90% (marked as “Pejus”) and 10% (marked as “Critical”) of the maximum AAS for each group and the intersection of these lines with the Fry Curves represent the pejus temperatures (Tpej) and critical temperatures (Tcrit). Owning to the similarity of AAS values at Tcrit, the “Critical” lines overlap with each other. Dash lines represent the fitting with polynomial quadratic function as an alternative way to extrapolate the rate transition temperatures. 59             Figure 2.4 Maximum heart rate (fH,max) in response to temperature increase in three family groups of rainbow trout. In all panels, red = Family Group 1, blue = Family Group 2, black open = Family Group 3.  A: Pharmacological-stimulated fH,max during incremental warming of three family groups of PFRC H-line rainbow trout. Mean (± s.e.m.) fH,max data for each group are connected by a solid line to indicate that all fish had regular heart rhythm (N = 12 for each group). After some fish developed 60  cardiac arrhythmia, the number of the remaining fish without cardiac arrhythmia (symbols with dash lines) are also given and their mean fH,max data is shown by the unconnected symbols.  B: Incremental Q10 values for fH,max during incremental warming of three family groups of H-line rainbow trout. Each line represents an individual fish responses. The rate transition temperature for Q10 (TQB) was arbitrarily set as the temperature beyond which the incremental Q10 for fH,max fell below 1.8 (the straight dash line). Specifically, TQB is the last temperature where the Q10 is above the dash line. C: Arrhenius plot of mean fH,max during incremental warming for three family groups of H-line rainbow trout. The mean fH,max were derived from the individual fish responses summarized in Panel A. It was possible to fit two segmental linear regressions to the data to identify the Arrhenius breakpoint temperature (TAB), which are indicated by the arrows.     61  Chapter 3: Intraspecific genomic variation and associations related to thermal performance in redband trout (Oncorhynchus mykiss gairdneri) Synopsis In the previous chapter, I used a combination of approaches to examine the thermal performance of a rainbow trout population that has been raised in a warm climate for over 19 generations. The following two chapters expand the study of a single domesticated population to three natural populations of inland rainbow trout (redband) populations (Oncorhynchus mykiss gairdneri) from different thermal habitats.  Beyond the same phenotypic measurements used in Chapter 2, I extend my exploration to genomic (Chapter 3) and cellular levels (Chapter 4). Using Columbia River redband trout from desert and montane streams, where large differences in summer water temperatures exist, this chapter shows intraspecific variations in CTMAX and AAS. As might be predicted from the climate of their natural habitat, CTMAX is significantly higher in the hot desert population (Little Jacks) compared to the cool montane populations (Keithley and Fawn). Moreover, Little Jacks has a broader optimum thermal window for AAS.  Genome-wide neutral loci suggest that geographical separation has caused genetic differentiations between desert and montane populations. Additionally, signs of directional selection were observed as indicated by the twenty-one “outlier” (i.e. potentially adaptive) loci showing population-specific allele frequencies. Furthermore, quantitative genetic markers for CTMAX and AAS are identified by the genome-wide association studies. A total of twenty-three loci are significantly associated with individual CTMAX. Results in this 62  chapter elucidate novel intraspecific variations of thermal performances and potential genetic mechanisms underlying thermal adaptation in rainbow trout.  3.1 Introduction In aquatic ectotherms, such as fishes, conspecific populations from different thermal regimes often differ in their thermal performance (Hart, 1952; Prosser, 1955; McCauley, 1958; Feminella and Matthews, 1984), which is caused by the evolutionary changes in biochemical and physiological functions, often as a result of reproductive isolation in diverse thermal regimes. Understanding the genetic components of such adaptive traits may help predict the potential impact of global warming on the survival and distribution of wild populations (Somero, 2010; Hoffmann and Sgro, 2011; Stillman and Armstrong, 2015). Nevertheless, while the physiological mechanisms of thermal adaptation have been studied in cellular membranes (Cossins and Prosser, 1978; Logue et al., 2000), functioning proteins (Hochachka and Somero, 2002) and aerobic metabolism (Pörtner and Farrell, 2008; Schulte, 2015), the underlying genetic mechanisms, although have been widely studied across taxa (Watt, 1977; Riehle et al., 2001; Sørensen et al., 2005; Knies et al., 2006; Gerken et al., 2015; Porcelli et al., 2015; Lendenmann et al., 2016), remain largely unclear. Advances in sequencing techniques, e.g. restriction site-associated DNA (RAD) sequencing, have made large-scale genome sequencing economical (Miller et al., 2007; Baird et al., 2008; Davey and Blaxter, 2010). In RAD sequencing, genome DNA was digested by a restriction enzyme (e.g., Sbf1) and then size selected to generate a reduced representation of genome, which is sequenced using any one of the several sequencing platforms (Davey et al., 2011). One advantage of RAD sequencing is the capacity to identify a large number of polymorphisms at relative low cost. Another advantage is that RAD 63  sequencing does not require a reference genome, thus it is useful in non-model organisms (Ekblom and Galindo, 2010; Baxter et al., 2011). As a result, high throughput sequencing has detected quantitative trait loci (QTLs) for migration traits (Hecht et al., 2013) and disease resistance traits (Campbell et al., 2014; Palti et al., 2015) in rainbow trout (Oncorhynchus mykiss). In redband rainbow trout (Oncorhynchus mykiss gairdneri), genome wide association study (GWAS) for the mortality at high temperatures has identified eighteen significant loci across the genome including three heat shock protein genes and one immune responsive gene (Narum et al., 2013b). QTLs for temperature tolerance have been studied in chinook salmon (Oncorhynchus tshawytscha) using RAD sequencing (Everett and Seeb, 2014; McKinney et al., 2016), but not in other salmonids species. Previously, microsatellite studies have mapped QTLs for upper thermal tolerance (UTT) in salmonids (Araneda et al., 2008), mostly in rainbow trout (Jackson et al., 1998; Danzmann et al., 1999; Perry et al., 2001). Microsatellite markers are simple sequence repeats that are commonly found across the genome. While microsatellite markers have the advantage of being highly heterozygous, they also have the disadvantage of low-density coverage across genome compared to the single nucleotide polymorphism markers (SNPs) (Liu and Cordes, 2004), which brings problems in pinpointing the relevant genes. Nevertheless, candidate genes for UTT have been suggested based on microsatellite QTLs in both rainbow trout (Coulibaly et al., 2006) and Arctic charr (Salvelinus alpinus) (Somorjai et al., 2003; Quinn et al., 2011). Beside the quantitative approaches, genomic regions that are associated with thermal adaptation can also be identified indirectly using population genetics (Hohenlohe et al., 2010; Stapley et al., 2010; Matala et al., 2011; Narum et al., 2013a; Hecht et al., 2015). 64  During thermal adaptation, genes that contribute to the variations of adaptive traits will change in allele frequency as a result of natural selection, such as the evolution of different functional isoforms of haemoglobin in Atlantic cod (Gadus morhua) (Andersen et al., 2009) and lactate dehydrogenase-b (LDH-B) in killifish (Fundulus heteroclitus) (Schulte et al., 2000). As a result, the functionally important genes often demonstrate genetic variations that are extremely divergent (“outlier”) among populations in relation to the rest of the genome (“neutral”) (Holderegger et al., 2006; Narum and Hess, 2011). Ideally, genes under selection in a population can be constantly monitored by comparing the allele frequencies before and after thermal adaptation across generations (Hoffmann and Willi, 2008; Franks and Hoffmann, 2012). However, it is rather ambitious to conduct such comparisons because of the large time-scale of adaptation relative to experimentation. Fortunately, there is a more realistic approach using conspecific populations that have been exposed to heterogeneous environments, sometimes without gene flow due to habitat fragmentation. Thus, by using geographic space as a proxy for time, allele frequencies can be compared among different thermally adapted populations. Columbia River redband trout is an inland form of rainbow trout, as opposed to the coastal rainbow trout (O. mykiss irideus) that are distributed on the west of Cascade and Coastal mountain ranges (Allendorf and Utter, 1979; Behnke, 2002; Currens et al., 2009). Redband trout populations have adapted to occupy montane and desert streams that have different thermal extremes during summer. For example, water temperatures in summer can reach over 29°C for some desert streams, yet some of the montane streams may not exceed 20°C (Zoellick, 1999; Rodnick et al., 2004; Meyer et al., 2010; Narum et al., 2010). The presence of physical barriers (e.g. waterfalls) in the Columbia River also creates reproductive 65  isolation between populations. Therefore, redband trout populations are good models to study thermal adaptation. Evidences for thermal adaptation of redband trout populations to different climate zones has been provided from both physiological (Gamperl et al., 2002; Rodnick et al., 2004) and genetic studies (Narum et al., 2010; Narum and Campbell, 2015; Garvin et al., 2015). Because of the habitat fragmentation, some redband trout populations have a restricted ability to disperse (Moyle et al., 2013) and thus may be forced to use their genetic potential to adjust for the climate change in the 21st century. Indeed, previous studies found the critical thermal maximum (CTMAX) of redband trout population to be around 29-30°C (Rodnick et al., 2004; Cassinelli and Moffitt, 2010), which is surprisingly close to the peak summer water temperatures in some of the desert streams that they inhabit. Thus, studying redband trout not only will provide mechanistic understandings of thermal adaptation, but also satisfy the urgency to better understand how to conserve wild stocks in the face of global warming (Nehlsen et al., 1991; Katz et al., 2013). In this chapter, I extend the physiological quantification of the thermal tolerance phenotype by measuring their CTMAX and aerobic scope, as well as considering three contrasting populations from desert and montane creeks in southern Idaho according to a previous population genetic study (Narum et al., 2010). In addition, a hybrid strain between a desert and a montane population is also investigated to provide insights into the relative effects of additive and dominance components of genetic variation in thermal performance. Because thermal performance traits are plastic to varying degrees (Beitinger and Bennett, 2000; Ferreira et al., 2014), all fish were reared in a common garden environment (15°C) to minimize the potential acclimation effect. 66  In addition to the characterization of thermal phenotypes, SNPs were generated from RAD sequencing to investigate the genetic differentiation among populations. Furthermore, GWAS was conducted to identify markers that are associated with CTMAX. My hypotheses are: (1) intraspecific differences in CTMAX and aerobic scope exist among redband trout populations; (2) redband trout populations are differentiated in part through adaptation to different climatic regions; (3) thermal performance is a polygenic trait and thus is associated with genetic variation at multiple loci. By studying the intraspecific differences at physiological and genetic levels, my aim is to provide a greater mechanistic understanding of thermal adaptation within a species at the population level.  3.2 Materials and methods All experiment procedures were approved by the University of British Columbia Committee on Animal Care in accordance with the Canadian Council on Animal Care (A10-0335) and by the University of Idaho (IACUC protocol 2013-80). 3.2.1 Fish culture and rearing conditions Columbia River Basin redband trout is native to southern Idaho, northern Nevada and eastern Oregon. Their habitat constitutes very different landscape types, including desert, montane and high plateau environments. This study targeted redband trout from three ecologically divergent populations in Snake River tributaries in southern Idaho, as determined from previous studies (Narum et al., 2010; Narum et al., 2013b): Little Jacks Creek (LJ, desert climate: 42.728700⁄−116.10516), Keithley Creek (K, cool montane climate: 44.553380⁄−116.88535) and Fawn Creek (F, cold montane climate: 44.382336⁄−116.05894). While Snake River redband trout spawn from May through September (Behnke, 2002), the three natural habitats in this study differ in many ways, including elevation, air temperature, 67  and riparian vegetation, as well as gradient of peak summer water temperature (Figure 3.1) (Meyer et al., 2010). In addition, genetic differentiation exists among these three redband trout populations, possibly due to the adaptive response to different climate regimes (Narum et al., 2010).  Also, watershed disturbances, such as the construction of dams, irrigation diversions and natural migration barriers, have not only altered the physical connectivity among resident redband trout populations but also reduced the genetic connectivity (Kozfkay et al., 2011). For my study, newly emerged fry (approximately three months old) were captured by electroshocking/dip-netting from each study stream during summer months and transported by truck to Hagerman Fish Culture Experiment Station (Hagerman, Idaho) where they were reared. Fry collection from each stream was approved by Idaho Department of Fish and Game (Permit F-13-06-13). All fry were reared in a common garden environment for approximately 6 weeks prior to the physiology studies. Rearing tanks were supplied with natural spring water at a constant temperature of 15°C, i.e. well below the summer thermal extremes in natural creeks for all populations. Fry were reared on a commercial diet (Rangen Inc., Buhl, Idaho, USA), feeding daily to satiation. Photoperiod was maintained at 14 hours light and 10 hours darkness.  Previous attempts to spawn each strain from captive adults (reared in the hatchery from juveniles collected in the previous generation) were generally unsuccessful. Despite that, a hybrid strain of half-sibling families from a directional cross between Keithley females and Little Jacks males (K×LJ) were successfully produced from captive adults and the F1 hybrid fry were also included as a study strain.  A reciprocal cross was not possible between Keithley males and Little Jacks females since eggs were unavailable from Little Jacks females (both wild and hatchery reared). 68  3.2.2 Phenotyping 3.2.2.1 Critical thermal maximum (CTMAX) I used the same protocol as detailed in Chapter 2, but a different heater/chiller unit was used in the experiment for this chapter (3016D heater/chiller; Fisher Scientific, Ottawa, Ontario, Canada). CTMAX was conducted once for each population. Post-testing fish mortality was <5% (3 of 63 fish). After each CTMAX measurement, fish were anaesthetized and euthanized in 80 mg L−1 MS222 buffered with 160 mg L−1 NaHCO3 to measure the body size and to obtain a fin clip for DNA extraction. Fin clips were dry stored on Whatman chromatography paper.  3.2.2.2 Routine and maximum metabolic rate This study used a repeated measurement design, i.e. six fish from each strain had their individual RMR and MMR repeatedly-measured at five test temperatures (12°C, 15°C, 18°C, 21°C and 24°C). Because it took four days to measure all four groups (one strain per day) at each temperature increment, each group had a minimum of three days to recover from the previous measurement before testing at the next temperature. Also, test temperatures were in the order of 15°C (acclimation temperature), 18°C, 12°C, 21°C and 24°C for each population to reduce the potential thermal acclimation effect. These measurements were completed within 30 days to minimize the influence of fish growth during the course of the metabolic rate measurements.  Briefly, six fish were randomly chosen and placed into separate ID-tagged floating boxes made from rigid plastic mesh (diameter 15 cm, height 20 cm). Floating boxes were kept in the same holding tank that reared other fish from the same strain. Before each metabolic rate measurements, fish were fasted for 48 h. Firstly, fish were individually 69  transferred to one of the six custom-made, intermittent-flow respirometers (water volume = 87.6-106.6 ml). Water temperature in respirometers was controlled with an in-line, recirculating heater/chiller unit (3016D, Fisher Scientific). Fish were left overnight for a minimum of 12 hours at the temperature of 15°C to adjust to the respirometer. Water temperature was then changed at the rate of 2°C hour−1 to the desired test temperature. After the test temperature was reached, fish were held for an additional 1-hour before RMR was measured by closing the respirometer and recording the depletion of O2 from the water with Loligo Systems (Loligo, PyroScience GmbH, Aachen, Germany). Dissolved oxygen was allowed to decrease by no more than 25% (i.e. maintain ≥75% air saturation), which generally took less than 10 min. Immediately following the RMR measurement, MMR was measured for each individual by removing one fish at a time from the respirometer and placing it into a 50 cm diameter circular tank that contained aerated water at the test temperature. The MMR state of a fish was achieved through a series of forced exercise until exhaustion, which took approximately five min by combining manual chasing with dip net and gentle tail pinches until the fish was unresponsive to touch (Norin and Clark, 2016). Then, exhausted fish was immediately transferred to a respirometer (within <1.5 min) and O2 uptake was continuously measured during recovery. During the measurements, water circulation through the respirometer was adjusted to the maximum level to ensure adequate water mixture. MMR was determined by taking the maximum rate of O2 removal from water for a minimum of 1.5 min during a 3-min record period. MMR typically occurred at the initial measurement period, i.e. immediately after the fish returned to the respirometer. After the MMR measurement, fish were weighed and transferred back to their floating mesh box to recover. Each fish was given a minimum of three days to recover, during which period 70  behavior was regularly monitored and feeding was resumed. Before the first metabolic rate experiment, mean body mass of tested fish was 2.8±0.9 g (mean±s.e.m, n=24). Body mass increased by an average of 7.9% during the experimentation period despite the food restriction, handling and testing. Calculation of RMR and MMR has been detailed in Chapter 2. Finclips were harvested after the measurements at last test temperatures. 3.2.3 Genotyping 3.2.3.1 DNA extraction DNA was extracted from dried caudal fin tissues using a DNeasy Blood and Tissue Kit (QIAGEN, Valencia, CA) according to the instruction manual. Extracted genomic DNA (gDNA) was quantified using Quant-iT PicoGreen dsDNA Assay kits (Invitrogen, Grand Island, NY, USA) and an Infinite M200 Pro Microplate reader (Tecan Group Ltd., männedorf, Switzerland, www.tecan.com). 3.2.3.2 Doubled haploid samples In addition to the samples from the CTMAX and metabolic rate measurements, five previously sequenced doubled haploid rainbow trout samples were included to help identify paralogous sequence variants (PSV), which are homologous regions of the genome created through local or global genome duplication events that cannot be differentiated easily. This is a common challenge when genotyping diploid salmonid fishes (Miller et al., 2012; Brieuc et al., 2014) given their high degree of genome duplication (Berthelot et al., 2014).  These doubled haploid rainbow trout samples were created using previously described methods (Parsons and Thorgaard, 1985; Scheerer et al., 1991) and detailed in previous studies (Hecht et al., 2012; Miller et al., 2012).  With doubled haploid samples, I expected the genotype at each locus to be homozygous, given their haploid genomic background.  Therefore, loci that 71  exhibit heterozygous genotypes in doubled haploid samples were considered to be putative PSVs or duplicated loci, and were filtered from further genetic and statistical analyses as commonly performed in salmonid studies (Hecht et al., 2012; Miller et al., 2012; Brieuc et al., 2014). 3.2.3.3 RAD library preparation and sequencing I employed RAD sequencing to simultaneously identify thousands of SNP markers distributed throughout the genome and to genotype samples at those loci (Miller et al., 2007; Baird et al., 2008). RAD libraries were prepared using methods previously outlined (Baird et al., 2008; Miller et al., 2012), wherein sample gDNA was digested using the restriction enzyme Sbf1-HF (NEB, Ipswich, MA, USA) and individually barcoded using a six-nucleotide barcode adapter sequence.  A total of 95 barcoded individuals were then divided and evenly pooled into two separate RAD libraries, where no two samples within a library were assigned the same barcode sequence and each barcode sequence within a library differed by at least two bases from another barcode sequence.  Libraries were mechanically sheared to generate DNA fragments with length between 200-700 bp using a Bioruptor 300 sonicator (Diagenode, Denville, NJ, USA). Fragments were size-selected and isolated using an Agencourt AMPure XP bead purification system (Beckman Coulter, Brea, CA, USA).  The remainder of the RAD library preparation followed the previously defined protocol in Hecht et al., (2012). Each library was run on a single lane of a single read 100 bp flow cell and sequenced on an Illumina HiSeq 1500 for 100 cycles. 3.2.3.4 De Novo SNP discovery and genotyping I identified and genotyped SNP loci de novo, by constructing a SNP catalog of individuals from the three source populations (Little Jacks, Keithley and Fawn).  This was 72  performed using the software pipeline Stacks (Catchen et al., 2011; Catchen et al., 2013).  Raw Illumina reads were first scrutinized for quality using the software program FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/).  It was determined that 25 bases of Illumina sequence reads on the 3’ end had reduced quality scores relative to the 5’ 75 base positions across my sequence data.  Therefore, I truncated sequence reads to 75 bases by removing the 3’ sequence that is most prone to error.  In addition to truncating, reads were quality filtered and de-multiplexed using the ‘process_radtags’ program of the Stacks pipeline and included options for cleaning the data by discarding any read with an uncalled base (-c), discarding reads with low quality scores (-q), and rescuing barcodes and partial restriction enzyme recognition sites (-r).  All other parameters and options were executed with the default parameters as outlined in the manual for the program (http://creskolab.uoregon.edu/stacks). After individual sample reads were quality filtered, trimmed and de-multiplexed, sequences for each sample were submitted to the ‘ustacks’ module of Stacks to identify loci.  In ‘ustacks’, the deleveraging (-d) and removal (-r) algorithms were applied to filter out those sequences that were likely to be paralogous and highly repetitive.  I required the minimum depth of coverage at a stack (-m) to be five and allowed a maximum distance (-M) of two between stacks, and four between secondary reads and primary stacks (-N).  SNP discovery was carried out using the ‘bounded’ SNP model with a lower and upper bound for epsilon of 0.001 and 0.01, respectively, and a chi-square significance level of 0.05 was set for the alpha parameter.  I created a de novo catalog of RAD tag SNPs using the ‘cstacks’ module by selecting 7 individuals from each population (Little Jacks, Keithley and Fawn) for a total of 21 fish with at least three million (M) reads to evenly represent genetic variation throughout 73  the source populations.  Individual samples were then aligned to the catalog using the module ‘sstacks’ and genotypes were exported using the ‘populations’ module.  Genotypes were filtered to exclude: 1) any RAD tag locus with more than four SNP sites to remove putative PSVs, hyper-variable, or poorly sequenced tags, 2) any RAD tag locus where one of the five doubled haploid samples was observed to be heterozygous at any of the SNP positions in order to remove putative PSVs, 3) any SNP marker with more than two alleles to remove SNPs with sequencing errors, putative PSVs, or loci that do not fit a bi-allelic statistical model, 4) any SNP marker missing more than 20% of the genotypes across all of the populations to limit the amount of missing data, and 5) any SNP marker with an average minor allele frequency falling below 0.05 in any of the populations in order to exclude spurious rare SNPs or sequencing errors.  Since physically linked SNP markers would bias downstream statistical models, I retained only one SNP marker per RAD tag, where the first SNP to occur within the RAD tag sequence was selected for downstream analysis.  Individual samples were also filtered from the dataset if they were missing more than 20% of genotypes across all filtered loci. 3.2.3.5 Annotation and mapping of RAD tag sequences In order to identify chromosome assignments and map positions for loci, RAD tag sequences were aligned to sequences from two previously generated RAD-based genetic linkage maps in rainbow trout (Hecht et al., 2012; Miller et al., 2012) using the software program Bowtie2 v.2.2.4 (Langmead et al., 2009).  RAD tag sequences were also aligned to the rainbow trout genome (Berthelot et al., 2014) using Bowtie 2 v.2.2.4 (Langmead and Salzberg, 2012) to assign sequences to rainbow trout chromosomes and obtain genomic positions.  By querying the rainbow trout genome for coding sequence within 5, 10, and 15 74  kb of RAD tag locus alignment sites, putatively linked genes were identified.  To identify gene functions and annotations, coding sequences were then queried against the NCBI nucleotide sequence database using the software program Blast2GO (Conesa et al., 2005).  3.2.4 Population genetics 3.2.4.1 Genetic diversity (heterozygosity) Measurements of genetic diversity in this study was estimated by expected heterozygosity (HE), which measures the proportion of individuals in the sample group that are heterozygous (Nei, 1978). Difference in HE between populations were tested using Monte Carlo procedure implemented in the R package ‘adegenet’ (Jombart, 2008). Deviation from Hardy–Weinberg (H-W) equilibrium for each combination of locus and population was estimated by Markov Chain Monte Carlo (MCMC) approximation of Fisher’s exact test (100 batches with 5000 iterations) in GENEPOP v. 4.2 (Raymond and Rousset, 1995). False discovery rate (FDR) of multiple comparisons was controlled by the BY-FDR method (Benjamini and Yekutieli, 2001; Narum, 2006). 3.2.4.2 Population differentiation (FST) FST values were used as a measure of genetic distance to compare the population divergence. Pairwise FST values (Weir and Cockerham, 1984) and tests of significance (Fisher’s exact test) for all the groups were obtained with GENEPOP V4.2 (Raymond and Rousset, 1995). 3.2.4.3 Outlier test Two different approaches were used to detect potential loci under positive adaptive selection. First, I used coalescent FDIST2 (Beaumont and Nichols, 1996) implemented in Lositan (Antao et al., 2008). Outliers were determined based on their higher genetic 75  differentiation than expected under a neutral hypothesis with confidence levels of 99.5% in 50000 simulations. In the first simulation, mean neutral FST was calculated by removing the potential selected loci. For the other simulations, mean neutral FST was used for a better estimation of the potential loci under selection. A locus was considered a candidate outlier when FST was over 0.5 with FDR below 0.1. I also used an FST that was based on a hierarchical Bayesian model (Beaumont and Balding, 2004) with a reversible jump MCMC, implemented in BayeScan V2.1 software (Foll and Gaggiotti, 2008). I did the analyses on 20 pilot runs with 5000 iterations, followed by 50 000 iterations with a burn-in length of 50 000 iterations. Using BayeScan, loci were considered to be candidate outliers based on a minimum Bayes factor of 3 and FDR of 0.1. 3.2.5 Genetic association In order to test for genome-wide associations between RAD-tag SNP genotypes and thermal tolerance related traits, I employ a unified mixed linear model (MLM) approach, also known as a “q-k” model, which simultaneously accounts for both population structure (q) and cryptic familial relatedness (k) among individuals (Balding, 2006; Yu et al., 2006).  It is known that both population structure and familial relatedness can cause spurious genotype-phenotype associations and should be accounted for in statistical models (Balding, 2006).  Association models were run in an R computing environment (R Core Team, 2013) using the package ‘GAPIT’ v.2.3.41 (Lipka et al., 2012), which employs the compressed mixed linear model approach (Zhang et al., 2010).  Within ‘GAPIT’, a principal component analysis (PCA) of all SNP markers was used to estimate the fixed effect of population structure (q), while the random effect of relatedness (k) between samples was provided in a kinship matrix estimated using VanRaden’s method (VanRaden, 2008) without compression.  Additionally 76  the fixed effect of body mass (g) at the time of sampling was also considered to be included as a cofactor in trait models.  The best fitting model was determined by using the model selection option in GAPIT, wherein the model with the largest BIC was selected, and confirmed by a qualitative assessment of the QQ plot.  In each case, cofactors tested included body mass at time of sampling, principal components of population structure (PC1-3), and the random effect of kinship.  Missing data was imputed as the major allele at each locus within the ‘GAPIT’ program.  Significant associations were determined by a P-value adjusted for multiple testing using a BY-FDR  adjustment on the initial level of α = 0.05 (Benjamini and Yekutieli, 2001).  This multiple test correction is less conservative than a Bonferroni adjustment, but is recommended as an alternative (Narum, 2006) in order to identify plausible associations and targets of selection, as opposed to holding the analysis so stringent that only the largest effects can be detected. This approach is also expected to reduce false positives over the standard FDR method (Benjamini and Hochberg, 1995). Results from GWAS testing were visualized in Manhattan plots drawn by the R package ‘qqman’ v.0.1.2 (Turner, 2014), where loci were ordered based on a partial rainbow trout genome assembly (Berthelot et al., 2014), and all unassigned loci were, for the purposes of visualization, arbitrarily assigned to the 31st or unknown chromosome. 3.2.6 Statistics All summary statistics were performed using SigmaPlot version 12.5 (Systat Software Inc., San Jose, CA, USA) except where specified. Body mass was compared using one-way ANOVA on ranks with Dunn’s post-hoc test at the level of p < 0·05. Habitat temperature, CTMAX and metabolic date were analyzed by one-way ANOVA with Turkey's HSD post-hoc analysis. To account for the effect of body size, CTMAX was also tested using 77  the analysis of covariance (SPSS version 21.0, IBM Corp., Armonk, NY, USA). Correlation between CTMAX and body size was analyzed by a simple linear regression and a two-segment linear regression. Metabolic rates were present as mass-independent values, i.e. mg O2 g−0.88 h−1. This mass exponent (0.88) was previously determined in a metabolic rate study with redband trout (Rodnick et al., 2004). Analysis of principal components was performed using the algorithm in R package ‘adegenet’ (Jombart, 2008) on a matrix of allele frequencies that was preprocessed by replacing the missing data with mean allele frequency. 3.3 Results 3.3.1 Habitat temperature Seasonal water temperatures for three creeks where the studied populations come from were logged every 20 min and are presented in Figure 3.1.  While the seasonal and diurnal fluctuations are evident in the data, diurnal fluctuations were less extreme in winter months. On the whole, summer water temperatures were significantly different (judged by daily maximum, minimum and mean, p < 0.05) among all three creeks. The desert climate (Little Jacks Creek) had the highest summer water temperatures, followed by the cool montane climate (Keithley Creek) and then the cold montane climate (Fawn Creek) (Table 3.1). Maximum daily summer temperature reached 26.0°C in Little Jacks Creek, 20.7°C in Keithley Creek, and 16.9°C in Fawn Creek. Daily temperature fluctuation in summer was as much as 8°C in Little Jacks Creek, 6.3°C in Keithley Creek and 5.7°C in Fawn Creek. All three creeks had the same minimum temperature (0°C) in winter, but Fawn Creek was around this temperature for a much longer period (from November to February; Figure 3.1). As a result of the lengthy winter, Fawn creek also has lower temperatures in spring, which 78  directly causes a delayed spawning timing in redband trout compared to the other populations from warmer climates. Because AAS was measured at the test temperatures of 12, 18, 21 and 24°C, as well as the acclimation temperature of 15°C, the amount of time fish spent in their natural habitat above each of these test temperatures is presented in Table 3.2. Acknowledging that water temperature will vary annually, Little Jacks Creek in 2014 was > 19°C for 583 h including 19 entire days, >21°C for several hours in some days (total = 160 h), and > 24°C for a limited time (total = 12 h).  In contrast, Keithley Creek in 2014 was never > 18°C for a full day (total = 236 h) or >21°C (total = 0 h).  Fawn Creek in 2015 was never > 18°C (total = 0 h). 3.3.2 Critical thermal maximum (CTMAX) Individual CTMAX values varied from the lowest of 28.3°C in Fawn to the highest of 30.0°C in Little Jacks. The difference in average CTMAX among all populations was < 1°C (Figure 3.2). Despite this limited numerical differences, the two montane populations (Keithley and Fawn) had significantly lower CTMAX values than the desert population (Little Jacks). K×LJ hybrid had a CTMAX value in between the two parental populations, but closer to Little Jacks.  Body mass varied significantly among populations (Table 3.3). Fawn fish were significantly smaller than the other populations.  Furthermore, CTMAX was also correlated with body mass (R2=0.396; linear regression), suggesting that larger juvenile fish have higher CTMAX than smaller juvenile fish. However, the relationship between CTMAX and body mass was better described using two segmented linear (two segment linear regression for body mass groupings centered at 1.5 g; R2=0.55, p < 0.001) (Appendix B.1) such that CTMAX became less mass-dependent when body mass was > 1.5 g.  After using body mass as a 79  covariate in an ANCOVA analysis, the statistical significance of CTMAX between populations persisted (p < 0.01), except for the difference between Little Jacks and K×LJ hybrid (p > 0.05). 3.3.3 Aerobic scope To control the effect of body mass, RMR and MMR were converted to mass-independent values using a scaling coefficient of 0.88, i.e. mg O2 g−0.88 h−1 (Figure 3.3), which was previous used to correct metabolic rate data in redband trout from Oregon, USA with body mass varied from 45 to 1,400 g (Gamperl et al., 2002; Rodnick et al., 2004). As expected, RMR increased significantly with test temperatures in all test groups, p < 0.05 (Figure 3.3 A; Appendix B.2). Interestingly, although not passing the statistically significant threshold (p > 0.05), RMR was numerically ranked as Fawn > Keithley > Little Jacks at all test temperatures (15-24°C) except 12°C, and this ranking became more apparent at warmer test temperatures.  As a result, Q10 values for RMR were: 2.0 in K×LJ hybrid, 2.2 in Little Jacks, 2.3 in Keithley and 2.4 in Fawn. All populations either increased or maintained MMR between 12°C and 21°C (Figure 3.3 A). Notably, only Little Jacks and K×LJ hybrid maintained MMR to temperatures beyond 21°C. All three wild populations maintained relatively constant AAS between 12 and 21°C, whereas K×LJ hybrid significantly increased AAS (p < 0.05) across this temperature range (Figure 3.3 B). Little Jacks was the only population that maintained AAS at 24°C (p > 0.05), whereas Fawn, Keithley and K×LJ hybrid significantly decreased AAS at 24°C. The numeric value for peak AAS occurred at 15.0-18.0°C for Little Jacks and Fawn (Table 3.4), and 21°C for Keithley and K×LJ, even though it is unwise to assign a discrete thermal optimum for AAS that has such a broad optimum thermal window (Appendix B.2), especially in the case 80  of Little Jacks.  Indeed, the optimum thermal window, which I define here as the test temperature range over which a population could maintain a nominal 90% of the peak value for AAS (Figure 3.3 D), was the entire 12-24°C test temperature range, and likely beyond, for Little Jacks.  In contrast, optimum thermal window was narrower in the other two populations (12-21.4°C for Fawn and 12.7-21.6°C for Keithley). The K×LJ hybrid had the narrowest optimum window (19.2-22.5°C). Thermal performance was also assessed by calculating the factorial aerobic scope (FAS = MMR/RMR; Figure 3.3 C), which is an index of the capacity to increase metabolic rate above RMR. All populations including the hybrid could triple the RMR at 12°C, 15°C and 18°C, and at least double RMR at 21°C. However, only Little Jacks and K×LJ hybrid had the capacity to double RMR at 24°C. 3.3.4 Genetic diversity After thermal phenotyping, genome-wide SNP variations were detected by individual genotyping with RAD sequencing in three natural populations and one hybrid strain of redband trout (n=95 fish).  Individuals were randomly assigned into two RAD sequencing libraries. The number of raw sequence reads for the two libraries were 132,766,888 (47 fish) and 135,934,577 (48 fish). After quality filtering, average retained read for each individual was 2,660,029 (range = 1,016,601 – 4,560,808). After all the quality control filters were applied, a set of 5,903 SNPs were retained for subsequent analyses. After correcting for multiple comparisons using BY-FDR, 24 of 23,612 locus/population combinations deviated from H-W equilibrium, which included 18 heterozygote deficits and 6 heterozygote excesses. There were no consistent deviations across loci or populations that violated assumptions of 81  H-W equilibrium. Equipped with the phenotype and genotypes, I conducted population genetics and GWAS for thermal tolerance.  A comparison of genetic diversity among populations suggested that mean HE was the highest in Little Jacks and lowest in Fawn (Table 3.5). Pairwise FST calculated from both putatively neutral loci and outlier loci (see below) suggested the least genetic differentiation between the two montane populations (Keithley and Fawn), while Little Jacks (desert) was strongly differentiated from both montane collections (Table 3.6). Also, Little Jacks (desert) was less differentiated from the Keithley Creek (cool montane) than the Fawn Creek (cold montane). The hybrid strain has similar genetic distance to its two parental populations. 3.3.5 Loci under positive adaptive selection Redband trout populations from different thermal regimes have showed significant differentiations in both phenotypes (thermal tolerance) and genotypes (pairwise FST). Thus, local thermal adaptation could have occurred and be responsible for intraspecific differences at adaptive loci as opposed to differences at neutral loci that occur due to random genetic drift. To identify adaptive loci under divergent selection, two commonly used outlier detection programs based on FST (Lositan and BayeScan) were used in the present study. In total, 21 potential outlier loci were identified (Table 3.7) with 20 outlier loci having an FST value >0.5 and > 99.5% quantiles of neutral expectations. BayeScan is considered as a more conservative method because of a lower Type I error for outlier detection (Narum and Hess, 2011). In the present study, BayeScan only identified three loci that were potential candidates under divergent selection. Two loci were identified as outliers by both tests (Figure 3.4) and thus can be considered as strong candidate markers under selection. Although these outlier loci may not be directly associated with thermal adaptation, they are 82  strong candidates considering the intraspecific variations in habitat temperature and thermal tolerance, especially with the strong association between outlier loci genotypes and CTMAX (Figure 3.5). However, there was no association between the genotypes of outlier loci and AAS (Figure 3.6). To evaluate the role of neutral and adaptive loci in shaping population structure, I applied PCA analyses using genotypes for each set of loci as input variables. A geographical pattern of population structure was observed in both analyses (Figure 3.7), suggesting the events of random drift and possible natural selection in redband trout populations. In the PCA analysis using neutral loci, additional variance in the K×LJ hybrid was shown by PC2 (Figure 3.7 A), which might be caused by the variance at family level. In the outlier loci PCA analysis, PC1 (50.8% of the total variation) was strong enough to separate desert and montane populations (Figure 3.7 B). However, PC1 did not separate the two montane populations, as indicated by the overlap for Fawn and Keithley. Adding PC2 for the outlier loci provided little additional separation among populations (6.6% of the total variation), suggesting a similar direction of selection in Fawn and Keithley. Among all the outlier loci, only the 11282_25 locus showed the same allele frequency between Keithley and Little Jacks (Table 3.7). It was also noted that the desert population had a larger individual variance compared to that of the montane population, which is in accordance with the HE result in Table 3.5.  3.3.6 Association analysis GWAS was also conducted to screen loci that were significantly associated with CTMAX. In total, 57 observations of CTMAX were included for analysis. GWAS was not conducted for Topt and Tpej because the broad plateau and large variance of AAS.  83  GWAS identified 12 loci with significant associations with CTMAX after correction for multiple comparison by BY-FDR (Table 3.8, Figure 3.8). None of these 12 loci were identified as candidates from the outlier tests. Of these 12 loci, three were on O. mykiss chromosome 5, one on chromosome 19, one on chromosome 13 and the rest were unassigned to chromosomes. Because of the correlation between CTMAX and body mass (R2=0.396), body mass was considered as a cofactor in GWAS model selection.  3.3.7 Annotation of candidate loci Alignment of candidate loci to the O. mykiss reference genome identified candidate genes for thermal adaptation (Table 3.9).  For the two markers that has been identified by both Lositan and BayeScan, Gene tdgf1 was identified within 5 kb of the marker 25279_58, but no genes were found within 15 kb of the marker 46551_55. Genes in the flanking regions (<15 kb) of other outlier loci included two potassium channels (kcnc1, srkc), two calcium calmodulin-dependent protein kinases (camk1, camkk2), lactate dehydrogenase (ldh-b) and a 40 kDa heat shock protein (dnajb6). 3.4 Discussion The genetic architecture underlying thermal performance and how it differs among conspecific populations is essential to understand thermal adaptation. However, identifying the underlying genetic mechanism is not a simple task. First, thermal performance is a polygenic trait, controlled by multiple interacting genes (Jackson et al., 1998; Danzmann et al., 1999; Perry et al., 2001), each of which quantitatively contributes to the overall phenotype. Furthermore, selection on genes during thermal adaptation depends on the gene effect to phenotype, standing genetic variation and the interaction with environmental factors other than temperature. In this study, I found intraspecific differences in CTMAX and aerobic 84  scope among redband trout populations from desert and montane climates, where peak summer temperatures differ significantly. Also, I identified putative SNP markers that are potentially under positive thermal selection among these redband trout populations. Lastly, I discovered SNP markers for thermal performance using GWAS. The results in this chapter, while provided novel insight into the genetic basis of thermal tolerance in locally adapted redband trout populations, may also have implications for other locally adapted populations from O. mykiss and for other fish species. 3.4.1 Intra-specific differences in redband trout Critical thermal maximum (CTMAX) Temperature tolerance is strongly affected by genetics and environment. Thus, the comparison of genetically distinct populations often uses an identical rearing condition. All fish in this study were reared in a common garden environment at 15°C since the fry stage to reduce thermal acclimation effect, which however leaves the thermal plasticity not evaluated. Redband trout in this study were sampled from three natural streams, which are separated by geographical distance as well as physical barriers. While Keithley and Fawn are typical montane populations, Fawn Creek has a colder climate because of the higher elevation (1,596 m versus 1,370 m) (Narum et al., 2010) and greater riparian shade. From the temperature logs in present study, peak water temperature in summer differed by 6°C among populations with the highest record in Little Jacks Creek and lowest in Fawn Creek.  The mean CTMAX values measured here for the redband trout populations from Idaho creeks (28.4-29.8°C) are comparable with those from Oregon creeks (29.0-29.7°C) (Rodnick et al., 2004). As expected, redband trout inhabiting warmer habitat in summer had significantly higher CTMAX than those from colder habitat. On average, CTMAX of the desert 85  redband trout (Little Jacks) was 1.0°C higher than that of the montane redband trout (Keithley and Fawn together). A high CTMAX may be important for temporary survival when redband trout are feeding in warm water surface or shallow portion of the creek at peak summer temperature. In natural creeks of southern Idaho, USA, redband trout were observed to be able to tolerate 26°C for up to 4.4 hours (Zoellick, 1999). A greater survival time at high temperatures means a better chance for the creek to cool down diurnally or for a fish to find cooler refugia (e.g. deep pools, cooler springs and upwelling groundwater). Although a difference of 1.0°C in CTMAX seems quantitatively small, it translates to a 10 min difference in survival time (giving a heating rate of 0.1°C min−1), which would be ample for short foraging activities or movement through shallow pools. Little Jacks had an average CTMAX of 29.8°C, which would represent a considerable safety margin relative to the highest water temperature recorded in 2014 (26.0°C). Although all populations had CTMAX values higher than the maximum habitat temperatures logged in this study, an exceptionally hot summer can easily boost water temperature close to or even over CTMAX in the creek, as seen in the historic temperatures for Little Jacks and other creeks (Zoellick, 2004; Rodnick et al., 2004). Fish would either have to acclimate to warmer water as summer progresses, or avoid these areas by seeking refugia in cooler reaches of the creek.   However, neither montane nor desert redband trout appear to be able to survive temperatures > 30°C, even with thermal acclimation (Cassinelli and Moffitt, 2010). Desert populations may have undergone temperature caused mortality/selection, because upper incipient lethal temperature (UILT) of redband trout is at temperatures below CTMAX (25.8-27.1°C when acclimated at 15°C; 26.4-27.7°C when acclimated at 23°C) (Sonski, 1982; Sonski, 1983). Compared to the thermally selected 86  Australian PFRC rainbow trout (Chapter 2), both Little Jacks and Keithley had higher CTMAX, suggesting a greater importance of CTMAX in natural streams than an artificially controlled environment. Previous studies on the correlation between CTMAX and body mass have provided mixed conclusions. While some studies failed to find a significant correlation between fish size and CTMAX  (Prodocimo and Freire, 2001; Ospina and Mora, 2004; Molony et al., 2004; Galbreath et al., 2006; Recsetar et al., 2012), other studies did (Galbreath et al., 2006; Recsetar et al., 2012; Underwood et al., 2012; Zhang and Kieffer, 2014). For those studies that did find significant correlations, both negative and positive correlations have been observed. My previous study on sockeye salmon (Oncorhynchus nerka) fry suggested that body size was positively correlated with CTMAX at early fry stage (Appendix A.1). In redband trout from Bridge Creek (Oregon), CTMAX was essentially the same between two size classes with 35-fold difference in body mass (Rodnick et al., 2004), which is much greater than the 7-fold difference in the present study. Despite that, I used ANCOVA to account for the effect of body mass on CTMAX and produced same statistical conclusion as ANOVA. Nevertheless, body mass may be a factor affecting CTMAX and needs to be examined in the future. Aerobic scope Thermal performance is also under natural selection (Eliason et al., 2011). Most thermal performance (e.g., growth and swimming) is purportedly shaped and limited by oxygen capacity according to the oxygen- and capacity-limited thermal tolerance (OCLTT) hypothesis (Pörtner and Knust, 2007; Pörtner and Farrell, 2008). In natural environment, temperature of the most occurrence for redband trout was between 13-17°C (Gamperl et al., 87  2002; Dauwalter et al., 2015), a thermal range that matches the summer temperatures in Keithley Creek and Fawn Creek, but not Little Jacks Creek (Table 3.1), which has an average minimum summer daily temperature of 16.7°C. Since redband trout from some warm environments have been observed actively feeding at temperatures between 26.2-28.3°C (Behnke, 1979; Sonski, 1983; Zoellick, 1999), and assuming that an Fry Curve provides insight into the thermal capabilities of a fish in a seasonally changing environment where summer temperatures approach their upper thermal limit, I predicted that Little Jacks would have better thermal performance in AAS. Indeed, Little Jacks has a wider optimum thermal window that allows 90% of peak aerobic performance up to a temperature that approaches the UILT (25.8-27.1°C) of other redband trout populations (Sonski, 1982; Sonski, 1983).  Therefore, together with the significantly higher CTMAX, I conclude that Little Jacks redband trout are better adapted to perform in warm temperatures than the montane conspecifics. Unlike the PFRC rainbow trout in Chapter 2, it was not possible to quantify a precise Topt for redband trout from present data because peak AAS had a wide plateau. Likewise, extrapolation of the Fry Curve to temperatures above 24°C suggests that Little Jacks trout would have a Tcrit that is higher than CTMAX (Appendix B.3), which is unlikely. Instead, I predict AAS in Little Jacks will abruptly decrease at temperatures above 24oC, which is exactly what I observed for the two montane populations, but between 21 and 24°C. In fact, I abandoned RMR and MMR measurements at the test temperature of 27°C because a > 3 h exposure to 27°C caused some mortality in the respirometer during preliminary tests.  Also, fish had a highly variable RMR, suggesting that the fish may have become glycolytic.  It seems that a broad optimum thermal window of Little Jacks comes at a cost of peak absolute AAS. Little Jacks had lower peak AAS compared to Keithley and its hybrid with 88  Keithley, but not Fawn Creek.  A previous study in redband trout from Oregon had similar results, a population from warm climate (Bridge Creek, maximum water temperature over 24°C) had lower AAS than a population from cool climate (Little Blitzen Creek, maximum water temperature below 18°C) at 12-14°C (Gamperl et al., 2002). However, at 24°C, fish from warm climate had higher AAS (and higher critical swimming speed, or Ucrit) than the fish from cool climate, which is a similar finding in present study. Thus, warm adaptation likely involves a trade-off in peak performances at optimum thermal ranges for the performance at critical temperatures. MMR was measured in this study by putting the fish in an O2 debt situation (Norin and Clark, 2016) rather than using a prolonged swimming test (e.g., Lee et al., 2003b), even though a fish also uses glycolytic metabolism as it reaches Ucrit.  MMR measurements yield similar values using either method (Norin and Clark, 2016).  Nevertheless, in addition to temperature, I cannot exclude that aerobic scope of redband trout may be shaped by hypoxia as well (McBryan et al., 2013), particularly with the intermittent occurrence of reduced oxygenation levels in streams inhabited by redband trout in Idaho (Vinson and Levesque, 1994) and declining relationship of O2 content with increasing water temperatures. The interaction between temperature and hypoxia in adaptation warrants empirical studies. I predicted the K×LJ hybrid would have intermediate performance in both CTMAX and aerobic scope compared to the parental populations. While this was true for CTMAX, aerobic scope demonstrated hybrid vigor (heterosis effect). Heterosis is a common phenomenon and has been observed for many other traits in salmonids, such as saltwater tolerance (McGeer et al., 1991; Bryden et al., 2004) and disease resistance (Beacham and Evelyn, 1992). CTMAX data suggested the component of additive heritability, which is not surprising given the 89  substantial heritability for upper thermal tolerance (range of h2=0.20 - 0.48)  (Ihssen, 1986; Meffe et al., 1995; Doyle et al., 2011). Aerobic scope heterosis is largely due to a lower RMR, or “RMR depression”, suggesting a decreased basal metabolic demand in the hybrid. The underlying mechanism is not clear, but may be related to the enzyme kinetics, mitochondria function and endocrine regulations. Because the K×LJ hybrid sample had a similar genetic diversity to both Keithley and Little Jacks, the phenotypic variance (and potentially phenotypic plasticity) of the hybrids is expected to be larger than the parental populations. Indeed, for instance, hybrid fish had a larger standard deviation (s.d. = 0.35) in CTMAX compared with Little Jacks and Keithley (both had s.d. = 0.29).  3.4.2 Population genetics Lower heterozygosity is expected for a population with smaller effective population size that is prone to inbreeding or random genetic drift. Previously, estimated effective population size of redband trout was higher in Little Jacks (133) than that in Keithley (49.5) and Fawn (26.1) using 65 neutral markers (Narum et al., 2010). Little Jacks Creek also has a higher redband trout density than Keithley Creek and Fawn Creek (Fesenmyer and Dauwalter, 2014), as well as a considerably larger drainage area (259.0 km2 for Little Jacks, 34.5 km2 for Keithley creek and 56.7 km2 for Fawn creek) (USGS and USFS).  In the present study, I found Little Jacks population that lives in a harsh desert environment had higher genetic diversity than the colder montane populations. Previously, inconsistent results were obtained for the genetic diversity of these three redband trout populations (Narum et al., 2010; Kozfkay et al., 2011). This could have arose because the sample size of either genetic markers or studied animals was too small to ensure accurate estimations of diversity.  90  By intentionally choosing populations from different thermal regimes and geographical locations, I expected greater similarity between the two montane populations given their shorter spatial distance, closer habitat thermal regimes, as well as the similar phenotypic data reported above. Genetic distance (FST) among the three redband trout populations in this study is in agreement with a previous study on the same populations (Narum et al., 2010). As predicted, I found that FST derived from both neutral and adaptive genetic variations was higher between the desert (Little Jacks) versus montane populations (Keithley and Fawn), and was lower between montane populations (Keithley versus Fawn). Compared to Keithley, Fawn was genetically more divergent from Little Jacks. Genetic distance is affected by the combined effects of genetic drift, gene flow, mutation and selection which are often influenced by geographical and ecological factors (Crispo et al., 2006). Geographically, Little Jacks is distant from Keithley (straight line distance of 270 km) and Fawn (183 km), while Keithley and Fawn are closer (68 km). Ecologically, while both Fawn Creek and Keithley Creek are both montane climates, Fawn Creek has a colder summer and winter environment than Keithley Creek. Signatures of putative positive adaptive selection in genes can be detected using population genetics approaches by identifying the “outlier loci”, where the nucleotide sequence variation deviates from the expectation of the neutral evolution model, showing extreme allele frequency divergences (Nosil et al., 2009). Recently developed genomic sequencing techniques greatly increases the number of genetic markers and provides high genomic coverage resolution for more precise mapping (Hohenlohe et al., 2010; Hohenlohe et al., 2011). From the total of 5903 SNP markers across the genome, the current study identified 21 loci as outliers. PCA clustering using outlier loci suggested an adaptive 91  divergence between montane and desert populations, although other demographic events could also generate similar patterns, such as bottleneck and founder effect (Costello et al., 2003; Hartl and Clark, 2007; Hofer et al., 2009). With the further discovery of strong association between outlier loci and CTMAX, the adaptive divergence can be specifically inferred as a result of thermal adaptation. However, the extent of the pleiotropic effects (indirect selection) is not known because genes are often involved in more than one phenotypes (Barrett et al., 2009). By searching the flanking regions of the outlier loci (within 15 kb ranges), I found more than one gene for most outliers. Here I suggest that genes beside the outlier loci may be associated with variations of adaptive traits in redband trout, especially the ability to tolerate critical temperatures, which is a hypothesis that needs further functional verification. A brief summary of selected gene functions is provided in followed paragraphs. Gene tdgf1 (teratocarcinoma-derived growth factor-like, also known as Cripto-1) is within 5 kb of a strong candidate outlier locus (25279_58) and encodes an epidermal growth factor-related protein that contains a cripto, FRL-1, and cryptic domain (Klauzinska et al., 2014). In mammalian embryo cells, it has been found that Cripto-1 contains a hypoxia-responsive elements within the promoter region and is directly regulated by hypoxia-inducible factor-1 (HIF-1) (Akukwe et al., 2007). The expression of Cripto-1 is essential for the normal differentiation of cardiac myocytes when mouse embryonic stem cells are exposed to hypoxia (Bianco et al., 2009). Furthermore, the expression of Cripto-1 has been used as a biomarker for cardiac hypoxia. Gene dnajb6 encodes the DnaJ homolog subfamily B member 6 (also known as heat shock protein 40 kD), which contains a highly conserved amino acid structure called “J-92  domain” (Qiu et al., 2006) that interacts with Hsp70 to suppress the aggregation of misfolded proteins in the cell (Fan et al., 2003; Hageman et al., 2010). Gene ldh-b (l-lactate dehydrogenase B-A chain like isoform x2) encodes a subunit of the lactate dehydrogenase enzyme. In low O2 environment, lactate dehydrogenase plays an important role to re-oxidize NADH to NAD+, which can be used for the oxidation of glyceraldehyde 3-phosphate (G3P). In previous studies with killifish populations, genotypes of ldh-b gene have been correlated with habitat thermal conditions (Schulte et al., 2000).  Gene camk1 (Calcium/calmodulin-dependent protein kinase) and camkk2 (CAMK kinase) encode important kinases in the CAMKK–CAMKI signaling pathway, which controls axonal elongation and dendritic branching (Davare et al., 2004; Neal et al., 2010). CAMKK–CAMKI signaling pathway compromises CAMKK and its downstream substrates, CAMK1 and CAMK4.  Potassium channels are ubiquitous and are also a complex class of voltage-gated ion channels that play important roles in cellular functions. In this study, two potassium channels from the Shaker and Shaw gene family have been identified as outliers. Gene kcnc1 encodes a Shaw related potassium voltage-gated channel subfamily c member 1 isoform x2, Kv3.1 (Ried et al., 1993). The Kv3.1 potassium channels have shown roles in neuronal signaling in trout (Henne and Jeserich, 2004) and mouse (Wang et al., 1998). Gene srkc (shaker-related potassium channel tsha2-like) is an ortholog to the kcna6 (potassium voltage-gated channel, shaker-related, subfamily, member 6, Kv1.6), which encodes a voltage gated potassium channel that belongs to the delayed rectifier class. In rainbow trout, SRKC has been found in glial cells (Nguyen and Jeserich, 1998).  93  3.4.3 Genome-wide association study The evolution of adaptive traits in the wild has a complex pattern because it is strongly affected by genetic × environment (G×E) and genetic × genetic (G×G) interactions (Danzmann et al., 1999; Kawecki and Ebert, 2004). Furthermore, this adaptation pattern is also very likely being population specific depending on the standing genetic variation (Barrett and Schluter, 2008) and selection pressure. Therefore, besides detecting outlier loci in thermally adapted populations, I also applied a genome-wide association study to identify markers that are quantitatively associated with CTMAX, which may or may not have undergone strong directional thermal selection. Previous studies in rainbow trout using microsatellites have identified QTLs that are related to upper thermal tolerance (i.e. time to loss equilibrium at 25.7°C) in rainbow trout (Jackson et al., 1998; Danzmann et al., 1999; Perry et al., 2001) and a total of six QTLs from five linkage groups were significantly associated with UTT (see a brief summary in Araneda et al., 2008). Knowledge of genes from the QTLs studies is limited given the large genomic regions between markers, which also made direct comparison between GWAS and QTL studies difficult.  Some studies have used dense SNPs markers from GWAS to validate QTLs (Palti et al., 2015) or used QTLs result to control the false positive SNP markers (Sonah et al., 2015). Results in my study suggested that temperature tolerance is a polygenic trait with large number of genes involved and each contributes small effect to the overall phenotype. The power to detect SNP markers in GWAS depends on many factors including marker density, sample size, population structure and genetic variation in studied population. The biggest limitation for my study is the relatively small sample size of fish from each population (Klein, 2007; Spencer et al., 2009; Bush and Moore, 2012). Despite this 94  limitation, thousands of SNPs generated from RAD sequencing provided an opportunity to discover potential candidate loci that may correlate with thermal tolerance. Beside sample size, an ideal population used for association mapping would have low population structure or relatedness (FST) and higher minor allele frequency (Rincent et al., 2014). However, population structure exists in most studies, especially with natural populations, but can be corrected using PCA analysis or mixed linear models. My study addressed this concern using the mixed linear model and accounted for both population structure and cryptic familial relatedness among individuals (Balding, 2006; Yu et al., 2006). Putting together the results from outlier loci and GWAS, chromosome number five (OMY5), OMY19 and OMY1 are candidate chromosomes that may contain QTLs for upper thermal tolerance (UTT). First, in this study, OMY5 contained the most candidate loci (five) in both GWAS and outlier selection. OMY5 is the fifth largest pair among the 30 chromosome pairs in rainbow trout (Phillips et al., 2006) and contains markers for spawning date (O’Malley et al., 2003) and migration related traits (Hecht et al., 2012). However, no QTLs for UTT from previous microsatellite study were located on OMY5 (Araneda et al., 2008). Second, OMY19 is also a strong candidate because it contains a significant QTL Ssa14DU (LG-14) for UTT (Jackson et al., 1998), which has also been found to be associated with thermal tolerance in Arctic charr (Somorjai et al., 2003). Moreover, OMY19 contains a homologue QTL that has been identified to be associated with UTT in chinook salmon (Everett and Seeb, 2014). In the present study, RAD tag 23616_10 on OMY19 emerged as significantly associated with CTMAX. Lastly, OMY1, which has a UTT QTL One14ASC (LG 9), contains an outlier locus in this study. For all the markers that identified 95  in this study, the distance to the existing UTT QTLs on the same chromosome were not measured. On OMY5, two genes are within 10 kb of the marker that was significantly associated with CTMAX. One gene is myoc that encodes the protein myocilin, which induces the formation of stress fibers in human trabecular meshwork through the wnt signaling pathway (Kwon et al., 2009). The other gene is mettl13 that encodes the enzyme methyltransferase like 13, whose specific function is unclear. But methyltransferases are generally involved in epigenetic modifications on DNA and histones, which have been documented in response to environmental temperatures in fish (Varriale and Bernardi, 2006) and may also affect gene expression (Tate and Bird, 1993). 3.4.4 Conclusion Taking advantage of the genome-wide SNP markers from high throughput sequencing, my results add knowledge to the genetic basis of thermal adaptation. RAD-sequencing identified outlier loci among redband populations from different climates. Moreover, quantitative trait loci for CTMAX and aerobic scope were discovered by GWAS. By locating the significant markers on rainbow trout genome, I pinpointed nearby genes that may be functionally important and contribute to the survival and reproductive fitness of an organism. These results could be used as putative markers to trace the genetic changes in wild populations for conservation-management purposes. To this end, further studies should (1) use more populations from diverse habitats, and (2) design mapping families to verify the results in this study and establish a more comprehensive understanding of the genetic basis of thermal adaptation.  96   Table 3.1 Summary of summer water temperatures logged for Little Jacks Creek, Keithley Creek and Fawn Creek in Idaho, USA. Creek Year Date Summer temperature (°C) Mean Daily min Daily max  Peak Little Jacks 2014 10 Jul - 10 Sept 18.5±2.1a 16.7±1.9a 21.3±2.6a  26.0 Keithley 2014 10 Jul - 10 Sept 15.7±1.9b 14.0±1.8b 17.5±2.0b  20.7 Fawn 2015 10 Jul - 10 Sept 11.4±2.2c 9.8±1.4c 13.9±1.6c  16.9 Note: Numbers are present as mean±s.d. All null hypothesis were tested by one way ANOVA and Tukey's post hoc test with significance level at α = 0.05.      Table 3.2 Total time of stream temperatures above the test temperatures in AAS measurements for three redband trout O. mykiss gairdneri habitats from July 10 to September 10. Temperature (°C) Little Jacks Creek (hour) Keithley Creek (hour) Fawn Creek (hour) 15 1335 973 99.3 18 719.3 236 0 21 160 0 0 24 12 0 0      Table 3.3 Body size of redband trout in CTMAX and AAS measurements. Population Body size of fish in CTMAX   Body size of fish in AAS  n Mass (g) Length (cm)  n Mass(g) Little Jacks (LJ) 14 3.4±0.4a 6.0±0.2a  6 3.4±0.3a K×LJ 14 2.2±0.2a 5.2±0.2a  6 3.0±0.3a Keithley (K) 18 1.9±0.1a 5.1±0.1a  6 2.4±0.2a Fawn (F) 17 0.7±0.1b 4.2±0.1b  6 1.1±0.1b Note: Numbers are present as mean±s.e.m. Body size comparisons were tested by ANOVA on ranks with Dunn’s post hoc analysis (p < 0.05).    97   Table 3.4 Topt amd Tpej for AAS in redband trout populations.  Maximum AAS  90% of maximum AAS Populations Topt  (°C) AAS  (mg O2 g−0.88 h−1)  Lower Tpej (°C) Upper Tpej (°C) Little Jacks (LJ) 15.0-18.0 0.39±0.03  < 12.0 > 24.0 K×LJ 21.0 0.63±0.05  19.2 22.5 Keithley (K) 21.0 0.52±0.08  15.6 21.6 Fawn (F) 15.0-18.0 0.42±0.03  < 12.0 21.4 Note: Topt were based on the maximum measured AAS; AAS are present as mean±s.e.m. Tpej were calculated from the linear connection between two adjacent data points.      Table 3.5 Average expected (HE) heterozygosity of redband trout populations. Different superscripted letters indicate significant differences. Population n HE Little Jacks (LJ) 20 0.261a K×LJ 22 0.251b Keithley (K) 25 0.219c Fawn (F) 21 0.196d      Table 3.6 Pairwise FST distance of redband trout populations. FST above diagonal blank cells were calculated based on candidate outlier loci (21 SNPs), while FST below diagonal line were calculated based on putatively neutral loci (5578 SNPs).   Little Jacks (LJ) K×LJ Keithley (K) Fawn (F) Little Jacks (LJ)  0.2519 0.6888 0.6769 K×LJ 0.1056  0.3363 0.3438 Keithley (K) 0.1512 0.1032  0.2010 Fawn (F) 0.1872 0.1549 0.0827      98  Table 3.7 Summary of outlier loci and the nearby genes (within 15 kb). Loci were ordered according to the reference allele frequency in Little Jacks population. Tag   Allele  Ref allele frequency  Nearby genes in reference genome ID Chr  Ref Alt  LJ hybrid K F  0-5 kb 5-10 kb 10-15 kb 25279_58* 5  A C  0.09 0.73 0.91 0.98  tdgf1   41368_19 UN  A G  0.09 0.34 0.85 0.95    rtpd 37609_72 7  A T  0.14 0.59 0.83 1.00  unidentified_1 cep63  11705_67 6  A T  0.19 0.79 0.96 1.00  unidentified_2   54211_35 15  A T  0.20 0.58 0.94 0.95  ephb1 aga  ptedp4 73456_35 2  A G  0.21 0.62 0.98 0.98  slc35b4 golt1b gp1 Ldh-b 46551_55* UN  A G  0.28 0.50 1.00 1.00     41738_48 27  A G  0.32 0.74 1.00 0.95  htr1b unidentified_3 fa2h  rasa2 unidentified_4 43135_9 14  G C  0.37 0.45  1.00 1.00  sgsm2   54575_55 UN  A G  0.43 0.83  1.00 1.00     84033_50 10  A G  0.53 0.09  0.00 0.00  kirrel unidentified_5 ypel1  pom121  23991_50 7  T C  0.53 0.38 0.00 0.00  kcnc1  kiaa1024l 18245_48 8  A G  0.53 0.25 0.00 0.00  dnajb6 trpa1  34587_35 5  A C  0.53 0.24 0.00 0.00  unidentified_6 camkk2 snrpd3   46815_9 1  A G  0.61  0.30  0.00  0.02   camk1  dhtkd1 18316_47 10  A T  0.62  0.15  0.00  0.00   wdr11 lyst  36015_64 UN  A G  0.64  0.13  0.00  0.00   grd1   33809_35 14  A T  0.66  0.38  0.02  0.00   st3gal1  unidentified_7 apmap 3723_61 7  A C  0.68  0.43  0.02  0.00   cadps   11282_25 14  A G  1.00 1.00 1.00 0.48  rad21 tc3a  99  Tag   Allele  Ref allele frequency  Nearby genes in reference genome ID Chr  Ref Alt  LJ hybrid K F  0-5 kb 5-10 kb 10-15 kb srkc trappc9 33984_22** UN  G C  1.00 0.72 0.17 0.39  greb1l  abhd3 Note: Loci without asterisks were identified by Lositan; *, loci identified by both Lositan and BayeScan; **, loci identified by BayeScan only; UN, unknown chromosome.     100  Table 3.8 Markers significantly associated with critical thermal maximum in GWAS analysis and the nearby genes within 15 kb range in reference genome. ID Ref Alt Maf Chr P value  0-5 kb 5-10 kb 10-15 kb 17679_36 T C 0.40  UN 0.00088      5870_32 G C 0.16  5 0.00149   myoc mettl13  23616_10 A G 0.12  19 0.00175      47691_73 T C 0.23  UN 0.00194      75319_60 A T 0.06  5 0.00204      79819_54 A G 0.36  4 0.00270   fnbp1l  bcar3 16091_71 A T 0.14  5 0.00394      39660_68 A T 0.24  UN 0.00428   umodl unidentified_8 myo1c 28241_36 T C 0.13  UN 0.00454     ppp1r12a 85563_46 T C 0.11  13 0.00456    hgsnat  82805_35 A T 0.17 UN 0.00460   vlcs hdc 51733_45 A T 0.19  UN 0.00537   unidentified_9 tal1  Note: Maf, global minor allele frequency; UN, unknown chromosome number.    101  Table 3.9 Annotation for genes within the 15 kb flanking region of significant markers from outlier tests and GWAS. Gene in reference Mean-similarity Gene symbol Protein GSONMT00067091001 93.2 abhd3 phospholipase abhd3 GSONMT00065867001 97.7 aga alpha- -mannosyl-glycoprotein 4-beta-n-acetylglucosaminylt GSONMT00064603001 92.6 apmap adipocyte plasma membrane-associated protein GSONMT00025767001 86.3 bcar3 breast cancer anti-estrogen resistance protein 3 isoform GSONMT00056967001 93.4 cadps calcium-dependent secretion activator 1 isoform x2 GSONMT00077302001 88 camk1 calcium calmodulin-dependent protein kinase type 1d GSONMT00034290001 90.6 camkk2 calcium calmodulin-dependent protein kinase kinase 2-like isoform  GSONMT00034289001 76.3 camkk2 calcium calmodulin-dependent protein kinase kinase 2 isoform x1 GSONMT00043985001 78.4 cep63 centrosomal protein of 63 kda GSONMT00077300001 91.9 dhtkd1 probable 2-oxoglutarate dehydrogenase e1 component mitochondrial GSONMT00073694001 91.7 dnajb6 dnaj homolog subfamily b member 6-like GSONMT00018503001 98.1 ephb1 ephrin type-b receptor 1-b isoform x1 GSONMT00063989001 90.55 fa2h fatty acid 2-hydroxylase GSONMT00025764001 93.6 fnbp1l formin-binding protein 1-like isoform x1 GSONMT00057972001 96.7 golt1b vesicle transport protein got1b GSONMT00057969001 80.2 gp1 vegetative cell wall protein gp1-like isoform x2 GSONMT00029740001 92.5 grd1 glutamate receptor delta-1-like isoform x2 GSONMT00067092001 86 greb1l greb1-like protein isoform x1 GSONMT00030939001 69.2 hdc histidine decarboxylase GSONMT00063063001 67.7 hgsnat heparan-alpha-glucosaminide n-acetyltransferase isoform x1 GSONMT00012511001 69.1 htr1b  5-hydroxytryptamine receptor 1d-like GSONMT00072062001 91 kcnc1 potassium voltage-gated channel subfamily c member 1 isoform x2 GSONMT00072063001 81.3 kcnc1 potassium voltage-gated channel subfamily c member 1 isoform x1 GSONMT00010840001 74 kiaa1024l upf0258 protein kiaa1024-like GSONMT00049185001 81.2 kirrel kin of irre-like protein 1 isoform x1 GSONMT00057974001 94.8 Ldh-b l-lactate dehydrogenase B-A chain like isoform x2 GSONMT00031485001 80.2 lyst lysosomal-trafficking regulator GSONMT00011640001 83.8 mettl13 methyltransferase-like protein 13 GSONMT00023741001 96.4 myo1c unconventional myosin-ic isoform x2 GSONMT00011643001 83.9 myoc myocilin 102  Gene in reference Mean-similarity Gene symbol Protein GSONMT00063951001 79.5 tdgf1 teratocarcinoma-derived growth factor-like GSONMT00035065001 69.2 pom121 nuclear envelope pore membrane protein pom 121 GSONMT00004697001 82.6 ppp1r12a protein phosphatase 1 regulatory subunit 12a-like isoform x3 GSONMT00067370001 60.6 ptedp4 piggybac transposable element-derived protein 4-like GSONMT00030393001 87.2 rad21 double-strand-break repair protein rad21 homolog GSONMT00028469001 88.5 rasa2 ras gtpase-activating protein 2 GSONMT00078073001 90.4 rmdn2 regulator of microtubule dynamics protein 2 GSONMT00037129001 98.7 rtpd receptor-type tyrosine-protein phosphatase delta-like isoform x3 GSONMT00001256001 82.5 sgsm2 small g protein signaling modulator 2 GSONMT00057971001 94 slc35b4 udp-xylose and udp-n-acetylglucosamine transporter GSONMT00034291001 95.7 snrpd3 small nuclear ribonucleoprotein sm d3 GSONMT00077280001 87.6 srkc shaker-related potassium channel tsha2-like GSONMT00023777001 84.6 st3gal1 cmp-n-acetylneuraminate-beta-galactosamide-alpha- -sialyltrans GSONMT00029529001 91.7 tal1 t-cell acute lymphocytic leukemia protein 2 GSONMT00018240001 79 tc3a transposable element tc3 transposase GSONMT00065866001 86.9 tcb1 transposable element tcb1 transposase GSONMT00029485001 74.2 trappc9 trafficking protein particle complex subunit 9 isoform x1 GSONMT00073696001 80.3 trpa1 transient receptor potential cation channel subfamily a member  GSONMT00023065001 78.7 umodl uromodulin-like GSONMT00057711001 N/A unidentified_1 Unidentified GSONMT00080796001 N/A unidentified_2 Unidentified GSONMT00028470001 N/A unidentified_3 Unidentified GSONMT00063990001 N/A unidentified_4 Unidentified GSONMT00075158001 N/A unidentified_5 Unidentified GSONMT00012958001 N/A unidentified_6 Unidentified GSONMT00064601001 N/A unidentified_7 Unidentified GSONMT00047832001 93.2 unidentified_8 Unidentified GSONMT00029530001 60.8 unidentified_9 Unidentified GSONMT00030937001 92.3 vlcs very long-chain acyl- synthetase-like GSONMT00030938001 89.3 vlcs very long-chain acyl- synthetase-like GSONMT00010361001 92.6 wdr11 wd repeat-containing protein 11 103  Gene in reference Mean-similarity Gene symbol Protein GSONMT00035067001 85.05 ypel1 protein yippee-like 1    104    Figure 3.1 Water temperatures of Little Jacks Creek, Keithley Creek and Fawn Creek in the Snake River tributary of southern Idaho.  Shaded areas indicate temperatures between daily minimum and maximum. Data for Little Jacks Creek and Keithley Creek was collected by Columbia River Inter-Tribal Fish Commission (CRITFC) between May 2014 and April 2015. Data for Fawn Creek was collected by CRITFC between Oct. 2015 and Sept. 2016. Dash line represents the temperature of 26°C. The dramatic drop of water temperature in Little Jacks and Keithley Creek in November 2014 was due to an Arctic cold surge.     105     Figure 3.2 CTMAX of redband trout populations. Date are in mean±s.e.m. ANOVA with Turkey's HSD post hoc analysis showed significant (p < 0.05) differences between groups (different lower-case letters). After include body mass as covariate in ANCOVA analysis, the statistical significance of CTMAX between populations persisted (p < 0.05) because body mass is not significantly affecting the differences in CTMAX (p=0.286). 106   Figure 3.3 Effect of temperature on metabolic rate and aerobic scope in redband trout populations. A: Routine metabolic rate (RMR) and maximum metabolic rate (MMR), mean ± s.e.m.; B: absolute aerobic scope (AAS), mean ± s.e.m.; C: factorial aerobic scope (FAS), mean ± s.e.m.; D: Percentage of AAS in relative to the maximum AAS for each tested temperatures in redband trout populations. Dashed line represents 90% of maximum AAS. Colors denotes different population: Little Jacks, ; hybrid: ; Keithley: ; Fawn: ). Statistics between groups and temperatures are detailed in Appendix B.2.    107      Figure 3.4 Outlier loci analyses in redband trout populations using Lositan and BayeScan.    108   Figure 3.5 Association between CTMAX and genotypes of the candidate outlier loci. Red brackets indicate significant differences at the level of α=0.05 in one way ANOVA on ranks with subsequent Dunn's post−hoc test. Numbers above genotypes represent the sample size.     109   Figure 3.6 Association between AAS and genotypes of the candidate outlier loci. For each individual, aerobic scope measured at five different temperatures has been summarized into one factor in SPSS. Larger number represents higher aerobic scope. No significant differences were detected at α=0.05.   110      Figure 3.7 Principal component analyses of genetic differentiation in redband trout using neutral loci (A) and outlier loci (B). Different color represent different populations (Little Jacks: ; Hybrid : ; Keithley: ; Fawn: ).         111   Figure 3.8 Summary of genome-wide association study results for CTMAX. Horizontal red line represents the genome-wide significant threshold after controlling false discovery rate (FDR < 0.05). Markers with unknown chromosome origin were not assigned chromosome numbers.     112  Chapter 4: Cardiac transcriptomic response to acute warming in redband trout populations (Oncorhynchus mykiss gairdneri) Synopsis In the previous chapter, evidence of intraspecific phenotypic variation (CTMAX and aerobic scope) among redband trout populations from desert and montane climates were presented. The underlying genetic basis was also examined using population genetics and genome-wide association studies. In this chapter, I will present the thermal performance of fH,max during acute warming and examine the associations between cardiac function and aerobic scope as well as CTMAX. In addition, cellular response to acute warming at the transcriptomic level is also studied to elucidate the mechanism underlying the capacity limitation of fH,max. Results demonstrate that desert redband trout (Little Jacks) have better cardiac performance than the two montane populations (Keithley and Fawn) by achieving the highest fH,max across all experimental temperatures. After accounting for the effect of body mass on fH,max, differences in fH,max among populations become more evident, especially between Little Jacks and Fawn. Using RNA sequencing, thermally induced gene expression is demonstrated in three redband trout populations after being acutely warmed from the rearing temperature (15°C) to critical temperatures until fH,max becomes arrhythmic (up to 28°C). Acute warming significantly induced the expression of a large number of genes from multiple pathways, including heat shock response, cardiac myocyte function and energy metabolism. Different patterns of gene regulation among populations suggest different intraspecific regulatory 113  strategies in response to acute warming. Also, mRNA abundance is significantly different among populations in genes with important functions. Furthermore, expression-trait association analysis revealed genes that may be related to the thermal induced cardiac arrhythmia, which interrupts the cardiac function and whole organismal thermal tolerance. Overall, results in this chapter provide strong candidate biomarkers for the molecular mechanisms of thermal tolerance and adaptation. 4.1 Introduction Aquatic ectotherms, including most fishes, have an internal body temperature closely matching the ambient temperature (Clausen, 1934) and they behaviorally select water temperatures by exploring between different thermal zones in their habitats (Reynolds and Casterlin, 1979). Nonetheless, fishes routinely encounter changes in water temperature, which can be transient (briefly foraging in a thermocline), acute (diurnal), prolonged (seasonal) or continuous (global warming). While thermal acclimatization and adaptation are possible over longer time periods, transient and acute changes leave little time for the compensatory mechanisms (Sidell et al., 1973; Bouchard and Guderley, 2003). For example, while acute exposure of redband trout to 28.3°C would be lethal, they have been observed briefly feeding in desert creeks at this temperature, possibly because it may be their only foraging option on a hot summer’s day (Behnke, 1979). Therefore, understanding how fish handle transient warming is as important as the adaptations, both of which allow populations to exploit different thermal habitats. In the present study, I characterize the physiological and cellular responses of the heart to acute warming and compare these responses across fish populations from different stream temperatures. 114  According to the theoretical oxygen- and capacity- limited thermal tolerance (OCLTT) hypothesis and empirical evidence, heart is the purported limiting factor in thermal tolerance. In fishes, the primary physiological response during acute warming is to increase cardiac output and gill ventilation so that the elevated O2 demand can be fulfilled. The increase in cardiac output is mostly supported by an increase in heart rate (fH) (Steinhausen et al., 2008; Farrell, 2009; Eliason et al., 2013), which has a Q10 around 2.0 within normal thermal ranges (Overgaard et al., 2004). At temperatures approaching the animal’s upper thermal tolerance limit (CTMAX), heart rate can no longer increase and ultimately can become arrhythmic (Vornanen, 2016; Badr et al., 2016). Thus, the capacity limitation in cardiac function is thought to play a role in setting the upper thermal tolerance of fish.  In addition, under stressful conditions, cardiac performance is supported by a neuro-endocrine response (Wendelaar Bonga, 1997): catecholamines stimulate and protect the cardiac contractility (Shiels and Farrell, 1997; Hanson et al., 2006), which would otherwise decrease at high contraction frequency. Although elevated temperatures increase the isometric force and Ca2+ sensitivity of cardiac myofilaments (Churcott et al., 1994), contractility often decreases due to the increased heart rate at Q10 around two (negative force-frequency relationships) and possibly the intracellular acidosis. Low contractility also decreases arteries blood pressure. Adrenergic stimulation has been shown to be able to increase sarcolemma Ca2+ influx (Shiels et al., 1998), which can improve the contractility over a broad range of temperatures (Shiels and Farrell, 1997). However, the cardiac sensitivity to adrenergic stimulation is lower at elevated temperatures (Shiels et al., 2003b). While the physiological responses of the heart to acute warming are reasonably well characterized in fishes, cardiac cellular responses are not.  For instance, I am unaware of any 115  study that has directly linked cardiac cellular responses with the thermal performance of the heart in salmonid fishes. Most importantly, cellular homeostasis needs to be maintained (CHR) to support the normal functioning of cardiac myocytes.   In addition, the stability and assembly of macromolecules needs to be protected by molecular chaperons through the heat shock response (HSR) (Iwama and Thomas, 1998; Feder and Hofmann, 1999; Iwama et al., 1999). Gene expression is critical for the synthesis of functional molecules and is responsive to environmental stressors. Thus, transcriptome has been studied in fish species from polar stenotherm to tropical eurytherm in various tissues (gill, white muscle, liver, heart) using different methodologies (candidate gene, microarray, RNA sequencing, and proteomics) (Gracey et al., 2004; Vornanen et al., 2005; Gracey, 2007; Hassinen et al., 2008; Castilho et al., 2009; Korajoki and Vornanen, 2012; Jayasundara et al., 2013; Anttila et al., 2014a; Jørgensen et al., 2014; Logan and Buckley, 2015; Tan et al., 2016). Among all the methodologies, RNA sequencing (Lister et al., 2008; Mortazavi et al., 2008; Nagalakshmi et al., 2008) is replacing microarray as a popular approach to study gene expression, due to some obvious advantages such as identifying low abundance transcripts  and transcripts isoforms (Zhao et al., 2014). The literature on transcriptome in response to temperature is accumulating in fishes (see review in Logan and Buckley, 2015). Literature using RNA Sequencing to study cardiac transcriptomic response to temperature changes is, however, limited (Tan et al., 2016). I focused on redband trout that inhabit desert and montane streams to characterize the physiological and cellular cardiac responses to acute warming. Redband trout prefer temperatures between 13-17°C (Gamperl et al., 2002; Dauwalter et al., 2015) and optimum growth occurs around 20°C (Sonski, 1982). Temperatures above 25.8°C are often lethal and CTMAX is 29-30°C (Rodnick et al., 2004; Cassinelli and Moffitt, 2010; Chapter 3). Fish in 116  natural streams frequently experience acute temperature fluctuations from optimal to critical levels. For instance, the diel temperature of redband trout dwelling streams was recorded fluctuating between 14 - 27°C (Rock Creek) and 19 - 30°C (12-mile Creek) in Oregon (Rodnick et al., 2004), and 17.9 - 29.9°C (Big Jacks Creek and Shoofly Creek) in Idaho (Zoellick, 1999). On hot days, streams are also thermally portioned (i.e. some regions are warmer than others) (Matthews and Berg, 1997). Redband trout has demonstrated an exceptional aerobic performance when warmed (Gamperl et al., 2002; Rodnick et al., 2004), but their cardiac performance is unstudied. Furthermore, by comparing the cardiac responses of different populations (hot desert vs cool montane) adaptive intraspecific differences can be explored, because transcriptomic response to temperature in gill tissues have suggested adaptive intraspecific differences in the HSR and metabolic processes between montane and desert populations (Narum and Campbell, 2015). In particular, a desert population significantly upregulated more gill oxidative phosphorylation subunits genes in fluctuating temperatures than the montane population (Garvin et al., 2015). Therefore, in this chapter, I used three fish populations from Idaho, USA: Fawn Creek - a cold montane climate; Keithley Creek - a cool montane climate and Little Jacks Creek - a desert climate. A hybrid between Keithley and Little Jacks (K×LJ) was also assessed.  Each fish was progressively warmed while monitoring the response of maximum heart rate (fH, max). Heart tissue was sampled at 15°C, 20°C and the temperature when the heart rate first became arrhythmic (TAR). I addressed three questions: (1) how does acute warming affect cardiac gene expression in redband trout; (2) how does the gene expression pattern differ among populations from different natural habitats; and (3) what cardiac gene expression is associated with TAR. 117  4.2 Materials and methods All experimental procedures were approved by the University of British Columbia Committee on Animal Care in accordance with the Canadian Council on Animal Care (A10-0335) and the University of Idaho (IACUC protocol 2013-80). 4.2.1 Fish culture and rearing condition All redband trout were collected and reared in a common garden environment from fry stage as detailed in Chapter 3 (Section 3.2.1).  4.2.2 Maximum heart rate Fish were given a minimum of four weeks to recover from the CTMAX measurements (previously described in Chapter 3) before being used for ƒH,max measurements.  Experimental setup and protocol for fH,max measurements were the same as the previously described in Chapter 2 (Section 2.2.4). A different heating unit (3016D heater/chiller; Fisher Scientific, Ottawa, Ontario, Canada) was used to heat the water at the same rate.  A different data acquisition system (BioPac MP-150; BIOPAC Systems, Goleta, California, USA ) was used for the electrocardiogram (ECG) recording. Ventricular tissue was harvested at strategic temperatures: 15°C (n=2-4 per strain), which acted as a control because it was the rearing temperature; 20°C (n=6 per strain), which is close to the thermal optimum for aerobic scope as determined in Chapter 3; and at critical temperature for fH (TAR) (n=6 per strain) Table 4.1). Tissue samples were preserved in 1.5 ml of RNAlater (Life Technologies Inc., Waltham, MA, USA) and stored at −80°C until RNA extraction. 118  4.2.3 RNA Sequencing 4.2.3.1 Total RNA extraction Whole ventricles (<10 mg) were used for RNA extraction to compensate for the low RNA yield in contractile tissues. Total RNA was extracted using RNeasy Mini Kit (Qiagen Inc., Valencia, CA, USA, www.qiagen.com) according to the instruction manual. Briefly, tissues were disrupted and homogenized in lysis buffer and the mixture was centrifuged at 12,000× g for 4 min at 4°C. After removing the clear supernatant and adding 70% ethanol, samples were centrifuged through an RNeasy spin column where genomic DNA were removed and RNA was selectively bound to the membrane. After removing contaminants in subsequent wash steps, high-quality RNA was eluted in 20 μL RNase-free water. Extracted RNA was quantified using Infinite M200 Pro Microplate reader (Tecan Group Ltd., Männedorf, Switzerland, www.tecan.com).  4.2.3.2 RNA library preparation Concentration of extracted total RNA was normalized to 50 μg/ul for RNA sequencing library preparation. Ribosome RNA (rRNA) was removed from total RNA using Ribo-ZeroTM Magnetic Gold Kit (Epicentre, Madison, WI, USA, www.epibio.com) following the manufacturer instructions. The rRNA-depleted RNA was used for the preparation of RNA Sequencing library using ScriptSeqTM v2 RNA-Seq Library Kit (Epicentre). First, RNA was randomly fragmented and then transcribed to complementary DNA (cDNA), followed by adapter ligations at both end. Then, the di-tagged cDNA was purified using MinElute PCR Purification Kit (Qiagen). PCR were performed for 12 cycles (95°C for 30 s, 55°C for 30 s, 68°C for 3 min) to amplify the library and incorporate a six-nucleotide individual-specific barcode (ScriptSeq index PCR primers, Epicentre) for sample 119  multiplexing in sequencing libraries. Each amplified library was then size selected (>200 nucleotides) using a solid phase reversible immobilization paramagnetic bead technology (Agencourt AMPure XP system, Danvers, MA, USA). The libraries were then eluted in 20 ul nuclease-free water containing 0.1% volume of tween-20. RNA libraries were quantified by qPCR using Power-Sybr master mix (Life Technologies) and standard Illumina primers (P5, P7) on an ABI 7900HT Real-Time PCR System (Life Technologies, Grand Island, NY, USA). Amplification conditions were as follows: 50°C for 2 min, 95°C for 10 min, 40 cycles of (95°C for 15 s, 60°C for 15 s) and 72°C for 1 min. Indexed libraries were normalized to either 3 nM or 5 nM depend on the lowest library concentration. After excluding three libraries with low concentration, 57 normalized libraries were pooled into five sequencing libraries, each of which consist of 10-12 samples. 4.2.3.3 Illumina sequencing Each of the pooled libraries was sequenced in two lanes of a single read 100 bp flow cell on an Illumina HiSeq 1500 platform for a total of 10 lanes. Data from each lane was demultiplexed by index sequence and reads were combined from both lanes for each sample.  4.2.3.4 Alignment to reference O. mykiss mRNA Sequence reads were trimmed to 60 bp to reduce sequencing errors at the end of reads and to remove the index used to split reads for each individual. Trimmed reads were subsequently mapped to the O. mykiss mRNA (Berthelot et al., 2014) using Bowtie2 local alignment (Langmead and Salzberg, 2012). Bowtie2 aligns the reads using the created mapping index between libraries and reference data based on the Burrows-Wheeler Transform (BWT). Maximum number of the allowed uncalled base per read was set to three, which ensured a minimum of 95% identical for a 60 bp transcript. This yielded an average of 120  ~13.3 million aligned reads per individual. Four failed samples were removed from further analyses, which left a total of 53 individuals available for downstream statistical analyses. 4.2.4 Differentially expressed genes analysis EdgeR package in R was used to calculate differential gene expression following the steps described in (Anders et al., 2013). Genes with weak expression (rare transcripts) or not aligned were excluded from analysis. Only genes with at least two counts per million (CPM) reads in at least nine samples were included in followed analyses. To compare the gene expression between populations and between temperature treatment groups, gene counts were normalized with trimmed mean of M-values (TMM) method (Robinson et al., 2010), which has been considered to be the most efficient and reliable method in RNA sequencing studies. Gene-wise dispersion and a general linear model (GLM) were used for differentially expressed (DE) transcript identification with a Benjamini–Hochberg multiple-testing adjusted P-value (false discovery rate, BH-FDR) of <0.05 (Benjamini and Hochberg, 1995). 4.2.5 Annotation and gene ontology enrichment analysis Functional annotation and gene ontology (GO) analysis were performed using Blast2Go version 3.1 (http://www.blast2go.com/b2ghome) (Conesa et al., 2005). Significantly regulated transcripts in each strain were annotated in Blast2GO using blastx with default settings. Enrichment analysis of DE genes was performed in Blast2GO, which uses a Fisher's exact test to determine significance after controlling for multiple tests with a BH-FDR at 0.05. All annotated genes were used as background to evaluate the overrepresented GO term categories. Enriched GO categories were classified into three major functional categories: biological process, molecular function and cellular component. Functionally similar GO terms were summarized in REVIGO by setting the cutoff value “C” 121  of 0.7 (minimum similarity) (http://revigo.irb.hr/) (Supek et al., 2011). Metabolic pathways genes were annotated with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis in Blast2GO, while other genes were analyzed using R package KEGGREST v1.12.2 (Tenenbaum, 2016). 4.2.6 Association transcriptomics Association between gene expression and TAR was measured using Random Forest analysis (Breiman, 2001; Diaz-Uriarte, 2007). First, genes that have relatively higher power to predict TAR were screened by linear or exponential regression. A total of 737 transcripts with a minimum R2 value of 0.3 were selected for the Random Forest analysis. Twenty-one fish were sub-divided into a low TAR group (< 24°C, n=10 fish) and a high TAR group (≥24°C, n=11).  Backward elimination Random Forest was used for gene selection and classification. The number of trees (ntree) was set to 5000. The number of variables randomly chosen at each split (ntry argument) was the square root of the total number of input variables. Variable importance was estimated for each transcripts abundance according to their usage in separating the low TAR and high low TAR group. Random Forest analysis was repeated 10 times and mean variable importance for each transcripts was calculated for ranking. 4.2.7 Statistics All summary data analysis was performed using SigmaPlot version 12.5 (Systat Software Inc., San Jose, CA, USA) except where specified. Body mass were compared using one-way ANOVA on ranks with Dunn’s post-hoc test at the level of p < 0.05. Four scaling exponents (−0.10, −0.15, −0.20, −0.25) were used to correct ƒH,max for body mass. The ƒH,max was compared using one-way ANOVA with Tukey’s post-hoc test at the level of α=0.05. 122  Arrhenius breakpoint temperature (TAB) analysis was performed according to Yeager and Ultsch, (1989) and calculates the intersection of two best-fit linear regression lines in an Arrhenius plot of the natural log of ƒH,max versus inverse of Kelvin temperatures. The Q10 breakpoint temperature (TQB) was determined as the temperature beyond which the incremental Q10 for ƒH,max decreased below and remained below 1.8. 4.3 Results 4.3.1 Maximum heart rate (ƒH,max) Significant differences in body mass existed among populations (Table 4.1). Fish from a cooler environment had significantly smaller body mass because spawning occurred later in these populations and fry were younger. Therefore, scaling of ƒH,max to body mass was explored using a range mass exponents of −0.10, −0.15, −0.20 and −0.25 (Figure 4.1).  For the uncorrected ƒH,max (Figure 4.1 A), redband trout had an average ƒH,max of 107.7 beats per minute (bpm) at 15°C, with individuals values varying between 96.4 bpm and 119.0 bpm. Warming from 15°C to 20°C, ƒH,max increased with a similar Q10 (1.90-1.96) in all populations. Overall, Little Jacks had significantly higher ƒH,max than Fawn over the temperature range from 15°C to 20°C (p < 0.05, Table 4.2).  After correction for body mass, e.g. using an exponent of −0.10, differences in ƒH,max between populations became more apparent and suggested a clearer relationship between cardiac performance and the thermal habitat of the population (Table 4.2; Figure 4.1 B-E), i.e. the population from warmer climate had higher ƒH,max.  Thermal performance of ƒH was estimated by the first Arrhenius breakpoint temperature (TAB) and Q10 breakpoint temperature (TQB) (Table 4.3). TAB was not significantly different among groups and ranged between 18.7±0.4 °C (Little Jacks) and 123  20.4±0.7°C (Hybrid). Similarly, TQB, which was arbitrary defined as the temperature beyond which Q10 decreased below 1.8, was similar for all populations. Keithley had the most dramatic drop in Q10 during acute warming, which was largely driven by the early plateau after temperature reached 23°C. The temperature of peak ƒH,max (TPEAK) was also not significantly different between populations, ranging between 23.2±0.7 to 23.8±0.7°C (Table 4.3). Most fish (82.6%) developed cardiac arrhythmia without having a distinct plateau in ƒH,max. Individual fish showed cardiac arrhythmia at temperatures from 22°C and beyond (up to 28°C) (Figure 4.2). All four groups showed cardiac arrhythmia by the temperature of 28°C and no significant differences in average arrhythmia temperature (TAR) were found among wild populations (24.5-25.1°C), or the hybrid group (23.3°C) (Table 4.3). One fish from Little Jack had an exceptional performance with peak ƒH,max occurred at 28°C, which is about 2°C below the maximum individual CTMAX (as measured in Chapter 3). 4.3.2 RNA sequencing  In total, ventricles of 60 individual fish from four redband trout strains were sampled for RNA sequencing. After the quality control for total RNA extraction, library preparation and alignment to the O. mykiss reference mRNA (Berthelot et al., 2014), 53 qualified samples were used for subsequent gene expression and association analysis (Appendix C.1). Total reads for individuals ranged from 11.8 to 47.1 million. On average, 61.3% of the total reads were aligned to the reference, with aligned reads in individual fish ranging from 6.1 to 27.7 million.  At the population level, mean percentage of alignments ranged from 60.0 to 61.8% and mean aligned reads ranged from 19.8 to 28.4 million. Among temperature treatments, mean percentage of alignments ranged from 59.1 to 64.5% and mean aligned 124  reads ranged from 21.4 to 23.8 million. In total, 46,586 unique transcripts were included for the subsequent analyses. 4.3.3 Effect of temperature on cardiac gene expression  4.3.3.1 Redband trout as a species Effect of temperature on gene expression was firstly examined at the species level by pooling data from all 53 fish. Warming individual fish from the control temperature of 15°C to their respective TAR, which took 42-78 min, resulted in a significant change in the regulation of 980 genes (711 up-regulated, 269 down-regulated) (Figure 4.3).  The number of significantly regulated genes and the magnitude of expression were both large considering the relatively short timeframe for a response (Appendix C.2). Among those significantly up-regulated genes, heat shock responsive genes (hsp70, hsp30 and dnajb1 (or hsp40) had the greatest fold changes (Figure 4.4). Expression of itpka and tmem56b genes were also highly up-regulated. Three genes were significantly down-regulated by over 4-fold (lfng, irx5 and sall1). All significantly regulated genes were broadly classified into three categories: biological process, molecular function and cellular component (Appendix C.3). To further identify the function of thermal regulated genes in cardiac myocytes, genes were mapped to pre-identified molecular pathways in KEGG Database (Kanehisa et al., 2016). A total of 263 pathways were identified and each of them was assigned at least one gene. Seven pathways were exclusively involved in cardiac muscle functions (see more details in Figure 4.5 and Appendix C.4), including adrenergic signaling in cardiomyocytes, cardiac muscle contraction, arrhythmogenic ventricular cardiomyopathy, dilated cardiomyopathy, and hypertrophic cardiomyopathy. Other common but important pathways are energy production 125  pathways (e.g. citrate cycle, pyruvate metabolism, galactose metabolism, oxidative phosphorylation), calcium signaling pathway and hypoxia-inducible factor 1 (HIF-1) signaling pathway (see more details in Figure 4.6 and Appendix C.4). Also, a number of signaling pathways were identified, e.g., PI3K-Akt, FoxO, MAPK and Wnt.  4.3.3.2 Intraspecific regulation of gene expression There were significant intraspecific variations in the number of thermally induced genes in redband trout (Table 4.4). In total, 4,050 transcripts were differentially expressed during acute warming in at least one population. Among these transcripts, 84.3% were only significantly regulated in Fawn (Figure 4.7 A), which is from the coldest summer climate among studied populations. Fawn had 3,647 significantly regulated genes (2,447 up-regulated and 1,200 down-regulated) during warming from 15°C to TAR. Little Jacks had 524 significant regulated genes (404 up-regulated and 120 down-regulated), while Keithley had only 26 significant regulated genes in total (24 up-regulated and 2 down-regulated). K×LJ hybrid had 188 significant regulated genes (161 up-regulated and 27 down-regulated), which was between the number for Little Jacks and Keithley. Despite the difference of population-specifically regulated genes, seven genes were significantly regulated in all three wild populations (only two genes if the hybrid is included). Four of these common genes were hsp70 (two transcripts), arip4 and tbsp2, and they were also the most strongly regulated with over 4-fold changes (Figure 4.7 B). The regulation of all detected heat shock protein genes were detailed in Appendix C.5. Similar to the analysis at the species level, regulation of most genes within populations did not become significant until temperatures reached >20°C. For treatments warming from 15°C to 20°C, all wild populations had < 10 significantly regulated genes 126  while the K×LJ hybrid had 21 up-regulated genes (Table 4.4). In contrast, when temperature reached TAR, the number of differentially regulated genes dramatically increased for all populations (454 for Little Jacks, 127 for K×LJ hybrid and 2,977 for Fawn) except for Keithley. Gene Ontology (GO) enrichment analyses revealed 66 enriched functional categories for the significantly up-regulated genes and 18 for the down-regulated genes (Figure 4.8). Little Jacks and K×LJ had similar numbers of enriched GO terms for both up- and down-regulated genes. Fawn had a significantly larger number of enriched GO terms (only the top eight GO terms for each category are present in Table 4.5). For up-regulated genes, most enriched GO terms were related to the transcription, translation, mitochondria and regulation of cell cycle. Little Jacks had three enriched pathways, which were also enriched either in hybrid K×LJ (cell cycle arrest, GO: 0007050; cyclin-dependent protein serine/threonine, GO: 0004861) or in Fawn (protein folding, GO: 0006457). In contrast, Fawn had more uniquely enriched functions than all other strains, such as nucleotide binding (GO: 0003723, GO: 0000166) in the categories of molecular function, mitochondria (GO: 0005743, GO: 0005739, GO: 0005753) and hemoglobin (GO: 0005833) in the categories of cellular component.  For down-regulated genes, GO terms related to nucleus (categories of cellular component) were found in all groups (Table 4.6). Little Jacks and Fawn also shared GO terms of transcription regulation (GO: 0006355). Fawn had the most unique enriched functions than all other groups. KEGG pathway enrichment analysis also demonstrated intraspecific differences in metabolic pathways, e.g. aminoacyl-tRNA biosynthesis, biosynthesis of antibiotics and purine metabolism (Appendix C.6).  No significant enriched 127  GO terms were discovered in Keithley because of the small number of both up- and down-regulated genes. 4.3.4 Differentially expressed (DE) genes between desert and montane populations Differentiation in gene expression was observed between populations (Table 4.7). For the pairwise comparisons among populations, Little Jacks and Fawn had the most DE genes (3330). Keithley had fewer DE genes with Fawn (1267) compared to Little Jacks (1714), which was a pattern in accordance with their genetic differentiations and habitat temperature differences (Figure 4.9). DE genes between Little Jacks and Keithley did not become significant (<100) until temperatures beyond 20°C (1033), while the difference between Little Jacks and Fawn was larger even at 20°C (1191). Interestingly, Keithley and Fawn maintained large number of DE genes across all temperatures, with 783 at 15°C, 753 at 20°C and 1243 at TAR. Hybrid had significant less DE genes with Keithley compared to Little Jacks (93 versus 1787). The large number of overall DE genes between Little Jacks and K×LJ hybrid was largely due to the difference at 20°C [15°C (0), 20°C (983), TAR (9)]. Differentially expressed genes between desert and montane populations allowed me to identify thermal adaptation related differences in the phenotypes of genes expression. Most DE genes were due to thermal-induced expression, while only several were constitutively different. I found five genes that were differentially expressed across all temperatures between Little Jacks and Fawn. Between Little Jacks and Keithley, only five genes were differentially expressed at 15 as well as 20°C, and none were constitutively different across all temperatures (Table 4.8). There were 85 DE genes between desert (Little Jacks) and montane (both Keithley and Fawn) redband trout, including some genes that may play important roles in thermal adaptation, such as heat shock protein and hemoglobin. 128  4.3.5 Association between gene expression and arrhythmia temperature (TAR) Genes whose expression levels were significantly associated with TAR (which ranged between 21 and 28°C) were screened using the Random Forest algorithm. As a result of setting the R2 threshold at 0.3 for the regression analysis of gene expression versus temperature, a total of 737 transcripts that satisfied the criteria were included in Random Forest analysis. To estimate the importance of each transcript, the effectiveness of the expression level in predicting low TAR (21-24°C) or high TAR (24-29°C) were tested and summarized as variable importance. After ten repeat runs of Random Forest, the average importance of the first 100 transcripts were sorted (Figure 4.10 A).  I chose the 30 most important transcripts and visualizes the relationship between gene expression and TAR using a heat map (Figure 4.10 B). Six genes that had been ranked the most important were pard6b, tram2, tef3, cnn3, clta and rasip1. Not surprisingly, three heat shock genes were also among the top genes, including hsp30, hsp70 and hsp90α (Figure 4.10 B). 4.4 Discussion Transcription is an important regulatory process in response to heat stress and the underlying mechanism improves our understanding on thermal adaptation. I focused on cardiac transcriptomic response in this thesis because of the limiting role of fH in upper thermal tolerance. Redband trout populations from desert and montane climates, as well as a hybrid strain, were used to provide insight into the cellular mechanisms of thermal adaptations. In less than 30 minutes, progressive warming of just 5°C (from rearing temperature of 15°C to 20°C) induced the regulation of some genes, and further warming to TAR triggered even greater gene expression changes in multiple pathways, including the heat shock response, cardiac muscle function and energy metabolism. Moreover, different sets of 129  gene were thermally induced in populations from different climates, which suggests intraspecific regulatory strategies in response to acute heat stress. Likewise, the abundance of mRNA was significantly different between populations for many genes (discussed below), especially at TAR, due to a combined effect of constitutively and thermally induced gene expression. Lastly, expression-trait association analyses identified genes that may be linked to the occurrence of TAR. 4.4.1 Maximum heart rate (ƒH,max) in response to warming In this study, redband trout populations are from habitats with different winter durations and, as a result, their spawning timing (often at temperatures reach over ~6°C in spring, (Muhlfeld, 2002)) are also different. Fawn Creek redband trout spawned >2 months later than Little Jacks fish because of the lengthy winter. Thus, body size differed among populations at the time I conducted this study. Therefore, I considered the effect of body mass on my ƒH,max results. Although the most used scaling exponent for ƒH in mammals is −0.25 (Schmidt-Nielsen, 1984), there is no prior knowledge of that for fishes. It is clear that mass scaling effect of ƒH is different between life stages. In zebrafish (Danio rerio) (Barrionuevo and Burggren, 1999), rainbow trout (Holeton, 1971; Miller et al., 2011), and Arctic charr (Salvelinus alpinus) (McDonald and McMahon, 1977), ƒH increases during the early embryo development until hatching and followed by a decrease period (likely as a result of the development of pacemaker cells and extrinsic innervation), which is the life stage of redband trout in present study and thus suggests a negative mass exponent. Interestingly, a positive correlation has been found in larger chinook salmon (Oncorhynchus tshawytscha) with body mass from 2.7 to 16.8 kg (Clark and Farrell, 2011).  130  Without a readily available scaling factor for salmonid species, I used four exponents (−0.10, −0.15, −0.20 and −0.25) from being conservative to being liberal. After applying the scaling law and correction for body mass, differences in ƒH,max among populations become even more obvious, e.g., Fawn had the lowest ƒH,max while Little Jacks had the highest ƒH,max. This result is in accordance with my predictions regarding the important role of ƒH in thermal tolerance (Pörtner, 2001; Pörtner and Farrell, 2008; Farrell, 2009; Pörtner, 2010).  Moreover, the K×LJ hybrid had ƒH,max values more closer to Keithley instead of Little Jacks, which suggests a dominance effect of Keithley in ƒH,max. Future study using common-sized fish at various life stages should be conducted to provide more solid information and to determine an exact scaling exponent. When only looking at the uncorrected values, redband trout demonstrated exceptional cardiac performances, which were observed here for the first time in the genus Oncorhynchus. Peak ƒH,max of redband trout (~184 bpm) is among the highest compared to other con-generic species such as ~133.2 bpm in rainbow trout from a colder climate (Blackwater Creek, British Columbia) (Verhille et al., 2013), ~ 140 bpm in coho salmon (Oncorhynchus. kisutch) (British Columbia) (Casselman et al., 2012), ~145 bpm in sockeye salmon (Oncorhynchus nerka) (British Columbia) (Chen et al., 2013).  Actually, peak ƒH,max of redband trout is closer to Atlantic salmon (Salmo salar) of a similar body mass, ~200 bpm (Anttila et al., 2014b), which has the highest CTMAX reported among salmonids (Beitinger et al., 2000). However, Anttila et al., (2014a) raised an important caveat with this comparison because peak ƒH,max increased from ~155 bpm for 12°C-acclimated fish to ~200 bpm for 20°C-acclimated fish (Anttila et al., 2014a). It is unknown whether such cardiac plasticity also exists in O. mykiss.  If the plasticity does exist, I predict redband trout may have the 131  same or even high peak ƒH,max as Atlantic salmon when redband trout are acclimated at 20°C instead of the 15°C that was used in this thesis. TPEAK averaged between 23.2 to 24.2°C, which is also the temperature at which absolute aerobic scope (AAS) collapsed (see Chapter 3), with the exception of Little Jacks, which had a wide optimal thermal window for AAS. Despite the intraspecific differences in ƒH,max, the rate transition temperatures for ƒH,max were not significantly different among populations, and hence did not reflect the significant differences in the rate transition temperatures for AAS (Chapter 3).  In montane populations, some thermal indices for ƒH,max were correlated with AAS presented in Chapter 3. Results showed a closeness between TAB and upper Tpej for AAS for Fawn, which was the same result in PFRC rainbow trout population from Western Australia in Chapter 2. However, Little Jacks is an exception because of the relative lower and thermally independent AAS, suggesting a lower AAS is less sensitive to the rate transition of ƒH,max, at least within the optimum thermal ranges. It should also be noted that ƒH,max was measured every 1°C and thus provided a higher resolution than the AAS measurements to generate rate transition temperatures.  4.4.2 Effect of temperature on cardiac gene expression Studying heat stress caused gene expression in fishes has been an active area in research (Gracey, 2007; Logan and Buckley, 2015). However, most studies on cardiac transcriptomes have focused on thermal acclimation (Vornanen et al., 2002; Gracey et al., 2004; Vornanen et al., 2005; Hassinen et al., 2008; Castilho et al., 2009; Korajoki and Vornanen, 2012; Jayasundara et al., 2013; Anttila et al., 2014a; Jørgensen et al., 2014). Acute responses to thermal stress have been limited to studies of liver, gill and muscle tissue (Buckley, 2006; Kassahn et al., 2007a; Kassahn et al., 2007b; Buckley and Somero, 2009; 132  Logan and Somero, 2010; Lewis et al., 2010; Thorne et al., 2010; Quinn et al., 2011; Newton et al., 2012). Only after I began my thesis has the cardiac transcriptome response to transient heat stress been studied in rainbow trout (Tan et al., 2016). Compared to a prolong thermal stress study using the same redband trout populations (Narum and Campbell, 2015), acute warming in the present study induced the expression of fewer genes, which is an expected result due to the shorter time of exposure. However, it remains clear that a 0.17°C min−1 change in water temperature over 42-78 minutes is sufficient to trigger transcription, suggesting the thermal sensitivity of the cellular regulation in heart. Acute warming also triggered large magnitude changes in gene expression (over 4-fold), which is in accordance with the previous acute warming studies (Buckley, 2006).   4.4.2.1 Important pathways in response to warming The heat shock response (HSR) is a widely recognized process in response to heat stress in fishes with the exception of the Antarctic fish (Trematomus bernacchii) (Hofmann et al., 2000). The HSR produces a family of heat shock proteins (Hsps) to facilitate protein assembly, rescue mis-foldings, translocation and degradation (Iwama and Thomas, 1998; Feder and Hofmann, 1999; Iwama et al., 1999). The HSR responds to thermal stress in a timely fashion and in large magnitude (Newton et al., 2012). In this study, I detected Hsps genes from all size classes (Hsp 90 kDa, 70 kDa, 40 kDa, 30 kDa and small Hsps). Among them, hsp30 showed the greatest change with a >200-fold increase in expression. Expression of hsp70 gene was also strongly induced with over 40-fold changes. Both hsp30 and hsp70 have been identified as thermally inducible genes in previous studies with rainbow trout, but in different tissues (Currie et al., 2000; Ojima et al., 2012). It is not surprising that dnaj (hsp40), which encodes a molecular chaperon of Hsp70, also increased over 20-fold. 133  Similarly, hsp90 gene was significantly up-regulated here (11-fold increase) and in previous studies using different tissues (Buckley, 2006; Fangue et al., 2006; Quinn et al., 2011). The onset temperature of HSR may have some ecological relevance (Fangue et al., 2011) so that energy allocation can be regulated. In this study, a significant HSR was triggered at temperatures > 20°C in all populations, which is in accordance with the maximum habitat temperatures. Cardiac specific gene expression also needs to be remodeled to meet the functional requirement of a heart in acute stress situations. In heart, stimulation of beta-2 adrenergic receptor (β2AR) increases the intracellular cyclic adenosine monophosphate (cAMP), which leads to the activation of the downstream cAMP-dependent protein kinase A (PKA). PKA phosphorylates several proteins that play key roles in cardiac function, e.g. sarcolemmal L-type Ca2+ channels (DHPR) (Lohse et al., 2003), potassium channels (Huang et al., 1994; Kurokawa et al., 2003), sarco(endo)plasmic reticulum Ca2+ pump (SERCA) regulator phospholamban (PLB) (MacLennan and Kranias, 2003), Na+/Ca2+ exchanger (NCX) and transcription factor Creb. In rainbow trout, adrenergic stimulations provide fully protection of cardiac function under hypoxia, acidotic and hyperlemic conditions (Hanson et al., 2006).  Here, I observed a significant up-regulation of the β2ar gene and some of its downstream effectors, which may protect the heart in acute stress situations. Significant up-regulation was observed in the potassium voltage-gated channel subfamily e member 1 (KCNE1) gene expression. KCNE1 is a slow-delayed rectifier K+ channel, pumping K+ to extracellular matrix during action potential depolarization and thus help setting the "plateau" period of cardiac action potentials. In mammals, up-regulation of kcne1 has been associated with the long QT syndrome (Watanabe et al., 2007). When a fish 134  heart beats at its peak rate, a long QT interval can be deleterious and increases the risk of ventricular arrhythmias or fibrillations. Thus, the exact purpose of the up-regulation of kcne1 during acute warming is not clear. Other common-known ion channel genes, such as dhpr, ncx, serca and ryr, were not significantly regulated in this study. However, serca and ncx have previously been shown to be significantly regulated after long-term thermal acclimation. For example, serca had lower gene expression in warm-acclimated rainbow trout (Korajoki and Vornanen, 2012) and endothermic Pacific bluefin tuna (Thunnus orientalis) (Jayasundara et al., 2013).  Myosin is a large protein containing two heavy chains and four light chains. Warm acclimation decreased the expression of myosin in rainbow trout (Vornanen et al., 2005) and Atlantic salmon (Jørgensen et al., 2014). Myosin heavy chain has several isoforms. In carp (Cyprinus carpio), different isoforms are expressed according to acclimation temperatures (Watabe, 2002). In present study, the down-regulation of myosin-7b, which encodes a slow-twitch isoform in cardiac muscle, might be part of the isoform remodeling process. Together with the previous observation of the up-regulation of fast-type myosin heavy chain in warm acclimation (Gerlach et al., 1990), expression of myosin isoforms may be important in compensating the thermal related tachycardia.  Titin is an elastic protein that distributed between the Z- and M- lines of sarcomere and is responsible for developing passive tension. Similar to myosin, cardiac titin has two isoforms, N2b (stiff) and N2ba (compliant), as seen in carp (Spierts et al., 1997) and mammals (Helmes et al., 1999; Cazorla et al., 2000). Although it is unclear whether a specific isoform is favored in thermal stress, it is known that tension changes may affect the type and amount of titin protein expression. In acute warming scenarios, heart rate 135  significantly increases and the ventricle filling time reduces, which results a reduced preload and active tension. Thus, it is reasonable to speculate that less titin is needed in warm hearts. Alternatively, the down regulation of myosin and titin could simply be the result of suppression for cell division, which might be related to the decrease in cardiac mass during warm acclimation of rainbow trout  (Farrell et al., 1988b; Graham and Farrell, 1989; Taylor et al., 1996; Klaiman et al., 2011). Other important pathways with significantly regulated genes include hypoxia inducible factor-1 (HIF-1) signaling pathway and energy metabolism pathways (pyruvate metabolism, galactose metabolism, citrate cycle and oxidative phosphorylation). All these pathways have been identified in previous studies with either prolonged or acute thermal stresses, but in different tissues. HIF-1 signaling pathway elements have been found inducible to thermal stress in Antarctic fishes’ liver (Thorne et al., 2010) and brain tissue (Beers and Sidell, 2011). By exposing redband trout to diel temperature cycles (17-28°C), genes in oxidative phosphorylation were significantly regulated in gill tissues in Little Jacks and Keithley, as well as their hybrids (Garvin et al., 2015). The activity of glycolytic and mitochondria enzymes are critically important for ATP production. Previous studies have showed the regulation of energy production genes or proteins in heart in response to changing temperatures (Vornanen et al., 2005; Jayasundara et al., 2013; Jayasundara et al., 2015). 4.4.2.2 Intraspecific pattern of gene expression Genes that are thermally inducible in one population could be unresponsive in other populations, particularly when they are from different thermal environments. Broadly, Antarctic fish have lost the heat shock response following, presumably, the adaptation to a 136  stable and frigid thermal environment (Hofmann et al., 2000; Buckley and Somero, 2009). Jeffries et al., (2014) found 14-time difference in the number of differentially expressed genes in response to high temperature between congeneric pink salmon (Oncorhynchus gorbuscha) and sockeye salmon. Similarly, in this study, Little Jacks significantly regulated 19-times more genes than Keithley. Previous studies also found that warm acclimation triggered more genes to be expressed in Little Jacks than Keithley (Narum and Campbell, 2015; Garvin et al., 2015). While I expected that redband trout from warm climate would have more regulated genes, it was not always the case. Fawn Creek is from the coldest climate but had the most significantly regulated genes. This inconsistent pattern in gene expression may be due to two inter-related factors. First is the thermal sensitivity (or timing) of cellular regulation. It has been found that the onset temperature of heat shock response is positively correlated with CTMAX (Fangue et al., 2011). Thus, a population with lower CTMAX would have a lower onset temperature for cellular response, and potentially would trigger more gene expression, which explains the large number of gene expression in Fawn Creek. Second is the magnitude of the regulation. By increasing constitutive gene expression, transcription only needs to be regulated to a smaller magnitude, which may not be statistically significant, especially with a small sample size. In addition, increased constitutive gene expression also increased the onset temperature of gene expression (Logan and Somero, 2011). Natural populations may have adopted a complex strategy to thrive the specific thermal regimes. Simply looking at the number of thermal induced genes may not provide much insight into the mechanism of thermal adaptation.  137  4.4.3 Gene expression among populations The purpose of transcription is to change the mRNAs abundance, which can further change the abundance of its functional products. Among locally adapted conspecific populations, intraspecific differences in the abundance of mRNA and protein have been observed, e.g. the lactate dehydrogenase-B (LDH-B) in common killifish (Fundulus heteroclitus) (Crawford and Powers, 1989; Schulte et al., 2000) and ventricular β-adrenergic receptors in Fraser River sockeye salmon (Eliason et al., 2011). Similarly, intraspecific differences have also been observed in the gene expression of Hsps (Fangue et al., 2006; Healy et al., 2010; Tan et al., 2016), mitochondria enzymes (Fangue et al., 2009) and ion transporter (Healy et al., 2010). I also found DE genes among redband trout populations, and the number tends to have a positive relationship with the differences in the genetic distance and habitat thermal regime. Among these DE genes, those differed between desert and montane populations are of particular interest, especially at critical temperatures. Between Little Jacks and Fawn, which differed the greatest in habitat temperatures, five constitutively expressed genes were found significant different across all test temperatures. All these genes were not found to be thermally inducible and thus may reflect the constitutive level differences in protein abundance (Newman et al., 2006). Neural cell adhesion gene expression differed about four-fold. Two lysozyme-related transcripts were found having higher expression in Fawn Creek. Lysozyme plays a key role in glycoside hydrolases to against bacterial infection. It is possible that Fawn fish acclimated in laboratory environment at constant 15°C, which is a higher temperature than Fawn Creek, constitutively triggered immune response, because it is well known that high temperature causes immune responses (Köllner et al., 2002). Lysozyme is also thermally inducible under warm 138  acclimation in Atlantic salmon heart (Jørgensen et al., 2014). In contrast, between Little Jacks and Keithley, no genes were found to be different in abundance across all temperatures. Only six genes differed at both 15°C and 20°C, but the differences diminished at TAR. Two heat shock protein mRNAs were more abundant in montane population (both Fawn and Keithley) than desert fish (Little Jacks). One is heat shock protein beta-7 (Hsp7b), also known as cardiovascular heat shock protein (CvHsp), which is a tissue-selective and highly expressed stress protein in heart (Krief et al., 1999). In contrast to the present study, the hsp7b gene was over two-fold more abundant in high temperature tolerant Atlantic salmon relative to the less tolerant ones (Quinn et al., 2011). The other is a heat shock protein co-chaperon Cdc37 (cell division cycle protein 37), which interacts with Hsp90 to form a complex. The Hsp90-Cdc37 complex then plays an import role in maturate a number of kinases, such as Tyrosine and Ser/Thr protein kinases (Shao et al., 2003; Wandinger et al., 2008; Hunter and Poon, 2016). The Hsp90 co-chaperon role is likely to be more important in thermal stress conditions. In channel catfish (Ictalurus punctatus), the cdc37 gene was significantly up-regulated after exposure to 36°C for 30 hours (Liu et al., 2013). For other heat shock genes that have been found to be thermally inducible (e.g. hsp70), contradictory results have been found, revealing difficulties in direct comparison between gene expression studies. Tan et al., (Tan et al., 2012; Tan et al., 2016) showed that thermal selected rainbow trout had exceptional higher expression of hsp70 compared to non-selected fish. In contrast, Narum et al., (2013) found that high temperature tolerant redband trout (Little Jacks) had lower hsp70 than Keithley. However, the present study did not find significant differences in 139  hsp70 expression between populations. The divergent results between studies may be due to the differences in fish strain or experiment thermal regimes. I also observed hemoglobin gene (both alpha and beta subunit) were more abundant in the desert population than the montane. Hemoglobin genes are responsive to temperature in brown trout (Salmo trutta) (Meier et al., 2014), Arctic charr (Quinn et al., 2011) and channel catfish  (Liu et al., 2013).  In my data, Little Jacks had twice as much myosin-binding protein fast-type mRNA as the montane populations. Myosin-binding protein fast-type gene is exclusively expressed in heart (Fougerousse et al., 1998) and is critical for the cardiac muscle contraction (Lecarpentier et al., 2008). Together with the aforementioned regulation on myosin isoforms, muscle contraction elements are closely regulated during thermal stress and their expression may have been altered under adaptive changes. 4.4.4 Association between gene expression and arrhythmia temperature (TAR) The relationship between mRNA abundance and phenotype variation is important to understand the molecular mechanism of complex traits. Thermal tolerance of heart, which was estimated by TAR in this study, has been linked to the whole animal thermal tolerance (Casselman et al., 2012). To identify the putative limiting genes underlying the occurrence of cardiac arrhythmia, I conducted an expression - trait association analysis using the Random Forest (RF) algorism(Diaz-Uriarte, 2007). In mammalian researches, RF has been widely used to identify gene expression-disease associations (Jiang et al., 2004; Díaz-Uriarte and Alvarez de Andrés, 2006; Griffith et al., 2013). The two most significant genes were pard6b (partitioning defective 6 homolog beta-like) and tram2 (translocating chain-associated membrane protein 2-like). However, both of 140  their functions are not well studied in fishes. As a member of the par6 gene family, pard6b encodes a protein with a PSD95/Discs-large/ZO1 (PDZ) domain, an OPR domain and a semi-Cdc42/Rac interactive binding (CRIB) domain. Pard6b is an adapter protein involved in asymmetrical cell division and cell polarization processes (Goldstein and Macara, 2007). Pard6b protein is probably also involved in the formation of epithelial tight junctions (Joberty et al., 2000). Although Pard6b has not been directly related to cardiac functions, changes in tight junction elements of heart have been found affecting arrhythmia via the propagation of action potentials (Lisewski et al., 2008). The potential role of Pard6b in developing cardiac arrhythmia needs to be studied in the future. Tram2 is a component of the translocon, which is a gated macromolecular channel that controls the posttranslational processing of nascent secretory and membrane proteins at the endoplasmic reticulum (ER) membrane (Stefanovic et al., 2004). Tram2 affects the molecular chaperon activities in collagen folding by interacting with the ER Ca2+ pump (SERCA2b), which also plays a critical role in removing cytoplasmic Ca2+ so that muscles can relax. Although the role of tram2 in thermal stress is unclear, it may also be important for other molecular chaperon activities, as well in dealing with unfolded proteins.  4.4.5 Conclusion This chapter revealed intraspecific patterns of cardiac transcriptome responses to acute warming among redband trout populations from different climates. A fast heating rate of 1°C every six minutes for 42-78 minutes triggered differential expression of a large number of genes (up to thousands) with a strong magnitude of change (up to over 200-fold). Thermally responsive genes were linked to heat stress responses, as well as cardiac-specific processes. Intraspecific patterns of regulation during acute warming suggested the potential 141  roles of transcriptome regulation in thermal adaptation. Due to the combined effect of constitutive and thermally induced expression, mRNA abundance was also observed to be different among populations for genes involved in important processes that may contribute to the thermal performance of heart and whole organism. Thus, my results provided putative biomarkers for future functional association studies. Taking that into consideration, I conducted association analysis between gene expression and arrhythmia temperatures and identified significant candidate genes. With the complexity of thermal tolerance and the limitations of transcriptomic responses in mind (Evans, 2015), further exploration on the biochemical and physiological functions of these candidate genes is needed.   142  Table 4.1 Number of heart samples harvested during acute warming for RNA sequencing.  Sample size (60 in total)  Mean mass  (s.e.m., g) 15ºC  20ºC  TAR (22-28ºC)  Little Jacks (LJ) 4  6  6  5.7±0.7a Hybrid (K×LJ) 2  6  6 (-2)  4.3±0.8a Keithley (K) 3 (-1)  6 (-1)  6 (-1)  3.1±0.4a Fawn (F) 3  6 (-2)  6  1.1±0.1b Note:  Numbers in parentheses indicate the decrease of sample size due to quality control in the library preparation and sequencing.  Body mass (mean and s.e.m.) without a common superscript lowercase letter are significantly different according to the one-way ANOVA on ranks with Dunn’s post hoc test at significance level of α=0.05.     Table 4.2 Maximum heart rate (ƒH,max) at 15ºC, 20ºC and the temperature with peak ƒH,max (TPEAK) for each redband trout population.   15°C   20°C  TPEAK (°C)  Max  n ƒH,max (bpm)  n ƒH,max (bpm)  n ƒH,max (bpm)  ƒH,max (bpm) Uncorrected         LJ 16 111.3±1.5a  12 154.7±2.2a  6 183.8±7.9  216.2 K×LJ 14 107.1±1.2ab  12 147.0±1.3b  6 170.6±6.6  193.2 K 14 108.4±1.3ab  11 148.3±2.2ab  6 164.8±4.2  178.3 F 15 104.9±1.3b  12 146.6±2.3b  6 184.4±6.3  205.8            Mass corrected (Mb−0.10)         LJ 16 131.2±2.6a  12 182.7±4.0a  6 213.5±9.1a  243.0 K×LJ 14 121.8±3.2ab  12 167.7±3.8bc  6 188.2±8.0ab  219.3 K 14 119.9±2.6b  11 165.8±3.9c  6 184.3±5.7b  200.0 F 15 106.1±1.5c  12 148.2±2.5d  6 186.9±6.0b  193.2  Note:  All numbers are in mean±s.e.m. Peak ƒH,max values represent the maximum of recorded individual ƒH,max. The ƒH,max values were also corrected to body mass of one gram using scaling factor of −0.10. Sample size decreased with warming because some fish were euthanized at 15ºC and 20ºC for ventricle sampling (see Table 4.1). Significances of intraspecific difference were tested using One Way ANOVA with Tukey’s post hoc test at the level of α=0.05.   143  Table 4.3 Thermal indices for maximum heart rate (ƒH,max) in redband trout.   n TAB (°C) TQB (°C) TPEAK (°C) Mean TAR (°C) Median TAR (°C) LJ 6 18.7±0.4 18.9 23.8±0.7 24.9±1.0 24.0 K×LJ 6 20.4±0.7* 19.9 23.5±0.9 24.5±1.0 22.8 K 6 19.3±0.7 19.0 23.2±0.7 23.3±1.0 23.5 F 6 19.8±1.1 19.6 24.2±0.5 25.1±0.7 24.5 Note:  TAB: Arrhenius breakpoint temperature (mean±s.e.m.);  TPEAK: temperature of peak ƒH,max; TAR: mean arrhythmia temperature (mean±s.e.m.). *: TAB was based on sample size of five, because of the imperceptible TAB for one fish. Q10 breakpoint temperature (TQB) was derived from mean ƒH,max for each group. No statistical significant difference was found in intraspecific comparisons using one-way ANOVA (p > 0.05).         Table 4.4  Number of significantly regulated genes during acute warming for each population (FDR < 0.05).  15-20°C  20°C -TAR  15°C -TAR  unique up down  up down  up down  up down Little Jacks(LJ) 7 2  202 40  357 97  404 120 K×LJ 21 0  63 17  114 13  161 27 Keithley (K) 0 1  24 1  0 0  24 2 Fawn 2 0  1542 726  2095 882  2447 1200   144  Table 4.5 Gene Ontology (GO) enrichment analysis for the significantly up-regulated genes. Category GO ID GO Description P-value Little Jacks (LJ)    BP GO:0006457† protein folding 5.2E-25  GO:0007050‡ cell cycle arrest 9.7E-06 MF GO:0004861‡ cyclin-dependent protein serine/threonine kinase inhibitor activity 1.9E-08     K×LJ    BP  GO:0006351 transcription, DNA-templated 9.4E-05  GO:0007050‡ cell cycle arrest 9.7E-06 MF GO:0004861‡ cyclin-dependent protein serine/threonine kinase inhibitor activity 1.9E-08  GO:0001071 nucleic acid binding transcription factor activity 1.7E-05  GO:0001071 transcription factor, sequence-specific binding 1.7E-05     Keithley (K)     None       Fawn    BP GO:0006412 translation 3.4E-59  GO:0006457† protein folding 5.2E-25  GO:0006413 translational initiation 1.8E-10  GO:0006418 tRNA aminoacylation for protein translation 0.1E-08  GO:0006415 translational termination 4.0E-08  GO:0006122 mitochondrial electron transport, ubiquinol to cytochrome c 5.8E-07  GO:0000413 protein peptidyl-prolyl isomerization 7.7E-07  GO:0000398 mRNA splicing, via spliceosome 2.2E-05 CC GO:0005840 ribosome 6.4E-37  GO:0005743 mitochondrial inner membrane 4.1E-24  GO:0005739 mitochondrion 9.0E-21  GO:0000502 proteasome complex 5.7E-14  GO:0005737 cytoplasm  1.2E-11  GO:0005753 mitochondrial proton-transporting ATP synthase complex 3.5E-09 145  Category GO ID GO Description P-value  GO:0005833 hemoglobin complex 3.1E-06  GO:0005839 proteasome core complex 7.6E-06 MF GO:0003735 structural constituent of ribosome 2.3E-35  GO:0005198 structural molecule activity 5.0E-25  GO:0004298 threonine-type endopeptidase activity 6.5E-14  GO:0004812 aminoacyl-tRNA ligase activity 5.0E-10  GO:0000166 nucleotide binding 9.3E-10  GO:0003743 translation initiation factor activity 2.4E-08  GO:0003723 RNA binding 1.9E-07  GO:0003755 peptidyl-prolyl cis-trans isomerase activity 7.7E-07 Note:  GO terms are categorized into biological processes (BP), cellular components (CC) and molecular functions (MF). †, enriched pathway in both Little Jacks and Fawn;  ‡, enriched pathway in both Little Jacks and K×LJ;  §, enriched pathway only in K×LJ;  Keithley does not have enriched GO term because of the less number of significantly regulated transcripts. Only the top eight were present for each category for Fawn.      146  Table 4.6 Gene Ontology (GO) enrichment analysis for the down-regulated DE genes. GO terms are categorized into biological process (BP), cellular component (CC) and molecular function (MF). Category GO ID GO Description P-value Little Jacks (LJ)    BP GO:0006355† regulation of transcription, DNA-templated 1.3E-06 CC GO:0005634† nucleus 1.6E-04     K×LJ    CC GO:0005638 lamin filament 7.0E-06  GO:0005637 nuclear inner membrane 4.2E-05     Keithley (K)     None   Fawn    BP GO:0006355† regulation of transcription, DNA-templated 1.3E-06  GO:0001946 lymphangiogenesis 3.9E-05  GO:0006461 protein complex assembly 9.7E-06  GO:0000902 cell morphogenesis 2.90-06 CC GO:0005634† nucleus 1.6E-04  GO:0000786 nucleosome 1.7E-05 MF GO:0003707 steroid hormone receptor activity 5.2E-07  GO:0004879 ligand-activated sequence-specific DNA binding RNA polymerase II transcription  factor activity 2.4E-07  GO:0000981 sequence-specific DNA binding RNA polymerase II transcription factor activity 9.3E-06  GO:0001071 nucleic acid binding transcription factor activity 5.9E-06  GO:0003700 sequence-specific DNA binding transcription factor activity 5.9E-06  GO:0005017 platelet-derived growth factor-activated receptor activity 8.2E-05  GO:0004672 protein kinase activity 1.6E-05  GO:0004719 protein-L-isoaspartate (D-aspartate) O-methyltransferase activity 1.6E-04 Note: †, enriched pathway in both Little Jacks and Fawn. Keithley does not have enriched GO term because of the less number of significantly regulated transcripts.147  Table 4.7 Number of differentially expressed genes between redband trout populations.  Little Jacks (LJ) K×LJ Keithley (K) Fawn (F) Overall     Little Jacks ---    K×LJ 1787 ---   Keithley 1714 93 ---  Fawn 3330 3854 1267 ---      15°C     Little Jacks ---    K×LJ 0 ---   Keithley 85 0 ---  Fawn 15 1153 783 ---      20°C     Little Jacks ---    K×LJ 983 ---   Keithley 77 16 ---  Fawn 1191 3813 753 ---      TAR (21-28°C)     Little Jacks  ---    K×LJ 9 ---   Keithley 1033 49 ---  Fawn 2170 550 1243 ---    148  Table 4.8 Differential gene expression between desert and montane populations.  First five genes significantly differed in abundance between Little Jacks (LJ) and Fawn (F) across all three temperatures. The followed six genes were significantly different in abundance between Little Jacks (LJ) and Fawn (F) at 15°C and 20°C. Rest genes were the ones with largest differences in abundance between LJ and the two montane populations (K and F) at arrhythmia temperature. Bold numbers represent significant differences with FDR<0.05.   Gene  LJ vs K (control)  LJ vs F (control) ID Protein name Log2CPM 15°C 20°C 22°C  15°C 20°C 22°C GSONMT00017097001 neural cell adhesion molecule 1-like isoform x2 ncam1 1.5 2.3 0.2 0.9  4.0 3.6 4.7 GSONMT00034947001 olfactomedin-like protein 3a-like olfml3a 5.1 0.0 0.6 0.6  2.3 1.6 2.4 GSONMT00021084001 lysozyme c ii precursor lyz2 4.0 –1.0 –1.4 –0.3  –2.4 –3.1 –1.9 GSONMT00021082001 af321519_1lysozyme variant  4.4 –2.1 –1.7 –0.6  –3.1 –3.4 –2.5 GSONMT00072650001 hepatic lectin-like isoform x3 hlx3 2.9 –0.7 –0.7 –0.1  –1.4 –1.4 –1.6 GSONMT00063156001 leucine-rich repeat and immunoglobulin-like domain-containing nogo receptor-interacting protein 1 lingo1 5.5 2.1 1.4 0.8  0.6 1.0 1.3 GSONMT00069458001 retrovirus-related pol polyprotein from transposon pol 2.6 2.1 1.8 0.3  2.2 2.2 0.4 GSONMT00082336001 prosaposin receptor gpr37-like gpr37 4.6 1.9 1.2 0.6  0.5 0.6 0.9 GSONMT00082194001 repressor of rna polymerase iii transcription maf1 homolog maf1 5.7 1.3 0.9 0.5  –0.1 0.0 1.0 GSONMT00030695001 ATP-dependent RNA helicase ddx3x-like isoform x1 ddx3x 4.1 –1.8 –1.3 –0.2  –0.1 –0.7 –0.9 GSONMT00058078001 heterogeneous nuclear ribonucleoprotein a0  4.8 –1.6 –1.2 0.0  0.3 –0.3 –0.4 GSONMT00012629001 hemoglobin subunit alpha-1 hba1 7.8 1.1 1.6 1.9  2.0 3.1 1.6 GSONMT00058133001 hemoglobin subunit beta-1 hbb1 8.1 0.9 1.7 2.0  2.0 3.0 1.6 GSONMT00023044001 unnamed protein product  8.1 –1.0 0.4 1.3  0.8 0.8 2.0 GSONMT00040798001 an1-type zinc finger protein 5 zfand5 7.4 0.1 0.1 1.2  0.0 0.0 1.3 GSONMT00067393001 vesicle-associated membrane protein 5-like vamp5 6.7 0.4 –0.2 1.5  0.0 –0.8 1.3 GSONMT00048302001 cysteine mitochondrial  7.4 0.1 0.4 0.8  0.4 0.4 0.8 GSONMT00017832001 immediate early response gene 5 protein ieg5 7.1 0.3 0.4 0.9  0.5 0.7 1.3 GSONMT00028770001 v-maf musculoaponeurotic fibrosarcoma oncogene-like vmaf 7.2 0.0 0.1 0.7  –0.1 –0.3 0.9 GSONMT00013076001 glucose 6 phosphatase g6pc 4.9 3.3 1.5 2.3  0.8 0.2 2.9 GSONMT00074714001 myosin-binding protein fast-type-like isoform x1 mybp 4.3 0.7 1.0 1.0  0.3 0.5 0.9 149    Gene  LJ vs K (control)  LJ vs F (control) ID Protein name Log2CPM 15°C 20°C 22°C  15°C 20°C 22°C GSONMT00030514001 heat shock protein beta-7-like hspb7 8.3 –0.7 –0.7 –1.3   0.0 –0.3 –1.2  GSONMT00075529001 alpha cardiac-like myosin heavy chain mhc 6.9 0.1 0.2 –0.9   0.6 0.1 –0.9  GSONMT00016489001 zinc finger protein 106  6.8 0.0 –0.3 –1.8   –0.1 0.1 –1.0  GSONMT00039290001 protein phosphatase 1 regulatory subunit 12a isoform x3 ppp1r12a 6.3 –0.3 –0.5 –1.0   –0.5 –0.4 –0.7  GSONMT00053924001 fructose-bisphosphate aldolase a aldoa 6.4 0.2 0.0 –1.1   0.3 0.0 –0.8  GSONMT00077490001 prolyl 4-hydroxylase subunit alpha-1 isoform x1 p4ha1 6.5 –0.6 –1.1 –1.2   0.0  0.6 –1.2  GSONMT00056920001 hsp90 co-chaperone cdc37 cdc37 1.2 –0.9 –0.3 –1.7  0.6 1.1 −1.9 GSONMT00074547001 potassium voltage-gated channel subfamily h member 7 kcnh7 1.1 0.5 −0.1 −1.8  –0.4 –0.2 −1.5   150    Figure 4.1 Correction of maximum heart rate for body mass using four scaling exponents.  Maximum heart rate (mean±s.e.m.) in panel A is uncorrected.  Maximum heart rate is corrected to body mass (Mb) of one gram using scaling exponent of −0.10 (B), −0.15 (C), −0.20 (D) and −0.25 (E). Little Jacks (LJ): ; K×LJ: ; Keithley (K): ; Fawn: .  151   Figure 4.2 Percentage of fish showing arrhythmia during acute warming for each redband trout population. Populations were from desert climate (Little Jacks Creek, ), montane climate (Keithley Creek, ; Fawn Creek, ) and a hybrid between Keithley and Little Jacks ( ). Intersection between dash line and color line indicates the median arrhythmia temperature (50% fish showed arrhythmia) for each population, which was between 23 and 24.5ºC. Note: No fish showed arrhythmia at 15 or 21ºC.   152   Up-regulated  Down-regulated Figure 4.3 Number of significantly regulated transcripts during acute warming in redband trout. Significant levels were controlled for false discovery rate at 0.05.     Figure 4.4 Genes that are most significantly up- and down- regulated during acute warming.  Expression levels are natural log transformed count per million (CPM). A-H: up-regulated genes (FDR<0.05); I-K: down-regulated genes (FDR<0.05). Color points and dashed lines denote different populations (Little Jacks: ; hybrid between Keithley and Little Jacks: ; Keithley: ; Fawn: ). Gene expression for fish beyond 21°C are summarized as one variable. Black points and lines are the overall mean. All data are in mean ± s.e.m.  153    Activation (direct)   Activation (indirect) +p Phosphorylation  Inhibition   Ion transport    Figure 4.5 Pathways of significantly regulated genes in cardiac myocyte functions.  Red color represents up-regulation, light blue represents down-regulation. This diagram is modified from KEGG pathways (map04260, map05412, map05414, map05410, map04261, map04810, map04020). Expression of colored genes are summarized in Appendix C.4.     154     Reaction step (single)   Reaction step (multiple) Figure 4.6 Pathway of significantly regulated genes in glycolysis and citric acid cycle. This diagram is modified from KEGG pathways (map00020, map00620, map00052). Red color represents up-regulation, light blue represents down-regulation. The expression of genes are summarized in Appendix C.4.    155      Figure 4.7 Number of significantly regulated transcripts for each population during acute warming.   A: all significantly regulated transcripts (FDR<0.05); B: transcripts with over four fold changes (FDR<0.05). K×LJ: hybrid between Keithley and Little Jacks.        Up-regulated  Down-regulated Figure 4.8 Number of enriched gene ontology terms in enrichment analysis. Specific pathways are presented in Table 4.5 and Table 4.6. Note: Keithley is not included for analysis because of the less number of significantly regulated transcripts.   156   Figure 4.9 Correlation between the number of overall differentially expressed (DE) genes and FST (A) as well as thermal regimes (B).  Note: K×LJ was included in panel A.   A B 157    Figure 4.10 Gene expression that significantly associated with cardiac arrhythmia temperature. A: Average variable importance for the top 100 genes in ten repeats of Random Forest analyses; B: Heat map depicting expression for the top 30 genes (left side of red dashed line in Panel A). Each row represents one gene, colors represent the expression levels (blue represents low level expression, red represents high level expression). Each column represents an individual fish in the order from the lowest arrhythmia temperature (TAR) to the highest TAR, as indicated by the numbers (unit in °C) below columns. Numbers in the brackets before the gene names are the R squared values from the linear regression of expression versus TAR.  Side bars above columns represent the grouping of fish based on the TAR above or below 24°C.  158  Chapter 5: Discussion and conclusions The overall objective of my thesis research was to examine the mechanisms of thermal adaptation in rainbow trout (Oncorhynchus mykiss). I hypothesized that populations from diverse thermal environments would divergent in thermal-related traits at whole organismal, physiological, cellular and genetic levels that best suit the local thermal regimes. My thesis characterized thermal tolerance using comprehensive measurements, including critical thermal maximum (CTMAX), absolute aerobic scope (AAS) and maximum heart rate (fH,max). Populations of O. mykiss had different CTMAX values, which were correlated with the peak summer water temperatures in their respective habitats. Optimum thermal window was significantly broader in the population from the warmest habitat. However, a broad optimum thermal window is associated with a cost of the peak aerobic performances (lower AAS). My results demonstrated an oxygen limitation during acute warming because AAS abruptly decreased at temperatures above 21°C for all populations, except for a desert redband trout population from the warmest environment. Furthermore, a capacity limitation in the cardiac performance also exists during acute warming as shown by the Arrhenius breakpoint and dysrhythmia of fH,max. In addition, intraspecific patterns of cellular regulation in heart were studied and the differentially regulated genes suggested the mechanistic limitations of cardiac capacity. Finally, using population genetics and genome-wide association study (GWAS) approaches, I identified candidate genes that were significantly associated with the phenotypic differences. 159  5.1 Study of thermal adaptation to local environments Local adaptation is an important process that allows populations to thrive in altered environments. Many species have a wide geographical distribution and comprise multiple reproductively isolated populations. Different environments and limited gene flow between habitats drive local genetic changes and adaptation through the selection of specific phenotypes. To survive in a particular thermal regime, an animal needs to be able to efficiently perform activities at the most frequently encountered temperatures, to tolerate the less frequent sub-optimum temperatures, and to resist the occasional extreme temperatures (Fry, 1947; Hofmann and Todgham, 2010). In natural conditions, the most frequent temperatures that fish encounter are generally within the optimum thermal window (optimum zone), hence maximum performance can be achieved to proceed through each life stage. When exposed for a short period to temperatures above or below optimum ranges (pejus zone), a fish can survive but has an impaired capacity for growth and reproduction, which if persists may affect fitness. Occasionally, temperature may reach a critical level (the resistant zone), where only temporary survival is possible. To thrive in a local thermal regime, animals need to have corresponding thermal zones to ensure fitness. In this thesis, I used a redband trout (O. mykiss gairdneri) population from warm desert (Idaho, USA) and a domesticated rainbow trout strain (O. mykiss irideus) that has been raised in a thermally challenging environment (Western Australia, Australia) to explore the potential adaptive consequences of the global warming for the species O. mykiss. Two redband trout populations from cool montane climate were used as control groups for the desert fish. There are advantages and disadvantages in using wild and domesticated populations to study the genetic mechanism of thermal adaptation (Angilletta, 2009). Using 160  wild populations can help identify patterns and mechanisms in an abiotic-biotic interacting environment, which reflects the actual requirement for adaptation under natural conditions. However, other environmental factors could also impose selection pressures and give misleading associations for thermal adaptation. Among the three wild redband trout populations in my thesis, the divergence in summer peak temperature is a relatively recent event (since the last glaciation) and has been one of the main drivers in local adaptation. Other factors that differed are elevation, water flow, prey and predators. In contrast to the natural process of selection, domestication is a selection process on reared animals in artificially controlled environments and consequently it is possible to manipulate temperature as the major selecting factor. Thus, changes in phenotypes and genetics in domesticated animals in thermally controlled environment can be attributed, to a larger degree, to thermal selection. A big difficulty of artificial selection is the large scale of experiment design that requires space, time and cost. In my thesis, I was able to use a rainbow trout strain that has been reared in a hatchery, where temperature is the main lethal factor and has caused high mortality, for over 19 generations. Unfortunately, a control group from a deep reservoir was unavailable at the time of experiment, despite significant fishing efforts. Nevertheless, my thesis used both natural and domesticated populations to provide a more thorough understanding of thermal adaptation. Thermal tolerance is affected by genetic × environment (G×E) interactions. In my thesis, I have raised fish since fry stage in a common garden environment at a constant water temperature of 15°C. Hence, phenotypic differences can be mostly attributed to genetics, but the possible contribution of early environmental and maternal effects cannot be discounted.  Also, a hybrid redband trout strain was produced by crossing montane females with desert 161  males. If additive genetic components are the main contributor to phenotypes, the hybrid should have intermediate performance between the two parental populations. Also, heterosis should not be a surprising result if the inferior alleles of some functionally important genes in parents are complemented in offspring, resulting in superior performances compared to parents.  While I have made the intraspecific comparison within this thesis, interpreting the comparison with existing literature needs caution because animals had very different thermal histories among studies.  5.2 Thermal adaptation of CTMAX Among the O. mykiss populations in this thesis, Little Jacks redband trout and PFRC domesticated rainbow trout live in warm environments. However, I predicted that Little Jacks fish would have better thermal performance than the PFRC fish because of a larger gene pool for the natural selection to act on. Thermal selection on PFRC rainbow trout has been restricted to the standing genetic variations in founder populations. Indeed, CTMAX was significantly higher in Little Jacks than the PFRC fish (Figure 5.1). Among the redband trout populations that I studied, CTMAX was positively correlated with the maximum habitat temperature. Despite the fact that the logged water temperatures in this thesis research was only for one year and cannot represent the entire historical conditions, the temperatures difference between habitats should persist over long time frames.  According to the existing literature for Little Jacks and my personal communications with PFRC staff, both habitats have experienced temperatures over 29°C. The reason that both populations have survived must be via either phenotypic plasticity or behavioral thermoregulation by seeking cool refugia (Ebersole et al., 2001). In summer conditions, all fish in the wild are acclimated to temperatures over 15°C (the acclimation temperature I used 162  for my thesis) and CTMAX in nature may be higher than the results in this thesis due to the phenotypic plasticity arising from higher acclimation temperatures. Redband trout occurrence in summer are found to be more abundant under shade in desert streams (Zoellick, 2004) and deep pools in montane streams (Muhlfeld et al., 2001), which is probably a result of behavioral thermoregulation. In PFRC hatchery (Western Australia), however, seeking cool refugia is not an option because of their shallow concrete rearing pond. Instead, refrigerators have been used to chill the water so that extirpation events could be avoided. Because CTMAX is a simple and reliable protocol, it has been extensively used to measure upper thermal limits in O. mykiss. CTMAX results in this thesis are consistent with previous studies that used similar acclimation temperatures (Strange et al., 1993; Currie et al., 1998; Myrick and Cech, 2005; Ineno et al., 2005; Galbreath et al., 2006). On average, CTMAX values of redband trout populations from Idaho in this thesis (28.4-29.8°C) were similar to those redband trout from Oregon, USA, that were acclimated at 14°C (29-29.7°C) (Rodnick et al., 2004). PFRC rainbow trout had lower average CTMAX than a thermally selected rainbow trout strain in Japan (29.0°C versus 29.7°C) (Ineno et al., 2005). Taking these data together, CTMAX appears to be an intraspecifically conserved trait with limited variance. Actually, an evolutionary constraint in the critical temperatures has been suggested in many species (Hoffmann et al., 2013). For all the fish that I measured in this thesis, individual CTMAX was between 28.1-29.7°C (375 fish) in PFRC rainbow trout and 28.3-30°C (63 fish) in redband trout. Some concerns are raised here for the future global warming, regarding the closeness between CTMAX and maximum summer temperatures, and the relatively low adaptability of CTMAX. It is possible that phenotypic plasticity and cooler 163  refugia in creeks will become the major options for fishes that currently live in warm environments to survive future increases in extreme temperatures. Although CTMAX did not vary appreciably, notable differences existed for AAS and the shape of the Fry Curves.  Therefore, thermal adaptation is probably more evident in physiological performance ranges rather than the critical limits, which has been previously shown in Fraser River sockeye salmon (Oncorhynchus nerka) (Lee et al., 2003b; Farrell et al., 2008; Eliason et al., 2011).  5.3 Thermal adaptation of Fry Curve Some important thermal indices can be generated from the Fry Curve: Topt (maximum AAS), optimum thermal window (> 90% of peak AAS), pejus or tolerance zone (< 90% but >10% of peak AAS), resistant zone (<10% of peak AAS) and Tcrit (AAS=0). The levels of 90% and 10% are arbitrarily set for the purposes of intraspecific comparison. In this thesis, because RMR and MMR were only measured at six test temperatures, I only obtained partial AAS data from the pejus zone and none from the resistant zone. My results showed that, compared to the CTMAX, optimum temperatures had much larger variance (Figure 5.1).  According to the OCLTT hypothesis, a broader optimum thermal window for AAS suggests the ability to maintain best performance across a wide range of temperatures.  I found that desert redband trout had a wider optimum thermal window which extended to a higher temperature than the montane redband trout populations. With heating from 21°C to 24°C, AAS in all montane populations abruptly decreased, and this decrease suggests an O2 limitation in these populations. Little Jacks, however, maintained a constant AAS between 15-24°C, which could allow Little Jacks to perform activities such as growth and reproduction at broader temperatures. On the other hand, a wide optimum thermal window 164  and an evolutionary conserved CTMAX result in a much narrower sub-optimal zone, which reduces their safety margin should there be future warming of their habitat. For populations that currently live in cooler summer environments such as Keithley and Fawn, there is a greater thermal buffer before CTMAX is reached. While this thesis only examined parameter differences at the upper thermal range of the Fry Curve, the lower thermal range will also be interesting to examine (e.g. see Yau and Taylor, (2014)) because winter temperatures also differ significantly and are apparently more extreme in terms of the duration of freezing temperatures, especially between Little Jacks and Fawn.  In this study, Little Jacks redband trout and PFRC rainbow trout represent two warm adapted populations and I predicted they should share some characteristics in their Fry Curves. Interestingly, Fry Curves significantly differed between Little Jacks and PFRC fish (Figure 5.2). While there are many factors affecting the evolution of Fry Curve, some insight can be obtained by comparing the thermal regime and condition of rearing.  While the selection for higher AAS is possible, natural condition seems prefer phenotypes of lower AAS with broad optimum thermal window.  Thus, the thermal adaptation of Little Jacks to desert environment may have involved tradeoffs between life history performances (e.g. growth and exercise) and thermal tolerance (Angilletta et al., 2003). The mechanistic basis of the tradeoff between peak AAS and optimum thermal window is not clear. The high overall AAS in PFRC rainbow trout is likely to be a combined result of selection on thermal tolerance and growth in hatchery culture (Molony et al., 2004). In the hatchery, fish are fed to satiation and growth is the most energy consuming activity. Fish after a large meal will significantly increase metabolic rate (i.e. specific dynamic action), which uses large portion of the AAS.  Although PFRC had narrow optimum range, the hatchery environment is much 165  simpler than wild streams. The absence of predators, static water flow and sufficient food supply may have decreased the metabolic demand for other activities during warming, thus preserving AAS for survival during extreme temperatures. When Fry Curves are compared with other fish species in previous studies (Figure 5.3), large inter- and intra-specific variances exist (Farrell, 2009). Despite that, there are clear differences between temperate salmonids and eurythermal fishes (Fry, 1947; Fry, 1948; Fry and Hart, 1948; Lee et al., 2003b). Within salmonids, the most obvious differences are in the peak AAS and the width of optimum thermal window. It also seems that fish with lower peak AAS tends to have broad optimum thermal windows. 5.4 Role of fH.max in limiting AAS The cause of oxygen limitation at temperatures above Tpej has been linked to a capacity limitation in the cardiorespiratory system. More specifically, fH has been proposed to be the weak link along the oxygen cascade pathway. According to the Fick Equation for oxygen consumption ( ?̇?𝑂2), fH has a positively relationship with  ?̇?𝑂2. In this thesis, I measured fH.max using pharmacological stimulation as a proxy for the fH during a MMR state (Figure 5.4) and predicted a positive relationship between post-exhaustion MMR and fH.max. Contrary to my predictions, Little Jacks with the highest fH.max had the lowest AAS. PFRC rainbow trout and Keithley redband trout had higher aerobic scopes than other populations, but their fH,max was the lowest. This result implies that other parameters in Fick Equation that I did not measure (such as stroke volume and hematological variables) could be different. This thesis is the first study that compared the intraspecific fH,max in O. mykiss. The only existing literature on O. mykiss fH,max was for a coastal population from Blackwater Creek, British Columbia (Canada) (Verhille et al., 2013), which has a colder climate than the 166  redband trout habitat in Idaho, USA. The British Columbia rainbow trout had lower fH,max than the PFRC rainbow trout and redband trout, which is a consistent result as the inter-population comparisons I made in redband trout. Despite fH.max not being a reliable predictor for MMR across redband trout populations, within a population fH.max could still limit AAS during acute warming because direct associations can be made between the rate transition temperatures for fH.max and MMR. I found in all populations, except Little Jacks, that TAB for fH.max is close to Topt for MMR. Casselman et al., (2012) showed that TAB for fH.max was the same as Topt for AAS, suggesting the rising phase of the Fry Curve is interrupted by the first rate transition in fH.max. However, this relationship may not be that tight according to the results in this thesis and other species (Ferreira et al., 2014), where TAB also occurs at lower and upper Tpej (i.e. TAB occurs within a range between Topt and Tpej of Fry Curve). In rainbow trout, MMR seems have a rate transition temperature before its plateau, which is probably not related to fH.max, and caused the plateau of Fry Curve. The reason for this rate transition of the rising phase MMR is not clear. Therefore, it is temporarily concluded here, at least for O. mykiss, that the rate transition of the rising phase MMR caused the plateau of Fry Curve, while the plateau of MMR (≈ TAB for fH.max) caused the decreasing phase of Fry Curve. It should be noted that the aerobic scope was measured every 3°C intervals and fH.max every 1°C, which made the precise correlation between AAS and fH.max difficult. Future studies need to measure RMR and MMR at more temperatures to provide better associations between Fry Curve, RMR, MMR and fH.max. At temperatures above optimum thermal window, fH.max either became arrhythmic without reaching a plateau, or became arrhythmic followed by the plateau period. Therefore, 167  an upper limitation in cardiac function must have been reached. On average, fH.max became arrhythmic at temperatures <4°C below CTMAX for all populations, and no individual had regular fH.max at temperatures at or above CTMAX. If CTMAX sets the upper limit for survival, failure in fH.max clearly indicates a capacity limitation.  In this study, the fH,max was measured without autonomic control to assess the maximum capacity of heart rate, but this response cannot be directly applied to fishes in natural environments.  Conscious fish in natural condition can theoretically regulate heart rate during warming, e.g. increased vagal tone could decrease fH to protect the heart from dysrhythmia. Nevertheless, in vivo studies with salmonids show that vagal tone is greatly reduced when a resting fish is warmed (Wood et al., 1979), and dysrhythmia has been observed during acute warming of conscious sockeye salmon (Steinhausen et al., 2008; Eliason et al., 2013) and chinook salmon (Clark et al., 2008).  Interestingly, sockeye salmon decreased fH when they began to swim at 26°C, rather than the normal increase observed at lower temperatures (Eliason et al., 2013), but it was not determined why this decrease occurred. 5.5 Cellular response to acute warming Gene expression during acute warming is an important regulatory strategy and was shown to be associated with the rate transition temperatures for fH.max in this thesis, particularly just prior to TAR, when there were large quantitative changes in gene expression. I used RNA sequencing to examine the gene expression in locally adapted redband trout populations that were acutely warmed from rearing (15°C), to TAB (~20°C) and critical temperatures (TAR). I observed that a warming rate of 1°C every six minutes significantly regulated large number of genes in multiple pathways, including heat shock response, cardiac 168  myocyte functions and energy metabolism. Compared to the control (i.e. the rearing temperature of 15°C), only small number of significantly regulated genes were triggered at 20°C in all populations. Taking HSPs as an example, among all the HSPs that I have identified in this thesis, none were significantly regulated at 20°C. The heat shock response is a fast process and occurs within minutes. In goby (Gillichthys mirabilis), hsp70 mRNA was significant regulated within 20 min after the heat shock (Buckley and Hofmann, 2004).  In my experiments, it took 30 minutes to increase from 15°C to 20°C.  Assuming the number of significantly regulated genes reflects the demand for functional proteins, I suggest that it was not until at temperatures above TAB (~20°C) that redband trout start manufacturing new proteins to deal with thermal stresses. This cellular observation is consistent with the AAS and fH.max results. All populations maintained >90% of AAS at 20°C and fH.max was increasing with a Q10 of ~1.8 at 20°C. Within each population, different numbers and categories of genes were differentially expressed suggesting strong intraspecific patterns of gene regulation. I predicted that the number of thermally induced genes would be correlated with climate, i.e. fish from warm habitat will trigger more differentially regulated genes. In contrast to my predictions, two populations from the montane climate had both the most (Fawn) and least (Keithley) significantly regulated genes. It was a surprising result given the genetic similarity between Keithley and Fawn. This result suggests that cellular response to temperature may be affected by other factors such as the abundance of the constitutively expressed mRNA and protein as well as the qualitative changes within them (e.g. amino acid changes in rate-limiting enzymes (Somero, 2003)). Indeed, differentially expressed (DE) genes between populations was correlated with the genetic distance between populations. Among the natural 169  populations, two montane populations had the least number of DE genes, while the desert population and the colder montane population had the most DE genes. The association between the number of DE genes and the difference in thermal regimes would be clearer if more natural populations were available for analysis. I only measured gene expression under acute warming conditions (6°C min−1), therefore, the results here do not imply a pattern for chronic warming or acclimation conditions.  In cardiac cells, the most immediate response to acute warming is the increase in work as reflected in the increase in fH (Chapter 4), which also means an increase in the frequency of action potentials and muscle contractions, i.e. cellular homeostasis response (CHR). My results suggest that cellular responses are regulated to support these functional requirement. For example, the beta-2 adrenergic receptor (β2ar) gene, which stimulates the cyclic-AMP signaling pathway, was significantly up-regulated. In Fraser River sockeye salmon, ventricular β-adrenergic receptor density has been associated with a broad AAS (Eliason et al., 2011). The downstream effectors of β2AR include multiple voltage-gated ion channels. However, only a potassium channel gene (kcne1) was up regulated. I also observed the over expression of genes that regulate intracellular Ca2+ concentration, which directly determines the force of cardiac contraction and must be tightly regulated in cardiomyocytes. However, voltage gated Ca2+ channel genes were not found to be regulated. It is possible that the regulation of Ca2+ cycling genes is a slow response (Korajoki and Vornanen, 2012; Jayasundara et al., 2013) and do not respond to acute thermal change. Both myosin and titin genes were down-regulated, indicating an acute regulation of contraction properties and muscle cell growth. Warm-acclimation results in a smaller ventricle in rainbow trout (Farrell et al., 1988b; Klaiman et al., 2011). Genes in some important metabolic pathways were 170  significantly up-regulated, including the energy related pathways such as citrate cycle, glycolysis / gluconeogenesis and oxidative phosphorylation. Also, I found several significantly regulated genes in hypoxia inducible factor -1 (HIF-1) pathway. Acute warming to high temperatures affected protein synthesis pathways. Often there are two strategies for cells to deal with damaged proteins. One is through the protein rescue with the facilitation of HSPs. The other is degradation through ubiquitin-proteasome pathway and lysosomal proteolysis. In this thesis, the average TAR was around 24°C, which should not cause severe damage to proteins. Among all the up-regulated genes during acute warming in this study, hsp30 had the most significant fold changes. Other heat shock protein genes were also among the highest (dnajb1, hsp90, hsp70). Gene expression-traits association analyses revealed genes that may be related to thermal induced cardiac arrhythmia, which further limits the whole animal temperature tolerance. However, without a clear relationship between mRNA and protein abundances, results in this study should be interpreted cautiously until verified at the protein level. Nevertheless, this study provided some candidate genes to explore the molecular-function relationships in thermal adaptation. 5.6 Genetic changes in thermal adaptation Earlier sections of this chapter have discussed the phenotypic differences at different organismal levels among populations from different thermal regimes. To ensure that the phenotypic differences are indeed the result of thermal adaptation, genetic mechanisms of population differentiation also needs to be studied. In this thesis, I discovered 5903 genome-wide single-nucleotide polymorphism (SNP) markers in redband trout populations. Using 171  these SNP markers, I applied population genetic approaches and genome-wide association study (GWAS) to identify genes that are putatively related to thermal adaptation. Previously, Narum et al., (2010) showed that genetic distance was larger between Little Jacks and Fawn when compare to that between Fawn and Keithley. I confirmed this result by using a larger number of SNP markers.  Narum et al., (2010) also identified markers that are significantly associated with habitat air temperatures using twelve redband trout populations from desert and montane environments. In this thesis, I identified twenty-one “outlier” loci that were potentially under positive thermal selection. For most of the outlier loci, the similarly minor allele frequencies were similar between the two montane populations, suggesting a divergence between desert and montane redband trout and the potential for them to be associated with thermal adaptation. To further assess whether these outlier markers are under positive thermal selection rather than due to other demographic factors (founder and bottleneck effect), I compared the CTMAX among genotypes of each outlier marker and found significant differences, suggesting that some genotypes have higher CTMAX than the other. Examining the flanking regions (15 kb) of the SNPs marker along genome, I identified candidate genes for thermal selection. Some of these genes are well known for their cellular roles, but most genes are poorly studied in fishes. Dnaj homolog subfamily b member 6 (DNAJB6) belongs to the heat shock protein 40 kDa family and has been found to be a superior suppressor for the misfolded protein aggregation (Hageman et al., 2010). L-lactate dehydrogenase B-A chain like (LDH-B) plays an important role in re-oxidizing NADH to NAD+ in low O2 environment. In killifish (Fundulus heteroclitus), populations from cold climate had significantly higher cardiac metabolism than those from warm climate 172  and 87% of the variance was explained by three enzymes, including LDH (Podrabsky et al., 2000).  Protein products of the camkk2 gene (Calcium/calmodulin-dependent protein kinase kinase) phosphorylate AMP-activated protein kinase (AMPK), which regulate multiple important cellular functions such as glucose uptake, fatty acids oxidation and mitochondria (Bergeron et al., 1999; Ojuka, 2004; Thomson et al., 2007).  Other genes that may be interesting to further study are the potassium channel genes (srkc and kcnc1) and transient receptor potential cation channel gene (TRPA1). TRPA1 is a heat sensor protein expressed in neuron cells (Viswanath et al., 2003; Saito et al., 2012) and is a member of the TRPs protein group that plays important role in thermoregulation (Seebacher, 2009).  The roles of these genes are not well studied in salmonids but are worth further functional verification. My data also suggested a candidate locus for cold adaptation. Fawn Creek has much lower spring temperatures, colder summer environment and a much longer period of winter at freezing temperature than Little Jacks Creek and Keithley Creek. I identified one locus (11282_25) that has similar minor allele frequency between Little Jacks and Keithley, in other words, divergent from Fawn. Both Little Jacks and Keithley had fixed A-allele, while Fawn had both A- and G-alleles. GG genotype had significantly lower CTMAX than the AA genotype. Previous study observed a unidirectional change in CTMIN and CTMAX (Fangue et al., 2006), therefore, I predict the GG genotype also has lower CTMIN. Four genes (rad21, srkc, tc3a and trappc9) were within 10kb of the 11282_25 marker and are worth further study for their roles in lower thermal tolerance.  Thermal adaptation relies on the genetic architecture of adaptive traits. Critical temperatures clearly have a genetic basis (Meffe et al., 1995; Doyle et al., 2011) and are able to respond rapidly to selection (Ihssen, 1986; Barrett et al., 2011). Genetic mapping of 173  thermal tolerance has been mostly studied in rainbow trout using microsatellite markers (Jackson et al., 1998; Danzmann et al., 1999). However, the further pinpointing of mapped loci to actual responsible genes has been limited (Somorjai et al., 2003; Quinn et al., 2011), largely due to the low resolution of microsatellite markers used in most previous mapping studies. In this thesis, I conducted GWAS analysis in redband trout and identified candidate SNP markers for CTMAX. A total of twelve loci were significantly associated with individual CTMAX. A similar GWAS study for mortality at high temperature in redband trout identified different groups of significant genes (Narum et al., 2013b), suggesting the complex genetic basis for different aspects of thermal tolerance. Along with the GWAS analysis for other important traits in rainbow trout (Hecht et al., 2013; Campbell et al., 2014; Palti et al., 2015), a more comprehensive understanding of the evolutionary mechanisms can be expected. Contrary to my expectations, no common locus was identified both by outlier tests and GWAS. Nevertheless, results in this study identified several candidate SNP markers for thermal tolerance that may be under selection now and under future global warming scenarios. 5.7 Conservation and management implications The 21st century is predicted to have a rapid temperature increase. For instance, the Columbia River, USA, is predicted to be 2.3-2.9°C warmer than the 1990s by 2045, and 4.5°C warmer by 2095 (Hamlet and Lettenmaier, 1999).  Thus, it has been questioned whether species, especially the freshwater ectotherms with limited migratory opportunity, will be able to adapt.  To address this question, this thesis studied populations that currently live in warm environments to examine the adaptive potential for O. mykiss and other conspecific populations. Although my results suggested that O. mykiss has some potential to 174  adapt to the predicted global warming trends, populations that currently live in warm environments may have a narrow safety margin and therefore need special conservation attention. Thus, phenotypic plasticity and behavioral thermoregulation, such as seeking thermal refugia (e.g. deep pools, cool springs and upwelling groundwater), are critically important in determining the tolerance of organisms experiencing environments with extreme temperatures. A good example is the effect of habitat damage and enhancement to the occurrence of cutthroat trout (Oncorhynchus clarkii) in the East Creek, British Columbia (Young et al., 1999). In 1974, logging removed the riparian canopy of East Creek and caused the increase of stream temperature to >30°C, which exceed the tolerable thermal range of cutthroat trout (McMahon et al., 2006; Bear et al., 2007; Underwood et al., 2012). Furthermore, the clearance of large woody debris likely removed the thermal refugia for cutthroat trout. As a result, the cutthroat trout population size greatly declined. The later habitat enhancement action between 1983 and 1997 brought down the stream temperature and allowed the recovery of the cutthroat trout population, suggesting the importance of thermal regime and refugia to a fish population. 5.8 Future directions This thesis explored some aspects of thermal adaptation in the whole organismal performance, physiological processes, genomic and transcriptome levels, and identified some directions for future studies that will deepen our understanding in this field.  Since thermal performance and tolerance are plastic and often change with thermal history (G×E interaction), I acclimated all fish in controlled environments at a constant rearing temperature of 15°C, so that the phenotypic differences could be attributed to a large extent to genetics. However, phenotypic plasticity is important and fish in natural conditions 175  use it to adjust for seasonal and other environmental fluctuations. Using different acclimation temperatures for rearing would provide an opportunity to estimate the intraspecific variations in phenotypic plasticity and improve our understanding of thermal adaptation in the context of G×E interactions. Given the physiological and cellular data when fish were acutely warmed to 20°C, it seems that acclimation to 20°C would be a good starting place for future studies in all these populations. To be even more cautious, trans-generation effect and epigenetics need to be considered, which requires common garden rearing conditions across generations.  Future study should also consider the complexity of natural conditions, where temperature is rarely the only stressor. Other biotic (e.g. competition, disease and predation) and abiotic (e.g. pH and hypoxia) factors also affect thermal performance. Therefore, thermal tolerance in the presence of other stressors will provide the understanding of thermal adaptation under an ecological context. I measured the AAS and found intraspecific differences in the width and height of Fry Curve. However, how AAS is partitioned and becomes a limiting factor for performance such as growth and gonad development is unknown. Future study on thermal performance at different development stages will provide a more in-depth assessment of OCLTT. I also suggest future study should measure aerobic scope at more temperatures to provide a higher resolution in estimating the rate transition temperatures for AAS, especially for more “eurythmic” fish.  For fH, I only measured the thermal performance of fH,max that represents the upper limits of fH. Future study should measure the thermal performance of fH at various levels between basal and maximum to provide a better understanding in the limiting role of fH and its regulation.  Furthermore, thermal performances at the lower thermal ranges also need to be examined as a compliment of the present study. 176  The outlier analysis and GWAS performed in this thesis had several limitations. First, having used a relatively small sample size (N = 14-17 per population), this study likely underestimated the association of some SNP markers with thermal tolerance. Future studies should consider a larger sample size (e.g. N>200), ideally using both a QTL approach with a family design and an association mapping approach with unrelated animals. Such an increase in sample size of course needs larger experiment design and higher cost of analysis, which, depending on the availability of resources, may require targeting fewer questions but analyzing those selected more deeply. Second, in rainbow trout, the apparent linkage between syntenic loci significantly decays when their distance exceeds 2 cM due to recombination. Therefore, it has been estimated that approximately at least 1,500 markers are needed to cover all the linkage groups in association analyses for rainbow trout (Rexroad and Vallejo, 2009). While the number of SNP markers in my study is almost quadruple the suggested number, fine gene mapping remains difficult because of large gaps between markers in many cases and blocks of linkage disequilibrium may be much smaller than 2 cM in some areas of the genome. In addition, linkage disequilibrium can be population specific in many cases, suggesting analyses need to be measured separately for each redband trout populations.  Third, while this study used strict standards to be conservative in SNPs discovery, future studies should consider increasing the genomic coverage of SNPs (With RAD-sequencing, this can be accomplished by cutting genomic DNA into more pieces during library preparation using either a combination of restriction enzymes or an enzyme that recognizes shorter sequences). Furthermore, with a limited budget, sequencing pools of individuals (Pool-seq) may also be a good alternative to individual sequencing for population genetics (Schlötterer et al., 2014). Ultimately, whole genome sequencing of many 177  individuals from multiple natural populations will provide the best coverage and allow family structure to be known in nature, thus enhance analytical power for understanding the adaptive capacity and evolution of thermal tolerance in redband trout.  Such an approach is rapidly becoming feasible and cost effective due to remarkable advances in genomic technology and bioinformatic capabilities. Results in this thesis demonstrated genetic differentiation among populations and provided several putative thermal adaptive loci. In nature, thermal adaptation is likely a common phenomenon and occurs in many populations. By studying more locally adapted and reproductively isolated populations, genomic regions that are repeatedly favored in multiple warm adapted populations (parallel thermal adaptation) will provide more compelling results. Future study can use a broader-scale experiment design to address this hypothesis. Landscape genetics (broad-scale among population surveys) also provide opportunities to identify potential markers for thermal adaptation (Dionne et al., 2008; Matala et al., 2011; Hecht et al., 2015). To identify common genetic signatures for thermal adaptation more broadly in O. mykiss, the PFRC rainbow trout and Little Jacks redband trout need to be studied in the context of native coastal rainbow trout. Although this thesis studied differential gene expression between temperatures and populations, the constitutive and induced changes at the protein level were not identified. Future study needs to use protein assays or proteomic analyses to assess these transcriptomic results. Taking together, mechanisms of thermal adaptation have been studied in O. mykiss using populations from diverse thermal conditions to provide valuable information for the conservation of threatened populations in a warming century. Adaptation to overcome sub-optimum thermal performance largely depends on environmental requirements, and more 178  likely will involve tradeoffs between performance and tolerance. Transcriptomic and genomic results demonstrated molecular and genetic variations among populations from different thermal regimes, suggesting complex mechanisms underlying thermal adaptation. The ultimate use of the results in this thesis would be in integrating ecology, physiology and genetics for a complete picture of thermal adaptation, which further can be applied to management measures for conservation.   179   Figure 5.1 Optimum and upper critical temperatures of O. mykiss populations.  Redband trout populations are from desert climate [Little Jacks (LJ)] and montane climate [Keithley (K) and Fawn (F)]. Pemberton Freshwater Research Centre (PFRC) rainbow trout is a thermal selected strain in Western Australia. Thermal indices include critical thermal maximum (CTMAX), pejus temperature (Tpej) and optimum temperature (Topt) for mass independent aerobic scope, Arrhenius breakpoint temperature (TAB) and cardiac arrhythmia temperature (TAR) for maximum heart rate.    180     Figure 5.2 RMR, MMR and AAS of O. mykiss populations. Panels (A), (C) and (E) are uncorrected value for RMR, MMR and AAS, respectively. (B), (D) and (F) are body mass (g−0.88 value. Pemberton Freshwater Research Center: , Redband trout from desert climate (Little Jacks: ) and montane climate (Keithley:  and Fawn ). 181   Figure 5.3 Fry Curves of several fish species. This figure is modified from Fig.2A in Farrell, (2009). Solid lines represent Fry Curves for rainbow trout (Oncorhynchus mykiss). Other dashed grey lines are sockeye salmon (Oncorhynchus nerka), coho salmon (Oncorhynchus kisutch) (Lee et al., 2003b), brown trout (Salmo trutta), lake trout (Salvelinus namaycush), brook trout (Salvelinus fontinalis) (Fry, 1948), goldfish (Carassius auratus) (Fry and Hart, 1948) and brown bullhead (Ameiurus nebulosus) (Fry, 1947). 182    Figure 5.4 Maximum heart rate of O. mykiss populations. (A): uncorrected value for fH,max. (B): corrected fH,max for body mass to one gram using scaling exponent of −0.1. Dash lines represent thermal ranges where fish showed cardiac arrhythmia.   183  References Adkison, M. D. (1995). Population differentiation in Pacific salmons: local adaptation genetic drift, or the environment? Can. J. Fish. Aquat. Sci. 52, 2762–2777. Aho, E. and Vornanen, M. (2001). Cold acclimation increases basal heart rate but decreases its thermal tolerance in rainbow trout (Oncorhynchus mykiss). J. Comp. Physiol. B 171, 173–179. Akukwe, B., Strizzi, L., Gonzales, M., Mancino, M., Watanabe, K., Hamada, S., Salomon, D. and Bianco, C. (2007). Human Cripto-1 is upregulated by hypoxia through hypoxia-inducible factor-1 α -dependent mechanism. Cancer Res. 67, 1375. Allendorf, F. W. and Utter, F. M. (1979). 8 - Population Genetics. In Fish Physiology, volume VIII: Bioenergetics and Growth Editor: Hoar, W. S. and Randall, D. J., pp. 407–454. Academic Press. Altimiras, J. and Larsen, E. (2000). Non-invasive recording of heart rate and ventilation rate in rainbow trout during rest and swimming. Fish go wireless! J. Fish Biol. 57, 197–209. Anders, S., McCarthy, D. J., Chen, Y., Okoniewski, M., Smyth, G. K., Huber, W. and Robinson, M. D. (2013). Count-based differential expression analysis of RNA sequencing data using R and Bioconductor. Nat. Protoc. 8, 1765–1786. Andersen, O., Wetten, O. F., De Rosa, M. C., Andre, C., Carelli Alinovi, C., Colafranceschi, M., Brix, O. and Colosimo, A. (2009). Haemoglobin polymorphisms affect the oxygen-binding properties in Atlantic cod populations. Proc. R. Soc. B 276, 833–841. Angilletta, M. (2009). Thermal adaptation: a theoretical and empirical synthesis. Oxford University Press. Angilletta, M. J., Wilson, R. S., Navas, C. A. and James, R. S. (2003). Tradeoffs and the evolution of thermal reaction norms. Trends Ecol. Evol. 18, 234–240. Antao, T., Lopes, A., Lopes, R. J., Beja-Pereira, A. and Luikart, G. (2008). LOSITAN: A workbench to detect molecular adaptation based on a Fst -outlier method. BMC Bioinforma. 9, 1–5. Anttila, K., Casselman, M. T., Schulte, P. M. and Farrell, A. P. (2013). Optimum temperature in juvenile salmonids: connecting subcellular indicators to tissue function and whole-organism thermal optimum. Physiol. Biochem. Zool. 86, 245–256. Anttila, K., Eliason, E. J., Kaukinen, K. H., Miller, K. M. and Farrell, A. P. (2014a). Facing warm temperatures during migration: cardiac mRNA responses of two adult Oncorhynchus nerka populations to warming and swimming challenges. J. Fish Biol. 84, 1439–1456. Anttila, K., Couturier, C. S., Overli, O., Johnsen, A., Marthinsen, G., Nilsson, G. E. and Farrell, A. P. (2014b). Atlantic salmon show capability for cardiac acclimation to warm temperatures. Nat. Commun. 5, 4252. Araneda, C., Neira, R., Lam, N. and Iturra, P. (2008). Salmonids. In Genome mapping and genomics in fishes and aquatic animals Editor: Kocher, T. and Kole, C., pp. 1–43. Berlin, Heidelberg: Springer. Armour, C. L. (1991). Guidance for evaluating and recommending temperature regimes to protect fish. Washington, D.C. 184  Badr, A., El-Sayed, M. F. and Vornanen, M. (2016). Effects of seasonal acclimatization on temperature-dependence of cardiac excitability in the roach, Rutilus rutilus. J. Exp. Biol. 1495–1504. Bagnyukova, T. V., Lushchak, O. V., Storey, K. B. and Lushchak, V. I. (2007). Oxidative stress and antioxidant defense responses by goldfish tissues to acute change of temperature from 3 to 23 oC. J. Therm. Biol. 32, 227–234. Bailey, R. M. (1955). Differential mortality from high temperature in a mixed population of fishes in southern Michigan. Ecology 36, 526–528. Baird, N. A., Etter, P. D., Atwood, T. S., Currey, M. C., Shiver, A. L., Lewis, Z. A., Selker, E. U., Cresko, W. A. and Johnson, E. A. (2008). Rapid SNP discovery and genetic mapping using sequenced RAD markers. PloS one 3, e3376. Balding, D. J. (2006). A tutorial on statistical methods for population association studies. Nat. Rev. Genet. 7, 781–791. Barrett, R. D. H. and Schluter, D. (2008). Adaptation from standing genetic variation. Trends Ecol. Evol. 23, 38–44. Barrett, R. D. H., Rogers, S. M. and Schluter, D. (2009). Environment specific pleiotropy facilitates divergence at the ectodysplasin locus in threespine stickleback. Evolution 63, 2831–2837. Barrett, R. D. H., Paccard, A., Healy, T. M., Bergek, S., Schulte, P. M., Schluter, D. and Rogers, S. M. (2011). Rapid evolution of cold tolerance in stickleback. Proc. R. Soc. B 278, 233–238. Barrionuevo, W. R. and Burggren, W. W. (1999). O2 consumption and heart rate in developing zebrafish (Danio rerio): influence of temperature and ambient O2. Am. J. Physiol. 276, R505–R513. Baxter, S. W., Davey, J. W., Johnston, J. S., Shelton, A. M., Heckel, D. G., Jiggins, C. D. and Blaxter, M. L. (2011). Linkage mapping and comparative genomics using next-generation rad sequencing of a non-model organism. PLoS ONE 6, e19315. Beacham, T. D. and Evelyn, T. P. T. (1992). Genetic variation in disease resistance and growth of chinook, coho, and chum salmon with respect to Vibriosis, Furunculosis, and Bacterial Kidney Disease. Trans. Am. Fish. Soc. 121, 456–485. Bear, E. A., McMahon, T. E. and Zale, A. V. (2007). Comparative thermal requirements of westslope cutthroat trout and rainbow trout: implications for species interactions and development of thermal protection standards. Trans. Am. Fish. Soc. 136, 1113–1121. Beaumont, M. A. and Balding, D. J. (2004). Identifying adaptive genetic divergence among populations from genome scans. Mol. Ecol. 13, 969–980. Beaumont, M. A. and Nichols, R. A. (1996). Evaluating loci for use in the genetic analysis of population structure. Proc. R. Soc. London B Biol. Sci. 263, 1619–1626. Becker, C. D. and Genoway, R. G. (1979). Evaluation of the critical thermal maximum for determining thermal tolerance of freshwater fish. Environ. Biol. Fishes 4, 245–256. Becker, C. D. and Wolford, M. G. (1980). Thermal resistance of juvenile salmonids sublethally exposed to nickel, determined by the critical thermal maximum method. Environ. Pollut. 21, 181–189. Beers, J. M. and Sidell, B. D. (2011). Thermal tolerance of Antarctic notothenioid fishes correlates with level of circulating hemoglobin. Physiol. Biochem. Zool. 84, 353–362. 185  Behnke, R. J. (1979). Monograph of the native trouts of the Genus Salmo Of western North America. Lakewood, Colorado: U.S. Fish and Wildlife Service & U.S. Forest Service. Behnke, R. J. (1992). Native trout of western North America. American Fisheries Society Monograph. Behnke, R. J. (2002). Trout and salmon of North America. New York: Free Press. Beitinger, T. L. and Bennett, W. A. (2000). Quantification of the role of acclimation temperature in temperature tolerance of fishes. Environ. Biol. Fishes 58, 277–288. Beitinger, T. L., Bennett, W. A. and McCauley, R. W. (2000). Temperature tolerances of North American freshwater fishes exposed to dynamic changes in temperature. Environ. Biol. Fishes 58, 237–275. Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B 57, 289–300. Benjamini, Y. and Yekutieli, D. (2001). The control of the false discovery rate in multiple testing under dependency. 1165–1188. Bennett, W. A. and Beitinger, T. L. (1997). Temperature tolerance of the sheepshead minnow, Cyprinodon variegatus. Copeia 1997, 77–87. Bergeron, R., Russell, R. R., Young, L. H., Ren, J. M., Marcucci, M., Lee, A. and Shulman, G. I. (1999). Effect of AMPK activation on muscle glucose metabolism in conscious rats. Am. J. Physiol. 276, E938–E944. Bernatchez, L. and Wilson, C. C. (1998). Comparative phylogeography of Nearctic and Palearctic fishes. Mol. Ecol. 7, 431–452. Berthelot, C., Brunet, F., Chalopin, D., Juanchich, A., Bernard, M., Noël, B., Bento, P., Da Silva, C., Labadie, K., Alberti, A., et al. (2014). The rainbow trout genome provides novel insights into evolution after whole-genome duplication in vertebrates. Nat. Commun. 5, 3657. Bianco, C., Cotten, C., Lonardo, E., Strizzi, L., Baraty, C., Mancino, M., Gonzales, M., Watanabe, K., Nagaoka, T., Berry, C., et al. (2009). Cripto-1 is required for hypoxia to induce cardiac differentiation of mouse embryonic stem cells. Am. J. Pathol. 175, 2146–2158. Birkedal, R. and Shiels, H. A. (2007). High [Na+]i in cardiomyocytes from rainbow trout. Am. J. Physiol. Regul. Integr. Comp. Physiol. 293, R861–R866. Black, E. C., Manning, G. T. and Hayashi, K. (1966). Changes in levels of hemoglobin, oxygen, carbon dioxide, pyruvate, and lactate in venous blood of rainbow trout (Salmo gairdneri) during and following severe muscular activity. J. Fish. Res. Board Canada 23, 783–795. Block, C. J., Spotila, J. R., Standora, E. A. and Gibbons, J. W. (1984). Behavioral thermoregulation of largemouth bass, Micropterus salmoides, and bluegill, Lepomis macrochirus, in a nuclear reactor cooling reservoir. Environ. Biol. Fishes 11, 41–52. Breiman, L. (2001). Random Forests. Mach. Learn. 45, 5–32. Brett, J. R. (1964). The respiratory metabolism and swimming performance of young sockeye salmon. J. Fish. Board Canada 21, 1183–1226. Brett, J. R. (1971). Energetic responses of salmon to temperature. A study of some thermal relations in the physiology and freshwater ecology of sockeye salmon (Oncorhynchus nerka). Am. Zool. 11, 99–113. 186  Brett, J. R. (1976). Scope for metabolism and growth of sockeye salmon, Oncorhynchus nerka, and some related energetics. J. Fish. Res. Board Canada 33, 307–313. Brett, J. R., Shelbourn, J. E. and Shoop, C. T. (1969). Growth rate and body composition of fingerling sockeye salmon, Oncorhynchus nerka, in relation to temperature and ration size. J. Fish. Res. Board Canada 26, 2363–2394. Brieuc, M. S. O., Waters, C. D., Seeb, J. E. and Naish, K. A. (2014). A dense linkage map for Chinook salmon (Oncorhynchus tshawytscha) reveals variable chromosomal divergence after an ancestral whole genome duplication event. G3 Genes, Genomes, Genet. 4, 447–460. Brooker, M. P., Morris, D. L. and Hemsworth, R. J. (1977). Mass mortalities of adult salmon, Salmo salar, in the R. Wye , 1976. J. Appl. Ecol. 14, 409–417. Brown, E. E. (1983). World Fish Farming: Cultivation and Economics. Second Edi. Boston, MA: Springer US. Bryden, C. A., Heath, J. W. and Heath, D. D. (2004). Performance and heterosis in farmed and wild Chinook salmon (Oncorhynchus tshawytscha) hybrid and purebred crosses. Aquaculture 235, 249–261. Buckley, B. A. (2006). The cellular response to heat stress in the goby Gillichthys mirabilis: a cDNA microarray and protein-level analysis. J. Exp. Biol. 209, 2660–2677. Buckley, B. a and Hofmann, G. E. (2004). Magnitude and duration of thermal stress determine kinetics of hsp gene regulation in the goby Gillichthys mirabilis. Physiol. Biochem. Zool. 77, 570–581. Buckley, B. A. and Somero, G. N. (2009). cDNA microarray analysis reveals the capacity of the cold-adapted Antarctic fish Trematomus bernacchii to alter gene expression in response to heat stress. Polar Biol. 32, 403–415. Bush, W. S. and Moore, J. H. (2012). Chapter 11: Genome-Wide Association Studies. PLoS Comput. Biol. 8, e1002822. Campbell, N. R., LaPatra, S. E., Overturf, K., Towner, R. and Narum, S. R. (2014). Association mapping of disease resistance traits in rainbow trout using restriction site associated DNA sequencing. G3 (Bethesda, Md.) 4, 2473–2481. Carline, R. F. and Machung, J. F. (2001). Critical thermal maxima of wild and domestic strains of trout. Trans. Am. Fish. Soc. 130, 1211–1216. Casselman, M. T., Anttila, K. and Farrell, A. P. (2012). Using maximum heart rate as a rapid screening tool to determine optimum temperature for aerobic scope in Pacific salmon Oncorhynchus spp. J. Fish Biol. 80, 358–377. Cassinelli, J. D. and Moffitt, C. M. (2010). Comparison of growth and stress in resident redband trout held in laboratory simulations of montane and desert summer temperature cycles. Trans. Am. Fish. Soc. 139, 339–352. Castilho, P. C., Buckley, B. A., Somero, G. and Block, B. A. (2009). Heterologous hybridization to a complementary DNA microarray reveals the effect of thermal acclimation in the endothermic bluefin tuna (Thunnus orientalis). Mol. Ecol. 18, 2092–2102. Catchen, J. M., Amores, A., Hohenlohe, P., Cresko, W. and Postlethwait, J. H. (2011). Stacks: building and genotyping Loci de novo from short-read sequences. G3 (Bethesda, Md.) 1, 171–182. 187  Catchen, J., Hohenlohe, P., Bassham, S., Amores, A. and Cresko, W. (2013). Stacks: an analysis tool set for population genomics. Mol. Ecol. 22, 3124–3140. Cazorla, O., Freiburg, A., Helmes, M., Centner, T., McNabb, M., Wu, Y., Trombitás, K., Labeit, S. and Granzier, H. (2000). Differential expression of cardiac titin isoforms and modulation of cellular stiffness. Circ. Res. 86, 59–67. Chabot, D., Steffensen, J. F. and Farrell, A. P. (2016). The determination of standard metabolic rate in fishes. J. Fish Biol. 88, 81–121. Chen, Z., Anttila, K., Wu, J., Whitney, C. K., Hinch, S. G. and Farrell, A. P. (2013). Optimum and maximum temperatures of sockeye salmon (Oncorhynchus nerka) populations hatched at different temperatures. Can. J. Zool. 274, 265–274. Churcott, C. S., Moyes, C. D., Bressler, B. H., Baldwin, K. M. and Tibbits, G. F. (1994). Temperature and pH effects on Ca2+ sensitivity of cardiac myofibrils: a comparison of trout with mammals. Am. J. Physiol. 267, R62-70. Clark, T. D. and Farrell, A. P. (2011). Effects of body mass on physiological and anatomical parameters of mature salmon: evidence against a universal heart rate scaling exponent. J. Exp. Biol. 214, 887–893. Clark, R. J. and Rodnick, K. J. (1998). Morphometric and biochemical characteristics of ventricular hypertrophy in male rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 201, 1541–1552. Clark, R. J. and Rodnick, K. J. (1999). Pressure and volume overloads are associated with ventricular hypertrophy in male rainbow trout. Am. J. Physiol. 277, R938-946. Clark, T. D., Sandblom, E., Cox, G. K., Hinch, S. G. and Farrell, A. P. (2008). Circulatory limits to oxygen supply during an acute temperature increase in the Chinook salmon (Oncorhynchus tshawytscha). Am. J. Physiol. Regul. Integr. Comp. Physiol. 295, R1631–R1639. Clark, T. D., Jeffries, K. M., Hinch, S. G. and Farrell, A. P. (2011). Exceptional aerobic scope and cardiovascular performance of pink salmon (Oncorhynchus gorbuscha) may underlie resilience in a warming climate. J. Exp. Biol. 214, 3074–3081. Clark, T. D., Sandblom, E. and Jutfelt, F. (2013). Aerobic scope measurements of fishes in an era of climate change: respirometry, relevance and recommendations. J. Exp. Biol. 216, 2771–2782. Clausen, R. G. (1934). Body temperature of fresh water fishes. Ecology 15, 139–144. Conesa, A., Götz, S., García-Gómez, J. M., Terol, J., Talón, M. and Robles, M. (2005). Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics 21, 3674–3676. Cossins, A. R. and Prosser, C. L. (1978). Evolutionary adaptation of membranes to temperature. Proc. Natl. Acad. Sci. USA 75, 2040–2043. Costello, A. B., Down, T. E., Pollard, S. M., Pacas, C. J. and Taylor, E. B. (2003). The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: an examination of microsatellite DNA variation in bull trout, salvelinus confluentus (pisces: salmonidae). Evolution 57, 328–344. Coulibaly, I., Danzmann, R. G., Palti, Y., Vallejo, R., Gahr, S. A., Yao, J. and Rexroad, C. E. (2006). Mapping of genes in a region associated with upper temperature tolerance in rainbow trout. Anim. Genet. 37, 598–599. 188  Crawford, D. L. and Powers, D. A. (1989). Molecular basis of evolutionary adaptation at the lactate dehydrogenase-B locus in the fish Fundulus heteroclitus. Proc. Natl. Acad. Sci. USA 86, 9365–9369. Crispo, E., Bentzen, P., Reznick, D. N., Kinnison, M. T. and Hendry, A. P. (2006). The relative influence of natural selection and geography on gene flow in guppies. Mol. Ecol. 15, 49–62. Currens, K. P., Schreck, C. and Li, H. W. (2009). Evolutionary ecology of redband trout. Trans. Am. Fish. Soc. 138, 797–817. Currie, R. J., Bennett, W. A. and Beitinger, T. L. (1998). Critical thermal minima and maxima of three freshwater game-fish species acclimated to constant temperatures. Environ. Biol. Fishes 51, 187–200. Currie, S., Moyes, C. D. and Tufts, B. L. (2000). The effects of heat shock and acclimation temperature on hsp70 and hsp30 mRNA expression in rainbow trout: in vivo and in vitro comparisons. J. Fish Biol. 56, 398–408. Danzmann, R. G., Jackson, T. R. and M. Ferguson, M. (1999). Epistasis in allelic expression at upper temperature tolerance QTL in rainbow trout. Aquaculture 173, 45–58. Dauwalter, D. C., Fesenmyer, K. A. and Bjork, R. (2015). Using aerial imagery to characterize redband trout habitat in a remote desert landscape. Trans. Am. Fish. Soc. 8487, 1322–1339. Davare, M. A., Saneyoshi, T., Guire, E. S., Nygaard, S. C. and Soderling, T. R. (2004). Inhibition of calcium/calmodulin-dependent protein kinase kinase by protein 14-3-3*. J. Biol. Chem. 279, 52191–52199. Davey, J. W. and Blaxter, M. L. (2010). RADSeq: next-generation population genetics. Briefings Funct. Genomics 9, 416–423. Davey, J. W., Hohenlohe, P. A., Etter, P. D., Boone, J. Q., Catchen, J. M. and Blaxter, M. L. (2011). Genome-wide genetic marker discovery and genotyping using next-generation sequencing. Nat. Rev. Genet. 12, 499–510. Davey, J. W., Cezard, T., Fuentes-Utrilla, P., Eland, C., Gharbi, K. and Blaxter, M. L. (2013). Special features of RAD Sequencing data: implications for genotyping. Mol. Ecol. 22, 3151–3164. Díaz-Uriarte, R. and Alvarez de Andrés, S. (2006). Gene selection and classification of microarray data using random forest. BMC Bioinforma. 7, 13. Diaz-Uriarte, R. (2007). GeneSrF and varSelRF: a web-based tool and R package for gene selection and classification using random forest. BMC Bioinforma. 8, 1–7. Dickson, I. W. and Kramer, R. H. (1971). Factors influencing scope for activity and active and standard metabolism of rainbow trout (Salmo gairdneri). J. Fish. Res. Board Canada 28, 587–596. Dietz, T. J. and Somero, G. N. (1992). The threshold induction temperature of the 90-kDa heat shock protein is subject to acclimatization in eurythermal goby fishes (genus Gillichthys). Proc. Natl. Acad. Sci. USA 89, 3389–3393. Dionne, M., Caron, F., Dodson, J. J. and Bernatchez, L. (2008). Landscape genetics and hierarchical genetic structure in Atlantic salmon: The interaction of gene flow and local adaptation. Mol. Ecol. 17, 2382–2396. 189  Doyle, C. M., Leberg, P. L. and Klerks, P. L. (2011). Heritability of heat tolerance in a small livebearing fish, Heterandria formosa. Ecotoxicol. (London, England) 20, 535–42. Drost, H. E., Carmack, E. C. and Farrell, A. P. (2014). Upper thermal limits of cardiac function for Arctic cod Boreogadus saida, a key food web fish species in the Arctic Ocean. J. Fish Biol. 84, 1781–1792. Drost, H. E., Fisher, J., Randall, F., Kent, D., Carmack, E. C. and Farrell, A. P. (2016). Upper thermal limits of the hearts of Arctic cod Boreogadus saida: adults compared with larvae. J. Fish Biol. 88, 718–726. Durham, B. W., Wilde, G. R. and Pope, K. L. (2006). Temperature-caused fish kill in a flowing Great Plains river. Southwest. Nat. 51, 397–401. Ebersole, J. L., Liss, W. J. and Frissell, C. A. (2001). Relationship between stream temperature, thermal refugia and rainbow trout Oncorhynchus mykiss abundance in arid-land streams in the northwestern United States. Ecol. Freshw. Fish 10, 1–10. Ekblom, R. and Galindo, J. (2010). Applications of next generation sequencing in molecular ecology of non-model organisms. Heredity 107, 1–15. Ekström, A., Jutfelt, F. and Sandblom, E. (2014). Effects of autonomic blockade on acute thermal tolerance and cardioventilatory performance in rainbow trout, Oncorhynchus mykiss. J. Therm. Biol. 44, 47–54. Ekström, A., Hellgren, K., Gräns, A., Pichaud, N. and Sandblom, E. (2016). Dynamic changes in scope for heart rate and cardiac autonomic control during warm acclimation in rainbow trout. J. Exp. Biol. 219, 1106–1109. Eliason, E. J. and Farrell, A. P. (2014). Effect of hypoxia on specific dynamic action and postprandial cardiovascular physiology in rainbow trout (Oncorhynchus mykiss). Comp. Biochem. Physiol. Part A 171, 44–50. Eliason, E. J., Clark, T. D., Hague, M. J., Hanson, L. M., Gallagher, Z. S., Jeffries, K. M., Gale, M. K., Patterson, D. A., Hinch, S. G. and Farrell, A. P. (2011). Differences in thermal tolerance among sockeye salmon populations. Science 332, 109–112. Eliason, E. J., Clark, T. D., Hinch, S. G. and Farrell, A. P. (2013). Cardiorespiratory collapse at high temperature in swimming adult sockeye salmon. Conserv. Physiol. 1, 1–19. Elshire, R. J., Glaubitz, J. C., Sun, Q., Poland, J. A., Kawamoto, K., Buckler, E. S. and Mitchell, S. E. (2011). A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS ONE 6, 1–10. Emerson, K. J., Merz, C. R., Catchen, J. M., Hohenlohe, P. A., Cresko, W. A., Bradshaw, W. E. and Holzapfel, C. M. (2010). Resolving postglacial phylogeography using high-throughput sequencing. Proc. Natl. Acad. Sci. USA 107, 16196–16200. Ern, R., Huong, D. T. T., Phuong, N. T., Wang, T. and Bayley, M. (2014). Oxygen delivery does not limit thermal tolerance in a tropical eurythermal crustacean. J. Exp. Biol. 217, 809–814. Evans, T. G. (2015). Considerations for the use of transcriptomics in identifying the “genes that matter” for environmental adaptation. J. Exp. Biol. 218, 1925–1935. Everett, M. V and Seeb, J. E. (2014). Detection and mapping of QTL for temperature tolerance and body size in Chinook salmon (Oncorhynchus tshawytscha) using genotyping by sequencing. Evol. Appl. 7, 480–492. 190  Fan, C. Y., Lee, S. and Cyr, D. M. (2003). Mechanisms for regulation of Hsp70 function by Hsp40. Cell Stress chaperones 8, 309–316. Fangue, N. A., Hofmeister, M. and Schulte, P. M. (2006). Intraspecific variation in thermal tolerance and heat shock protein gene expression in common killifish, Fundulus heteroclitus. J. Exp. Biol. 209, 2859–2872. Fangue, N. A., Richards, J. G. and Schulte, P. M. (2009). Do mitochondrial properties explain intraspecific variation in thermal tolerance? J. Exp. Biol. 212, 514–522. Fangue, N. A., Osborne, E. J., Todgham, A. E. and Schulte, P. M. (2011). The onset temperature of the heat-shock response and whole-organism thermal tolerance are tightly correlated in both laboratory-acclimated and field-acclimatized tidepool sculpins (Oligocottus maculosus). Physiol. Biochem. Zool. 84, 341–352. Farrell, A. P. (1996). Effects of temperature on cardiovascular performance. In Global Warming: Implications for freshwater and marine fish Editor: Wood, C. M. and McDonald, D. G., pp. 135–158. Cambridge University Press. Farrell, A. P. (2007). Cardiorespiratory performance during prolonged swimming tests with salmonids: a perspective on temperature effects and potential analytical pitfalls. Philos. Trans. R. Soc. London. Ser. B, Biol. Sci. 362, 2017–2130. Farrell, A. P. (2009). Environment, antecedents and climate change: lessons from the study of temperature physiology and river migration of salmonids. J. Exp. Biol. 212, 3771–3780. Farrell, A. P. (2013). Aerobic scope and its optimum temperature: clarifying their usefulness and limitations–correspondence on J. Exp. Biol. 216, 2771-2782. J. Exp. Biol. 4493–3397. Farrell, A. P. (2016). Pragmatic perspective on aerobic scope: peaking, plummeting, pejus and apportioning. J. Fish Biol. 88, 322–343. Farrell, A. P. and Jones, D. R. (1992). 1 - The Heart. In Fish Physiology, Volume XII: The Cardiovascular System Editor: Hoar, W. S., Randall, D. J., and Farrell, A. P., pp. 1–88. Academic Press. Farrell, A. P. and Milligan, C. L. (1986). Myocardial intracellular pH in a perfused rainbow trout heart during extracellular acidosis in the presence and absence of adrenaline. J. Exp. Biol. 125, 347–359. Farrell, A. P., MacLeod, K. R., Driedzic, W. R. and Wood, S. (1983). Cardiac performance in the in situ perfused fish heart during extracellular acidosis: interactive effects of adrenaline. J. Exp. Biol. 107, 415–429. Farrell, A. P., Hart, T., Wood, S. and Driedzic, W. R. (1984). The effect of extracellular calcium and preload on a teleost heart during extracellular hypercapnic acidosis. Can. J. Zool. 62, 1429–1435. Farrell, A. P., Macleod, K. R. and Scott, C. (1988a). Cardiac performance of the trout (Salmo gairdneri) heart during acidosis: effects of low bicarbonate, lactate and cortisol. Comp. Biochem. Physiol. 91A, 271–277. Farrell, A. P., Hammons, A. M., Graham, M. S. and Tibbits, G. F. (1988b). Cardiac growth in rainbow trout, Salmo gairdneri. Can. J. Zool. 66, 2368–2373. Farrell, A. P., Thorarensen, H., Axelsson, M., Crocker, C. E., Gamperl, A. K. and Cech, J. J. (2001). Gut blood flow in fish during exercise and severe hypercapnia. Comp. Biochem. Physiol. Part A 128, 551–563. 191  Farrell, A. P., Hinch, S. G., Cooke, S. J., Patterson, D. A., Crossin, G. T., Lapointe, M. and Mathes, M. T. (2008). Pacific salmon in hot water: applying aerobic scope models and biotelemetry to predict the success of spawning migrations. Physiol. Biochem. Zool. 81, 697–709. Feder, M. E. and Hofmann, G. E. (1999). Heat-shock proteins, molecular chaperones, and the stress response. Annu. Rev. Physiol. 61, 243–282. Feminella, J. W. and Matthews, W. J. (1984). Intraspecific differences in thermal tolerance of Etheostoma spectabile (Agassiz) in constant versus fluctuating environments. J. Fish Biol. 25, 455–461. Ferreira, E. O., Anttila, K. and Farrell, A. P. (2014). Thermal optima and tolerance in the eurythermic goldfish (Carassius auratus): relationships between whole-animal aerobic capacity and maximum heart rate. Physiol. Biochem. Zool. 87, 599–611. Fesenmyer, K. and Dauwalter, D. (2014). Redband trout habitat assessment: Owyhee, Bruneau-Jarbidge, and Salmon Falls Creek Basins. Final report to Nevada State Office, U.S. Bureau of Land Management. Trout Unlimited, Arlington, Virginia. Foll, M. and Gaggiotti, O. (2008). A genome-scan method to identify selected loci appropriate for both dominant and codominant markers: a Bayesian perspective. Genetics 180, 977–993. Fougerousse, F., Delezoide, A.-L., Fiszman, M. Y., Schwartz, K., Beckmann, J. S. and Carrier, L. (1998). Cardiac myosin binding protein c gene is specifically expressed in heart during murine and human development. Circ. Res. 82, 130–133. Franklin, C. E., Davison, W. and Seebacher, F. (2007). Antarctic fish can compensate for rising temperatures: thermal acclimation of cardiac performance in Pagothenia borchgrevinki. J. Exp. Biol. 210, 3068–3074. Franks, S. J. and Hoffmann, A. A. (2012). Genetics of climate change adaptation. Annu. Rev. Genet. 46, 185–208. Fraser, D. J., Weir, L. K., Bernatchez, L., Hansen, M. M. and Taylor, E. B. (2011). Extent and scale of local adaptation in salmonid fishes: review and meta-analysis. Heredity 106, 404–420. Fry, F. E. J. (1947). Effects of the environment on animal activity. Publ. Ontario Fish. Res. Lab. 55, 1–62. Fry, F. E. J. (1948). Temperature relations of salmonids. Proc. 1st Meet. Can Commi Freshw. Fish. Res. 5. Fry, F. E. J. and Hart, J. S. (1948). The relation of temperature to oxygen consumption in the gold fish. Biol. Bull. 94, 66–77. Fu, S., Xie, X. and Cao, Z. (2005). Effect of feeding level and feeding frequency on specific dynamic action in Silurus meridionalis. J. Fish Biol. 67, 171–182. Fulda, S., Gorman, A. M., Hori, O. and Samali, A. (2010). Cellular stress responses: cell survival and cell death. Int. J. cell Biol. 2010, 214074. Fuller, R. C., Baer, C. F. and Travis, J. (2005). How and when selection experiments might actually be useful. Integr. Comp. Biol. 45, 391–404. Galbreath, P. F., Adams, N. D., Sherrill, L. W. and Martin, T. H. (2006). Thermal tolerance of diploid versus triploid rainbow trout and brook trout assessed by time to chronic lethal maximum. Environ. Biol. Fishes 75, 183–193. 192  Gallaugher, P. and Farrell, A. P. (1998). Hematocrit and blood oxygen-carrying capacity. In Fish Physiology Editor: Perry, S. F. and Tufts, B., pp. 185–227. San Diego: Academic Press. Gamperl, A. K., Rodnick, K. J., Faust, H. A., Venn, E. C., Bennett, M. T., Crawshaw, L. I., Keeley, E. R., Powell, M. S. and Li, H. W. (2002). Metabolism, swimming performance, and tissue biochemistry of high desert redband trout (Oncorhynchus mykiss ssp.): evidence for phenotypic differences in physiological function. Physiol. Bochemical Zool. 75, 413–431. Garvin, M. R., Thorgaard, G. H. and Narum, S. R. (2015). Differential expression of genes that control respiration contribute to thermal adaptation in redband trout (Oncorhynchus mykiss gairdneri). Genome Biol. Evol. 7, 1404–1414. Gerken, A. R., Eller, O. C., Hahn, D. a and Morgan, T. J. (2015). Constraints, independence, and evolution of thermal plasticity: Probing genetic architecture of long- and short-term thermal acclimation. Proc. Natl. Acad. Sci. USA 112, 4399–404. Gerlach, G. F., Turay, L., Malik, K. T., Lida, J., Scutt, A., Goldspink, G., Turay, L., Malik, K. T. A., Lida, J. and Goldspink, G. (1990). Mechanisms of temperature acclimation in the carp : a molecular biology approach. Am. J. Physiol. Regul. Integr. Comp. Physiol. 259, R237–R244. Glucksman, J., West, G. and Berra, T. M. (1976). The introduced fishes of Papua New Guinea with special reference to Tilapia mossambica. Biol. Conserv. 9, 37–44. Goldstein, B. and Macara, I. G. (2007). The PAR proteins: fundamental players in animal cell polarization. Dev. cell 13, 609–622. Gollock, M. J., Currie, S., Petersen, L. H. and Gamperl, A. K. (2006). Cardiovascular and haematological responses of Atlantic cod (Gadus morhua) to acute temperature increase. J. Exp. Biol. 209, 2961–2970. Gracey, A. Y. (2007). Interpreting physiological responses to environmental change through gene expression profiling. J. Exp. Biol. 210, 1584–1592. Gracey, A. Y., Fraser, E. J., Li, W., Fang, Y., Taylor, R. R., Rogers, J., Brass, A. and Cossins, A. R. (2004). Coping with cold: An integrative, multitissue analysis of the transcriptome of a poikilothermic vertebrate. Proc. Natl. Acad. Sci. USA 101, 16970–16975. Graham, M. S. and Farrell, A. P. (1989). The effect of temperature acclimation and adrenaline on the performance of a perfused trout heart. Physiol. Zool. 62, 38–61. Gräns, A., Jutfelt, F., Sandblom, E., Jönsson, E., Wiklander, K., Seth, H., Olsson, C., Dupont, S., Ortega-Martinez, O., Einarsdottir, I., et al. (2014). Aerobic scope fails to explain the detrimental effects on growth resulting from warming and elevated CO2 in Atlantic halibut. J. Exp. Biol. 217, 711–717. Griffith, O. L., Pepin, F., Enache, O. M., Heiser, L. M., Collisson, E. A., Spellman, P. T. and Gray, J. W. (2013). A robust prognostic signature for hormone-positive node-negative breast cancer. Genome Med. 5, 92. Hageman, J., Rujano, M. A., van Waarde, M. A. W. H., Kakkar, V., Dirks, R. P., Govorukhina, N., Oosterveld-Hut, H. M. J., Lubsen, N. H. and Kampinga, H. H. (2010). A DNAJB chaperone subfamily with HDAC-dependent activities suppresses toxic protein aggregation. Mol. Cell 37, 355–369. 193  Hall, A. E. (1992). Breeding for heat tolerance. In Plant Breeding Reviews, pp. 129–168. John Wiley & Sons, Inc. Hamlet, A. F. and Lettenmaier, D. P. (1999). Effects of climate change on hydrology and water resources in the columbia river basin. J. Am. Water Resour. Assoc. 35, 1597–1623. Hanson, L. M., Obradovich, S., Mouniargi, J. and Farrell, A. P. (2006). The role of adrenergic stimulation in maintaining maximum cardiac performance in rainbow trout (Oncorhynchus mykiss) during hypoxia, hyperkalemia and acidosis at 10 degrees C. J. Exp. Biol. 209, 2442–2451. Harper, A., Newton, I. and Watt, P. (1995). The effect of temperature on spontaneous action potential discharge of the isolated sinus venosus from winter and summer plaice (Pleuronectes platessa). J. Exp. Biol. 198, 137–140. Hart, J. S. (1952). Geographic variations of some physiological and morphological characters in certain freshwater fish. Publ. Ontario Fish. Res. Lab. 72, 1–79. Hartl, D. L. and Clark, A. G. (2007). Principles of population genetics. Sunderland, Mass: Sinauer Associates. Hassinen, M., Haverinen, J. and Vornanen, M. (2008). Electrophysiological properties and expression of the delayed rectifier potassium (ERG) channels in the heart of thermally acclimated rainbow trout. Am. J. Physiol. Regul. Integr. Comp. Physiol. 295, R297–R308. Hayes, J. D. and McLellan, L. I. (1999). Glutathione and glutathione-dependent enzymes represent a co-ordinately regulated defence against oxidative stress. Free Radic. Res. 31, 273–300. Healy, T. M. and Schulte, P. M. (2012). Thermal acclimation is not necessary to maintain a wide thermal breadth of aerobic scope in the common killifish (Fundulus heteroclitus). Physiol. Biochem. Zool. 85, 107–119. Healy, T. M., Tymchuk, W. E., Osborne, E. J. and Schulte, P. M. (2010). Heat shock response of killifish (Fundulus heteroclitus): candidate gene and heterologous microarray approaches. Physiol. Genomics 41, 171–184. Hecht, B. C., Thrower, F. P., Hale, M. C., Miller, M. R. and Nichols, K. M. (2012). Genetic architecture of migration-related traits in rainbow and steelhead trout, Oncorhynchus mykiss. G3 Genes, Genomes, Genet. 2, 1113–1127. Hecht, B. C., Campbell, N. R., Holecek, D. E. and Narum, S. R. (2013). Genome-wide association reveals genetic basis for the propensity to migrate in wild populations of rainbow and steelhead trout. Mol. Ecol. 22, 3061–3076. Hecht, B. C., Matala, A. P., Hess, J. E. and Narum, S. R. (2015). Environmental adaptation in Chinook salmon (Oncorhynchus tshawytscha) throughout their North American range. Mol. Ecol. 24, 5573–5595. Helmes, M., Trombitás, K., Centner, T., Kellermayer, M., Labeit, S., Linke, W. A. and Granzier, H. (1999). Mechanically driven contour-length adjustment in rat cardiac titin’s unique N2B sequence: titin is an adjustable spring. Circ. Res. 84, 1339–1352. Hemmer-Brepson, C., Replumaz, L., Romestaing, C., Voituron, Y. and Daufresne, M. (2014). Non-stressful temperature effect on oxidative balance and life history traits in adult fish (Oryzias latipes). Clin. Cancer Res. 217, 274–282. 194  Henne, J. and Jeserich, G. (2004). Maturation of spiking activity in trout retinal ganglion cells coincides with upregulation of kv3.1- and BK-related potassium channels. J. Neurosci. Res. 75, 44–54. Hochachka, P. W. and Somero, G. N. (2002). Biochemical adaptation: mechanism and process in physiological evolution. Oxford University Press. Hofer, T., Ray, N., Wegmann, D. and Excoffier, L. (2009). Large allele frequency differences between human continental groups are more likely to have occurred by drift during range expansions than by selection. Ann. Hum. Genet. 73, 95–108. Hoffmann, A. A. and Sgro, C. M. (2011). Climate change and evolutionary adaptation. Nature 470, 479–485. Hoffmann, A. A. and Willi, Y. (2008). Detecting genetic responses to environmental change. Nat. Rev. Genet. 9, 421–432. Hoffmann, A. A., Chown, S. L. and Clusella-Trullas, S. (2013). Upper thermal limits in terrestrial ectotherms: How constrained are they? Funct. Ecol. 27, 934–949. Hofmann, G. E. and Todgham, A. E. (2010). Living in the now: physiological mechanisms to tolerate a rapidly changing Environment. Annu. Rev. Physiol. 72, 127–145. Hofmann, G. E., Buckley, B. a, Airaksinen, S., Keen, J. E. and Somero, G. N. (2000). Heat-shock protein expression is absent in the antarctic fish Trematomus bernacchii (family Nototheniidae). J. Exp. Biol. 203, 2331–2339. Hohenlohe, P. A., Bassham, S., Etter, P. D., Stiffler, N., Johnson, E. A. and Cresko, W. A. (2010). Population genomics of parallel adaptation in threespine stickleback using sequenced RAD tags. PLoS Genet. 6, e1000862. Hohenlohe, P. A., Amish, S. J., Catchen, J. M., Allendorf, F. W. and Luikart, G. (2011). Next-generation RAD sequencing identifies thousands of SNPs for assessing hybridization between rainbow and westslope cutthroat trout. Mol. Ecol. Resour. 11 Suppl 1, 117–122. Hokanson, K. E. F., Kleiner, C. F. and Thorslund, T. W. (1977). Effects of constant temperatures and diel temperature fluctuations on specific growth and mortality rates and yield of juvenile rainbow trout, Salmo gairdneri. J. Fish. Res. Board Canada 34, 639–648. Holderegger, R., Kamm, U. and Gugerli, F. (2006). Adaptive vs. neutral genetic diversity: Implications for landscape genetics. Landsc. Ecol. 21, 797–807. Holeton, G. F. (1971). Respiratory and circulatory responses of rainbow trout larvae to carbon monoxide and to hypoxia. J. Exp. Biol. 55, 683–694. Hove-Madsen, L. (1992). The influence of temperature on ryanodine sensitivity and the force-frequency relationship in the myocardium of rainbow trout. J. Exp. Biol. 167, 47–60. Hove-Madsen, L., Llach, A. and Tort, L. (1998). Quantification of Ca2+ uptake in the sarcoplasmic reticulum of trout ventricular myocytes. Am. J. Physiol. Regul. Integr. Comp. Physiol. 275, R2070–R2080. Huang, X. Y., Morielli, A. D. and Peralta, E. G. (1994). Molecular basis of cardiac potassium channel stimulation by protein kinase A. Proc. Natl. Acad. Sci. USA 91, 624–628. Hunter, T. and Poon, R. Y. C. (2016). Cdc37: a protein kinase chaperone? Trends Cell Biol. 7, 157–161. 195  Huntsman, A. G. (1942). Death of salmon and trout with high temperature. J. Fish. Res. Board Canada 5, 485–501. Huntsman, A. G. (1946). Heat stroke in canadian maritime stream fishes. J. Fish. Res. Board Canada 6e, 476–482. Iftikar, F. I. and Hickey, A. J. R. (2013). Do mitochondria limit hot fish hearts? Understanding the role of mitochondrial function with heat stress in Notolabrus celidotus. PLoS ONE 8, e64120. Iftikar, F. I., MacDonald, J. R., Baker, D. W., Renshaw, G. M. C. and Hickey,  a. J. R. (2014). Could thermal sensitivity of mitochondria determine species distribution in a changing climate? J. Exp. Biol. 217, 2348–2357. Ihssen, P. E. (1986). Selection of fingerling rainbow trout for high and low tolerance to high temperature. Aquaculture 57, 370. Ineno, T., Tsuchida, S., Kanda, M. and Watabe, S. (2005). Thermal tolerance of a rainbow trout Oncorhynchus mykiss strain selected by high-temperature breeding. Fish. Sci. 71, 767–775. Iwama, G. and Thomas, P. (1998). Heat shock protein expression in fish. Rev. Fish Biol. Fish. 8, 35–56. Iwama, G. K., Vijayan, M. M., Forsyth, R. O. B. B. and Ackerman, P. A. (1999). Heat shock proteins and physiological stress in fish. Am. Zool. 39, 901–909. Jackson, T. R., Ferguson, M. M., Danzmann, R. G., Fishback, A. G., Ihssen, P. E., O’Connell, M. and Crease, T. J. (1998). Identification of two QTL influencing upper temperature tolerance in three rainbow trout (Oncorhynchus mykiss) half-sib families. Heredity 80, 143–151. Jayasundara, N., Gardner, L. D. and Block, B. A. (2013). Effects of temperature acclimation on Pacific bluefin tuna (Thunnus orientalis) cardiac transcriptome. Am. J. Physiol. Regul. Integr. Comp. Physiol. 305, R1010–R1020. Jayasundara, N., Tomanek, L., Dowd, W. W. and Somero, G. N. (2015). Proteomic analysis of cardiac response to thermal acclimation in the eurythermal goby fish Gillichthys mirabilis. J. Exp. Biol. 218, 1359–1372. Jeffries, K. M., Hinch, S. G., Sierocinski, T., Clark, T. D., Eliason, E. J., Donaldson, M. R., Li, S., Pavlidis, P. and Miller, K. M. (2012). Consequences of high temperatures and premature mortality on the transcriptome and blood physiology of wild adult sockeye salmon (Oncorhynchus nerka). Ecol. Evol. 2, 1747–1764. Jeffries, K. M., Hinch, S. G., Sierocinski, T., Pavlidis, P. and Miller, K. M. (2014). Transcriptomic responses to high water temperature in two species of Pacific salmon. Evol. Appl. 7, 286–300. Jiang, H., Deng, Y., Chen, H.-S., Tao, L., Sha, Q., Chen, J., Tsai, C.-J. and Zhang, S. (2004). Joint analysis of two microarray gene-expression data sets to select lung adenocarcinoma marker genes. BMC Bioinforma. 5, 81. Jinadasa, J., Kothalawala, A. and Herath, H. (2005). Status of rainbow trout (Oncorhynchus mykiss-Waldaum 1792) population after the cessation of stocking in waterways of Horton Plains, Sri Lanka. Vidyodaya J. Sci. 12, 9–27. Joberty, G., Petersen, C., Gao, L. and Macara, I. G. (2000). The cell-polarity protein Par6 links Par3 and atypical protein kinase C to Cdc42. Nat Cell Biol 2, 531–539. 196  Jombart, T. (2008). adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–1405. Jørgensen, S. M., Castro, V., Krasnov, A., Torgersen, J., Timmerhaus, G., Hevrøy, E. M., Hansen, T. J., Susort, S., Breck, O. and Takle, H. (2014). Cardiac responses to elevated seawater temperature in Atlantic salmon. BMC Physiol. 14, 2. Ju, Z., Dunham, R. and Liu, Z. (2002). Differential gene expression in the brain of channel catfish (Ictalurus punctatus) in response to cold acclimation. Mol. Genet. Genomics 268, 87–95. Kanehisa, M., Sato, Y., Kawashima, M., Furumichi, M. and Tanabe, M. (2016). KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 44, D457–D462. Kassahn, K., Crozier, R., Ward, A., Stone, G. and Caley, M. J. (2007a). From transcriptome to biological function: environmental stress in an ectothermic vertebrate, the coral reef fish Pomacentrus moluccensis. BMC Genomics 8,. Kassahn, K. S., Caley, M. J., Ward, A. C., Connolly, A. R., Stone, G. and Crozier, R. H. (2007b). Heterologous microarray experiments used to identify the early gene response to heat stress in a coral reef fish. Mol. Ecol. 16, 1749–1763. Kassahn, K. S., Crozier, R. H., Pörtner, H. O. and Caley, M. J. (2009). Animal performance and stress: Responses and tolerance limits at different levels of biological organisation. Biol. Rev. 84, 277–292. Katz, J., Moyle, P. B., Quiñones, R. M., Israel, J. and Purdy, S. (2013). Impending extinction of salmon, steelhead, and trout (Salmonidae) in California. Environ. Biol. Fishes 96, 1169–1186. Kawecki, T. J. and Ebert, D. (2004). Conceptual issues in local adaptation. Ecol. Lett. 7, 1225–1241. Kaya, C. (1978). Thermal resistance of rainbow trout from a permanently heated stream, and of two hatchery strains. Progress. Fish-Culturist 40, 37–41. Keen, A. N., Fenna, A. J., McConnell, J. C., Sherratt, M. J., Gardner, P. and Shiels, H. A. (2015). The dynamic nature of hypertrophic and fibrotic remodeling of the fish ventricle. Front. Physiol. 6, 427. Kelley, A. L., de Rivera, C. E. and Buckley, B. A. (2011). Intraspecific variation in thermotolerance and morphology of the invasive European green crab, Carcinus maenas, on the west coast of North America. J. Exp. Mar. Biol. Ecol. 409, 70–78. Klaiman, J. M., Fenna, A. J., Shiels, H. A., Macri, J. and Gillis, T. E. (2011). Cardiac remodeling in fish: strategies to maintain heart function during temperature Change. PloS one 6, e24464. Klauzinska, M., Castro, N. P., Rangel, M. C., Spike, B. T., Gray, P. C., Bertolette, D., Cuttitta, F. and Salomon, D. (2014). The multifaceted role of the embryonic gene Cripto-1 in cancer, stem cells and epithelial-mesenchymal transition. Semin. Cancer Biol. 29, 51–58. Klein, R. J. (2007). Power analysis for genome-wide association studies. BMC Genet. 8, 58. Knies, J. L., Izem, R., Supler, K. L., Kingsolver, J. G. and Burch, C. L. (2006). The genetic basis of thermal reaction norm evolution in lab and natural phage populations. PLoS Biol. 4, e201. 197  Köllner, B., Wasserrab, B., Kotterba, G. and Fischer, U. (2002). Evaluation of immune functions of rainbow trout (Oncorhynchus mykiss)--how can environmental influences be detected? Toxicol. Lett. 131, 83–95. Korajoki, H. and Vornanen, M. (2012). Expression of SERCA and phospholamban in rainbow trout (Oncorhynchus mykiss) heart: comparison of atrial and ventricular tissue and effects of thermal acclimation. J. Exp. Biol. 215, 1162–1169. Kozfkay, C. C., Campbell, M. R., Meyer, K. A. and Schill, D. J. (2011). Influences of habitat and hybridization on the genetic structure of redband trout in the upper Snake River basin, Idaho. Trans. Am. Fish. Soc. 140, 282–295. Kregel, K. C. (2002). Heat shock proteins: modifying factors in physiological stress responses and acquired thermotolerance. J. Appl. Physiol. 92, 2177–2186. Krief, S., Faivre, J.-F., Robert, P., Le Douarin, B., Brument-Larignon, N., Lefrère, I., Bouzyk, M. M., Anderson, K. M., Greller, L. D., Tobin, F. L., et al. (1999). Identification and characterization of cvHsp: a novel human small stress protein selectively expressed in cardiovascular and insulin-sensitive tissues. J. Biol. Chem. 274, 36592–36600. Kültz, D. (2005). Molecular and evolutionary basis of the cellular stress response. Annu. Rev. Physiol. 67, 225–257. Kurokawa, J., Chen, L. and Kass, R. S. (2003). Requirement of subunit expression for cAMP-mediated regulation of a heart potassium channel. Proc. Natl. Acad. Sci. USA 100, 2122–2127. Kwon, H.-S., Lee, H.-S., Ji, Y., Rubin, J. S. and Tomarev, S. I. (2009). Myocilin is a modulator of Wnt signaling. Mol. Cell. Biol. 29, 2139–2154. Lakatta, E. G., Maltsev, V. A. and Vinogradova, T. M. (2010). A Coupled SYSTEM of intracellular Ca2+ clocks and surface membrane voltage clocks controls the timekeeping mechanism of the heart’s pacemaker. Circ. Res. 106, 659–673. Langmead, B. and Salzberg, S. L. (2012). Fast gapped-read alignment with Bowtie 2. Nat Meth 9, 357–359. Langmead, B., Trapnell, C., Pop, M. and Salzberg, S. L. (2009). Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25.1-R25.10. LeBlanc, S., Middleton, S., Gilmour, K. M. and Currie, S. (2011). Chronic social stress impairs thermal tolerance in the rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 214, 1721–1731. Lecarpentier, Y., Vignier, N., Oliviero, P., Guellich, A., Carrier, L. and Coirault, C. (2008). Cardiac myosin-binding protein c modulates the tuning of the molecular motor in the heart. Biophys. J. 95, 720–728. Lee, R. M. and Rinne, J. N. (1980). Critical thermal maxima of five trout species in the southwestern United States. Trans. Am. Fish. Soc. 109, 632–635. Lee, C. G., Devlin, R. H. and Farrell, A. P. (2003a). Swimming performance, oxygen consumption and excess post‐exercise oxygen consumption in adult transgenic and ocean‐ranched coho salmon. J. Fish Biol. 62, 753–766. Lee, C. G., Farrell, A. P., Lotto, A., MacNutt, M. J., Hinch, S. G. and Healey, M. C. (2003b). The effect of temperature on swimming performance and oxygen consumption 198  in adult sockeye (Oncorhynchus nerka) and coho (O. kisutch) salmon stocks. J. Exp. Biol. 206, 3239–3251. Leimu, R. and Fischer, M. (2008). A meta-analysis of local adaptation in plants. PLoS ONE 3, e4010. Lendenmann, M. H., Croll, D., Palma-Guerrero, J., Stewart, E. L. and McDonald, B. A. (2016). QTL mapping of temperature sensitivity reveals candidate genes for thermal adaptation and growth morphology in the plant pathogenic fungus Zymoseptoria tritici. Heredity 116, 384–94. Lewis, J. M., Hori, T. S., Rise, M. L., Walsh, P. J. and Currie, S. (2010). Transcriptome responses to heat stress in the nucleated red blood cells of the rainbow trout (Oncorhynchus mykiss). Physiol Genomics 42,. Lipka, A. E., Tian, F., Wang, Q., Peiffer, J., Li, M., Bradbury, P. J., Gore, M. A., Buckler, E. S. and Zhang, Z. (2012). GAPIT: genome association and prediction integrated tool. Bioinforma. (Oxford, England) 28, 2397–2399. Lisewski, U., Shi, Y., Wrackmeyer, U., Fischer, R., Chen, C., Schirdewan, A., Jüttner, R., Rathjen, F., Poller, W., Radke, M. H., et al. (2008). The tight junction protein CAR regulates cardiac conduction and cell–cell communication. J. Exp. Med. 205, 2369–2379. Lissandron, V. and Zaccolo, M. (2006). Compartmentalized cAMP/PKA signalling regulates cardiac excitation-contraction coupling. J. Muscle Res. Cell Motil. 27, 399–403. Lister, R., O’Malley, R. C., Tonti-Filippini, J., Gregory, B. D., Berry, C. C., Millar, A. H. and Ecker, J. R. (2008). Highly integrated single-base resolution maps of the epigenome in Arabidopsis. Cell 133, 523–536. Liu, Z. J. and Cordes, J. F. (2004). DNA marker technologies and their applications in aquaculture genetics. Aquaculture 238, 1–37. Liu, S., Wang, X., Sun, F., Zhang, J., Feng, J., Liu, H., Rajendran, K. V, Sun, L., Zhang, Y., Jiang, Y., et al. (2013). RNA-Seq reveals expression signatures of genes involved in oxygen transport, protein synthesis, folding, and degradation in response to heat stress in catfish. Physiol. genomics 45, 462–476. Logan, C. A. and Buckley, B. A. (2015). Transcriptomic responses to environmental temperature in eurythermal and stenothermal fishes. J. Exp. Biol. 218, 1915–1924. Logan, C. A. and Somero, G. N. (2010). Transcriptional responses to thermal acclimation in the eurythermal fish Gillichthys mirabilis ( Cooper 1864 ). Am. J. Physiol. Integr. Comp. Physiol. 299, R843–R852. Logan, C. A. and Somero, G. N. (2011). Effects of thermal acclimation on transcriptional responses to acute heat stress in the eurythermal fish Gillichthys mirabilis (Cooper). Am. J. Physiol. Regul. Integr. Comp. Physiol. 300, R1373-1383. Logue, J. A., de Vries, A. L., Fodor, E. and Cossins, A. R. (2000). Lipid compositional correlates of temperature-adaptive interspecific differences in membrane physical structure. J. Exp. Biol. 203, 2105–2115. Lohse, M. J., Engelhardt, S. and Eschenhagen, T. (2003). What is the role of β-adrenergic signaling in heart failure? Circ. Res. 93, 896–906. Lushchak, V. I. (2011). Environmentally induced oxidative stress in aquatic animals. Aquat. Toxicol. 101, 13–30. 199  Lutterschmidt, W. and Hutchison, V. (1997a). The critical thermal maximum: data to support the onset of spasms as the definitive end point. Can. J. Zool. 75, 1553–1560. Lutterschmidt, W. I. and Hutchison, V. H. (1997b). The critical thermal maximum: history and critique. Can. J. Zool. 75, 1561–1574. MacCrimmon, H. R. (1971). World distribution of rainbow trout (Salmo gairdneri). J. Fish. Res. Board Canada 28, 663–704. MacLennan, D. H. and Kranias, E. G. (2003). Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol 4, 566–577. Mangoni, M. E. and Nargeot, J. (2008). Genesis and regulation of the heart automaticity. Physiol. Rev. 88, 919–982. Martínez-Álvarez, R. M., Morales, A. E. and Sanz, A. (2005). Antioxidant defenses in fish: biotic and abiotic factors. Rev. Fish Biol. Fish. 15, 75–88. Matala, A. P., Hess, J. E. and Narum, S. R. (2011). Resolving adaptive and demographic divergence among Chinook salmon populations in the Columbia River Basin. Trans. Am. Fish. Soc. 140, 783–807. Matthews, K. R. and Berg, N. H. (1997). Rainbow trout responses to water temperature and dissolved oxygen stress in two southern California stream pools. J. Fish Biol. 50, 50–67. McBryan, T. L., Anttila, K., Healy, T. M. and Schulte, P. M. (2013). Responses to temperature and hypoxia as interacting stressors in fish: Implications for adaptation to environmental change. Integr. Comp. Biol. 53, 648–659. McCauley, R. (1958). Thermal relations of geographic races of Salvelinus. Can. J. Zool. 36, 655–662. McCauley, R. W. and Pond, W. L. (1971). Temperature selection of rainbow trout (Salmo gairdneri) fingerlings in vertical and horizontal gradients. J. Fish. Res. Board Canada 28, 1801–1804. McCullough, D., Spalding, S., Sturdevant, D. and Hicks, M. (2001). Issue Paper 5 Summary of technical literature examining the physiological effects of temperature on salmonids. McDonald, D. G. and McMahon, B. R. (1977). Respiratory development in Arctic char Salvelinus alpinus under conditions of normoxia and chronic hypoxia. Can. J. Zool. 55, 1461–1467. McGeer, J. C., Barnyi, L. and Iwama, G. K. (1991). Physiological responses to challenge tests in six stocks of coho salmon (Oncorhynchus kisutch). Can. J. Fish. Aquat. Sci. 48, 1761–1771. McKinney, G., Seeb, L., Larson, W., Gomez-Uchida, D., Limborg, M., Brieuc, M., Everett, M., Naish, K., Waples, R. and Seeb, J. (2016). An integrated linkage map reveals candidate genes underlying adaptive variation in Chinook salmon (Oncorhynchus tshawytscha). Mol. Ecol. Resour. 16, 769–783. McMahon, T. E., Bear, B. A. and Zale, A. V. (2006). Comparative thermal preferences of westslope cutthroat trout and rainbow trout. Wild Fish Habitat Initiative Montana Water Center at Montana State University-Bozeman Partners for Fish and Wildlife Program, U.S. Fish and Wildlife Service. McMahon, B. J., Teeling, E. C. and Hoglund, J. (2014). How and why should we implement genomics into conservation? Evol. Appl. 7, 999–1007. 200  McMurchie, E. J., Raison, J. K. and Cairncross, K. D. (1973). Temperature-induced phase changes in membranes of heart: A contrast between the thermal response of poikilotherms and homeotherms. Comp. Biochem. Physiol. -- Part B Biochem. 44, 1017–1026. Meffe, G. K., Weeks, S. C., Mulvey, M. and Kandl, K. L. (1995). Genetic differences in thermal tolerance of eastern mosquitofish (Gambusia holbrooki; Poeciliidae) from ambient and thermal ponds. Can. J. Fish. Aquat. Sci. 52, 2704–2711. Meier, K., Hansen, M. M., Normandeau, E., Mensberg, K.-L. D., Frydenberg, J., Larsen, P. F., Bekkevold, D. and Bernatchez, L. (2014). Local adaptation at the transcriptome level in brown trout: evidence from early life history temperature genomic reaction norms. PLoS ONE 9, e85171. Meka, J. M. and McCormick, S. D. (2005). Physiological response of wild rainbow trout to angling: Impact of angling duration, fish size, body condition, and temperature. Fish. Res. 72, 311–322. Meyer, K. A., Lamansky, J. A. and Schill, D. J. (2010). Biotic and abiotic factors related to redband trout occurrence and abundance in desert and montane streams. West. North Am. Nat. 70, 77–91. Miller, M. R., Dunham, J. P., Amores, A., Cresko, W. A. and Johnson, E. A. (2007). Rapid and cost-effective polymorphism identification and genotyping using restriction site associated DNA ( RAD ) markers. Genome Res. 17, 240–248. Miller, S. C., Gillis, T. E. and Wright, P. A. (2011). The ontogeny of regulatory control of the rainbow trout (Oncorhynchus mykiss) heart and how this is influenced by chronic hypoxia exposure. J. Exp. Biol. 214, 2065–2072. Miller, M. R., Brunelli, J. P., Wheeler, P. A., Liu, S., Rexroad, C. E., Palti, Y., Doe, C. Q. and Thorgaard, G. H. (2012). A conserved haplotype controls parallel adaptation in geographically distant salmonid populations. Mol. Ecol. 21, 237–249. Molony, B. (2001). Environmental requirements and tolerances of Rainbow trout (Oncorhynchus mykiss)and Brown trout (Salmo trutta) with special reference to Western Australia : A review. Western Australia: Department of Fisheries. Molony, B. W., Church, A. R. and Maguire, G. B. (2004). A comparison of the heat tolerance and growth of a selected and non-selected line of rainbow trout, Oncorhynchus mykiss, in Western Australia. Aquaculture 241, 655–665. Morimoto, R. I., Sarge, K. D. and Abravaya, K. (1992). Transcriptional regulation of heat shock genes. A paradigm for inducible genomic responses. J. Biol. Chem. 267, 21987–21990. Morrissy, N. M. (1973). Comparison of strains of Salmo gairdneri Richardson from New South Wales, Victoria and Western Australia. Aust. Soc. Limnol. Bull. 5, 11–20. Morrissy, N., Hambleton, S., Gill, H. and Morgan, D. (2002). The translocation of Brown trout (Salmo trutta) and Rainbow trout (Oncorhynchus mykiss) into and within Western Australia. Perth: Department of Fisheries, Western Australia. Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L. and Wold, B. (2008). Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods 5, 621–628. Moyle, P. B., Kiernan, J. D., Crain, P. K. and Quiñones, R. M. (2013). Climate change vulnerability of native and alien freshwater fishes of California: a systematic assessment approach. PLoS ONE 8, e63883. 201  Muhlfeld, C. C. (2002). Spawning characteristics of redband trout in a headwater stream in montana. North Am. J. Fish. Manag. 22, 1314–1320. Muhlfeld, C., Bennett, D. and Marotz, B. (2001). Summer habitat use by Columbia River redband trout in the Kootenai River drainage, Montana. North Am. J. Fish. Manag. 21, 223–235. Mundahl, N. D. (1990). Heat death of fish in shrinking stream pools. Am. Midl. Nat. 123, 40–46. Muñoz, N. J., Anttila, K., Chen, Z., Heath, J. W., Farrell, A. P. and Neff, B. D. (2014). Indirect genetic effects underlie oxygen-limited thermal tolerance within a coastal population of chinook salmon. Proc. R. Soc. B 281, 20141082. Muñoz, N. J., Farrell, A. P., Heath, J. W. and Neff, B. D. (2015). Adaptive potential of a Pacific salmon challenged by climate change. Nat. Clim. Chang. 5, 163–166. Myrick, C. A. and Cech, J. J. (2000a). Temperature influences on California rainbow trout physiological performance. Fish Physiol. Biochem. 22, 245–254. Myrick, C. A. and Cech, J. J. (2000b). Swimming performances of four California stream fishes: temperature effects. Environ. Biol. Fishes 58, 289–295. Myrick, C. and Cech, J. (2005). Effects of temperature on the growth, food consumption, and thermal tolerance of age-0 Nimbus-Strain steelhead. North Am. J. Aquac. 67, 324–330. Myrick, C. A., Folgner, D. K. and Cech, J. J. (2004). An annular chamber for aquatic animal preference studies. Trans. Am. Fish. Soc. 133, 427–433. Nadeau, N. J. and Jiggins, C. D. (2010). A golden age for evolutionary genetics? Genomic studies of adaptation in natural populations. Trends Genet. 26, 484–492. Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., Gerstein, M. and Snyder, M. (2008). The transcriptional landscape of the yeast genome defined by RNA sequencing. Science 320, 1344–1349. Narum, S. R. (2006). Beyond Bonferroni: less conservative analyses for conservation genetics. Conserv. Genet. 7, 783–787. Narum, S. R. and Campbell, N. R. (2015). Transcriptomic response to heat stress among ecologically divergent populations of redband trout. BMC genomics 16, 103. Narum, S. R. and Hess, J. E. (2011). Comparison of FST outlier tests for SNP loci under selection. Mol. Ecol. Resour. 11, 184–194. Narum, S. R., Campbell, N. R., Kozfkay, C. C. and Meyer, K. A. (2010). Adaptation of redband trout in desert and montane environments. Mol. Ecol. 19, 4622–4637. Narum, S. R., Buerkle, C. A., Davey, J. W., Miller, M. R. and Hohenlohe, P. A. (2013a). Genotyping-by-sequencing in ecological and conservation genomics. Mol. Ecol. 22, 2841–2847. Narum, S. R., Campbell, N. R., Meyer, K. A., Miller, M. R. and Hardy, R. W. (2013b). Thermal adaptation and acclimation of ectotherms from differing aquatic climates. Mol. Ecol. 22, 3090–3097. Neal, A. P., Molina-Campos, E., Marrero-Rosado, B., Bradford, A. B., Fox, S. M., Kovalova, N. and Hannon, H. E. (2010). CaMKK-CaMKI signaling pathways differentially control axon and dendrite elongation in cortical neurons. J. Neurosci. 30, 2807–2809. 202  Nehlsen, W., Williams, J. E. and Lichatowich, J. A. (1991). Pacific salmon at the crossroads: stocks at risk from California, Oregon, Idaho, and Washington. Fisheries 16, 4–21. Nei, M. (1978). Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89, 583–590. Newman, J. R. S., Ghaemmaghami, S., Ihmels, J., Breslow, D. K., Noble, M., DeRisi, J. L. and Weissman, J. S. (2006). Single-cell proteomic analysis of S. cerevisiae reveals the architecture of biological noise. Nature 441, 840–846. Newton, J. R., De Santis, C. and Jerry, D. R. (2012). The gene expression response of the catadromous perciform barramundi Lates calcarifer to an acute heat stress. J. Fish Biol. 81, 81–93. Ngugi, C. C. and Green, J. M. (2007). Anthropogenic influences on population sizes, age and growth of naturalized rainbow trout, Oncorhynchus mykiss, in Kenya. African J. Ecol. 45, 86–93. Nguyen, T. D. and Jeserich, G. (1998). Molecular structure and expression of shaker type potassium channels in glial cells of trout CNS. J. Neurosci. Res. 51, 284–292. Nielsen, J. L., Lisle, T. E. and Ozaki, V. (1994). Thermally stratified pools and their use by steelhead in Northern California streams. Trans. Am. Fish. Soc. 123, 613–626. Nilsson, G. E., Crawley, N., Lunde, I. G. D. A. G. and Munday, P. L. (2009). Elevated temperature reduces the respiratory scope of coral reef fishes. Glob. Chang. Biol. 15, 1405–1412. Norin, T. and Clark, T. D. (2016). Measurement and relevance of maximum metabolic rate in fishes. J. Fish Biol. 88, 122–151. Norin, T., Malte, H. and Clark, T. D. (2014). Aerobic scope does not predict the performance of a tropical eurythermal fish at elevated temperatures. J. Exp. Biol. 217, 244–251. Nosil, P., Funk, D. J. and Ortiz-Barrientos, D. (2009). Divergent selection and heterogeneous genomic divergence. Mol. Ecol. 18, 375–402. O’Malley, K. G., Sakamoto, T., Danzmann, R. G. and Ferguson, M. M. (2003). Quantitative trait loci for spawning date and body weight in rainbow trout: Testing for conserved effects across ancestrally duplicated chromosomes. J. Hered. 94, 273–284. Ojima, N., Mekuchi, M., Ineno, T., Tamaki, K., Kera, A., Kinoshita, S., Asakawa, S. and Watabe, S. (2012). Differential expression of heat-shock proteins in F2 offspring from F1 hybrids produced between thermally selected and normal rainbow trout strains. Fish. Sci. 78, 1051–1057. Ojuka, E. O. (2004). Role of calcium and AMP kinase in the regulation of mitochondrial biogenesis and GLUT4 levels in muscle. Proc. Nutr. Soc. 63, 275–278. Ospina, A. F. and Mora, C. (2004). Effect of body size on reef fish tolerance to extreme low and high temperatures. Environ. Biol. Fishes 70, 339–343. Overgaard, J., Stecyk, J. A. W., Gesser, H., Wang, T. and Farrell, A. P. (2004). Effects of temperature and anoxia upon the performance of in situ perfused trout hearts. J. Exp. Biol. 207, 655–665. Overgaard, J., Andersen, J. L., Findsen, A., Pedersen, P. B. M., Hansen, K., Ozolina, K. and Wang, T. (2012). Aerobic scope and cardiovascular oxygen transport is not 203  compromised at high temperatures in the toad Rhinella marina. J. Exp. Biol. 215, 3519–3526. Palti, Y., Vallejo, R. L., Gao, G., Liu, S., Hernandez, A. G., Rexroad III, C. E. and Wiens, G. D. (2015). Detection and validation of QTL affecting bacterial cold water disease resistance in rainbow troutusing restriction-site associated DNA sequencing. PLoS ONE 10, e0138435. Parsons, J. and Thorgaard, G. (1985). Production of androgenetic diploid rainbow trout. J. Hered. 76, 177–181. Pearson, M. P. and Stevens, E. D. (1991). Size and hematological impact of the splenic erythrocyte reservoir in rainbow trout, Oncorhynchus mykiss. Fish Physiol. Biochem. 9, 39–50. Perry, G. M. L., Danzmann, R. G., Ferguson, M. M. and Gibson, J. P. (2001). Quantitative trait loci for upper thermal tolerance in outbred strains of rainbow trout (Oncorhynchus mykiss). Heredity 86, 333–341. Perry, G. M. L., Martyniuk, C. M., Ferguson, M. M. and Danzmann, R. G. (2005). Genetic parameters for upper thermal tolerance and growth-related traits in rainbow trout (Oncorhynchus mykiss). Aquaculture 250, 120–128. Phillips, R. B., Nichols, K. M., DeKoning, J. J., Morasch, M. R., Keatley, K. A., Rexroad, C., Gahr, S. A., Danzmann, R. G., Drew, R. E. and Thorgaard, G. H. (2006). Assignment of rainbow trout linkage groups to specific chromosomes. Genetics 174, 1661–1670. Podrabsky, J. E. and Somero, G. N. (2004). Changes in gene expression associated with acclimation to constant temperatures and fluctuating daily temperatures in an annual killifish Austrofundulus limnaeus. J. Exp. Biol. 207, 2237–2254. Podrabsky, J. E., Javillonar, C., Hand, S. C. and Crawford, D. L. (2000). Intraspecific variation in aerobic metabolism and glycolytic enzyme expression in heart ventricles. Am. J. Physiol. Integr. Comp. Physiol. 279, R2344–R2348. Porcelli, D., Butlin, R. K., Gaston, K. J., Joly, D. and Snook, R. R. (2015). The environmental genomics of metazoan thermal adaptation. Heredity 114, 502–514. Pörtner, H. O. (2001). Climate change and temperature-dependent biogeography: oxygen limitation of thermal tolerance in animals. Die Naturwissenschaften 88, 137–146. Pörtner, H. O. (2002). Climate variation and physiology basis to temperature dependent biogeography: sistemic to molecular hierarchy of thermal tolerance in animals. Comp. Biochem. Physiol. Part A 132, 739–761. Pörtner, H. O. (2010). Oxygen-and capacity-limitation of thermal tolerance: a matrix for integrating climate-related stressor effects in marine ecosystems. J. Exp. Biol. 213, 881–893. Pörtner, H. O. and Farrell, A. P. (2008). ECOLOGY: Physiology and Climate Change. Science 322, 690–692. Pörtner, H. O. and Giomi, F. (2013). Nothing in experimental biology makes sense except in the light of ecology and evolution – correspondence on J. Exp. Biol. 216, 2771-2782. J. Exp. Biol. 216, 4494–4495. Pörtner, H. O. and Knust, R. (2007). Climate change affects marine fishes through the oxygen limitation of thermal tolerance. Science 315, 95–97. 204  Pörtner, H. O., Peck, L. and Somero, G. (2007). Thermal limits and adaptation in marine Antarctic ectotherms : an integrative view. Society 362, 2233–2258. Prodocimo, V. and Freire, C. A. (2001). Criticai thermal maxima and minima of the platyfish Xiphophorus maculatus Günther ( Poecillidae , Cyprinodontiformes ) - a tropical species of ornamental freshwater fish. Revta Btrs. Zool. 18, 97–106. Prosser, C. L. (1955). Physiological Variation In Animals. Biol. Rev. 30, 229–261. Qiu, X. B., Shao, Y. M., Miao, S. and Wang, L. (2006). The diversity of the DnaJ/Hsp40 family, the crucial partners for Hsp70 chaperones. Cell. Mol. Life Sci. 63, 2560–2570. Quinn, T. P. (2005). The Behavior and Ecology of Pacific Salmon and Trout. 391. Quinn, N. L., McGowan, C. R., Cooper, G. A., Koop, B. F. and Davidson, W. S. (2011). Identification of genes associated with heat tolerance in Arctic charr exposed to acute thermal stress. Physiol. genomics 43, 685–696. R Core Team (2013). R Development Core Team. Vienna, Austria: R Foundation for Statistical Computing. Randall, D. J. and Daxboeck, C. (1982). Cardiovascular changes in the rainbow trout (Salmo Gairdneri Richardson) during exercise. Can. J. Zool. 60, 1135–1140. Raymond, M. and Rousset, F. (1995). GENEPOP (Version 1.2): Population genetics software for exact tests and ecumenicism. J. Hered. 86, 248–249. Recsetar, M. S., Zeigler, M. P., Ward, D. L., Bonar, S. A. and Caldwell, C. A. (2012). Relationship between fish size and upper thermal tolerance. Trans. Am. Fish. Soc. 141, 1433–1438. Reidy, S. P., Nelson, J., Tang, Y. and Kerr, S. R. (1995). Postexercise metabolic rate in Atlantic cod and its dependence upon the method of exhaustion. J. Fish Biol. 47, 377–386. Rexroad, C. E. and Vallejo, R. L. (2009). Estimates of linkage disequilibrium and effective population size in rainbow trout. BMC Genet. 10, 83. Reynolds, W. W. and Casterlin, M. E. (1979). Behavioral thermoregulation and the “final preferendum” paradigm. Am. Zool. 19, 211–224. Richter, A. and Kolmes, S. A. (2005). Maximum temperature limits for chinook, coho, and chum salmon, and steelhead trout in the Pacific Northwest. Rev. Fish. Sci. 13, 23–49. Ried, T., Rudy, B., de Miera, E. V.-S., Lau, D., Ward, D. C. and Sen, K. (1993). Localization of a highly conserved human potassium channel gene (NGK2-KV4; KCNC1) to chromosome 11p15. Genomics 15, 405–411. Riehle, M. M., Bennett, A. F. and Long, A. D. (2001). Genetic architecture of thermal adaptation in Escherichia coli. Proc. Natl. Acad. Sci. 98, 525–30. Rincent, R., Moreau, L., Monod, H., Kuhn, E., Melchinger, A. E., Malvar, R. A., Moreno-Gonzalez, J., Nicolas, S., Madur, D., Combes, V., et al. (2014). Recovering power in association mapping panels with variable levels of linkage disequilibrium. Genetics 197, 375–387. Rissanen, E., Tranberg, H. K., Sollid, J., Nilsson, G. E. and Nikinmaa, M. (2006). Temperature regulates hypoxia-inducible factor-1 (HIF-1) in a poikilothermic vertebrate, crucian carp (Carassius carassius). J. Exp. Biol. 209, 994–1003. Ritossa, F. (1962). A new puffing pattern induced by temperature shock and DNP in drosophila. Experientia 18, 571–573. 205  Robinson, M. D., McCarthy, D. J. and Smyth, G. K. (2010). edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140. Roden, D. M., Balser, J. R., George, A. L. and Anderson, M. E. (2002). Cardiac ion channels. Annu. Rev. Physiol. 64, 431–475. Rodnick, K. J., Gamperl, A. K., Lizars, K. R., Bennett, M. T., Rausch, R. N. and Keeley, E. R. (2004). Thermal tolerance and metabolic physiology among redband trout populations in south - eastern Oregon. J. Fish Biol. 64, 310–335. Rodnick, K. J., Gamperl, A. K., Nash, G. W. and Syme, D. A. (2014). Temperature and sex dependent effects on cardiac mitochondrial metabolism in Atlantic cod (Gadus morhua L.). J. Therm. Biol. 44, 110–118. Rummer, J. L., Couturier, C. S., Stecyk, J. a W., Gardiner, N. M., Kinch, J. P., Nilsson, G. E. and Munday, P. L. (2014). Life on the edge: thermal optima for aerobic scope of equatorial reef fishes are close to current day temperatures. Glob. Chang. Biol. 20, 1055–1066. Saito, S., Nakatsuka, K., Takahashi, K., Fukuta, N., Imagawa, T., Ohta, T. and Tominaga, M. (2012). Analysis of transient receptor potential ankyrin 1 (TRPA1) in frogs and lizards illuminates both nociceptive heat and chemical sensitivities and coexpression with TRP vanilloid 1 (TRPV1) in ancestral vertebrates. J. Biol. Chem. 287, 30743–30754. Sandblom, E. and Axelsson, M. (2007). Venous hemodynamic responses to acute temperature increase in the rainbow trout (Oncorhynchus mykiss). Am. J. Physiol. - Regul. Integr. Comp. Physiol. 292, R2292–R2298. Sandblom, E. and Axelsson, M. (2011). Autonomic control of circulation in fish: A comparative view. Auton. Neurosci. Basic Clin. 165, 127–139. Scheerer, P. D., Thorgaard, G. H. and Allendorf, F. W. (1991). Genetic analysis of androgenetic rainbow trout. J. Exp. Zool. 260, 382–390. Schlötterer, C., Tobler, R., Kofler, R. and Nolte, V. (2014). Sequencing pools of individuals — mining genome-wide polymorphism data without big funding. Nat. Rev. Genet. 15, 749–763. Schmidt-Nielsen, K. (1984). Scaling, why is animal size so important? Cambridge; New York; Cambridge University Press. Schulte, P. M. (2015). The effects of temperature on aerobic metabolism: towards a mechanistic understanding of the responses of ectotherms to a changing environment. J. Exp. Biol. 218, 1856–1866. Schulte, P. M., Glémet, H. C., Fiebig, A. A. and Powers, D. A. (2000). Adaptive variation in lactate dehydrogenase-B gene expression: Role of a stress-responsive regulatory element. Proc. Natl. Acad. Sci. USA 97, 6597–6602. Schulte, P. M., Healy, T. M. and Fangue, N. A. (2011). Thermal performance curves, phenotypic plasticity, and the time scales of temperature exposure. Integr. Comp. Biol. 51, 691–702. Scott, M. A., Dhillon, R. S., Schulte, P. M., Richards, J. G. and Magnan, P. (2015). Physiology and performance of wild and domestic strains of diploid and triploid rainbow trout ( Oncorhynchus mykiss ) in response to environmental challenges. Can. J. Fish. Aquat. Sci. 72, 125–134. 206  Seebacher, F. (2009). Responses to temperature variation: integration of thermoregulation and metabolism in vertebrates. J. Exp. Biol. 212, 2885–2891. Shao, J., Prince, T., Hartson, S. D. and Matts, R. L. (2003). Phosphorylation of serine 13 Is required for the proper function of the Hsp90 co-chaperone, Cdc37. J. Biol. Chem. 278, 38117–38120. Shiels, H. A. and Farrell, A. P. (1997). The effect of temperature and adrenaline on the relative importance of the sarcoplasmic reticulum in contributing Ca2+ to force development in isolated ventricular trabeculae from rainbow trout. J. Exp. Biol. 200, 1607–1621. Shiels, H., Stevens, E. and Farrell, A. (1998). Effects of temperature, adrenaline and ryanodine on power production in rainbow trout oncorhynchus mykiss ventricular trabeculae. J. Exp. Biol. 201, 2701–2710. Shiels, H. A., Vornanen, M. and Farrell, A. P. (2003a). Acute temperature change modulates the response of ICa to adrenergic stimulation in fish cardiomyocytes. Physiol. Biochem. Zool. 76, 816–24. Shiels, H. A., Vornanen, M. and Farrell, A. P. (2003b). Acute temperature change modulates the response of ICa to adrenergic stimulation in fish cardiomyocytes. Physiol. Biochem. Zool. 76, 816–824. Sidhu, R., Anttila, K. and Farrell, A. P. (2014). Upper thermal tolerance of closely related Danio species. J. Fish Biol. 84, 982–995. Somero, G. N. (2003). Protein adaptations to temperature and pressure: Complementary roles of adaptive changes in amino acid sequence and internal milieu. Comp. Biochem. Physiol. - B Biochem. Mol. Biol. 136, 577–591. Somero, G. N. (2010). The physiology of climate change: how potentials for acclimatization and genetic adaptation will determine “winners” and “losers”. J. Exp. Biol. 213, 912–920. Somorjai, I. M. L., Danzmann, R. G. and Ferguson, M. M. (2003). Distribution of temperature tolerance quantitative trait loci in Arctic charr (Salvelinus alpinus) and inferred homologies in rainbow trout (Oncorhynchus mykiss). Genetics 165, 1443–1456. Sonah, H., O’Donoughue, L., Cober, E., Rajcan, I. and Belzile, F. (2015). Identification of loci governing eight agronomic traits using a GBS-GWAS approach and validation by QTL mapping in soya bean. Plant Biotechnol. J. 13, 211–221. Sonski, A. J. (1982). Heat tolerance of redband trout. Annual Proceedings of the Texas Chapter American Fisheries Society. Sonski, A. J. (1983). comparson of heat tolerances of redband trout, firehole river rainbow trout and Whtheville rainbow trout. Ingram, Texas: Texas Parks and Wildlife Department. Sørensen, J. G., Norry, F. M., Scannapieco, A. C. and Loeschcke, V. (2005). Altitudinal variation for stress resistance traits and thermal adaptation in adult Drosophila buzzatii from the New World. J. Evol. Biol. 18, 829–37. Spencer, C. C. A., Su, Z., Donnelly, P. and Marchini, J. (2009). Designing genome-wide association studies: Sample size, power, imputation, and the choice of genotyping chip. PLoS Genet. 5, e1000477. 207  Spierts, Y. I. L., Akster, A. H. and Granzier, L. H. (1997). Expression of titin isoforms in red and white muscle fibres of carp (Cyprinus carpio L.) exposed to different sarcomere strains during swimming. J. Comp. Physiol. B 167, 543–551. Stapley, J., Reger, J., Feulner, P. G. D., Smadja, C., Galindo, J., Ekblom, R., Bennison, C., Ball, A. D., Beckerman, A. P. and Slate, J. (2010). Adaptation genomics: the next generation. Trends Ecol. Evol. 25, 705–712. Stefanovic, B., Stefanovic, L., Schnabl, B., Bataller, R. and Brenner, D. A. (2004). TRAM2 protein interacts with endoplasmic reticulum Ca2+ pump Serca2b and is necessary for collagen type I synthesis. Mol. Cell. Biol. 24, 1758–1768. Steinhausen, M. F., Sandblom, E., Eliason, E. J., Verhille, C. and Farrell, A. P. (2008). The effect of acute temperature increases on the cardiorespiratory performance of resting and swimming sockeye salmon (Oncorhynchus nerka). J. Exp. Biol. 211, 3915–3926. Stillman, J. H. and Armstrong, E. (2015). Genomics are transforming our understanding of responses to climate change. BioScience 65, 237–246. Strange, R. J., Petrie, R. B. and Cech, J. J. (1993). Slight stress does not lower critical thermal maximums in hatchery-reared rainbow trout. FOLIA Zool. 42, 251–256. Stuenkel, E. L. and Hillyard, S. D. (1981). The effects of temperature and salinity acclimation on metabolic rate and osmoregulation in the pupfish Cyprinodon salinus. Copeia 1981, 411–417. Supek, F., Bošnjak, M., Škunca, N. and Šmuc, T. (2011). REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS ONE 6, e21800. Tamkee, P., Parkinson, E. and Taylor, E. B. (2010). The influence of Wisconsinan glaciation and contemporary stream hydrology on microsatellite DNA variation in rainbow trout (Oncorhynchus mykiss). Can. J. Fish. Aquat. Sci. 67, 919–935. Tan, E., Wongwarangkana, C., Kinoshita, S., Suzuki, Y., Oshima, K., Hattori, M., Ineno, T., Tamaki, K., Kera, A., Muto, K., et al. (2012). Global gene expression analysis of gill tissues from normal and thermally selected strains of rainbow trout. Fish. Sci. 78, 1041–1049. Tan, E., Kinoshita, S., Suzuki, Y., Ineno, T., Tamaki, K., Kera, A., Muto, K., Yada, T., Kitamura, S., Asakawa, S., et al. (2016). Different gene expression profiles between normal and thermally selected strains of rainbow trout, Oncorhynchus mykiss, as revealed by comprehensive transcriptome analysis. Gene 576, 637–643. Tate, P. H. and Bird, A. P. (1993). Effects of DNA methylation on DNA-binding proteins and gene expression. Curr. Opin. Genet. Dev. 3, 226–231. Taylor, B. T. (1991). A review of local adaptation in Saimonidae with particular reference to Pacific and Atlantic salmon. Aquaculture 98, 185–207. Taylor, S., Egginton, S. and Taylor, E. (1996). Seasonal temperature acclimatisation of rainbow trout: cardiovascular and morphometric influences on maximal sustainable exercise level. J. Exp. Biol. 199, 835–845. Tenenbaum, D. (2016). KEGGREST: Client-side REST access to KEGG. R Packag. version 1.12.2. Tepolt, C. K. and Somero, G. N. (2014). Master of all trades: thermal acclimation and adaptation of cardiac function in a broadly distributed marine invasive species, the European green crab, Carcinus maenas. J. Exp. Biol. 217, 1129–1138. 208  Thomson, D. M., Porter, B. B., Tall, J. H., Kim, H.-J., Barrow, J. R. and Winder, W. W. (2007). Skeletal muscle and heart LKB1 deficiency causes decreased voluntary running and reduced muscle mitochondrial marker enzyme expression in mice. Am. J. Physiol. Endocrinol. Metab. 292, E196–E202. Thorne, M. A. S., Burns, G., Fraser, K. P. P., Hillyard, G. and Clark, M. S. (2010). Transcription profiling of acute temperature stress in the Antarctic plunderfish Harpagifer antarcticus. Mar. genomics 3, 35–44. Tomalty, K. M. H., Meek, M. H., Stephens, M. R., Rincón, G., Fangue, N. A., May, B. P. and Baerwald, M. R. (2015). Transcriptional response to acute thermal exposure in juvenile chinook salmon determined by RNAseq. G3 (Bethesda, Md.) 5, 1335–1349. Tomanek, L. (2015). Proteomic responses to environmentally induced oxidative stress. J. Exp. Biol. 218, 1867–1879. Tomanek, L. and Somero, G. N. (2000). Time course and magnitude of synthesis of heat-shock proteins in congeneric marine snails (Genus Tegula) from different tidal heights. Physiol. Biochem. Zool. 73, 249–256. Turner, S. D. (2014). qqman: an R package for visualizing GWAS results using Q-Q and manhattan plots. bioRxiv. Underwood, Z. E., Myrick, C. A. and Rogers, K. B. (2012). Effect of acclimation temperature on the upper thermal tolerance of Colorado River cutthroat trout Oncorhynchus clarkii pleuriticus: thermal limits of a North American salmonid. J. fish Biol. 80, 2420–2433. VanRaden, P. M. (2008). Efficient methods to compute genomic predictions. J. dairy Sci. 91, 4414–4423. Varriale, A. and Bernardi, G. (2006). DNA methylation and body temperature in fishes. Gene 385, 111–121. Verhille, C. and Farrell, A. P. (2012). The in vitro blood-O2 affinity of triploid rainbow trout Oncorhynchus mykiss at different temperatures and CO2 tensions. J. Fish Biol. 81, 1124–1132. Verhille, C., Anttila, K. and Farrell, A. P. (2013). A heart to heart on temperature: Impaired temperature tolerance of triploid rainbow trout (Oncorhynchus mykiss) due to early onset of cardiac arrhythmia. Comp. Biochem. Physiol. Part A, Mol. Integr. Physiol. 164, 653–657. Verhille, C. E., English, K. K., Cocherell, D. E., Farrell, A. P. and Fangue, N. A. (2016). High thermal tolerance of a rainbow trout population near its southern range limit suggests local thermal adjustment. Physiol. Biochem. Zool. In press,. Vinagre, C., Madeira, D., Narciso, L., Cabral, H. N. and Diniz, M. (2012). Effect of temperature on oxidative stress in fish: Lipid peroxidation and catalase activity in the muscle of juvenile seabass, Dicentrarchus labrax. Ecol. Indic. 23, 274–279. Vinson, M. and Levesque, S. (1994). Redband trout response to hypoxia in a natural environment. Gt. Basin Nat. 54, 150–155. Viswanath, V., Story, G. M., Peier, A. M., Petrus, M. J., Lee, V. M., Hwang, S. W., Patapoutian, A. and Jegla, T. (2003). Ion channels: opposite thermosensor in fruitfly and mouse. Nature 423, 822–823. 209  Vornanen, M. (1997). Sarcolemmal Ca influx through L-type Ca channels in ventricular myocytes of a teleost fish. Am. J. Physiol. - Regul. Integr. Comp. Physiol. 272, R1432–R1440. Vornanen, M. (2016). The temperature dependence of electrical excitability in fish hearts. J. Exp. Biol. 219, 1941–1952. Vornanen, M., Ryökkynen, A. and Nurmi, A. (2002). Temperature-dependent expression of sarcolemmal K+ currents in rainbow trout atrial and ventricular myocytes. Am. J. Physiol. Regul. Integr. Comp. Physiol. 282, R1191–R1199. Vornanen, M., Hassinen, M., Koskinen, H. and Krasnov, A. (2005). Steady-state effects of temperature acclimation on the transcriptome of the rainbow trout heart. Am. J. Physiol. Regul. Integr. Comp. Physiol. 289, R1177–R1184. Vornanen, M., Haverinen, J. and Egginton, S. (2014). Acute heat tolerance of cardiac excitation in the brown trout (Salmo trutta fario). Clin. Cancer Res. 217, 299–309. Wandinger, S. K., Richter, K. and Buchner, J. (2008). The Hsp90 chaperone machinery. J. Biol. Chem. 283, 18473–18477. Wang, L. Y., Gan, L., Forsythe, I. D. and Kaczmarek, L. K. (1998). Contribution of the Kv3.1 potassium channel to high frequency firing in mouse auditory neurones. J. Physiol. 509, 183–194. Wang, S., Meyer, E., McKay, J. K. and Matz, M. V (2012). 2b-RAD: a simple and flexible method for genome-wide genotyping. Nat Meth 9, 808–810. Waples, R. S., Pess, G. R. and Beechie, T. J. (2008). Evolutionary history of Pacific salmon in dynamic environments. Evol. Appl. 1, 189–206. Ward, R. D., Jorstad, K. E. and Maguire, G. B. (2003). Microsatellite diversity in rainbow trout (Oncorhynchus mykiss) introduced to Western Australia. Aquaculture 219, 169–179. Watabe, S. (2002). Temperature plasticity of contractile proteins in fish muscle. J. Exp. Biol. 205, 2231–2236. Watanabe, E., Yasui, K., Kamiya, K., Yamaguchi, T., Sakuma, I., Honjo, H., Ozaki, Y., Morimoto, S., Hishida, H. and Kodama, I. (2007). Upregulation of KCNE1 induces QT interval prolongation in patients with chronic heart failure. Circ. J.  Off. J. Japanese Circ. Soc. 71, 471–478. Watt, W. B. (1977). Adaptation at specific loci. I. Natural selection on phosphoglucose isomerase of Colias butterflies: Biochemical and population aspects. Genetics 87, 177–94. Wdziczak, J., Zaleśna, G., Wujec, E. and Pérès, G. (1982). Comparative studies on superoxide dismutase, catalase and peroxidase levels in erythrocytes and livers of different freshwater and marine fish species. Comp. Biochem. Physiol. 73B, 361–365. Weir, B. S. and Cockerham, C. C. (1984). Estimating F-statistics for the analysis of population structure. Evolution 38, 1358–1370. Wendelaar Bonga, S. E. (1997). The stress response in fish. Physiol. Rev. 77, 591–625. Wieser, W. (1985). Developmental and metabolic constraints of the scope for activity in young rainbow trout (Salmo gairdneri). J. Exp. Biol. 142, 133–142. Wilson, C. M. and Farrell, A. P. (2013). Pharmacological characterization of the heartbeat in an extant vertebrate ancestor, the Pacific hagfish, Eptatretus stoutii. Comp. Biochem. Physiol. Part A, Mol. Integr. Physiol. 164, 258–263. 210  Wood, C. M., Pieprzak, P. and Trott, J. N. (1979). The influence of temperature and anaemia on the adrenergic and cholinergic mechanisms controlling heart rate in the rainbow trout. Can. J. Zool. 57, 2440–2447. Yau, M. M. and Taylor, E. B. (2014). Cold tolerance performance of westslope cutthroat trout ( Oncorhynchus clarkii lewisi ) and rainbow trout (Oncorhynchus mykiss) and its potential role in influencing interspecific hybridization. Can. J. Zool. 92, 777–784. Yeager, D. P. and Ultsch, G. R. (1989). Physiological regulation and conformation: a BASIC program for the determination of critical points. Physiol. Zool. 62, 888–907. Young, K. A., Hinch, S. G. and Northcote, T. G. (1999). Status of resident coastal cutthroat trout and their habitat twenty-five years after riparian logging. North Am. J. Fish. Manag. 19, 901–911. Yu, J., Pressoir, G., Briggs, W. H., Vroh Bi, I., Yamasaki, M., Doebley, J. F., McMullen, M. D., Gaut, B. S., Nielsen, D. M., Holland, J. B., et al. (2006). A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat. Genet. 38, 203–208. Zhang, Y. and Kieffer, J. D. (2014). Critical thermal maximum (CTmax) and hematology of shortnose sturgeons (Acipenser brevirostrum) acclimated to three temperatures. Can. J. Zool. 92, 215–221. Zhang, Z., Ersoz, E., Lai, C. Q., Todhunter, R. J., Tiwari, H. K., Gore, M. A., Bradbury, P. J., Yu, J., Arnett, D. K., Ordovas, J. M., et al. (2010). Mixed linear model approach adapted for genome-wide association studies. Nat. Genet. 42, 355–360. Zhao, S., Fung-Leung, W.-P., Bittner, A., Ngo, K. and Liu, X. (2014). Comparison of RNA-Seq and microarray in transcriptome profiling of activated T cells. PLoS ONE 9, e78644. Zoellick, B. W. (1999). Stream temperatures and the elevational distribution of redband trout in southwestern Idaho. Gt. Basin Nat. 59, 136–143. Zoellick, B. W. (2004). Density and biomass of redband trout relative to stream shading and temperature in southwestern Idaho. West. North Am. Nat. 64, 18–26.  211  Appendices Appendix A  Additional tables and figures for Chapter 1.  A.1 Critical thermal maximum (CTMAX) for nine sockeye salmon populations  Association between critical thermal maximum (CTMAX) and body mass for nine sockeye salmon populations from British Columbia, Canada (Chen et al., 2013): (■) Harrison River, (●) Weaver Creek, (▲) Gates Creek, (▼) Scotch Creek, (◆) Adams River, ( ) Chilko River, ( ) Horsefly River, (★) Stellako River, ( ) Okanagan River. CTMAX was measured twice: first at 90 days post hatch with large body mass differences (data excluded from the box) and second at 135 to 214 days post-hatch with similar body mass (data in the box). Egg incubation was at 10°C (open symbols), 14°C (solid symbols) or 16°C (half-solid symbols). The regression lines are fitted for the fish that were incubated at 10 and 14°C.  212   A.2 The CTMAX for various populations of O. mykiss Tacclim (°C) CTMAX (°C) Heating rate (°C min-1) Mass (g) Length (cm) Strain Source Reference 8 26.9 ± 0.12 0.1   11.0 - 18.0 Washington Becker and Wolford, 1980 9.8 27.9 ± 0.05 0.3  15.3 ± 0.25 Pennsylvania Carline and Machung, 2001 10 28.5 ± 0.28 0.02  15.0 - 20.0 Arizona Lee and Rinne, 1980 10 28.0 ± 0.12 0.3 ~15.0 ~10.0 Missouri Currie et al., 1998 10 27.7 ±0.08 0.3 12.9 ± 0.6  California Myrick and Cech, 2000 11 ~27.5 0.3 8.0 ± 1.6  California Myrick and Cech, 2005 11I 29.0 ± 0.05 0.3 2.4 ± 0.5  British Columbia Scott et al., 2015 13 27.9 ± 0.14 0.33 21.8 ± 0.4  Ontario LeBlanc et al., 2011 14 28.5± 0.11 0.3 13.8 ± 0.8  California Myrick and Cech, 2000 14 29.4 ± 0.1 0.033II 41.0-140.0  Oregon Rodnick et al., 2004 15 29.4 ± 0.08 0.3    Strange et al., 1993 15 29.1 ± 0.09 0.3 ~15.0 ~10.0 Missouri Currie et al., 1998 15 27.7 ± 0.03 0.0014 III 89.9 ± 5.4 19.9 ± 3.6 North Carolina Galbreath et al., 2006 15 ~28.4 0.3 9.3 ± 2.0  California Myrick and Cech, 2005 15 ~29.7 0.083 IV   Miyazaki, Japan Ineno et al., 2005 18 ~31.2 0.3  4.1 - 20 Arizona Recsetar et al., 2012 19 ~29.6 0.3 14.3 ± 2.9  California Myrick and Cech, 2005 19 29.9 ± 0.17 0.3 11.8 ± 0.7  California Myrick and Cech, 2000 20 29.4 ± 0.19 0.02  15 - 20 Arizona Lee and Rinne, 1980 20 29.8 ± 0.12 0.3 ~2.0 ~4.07 Missouri Currie et al., 1998 20 ~30.4 0.083 IV   Miyazaki, Japan Ineno et al., 2005 22 30.9 ± 0.13 0.3 9.29 ± 0.99  California Myrick and Cech, 2000 25 31.8 ± 0.1 0.3 6.1 ± 0.63  California Myrick and Cech, 2000 Note: I fish held at 10 ~12°C. II temperature was increased at 2°C h-1. III Temperature was increased at 2°C day-1.  IV temperature was increased at 5°C h-1. “~” indicates an estimated or the calculated mean by original author. Values separated by “-” represent the range of the traits. Other values were given as mean ± s.e.m. In all studies, loss of equilibrium was used as the endpoint of CTMAX.   213  Appendix B  Additional tables and figures for Chapter 3. B.1 Correlation between CTMAX and body mass    Correlation between critical thermal maximum (CTMAX) and body mass in redband trout in Idaho. Each point represents an individual and color represent its origin population (Little Jacks: ; hybrid: ; Keithley: ; Fawn: ). The breakpoint is at 1.5 g using two segment linear regression (R2=0.55, p < 0.001).     214  B.2 Body mass corrected RMR, MMR, AAS and FAS of redband trout.   Temp (°C) Little Jacks (LJ) K×LJ Keithley (K) Fawn (F) RMR 12 b0.15±0.02 b0.13±0.02 c0.17±0.01 c0.16±0.01 (mg O2 g−0.88 h−1) 15 b0.17±0.02 b0.17±0.03 bc0.20±0.02 bc0.21±0.03  18 b0.23±0.03 b0.15±0.02 bc0.22±0.03 bc0.25±0.04  21 ab0.27±0.04 ab0.24±0.03 ab0.32±0.05 ab0.34±0.04  24 a0.37±0.04 a0.30±0.05 a0.45±0.04 a0.46±0.07       MMR 12 c0.51±0.01B d0.49±0.02B b0.61±0.03A c0.55±0.02AB (mg O2 g−0.88 h−1) 15 bc0.56±0.03 cd0.61±0.04 b0.67±0.02 bc0.64±0.02  18 abc0.62±0.02 bc0.67±0.04 ab0.72±0.03 ab0.67±0.01  21 ab0.66±0.06B a0.84±0.04A a0.83±0.04A a0.74±0.03AB  24 a0.75±0.04 ab0.81±0.03 ab0.71±0.04 ab0.66±0.04       AAS 12 0.36±0.02 b0.36±0.03 ab0.44±0.03 a0.39±0.02 (mg O2 g−0.88 h−1) 15 0.39±0.03 ab0.44±0.07 a0.46±0.02 a0.42±0.03  18 0.39±0.04 ab0.52±0.04 a0.50±0.03 a0.42±0.03  21 0.38±0.05B a0.63±0.05A a0.52±0.08AB a0.41±0.04AB  24 0.38±0.04AB ab0.51±0.07A b0.26±0.06B b0.20±0.08B       FAS 12 3.94±0.84 4.24±0.77 a3.78±0.32 a3.54±0.32 (MMR/RMR) 15 3.53±0.46 4.42±1.00 a3.48±0.43 a3.17±0.36  18 3.01±0.55 5.02±0.94 a3.61±0.50 ab2.97±0.45  21 2.72±0.47 3.96±0.58 ab3.09±0.57 ab2.39±0.32  24 2.13±0.20AB 3.02±0.45A b1.65±0.18B b1.62±0.28B Note: Superscripted lowercase letters represent significant differences between temperatures within population. Superscripted uppercase letters represent significant differences between populations at the same temperature. All null hypothesis were tested by one way ANOVA and Tukey's post hoc test with significance level at α = 0.05.    215  B.3 Quadratic fitting analysis for AAS in redband trout populations.   For each strain, color shaded area represents over 90% of maximum absolute aerobic scope (AAS).      B.4 Parameters in quadratic fitting analysis for aerobic scope  Quadratic fitting  Optimum  Pejus  y0/a/b R2 p  Temp  (°C) AAS  Lower (°C) Upper (°C) AAS LJ 0.19807/0.02057/−0.00055 0.01 0.865  20.6 0.41  10.3 27.2 0.37 K×LJ −0.6982/0.1225/−0.0031 0.29 0.011  19.8 0.51  15.7 23.8 0.46 K −0.8043/0.1573/−0.0047 0.29 0.012  16.1 0.44  12.7 19.4 0.44 F −0.5732/0.1255/−0.0039 0.33 0.004  16.7 0.51  13.4 20.0 0.51 Note: LJ, Little Jacks; F, Fawn; K, Keithley; K×LJ, hybrid between Keithley and Little Jacks. Quadratic functions is y=y0+a*x+b*x^2. The unit of AAS is mg (O2) g−0.88 (fish) h−1 (hour).    216  Appendix C  Additional tables and figures for Chapter 4.  C.1 Alignment of RNA sequencing reads to a reference mRNA of rainbow trout Alignment of RNA sequencing reads to a reference mRNA of O. mykiss (Berthelot et al., 2014) for redband trout from Little Jacks Creek (LJ), Keithley Creek (K), Fawn Creek (F) and a hybrid between Keithley and Little Jacks (K×LJ). Strain ID Temperature (°C) Total reads (Million) Aligned (%) Mean Temp (°C) Mean Aligned (%) F 13 15.0 24.6 64.8 15.0 63.3 F 14 15.0 20.9 63.1   F 15 15.0 17.5 62.1   F 07 20.0 27.9 63.3 20.0 63.1 F 09 20.0 16.8 64.4   F 10 20.0 21.0 62.5   F 12 20.0 19.4 62.1   F 04 23.5 19.9 58.0 25.0  56.3 F 03 23.8 15.6 39.4   F 01 24.6 17.9 63.3   F 02 24.9 15.8 60.6   F 06 25.6 21.7 53.5   F 05 27.8 18.1 63.2   K×LJ 12 15.0 29.9 69.8 15.0 66.4 K×LJ 15 15.0 17.6 63.0   K×LJ 08 20.0 15.7 63.8 20.0 61.3 K×LJ 09 20.0 24.7 62.3   K×LJ 10 20.0 17.9 59.6   K×LJ 11 20.0 15.6 66.1   K×LJ 13 20.0 22.8 66.2   K×LJ 01 20.4 21.0 50.1   K×LJ 06 22.0 23.0 64.8 23.6  60.1 K×LJ 04 22.4 23.4 64.6   K×LJ 02 24.0 18.2 52.3   K×LJ 05 25.9 18.9 58.7   K 15 15.0 14.8 63.6 15.0 63.3 K 16 15.0 19.5 62.9   K 08 20.0 27.3 61.0 20.0 62.4 K 09 20.0 25.6 57.7   K 11 20.0 25.0 65.2   K 12 20.0 22.6 64.6   K 13 20.0 27.3 63.4   K 06 21.0 13.9 54.4 24.1  59.7 K 03 22.6 18.7 62.5   217  Strain ID Temperature (°C) Total reads (Million) Aligned (%) Mean Temp (°C) Mean Aligned (%) K 07 23.4 21.1 65.5   K 02 26.0 47.1 49.8   K 01 27.6 16.3 66.5   LJ 13 15.0 16.3 65.0 15.0 65.2 LJ 14 15.0 24.0 66.4   LJ 15 15.0 31.9 61.3   LJ 16 15.0 18.2 68.1   LJ 07 20.0 20.5 55.3 20.0 61.3 LJ 08 20.0 38.0 58.8   LJ 09 20.0 28.5 62.7   LJ 10 20.0 29.3 66.0   LJ 11 20.0 20.4 61.3   LJ 12 20.0 43.5 63.7   LJ 04 22.5 11.8 66.0 24.9 60.0 LJ 03 23.3 21.1 57.1   LJ 01 23.6 28.0 55.7   LJ 05 24.8 17.3 58.3   LJ 02 26.2 22.4 65.8   LJ 06 29.0 33.1 57.2      218  C.2 Fold change of genes during warming for each redband trout population     Fold change of genes with temperature increase from control (15°C) to arrhythmia temperature (TAR) in redband trout. All colored points (exclude grey) are statistically significant after correction for false discovery rate at 0.05. Blue points represent less than 2 fold changes. Yellow points represent 2-4 fold changes. Red points represent over 4 fold changes. Both fold changes and expression levels (count per million, CPM) are log transformed.  219  C.3 Gene ontology distribution of significantly regulated genes                     Gene ontology distribution of significantly up- (red) and down- (blue) regulated genes in Appendix C.2. (a) Biological process, (b) molecular function, (c) cellular component. (c) 220  C.4 Summary of significantly regulated genes in selected pathways. Bold numbers denote significant fold changes (FDR<0.05).   Gene CPM Fold change Location Protein Name [log2(*)] 15-20°C 20-22°C 15-22°C Cardiac muscle Contraction, map04260 GSONMT00065371001 myosin-7b myosin 7.7  -0.5  -0.5  -1.0  Arrhythmogenic right ventricular cardiomyopathy (ARVC), map05412  GSONMT00009120001 integrin alpha-6 isoform x1 itga 1.0  -0.4  1.0  0.6  GSONMT00079647001 desmocollin-1-like isoform x1 dsc 3.4  0.0  0.8  0.9  GSONMT00026322001 plakophilin-2 pp 3.7  0.2  0.7  0.9  Dilated cardiomyopathy, map05414 GSONMT00009120001 integrin alpha-6 isoform x1 itga 1.0  -0.4  1.0  0.6  GSONMT00075800001 transforming growth factor beta-3 tgf-β 4.8  0.1  0.6  0.7  GSONMT00073220001 titin titin 6.8  -0.4  -0.2  -0.6  Vascular smooth muscle contraction, map04270 GSONMT00037067001 guanine nucleotide-binding protein alpha-13 subunit gna13 1.2  -0.1  0.9  0.8  GSONMT00032856001 guanine nucleotide-binding protein g (q) subunit alpha gq 2.5  0.1  0.6  0.7  GSONMT00043870001 protein kinase c epsilon type prkce 2.8  0.2  0.4  0.6  Hypertrophic cardiomyopathy (HCM), map05410 GSONMT00009120001 integrin alpha-6 isoform x1 itga 1.0  -0.4  1.0  0.6  GSONMT00000701001 amp-acitvated protein kinase gamma 2 isoform prkag 1.2  -0.1  0.7  0.6  GSONMT00075800001 transforming growth factor beta-3 tgf-β 4.8  0.1  0.6  0.7  GSONMT00073220001 titin titin 6.8  -0.4  -0.2  -0.6  Adrenergic signaling in cardiomyocytes, map04261 GSONMT00064390001 camp-responsive element modulator isoform x2 crem 4.9  1.5  1.7  3.2  GSONMT00038526001 camp-responsive element modulator isoform x1 crem 5.1  0.9  0.9  1.9  GSONMT00064102001 cyclic amp-responsive element-binding protein 3-like protein 1 isoform x3 creb 6.0  -0.3  1.6  1.3  GSONMT00077072001 potassium voltage-gated channel subfamily e member 1-like kcne1 0.9  -0.8  1.4  0.6  GSONMT00036180001 beta-2 adrenergic receptor-like β2ar 4.3  0.4  0.7  1.1  GSONMT00032856001 guanine nucleotide-binding protein g (q) subunit alpha gq 2.5  0.1  0.6  0.7  GSONMT00006202001 phosphatidylinositol 3-kinase regulatory subunit beta pi3k 6.1  0.3  -0.4  -0.1  GSONMT00028018001 guanine nucleotide-binding protein g (i) subunit alpha-2-like gi 5.7  0.3  -0.4  -0.1  GSONMT00065371001 myosin-7b myosin 7.7  -0.5  -0.5  -1.0  221    Gene CPM Fold change Location Protein Name [log2(*)] 15-20°C 20-22°C 15-22°C Calcium signaling pathway, map04020 GSONMT00079508001 inositol-trisphosphate 3-kinase c-like itpk 4.7  0.9  1.1  2.0  GSONMT00036180001 beta-2 adrenergic receptor-like β2AR 4.3  0.4  0.7  1.1  GSONMT00032856001 guanine nucleotide-binding protein g (q) subunit alpha gq 2.5  0.1  0.6  0.7  GSONMT00044011001 receptor tyrosine-protein kinase erbb-2 isoform x2 erbb2 4.3  0.1  0.5  0.6  GSONMT00037379001 p2x purinoceptor 5 p2rx5 6.8  0.4  0.5  0.9  GSONMT00023327001 platelet-derived growth factor receptor alpha pdgfra 6.1  -0.3  -0.3  -0.6  Citrate cycle (TCA cycle), map00020 GSONMT00050996001 2-oxoglutarate mitochondrial- partial ogdh 4.5  -0.1  0.5  0.4  GSONMT00044109001 succinyl- ligase sucla 5.7  0.2  -0.4  -0.3  Pyruvate metabolism, map00620 GSONMT00035052001 acyl-coenzyme a thioesterase 12 acot12 5.2  0.1  0.7  0.9  Galactose metabolism, map00052 GSONMT00053165001 udp-glucose 4-epimerase gale 3.7  0.6  0.7  1.4  Oxidative phosphorylation, map00190 GSONMT00007417001 nadh dehydrogenase subunit 1 nd1 13.3  0.6  1.1  1.8  GSONMT00081520001 atp synthase subunit mitochondrial atp5 9.4  0.3  0.6  0.9  GSONMT00012176001 v-type proton atpase 21 kda proteolipid subunit-like atp6v0b 4.6  0.3  0.2  0.5  HIF-1 signaling pathway, map04066 GSONMT00051468001 cyclin-dependent kinase inhibitor 1b cdkn1ba 6.3  0.6  0.1  0.8  GSONMT00009736001 egl nine homolog 2 egln2 5.6  0.4  1.0  1.4  GSONMT00044011001 receptor tyrosine-protein kinase erbb-2 isoform x2 erbb2 4.3  0.1  0.5  0.6  GSONMT00016563001 histone acetyltransferase p300 ep300 3.7  -0.1  0.4  0.3  GSONMT00017548001 vascular endothelial growth factor receptor 1 isoform x2 vegf 1.9  0.6  0.7  1.4  GSONMT00030452001 erythropoietin epo 5.5  0.8  0.3  1.1  GSONMT00051064001 transcription elongation factor b polypeptide 1 tceb1 6.1  0.3  0.3  0.6  GSONMT00006202001 phosphatidylinositol 3-kinase regulatory subunit beta pik3r2 6.1  0.3  -0.4  -0.1      222  C.5 Fold change of heat shock proteins mRNA in redband trout populations Log transformed (base of two) fold changes of heat shock proteins (Hsps) genes at 20°C versus 15°C (A), arrhythmia temperature (TAR) versus 20°C (B), arrhythmia temperature (TAR) versus 15°C (C) for each of the redband trout groups. Shaded bold numbers represent significant fold changes at FDR<0.05. Some transcripts were annotated with the same gene name. Gene Little Jacks (LJ) K × LJ Keithley (K) Fawn  A B C A B C A B C A B C heat shock 90 kda protein 0.9  2.3  3.3  0.2  -0.3  -0.1  -0.9  0.2  -0.6  0.9  3.4  4.3  heat shock 90 kda protein alpha 0.9  5.1  5.9  1.0  1.1  2.1  -1.0  2.4  1.4  1.3  4.5  5.8  heat shock 90 kda protein -alpha 1 0.4  2.9  3.3  0.3  -0.3  -0.1  -0.6  0.7  0.1  0.7  3.1  3.9  heat shock 90 kda protein -alpha 3 0.3  1.8  2.1  0.1  -0.7  -0.6  -0.6  -0.2  -0.8  0.7  2.7  3.4  heat shock 90 kda protein -1 beta isoform a -0.1  0.6  0.5  0.3  0.1  0.4  -0.1  -0.1  -0.2  0.1  0.0  0.2  heat shock 90 kda protein -1 beta isoform b 0.5  0.5  1.0  0.2  -0.4  -0.2  -0.6  -0.2  -0.9  0.7  1.1  1.8  activator of 90 kda heat shock protein atpase homolog 1 0.3  0.6  1.0  0.1  -0.5  -0.3  -0.2  -0.3  -0.5  0.6  0.4  1.0  0.5  0.3  0.7  0.0  -0.5  -0.5  -0.5  -0.4  -0.9  0.5  1.3  1.8               heat shock 75 kda protein mitochondrial-like -0.1  -0.8  -0.9  0.2  -1.1  -0.9  -0.1  0.0  -0.1  0.1  0.4  0.5  heat shock 70 kda protein 0.6  5.9  6.5  1.2  2.7  3.9  -0.2  3.4  3.2  1.4  3.8  5.2  0.5  6.1  6.6  1.3  2.6  3.8  0.0  3.4  3.4  1.5  4.2  5.7  heat shock 70 kda protein 4 0.3  1.5  1.8  -0.1  -0.3  -0.4  -0.8  -0.1  -0.9  0.8  1.7  2.5  0.4  -0.1  0.3  0.1  -1.1  -1.0  -0.4  -0.2  -0.6  0.0  1.5  1.5  0.7  -0.5  0.2  0.2  -1.6  -1.4  -0.2  0.2  -0.1  0.7  2.7  3.5  0.4  1.1  1.5  0.2  -0.4  -0.2  -0.6  -0.3  -0.9  0.7  1.1  1.8  heat shock 70 kda protein 4l -0.3  1.5  1.2  -0.9  -0.1  -0.9  -1.1  -0.1  -1.2  1.1  0.9  2.0  0.1  0.9  1.0  -0.4  0.3  -0.2  -0.5  -0.2  -0.7  0.3  0.4  0.7  heat shock 70 kda protein 4-like isoform x1 0.6  -0.2  0.4  0.1  -1.5  -1.5  -0.5  -0.6  -1.1  1.4  2.0  3.4  heat shock 70 kda protein 4-like isoform x2 1.3  -0.5  0.8  0.4  -1.7  -1.3  -0.7  -0.7  -1.4  1.7  2.1  3.7  heat shock 70 kda protein 8 isoform b 0.1  0.2  0.3  -0.2  -0.4  -0.6  -0.1  0.0  -0.1  0.0  0.2  0.2  heat shock 70 kda protein 12a-like -0.1  0.0  -0.1  0.1  0.0  0.1  -0.3  0.0  -0.3  0.1  -0.3  -0.1  heat shock 70 kda protein 12a-like isoform x3 0.1  -0.2  -0.1  0.3  0.1  0.3  -0.2  0.1  -0.1  -0.1  -0.3  -0.4  heat shock 70 kda protein 12b isoform x3 0.5  -0.2  0.2  0.2  -0.4  -0.2  0.1  0.5  0.6  0.0  0.3  0.3  heat shock 70 kda protein 14 0.4  0.4  0.8  0.1  -0.2  -0.1  -0.6  -0.7  -1.3  0.8  1.0  1.8               223  Gene Little Jacks (LJ) K × LJ Keithley (K) Fawn  A B C A B C A B C A B C heat shock cognate 71 kda 0.3  2.3  2.5  0.1  0.7  0.7  -0.2  0.8  0.6  0.0  1.9  1.9  heat shock cognate 70 kda protein 0.6  1.0  1.6  0.1  -0.1  0.0  -0.8  0.1  -0.7  0.6  1.4  2.0   0.0  0.1  0.1  -0.1  -0.4  -0.4  0.1  -0.3  -0.2  -0.1  0.2  0.2   0.0  0.1  0.2  0.0  -0.2  -0.3  0.1  -0.1  0.0  -0.1  0.3  0.1   0.1  2.0  2.1  0.0  0.5  0.5  -0.5  0.4  -0.1  0.2  1.4  1.6   -0.2  1.6  1.4  0.4  0.4  0.8  -1.2  -0.3  -1.4  0.2  1.1  1.4   0.3  0.5  0.8  0.3  -0.2  0.1  0.2  0.0  0.2  0.3  0.2  0.5   0.0  0.8  0.8  -0.1  -0.2  -0.3  0.2  0.2  0.4  -0.1  0.4  0.3               heat shock protein 67b2 0.6  -0.1  0.5  -0.1  0.0  0.0  -0.4  0.1  -0.3  0.2  0.3  0.5  60 kda heat shock mitochondrial precursor 0.5  0.6  1.1  -0.2  -0.3  -0.5  -0.3  0.5  0.1  0.5  1.5  2.0  -0.1  1.1  0.9  -0.1  -0.6  -0.8  -0.3  -0.2  -0.5  0.6  1.9  2.5  stress 70 protein chaperone microsome-associated 60 kda protein precursor 0.0  0.7  0.7  0.0  -0.3  -0.3  0.5  -0.6  -0.1  1.1  1.1  2.1  0.0  0.8  0.8  0.1  0.4  0.4  -0.2  -0.3  -0.5  0.4  0.4  0.8               serpin h1 isoform x1 0.3  1.8  2.1  0.5  0.3  0.8  0.1  0.9  1.0  0.7  1.4  2.1  serpin h1-like isoform x1 0.8  1.7  2.5  0.3  0.2  0.5  0.1  1.3  1.4  0.5  2.4  2.9  dnaj homolog subfamily a member 1-like 0.2  0.2  0.4  -0.1  -0.1  -0.2  -0.3  -0.2  -0.5  0.2  0.1  0.3  dnaj homolog subfamily a member 2-like 0.3  0.4  0.8  -0.1  0.0  -0.1  -0.3  0.0  -0.3  -0.1  1.2  1.1   0.1  0.4  0.5  0.0  0.4  0.4  -0.1  0.5  0.5  -0.2  0.2  0.0  dnaj homolog subfamily a member 4-like 0.0  1.7  1.8  -0.3  0.1  -0.2  -1.0  -0.3  -1.3  0.7  1.4  2.1  dnaj homolog subfamily a member mitochondrial isoform x1 0.2  -0.2  0.0  0.4  -0.8  -0.4  -0.3  -0.1  -0.5  -0.1  0.7  0.7  dnaj subfamily member 2 0.4  0.0  0.4  0.1  -0.3  -0.2  -0.3  -0.6  -0.9  0.2  0.9  1.1   0.3  0.1  0.4  0.2  -0.3  -0.1  0.0  -0.4  -0.4  0.1  0.5  0.7  dnaj homolog subfamily a member 1 -0.2  0.6  0.4  -0.3  0.5  0.2  -0.5  0.2  -0.3  0.0  1.0  1.0               heat shock protein 30 0.7  9.0  9.7  0.7  2.4  3.1  0.1  3.1  3.3  0.1  7.5  7.6               heat shock protein beta-11 1.0  -0.6  0.3  -0.8  -0.2  -1.1  -1.3  -0.4  -1.7  -0.3  1.5  1.2   1.4  0.1  1.5  -1.1  -0.2  -1.3  -1.0  -0.6  -1.6  -0.4  2.5  2.1  10 kda heat shock mitochondrial 0.9  -0.1  0.8  0.2  -0.2  0.0  -0.2  -0.5  -0.7  0.5  1.9  2.3  224  Gene Little Jacks (LJ) K × LJ Keithley (K) Fawn  A B C A B C A B C A B C  0.8  -0.3  0.4  0.3  -0.6  -0.3  -0.3  -0.3  -0.6  1.1  1.8  2.9  mob-like protein phocein isoform x1 0.0  -0.1  0.0  0.3  0.3  0.6  0.4  0.1  0.5  0.0  -0.6  -0.6  heat shock protein beta-8 0.5  0.5  1.0  -0.1  -0.3  -0.4  -0.3  0.3  0.0  0.7  1.0  1.6   0.4  0.6  0.9  0.2  0.1  0.3  0.1  0.1  0.2  -0.3  1.2  0.9  heat shock protein beta-7-like 0.1  -0.5  -0.4  -0.2  -0.6  -0.8  0.2  0.0  0.2  -0.2  0.7  0.5   0.3  -0.2  0.2  -0.2  0.1  -0.2  0.7  -0.2  0.4  0.3  0.9  1.3  small heat shock 0.8  -0.1  0.7  -0.4  0.4  0.0  -0.4  0.2  -0.2  0.1  0.7  0.8  0.2  0.5  0.7  -0.6  0.6  0.0  -0.6  -0.2  -0.8  0.2  0.7  0.9               heat shock factor protein 2 0.5  -0.3  0.2  0.0  0.4  0.3  0.2  -0.1  0.2  0.2  -1.2  -1.0  -0.1  0.1  -0.1  0.1  0.4  0.5  0.6  -0.3  0.2  -0.1  -0.4  -0.5  heat shock factor-binding protein 1 0.6  0.2  0.8  -0.1  0.1  0.0  -0.2  -0.4  -0.6  0.0  1.0  1.0  heat shock factor-binding protein 1-like 0.2  -0.2  0.0  0.4  -0.1  0.4  -0.2  0.0  -0.2  -0.2  0.5  0.3  heat shock transcription factor 1a 0.3  -0.3  0.0  -0.1  0.4  0.3  -0.1  0.0  -0.1  0.0  -0.2  -0.1  0.4  0.0  0.4  0.2  0.6  0.9  0.1  -0.4  -0.3  0.1  0.2  0.3  heat shock transcription factor 1b 0.0  0.0  0.0  -0.1  0.1  0.0  0.1  0.0  0.1  -0.2  -0.1  -0.2  clathrin heavy chain 1- partial -0.1  0.2  0.1  -0.1  -0.2  -0.3  -0.3  0.0  -0.3  0.4  0.4  0.8  heat shock alpha-crystallin- 1 0.0  -0.6  -0.7  -0.3  -0.5  -0.8  0.2  0.2  0.4  -0.5  0.6  0.1  heat shock protein hsp- hsp- 0.3  -0.4  -0.1  -0.1  -0.6  -0.7  -0.2  0.2  0.0  -0.2  0.2  -0.1  heat shock protein partial 0.1  -0.6  -0.5  -0.4  -0.2  -0.6  0.5  0.0  0.6  -0.5  0.7  0.2  mob-like protein phocein isoform x1 -0.2  0.1  -0.1  -0.1  0.2  0.1  0.0  0.1  0.1  0.0  0.0  0.0  mob-like protein phocein-like isoform x2 -0.1  -0.2  -0.3  -0.1  -0.2  -0.3  0.4  0.0  0.4  -0.1  -0.2  -0.3  protein ssuh2 homolog isoform x3 -0.3  0.2  -0.1  -0.2  0.1  -0.1  0.1  0.1  0.2  -0.1  0.3  0.3    225  C.6 KEGG pathway analysis for metabolism pathways Number of differentially expressed genes in metabolic pathways during acute warming in redband trout populations.  Metabolism Pathways LJ F1 K F Aminoacyl-tRNA biosynthesis 16   16 Carbon fixation pathways 6 6  6 Biosynthesis of antibiotics 5   31 Purine metabolism 5 4  22 Oxidative phosphorylation 5 5  5 Glutathione metabolism 5 5  5 One carbon pool by folate 5 5  5 Lysine degradation 5 5  5 Cysteine and methionine metabolism 5   5 Phenylalanine metabolism 5 5  5 Tyrosine metabolism 5 5  5 Tropane, piperidine and pyridine alkaloid biosynthesis 4 4  4 Novobiocin biosynthesis 4 4  4 Phenylalanine, tyrosine and tryptophan biosynthesis 4 4  4 Glycine, serine and threonine metabolism 4 4  4 Alanine, aspartate and glutamate metabolism 3   6 Arginine biosynthesis 3   5 Pyruvate metabolism 3 3  4 Inositol phosphate metabolism 3   4 Pentose phosphate pathway 3 3  3 Isoquinoline alkaloid biosynthesis 3 3  3 Selenocompound metabolism 3 3  3 Propanoate metabolism 3 3  3 Amino sugar and nucleotide sugar metabolism 2   9 Fructose and mannose metabolism 2   8 Drug metabolism - other enzymes 2   7 Glycolysis / Gluconeogenesis 2   6 Aminobenzoate degradation 2   4 Tryptophan metabolism 2   4 Phosphatidylinositol signaling system 2   4 Pyrimidine metabolism 1   17 Citrate cycle (TCA cycle) 1   6 Starch and sucrose metabolism 1   6 Arginine and proline metabolism 1   5 Porphyrin and chlorophyll metabolism 1   5 Nicotinate and nicotinamide metabolism 1   4 Sphingolipid metabolism 1   4 Glycerolipid metabolism 1   4 Galactose metabolism 1   3    226  C.7 Differential gene expression between redband trout populations    Differential gene expression between populations in redband trout using data across all temperatures (overall). All colored points (exclude grey) are statistically significant after the correction for false discovery rate at 0.05. Blue points represent less than 2 fold changes. Yellow points represent 2-4 fold changes. Red points represent over 4 fold changes. Expression levels are log transformed count per million (CPM)    227  C.8 Genes that are associated with cardiac arrhythmia temperature.  Location Gene Gene name Score GSONMT00039842001 partitioning defective 6 homolog beta-like pard6b 0.0554 GSONMT00011526001 translocating chain-associated membrane protein 2-lik tram2 0.0549 GSONMT00026877001 transcriptional enhancer factor tef-3 isoform x3 tef3 0.0402 GSONMT00066471001 calponin-3-like isoform x1 cnn3 0.0392 GSONMT00074055001 clathrin light chain a isoform x1 clta 0.0314 GSONMT00024226001 ras-interacting protein 1 rasip1 0.0231 GSONMT00037923001 neuroepithelial cell-transforming gene 1 protein isoform x1 net1 0.0204 GSONMT00020438001 serpin h1 isoform x1 serpinh1 0.0179 GSONMT00020100001 inducible heat shock protein 70 hsp70 0.0179 GSONMT00040147001 tryptase-2 partial tpsb2 0.0173 GSONMT00030919001 gtp-binding protein rad rrad 0.0165 GSONMT00001512001 frizzled- partial fzd 0.0161 GSONMT00081507001 heat shock protein 30 hsp30 0.0157 GSONMT00074867001 transmembrane protein 87a-like isoform x3 tmem87 0.0156 GSONMT00080903001 heat shock protein 90 alpha hsp90aa1 0.0155 GSONMT00067502001 camp-responsive element modulator-like isoform x3 crem 0.0152 GSONMT00066470001 tissue factor tf 0.0146 GSONMT00069859001 cathepsin o precursor ctso 0.0131 GSONMT00033669001 low quality protein: myotubularin-related protein 12 mtmr12 0.0126 GSONMT00017643001 leucine-rich repeat-containing protein 58 lrrc58 0.0126 GSONMT00033908001 snw domain-containing protein 1 snw1 0.0125 GSONMT00004377001 unnamed protein product unknown 0.0124 GSONMT00043283001 rho guanine nucleotide exchange factor 2 isoform x1 gef2 0.0124 GSONMT00033190001 unnamed protein product unknown 0.0117 GSONMT00065369001 myosin-7b isoform x1 myh7b 0.0114 GSONMT00076574001 ben domain-containing protein 3 bend3 0.0111 GSONMT00061401001 lysine-specific demethylase 8 kdm8 0.0096 GSONMT00053551001 kinase d-interacting substrate of 220 kda-like isoform partial kidins220 0.0086 GSONMT00035844001 udp-glucuronosyltransferase 1 family polypeptide a1 isoform 1 ugt1a1 0.0081   

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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0340726/manifest

Comment

Related Items