UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Physiological, behavioural, and fitness consequences of introgression of domestic genotypes into wild… Tymchuk, Wendy Elizabeth Vandersteen 2006

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

Item Metadata

Download

Media
831-ubc_2006-201138.pdf [ 12.79MB ]
Metadata
JSON: 831-1.0093023.json
JSON-LD: 831-1.0093023-ld.json
RDF/XML (Pretty): 831-1.0093023-rdf.xml
RDF/JSON: 831-1.0093023-rdf.json
Turtle: 831-1.0093023-turtle.txt
N-Triples: 831-1.0093023-rdf-ntriples.txt
Original Record: 831-1.0093023-source.json
Full Text
831-1.0093023-fulltext.txt
Citation
831-1.0093023.ris

Full Text

PHYSIOLOGICAL, B E H A V I O U R A L , A N D FITNESS CONSEQUENCES OF INTROGRESSION OF DOMESTIC GENOTYPES INTO WILD SALMONID POPULATIONS by W E N D Y E L I Z A B E T H V A N D E R S T E E N T Y M C H U K B.Sc , Queen's University, 1997 M.Sc , The University of Manitoba, 2001 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Zoology) THE UNIVERSITY OF BRITISH C O L U M B I A September 2006 © Wendy Elizabeth Vandersteen Tymchuk, 2006 A B S T R A C T Within the natural environment, growth rates of an animal can be maintained at an adaptive level determined by cost-benefit tradeoffs. Due to an altered selection regime, fish selectively bred for the aquaculture industry may not be as adapted to the natural environment as wild fish. To increase understanding of the genetic changes underlying selection for enhanced growth which results in phenotypic differentiation of domestic (farmed) from wild Pacific salmon, multiple generations of pure and hybrid families were generated for coho salmon (Oncorhynchus kisutch) and rainbow trout (Oncorhynchus mykiss), including pure domestic (D), pure native (W), Fi and F2 hybrids, Fj x wild (Bi), and Bi x wild (B 2) backcross genotypes. The family groups were reared in: 1) culture conditions, 2) semi-natural conditions with either competition, risk of predation, or both, and 3) nature. Under culture and most semi-natural environments there was a significant genotype effect on growth performance (mass and length), with a strong linear relationship between the proportion of domestic alleles within the genotype and size. This rank remained the same for both species. Behavioural differences were observed among the families, with fast-growing domestic families showing a reduced anti-predator response relative to slow-growing wild families. Expression of the phenotypic differences in the hybrids and backcrosses, as well as results from a joint-scale analysis on line means, suggests that additive genetic effects contribute significantly to the divergence between the fast- and slow-growing strains. Survival of the fry in a semi-natural environment with competition fit an additive model of gene action with the domestic fish having the highest survival and the wild fish the lowest, but under risk of predation outbreeding depression was suggested by low survival of the B2 lines. Evidence of a tradeoff in growth and survival under risk of predation along with observations of genetically-determined behavioural differences among the strains may provide some explanation for the observed differences in survival among the strains. Endocrine analysis indicates that an increase in circulating thyroid hormone and insulin-like growth factor I, correlated with the proportion of domestic alleles in the genome, may explain some of the phenotypic differences observed between domestic and wild strains of salmonids. This information is relevant to improving our evolutionary understanding of the interaction among genomes, and the influence of environment, during hybridization events. Results from these experiments indicate that alteration of phenotype likely plays a prominent role in the reduced fitness experienced by progeny produced after three generations of introgression, supporting theory that disruption of genotypes selected for adaptation to local conditions may be a primary cause of outbreeding depression in species such as salmon. I V T A B L E O F C O N T E N T S ABSTRACT : ii T A B L E OF CONTENTS iv LIST OF TABLES ..' viii LIST OF FIGURES^ ix GLOSSARY xii ACKNOWLEDGEMENTS xiv DEDICATION xvi Co-AUTHORSHIP STATEMENT xvii CHAPTER I : A REVIEW OF T H E R O L E OF GENOTYPE AND ENVIRONMENT IN PHENOTYPIC DIFFERENTIATION AMONG W I L D AND CULTURED SALMONIDS 1 1.1 I N T R O D U C T I O N 2 1.2 D E V E L O P M E N T O F C U L T U R E D S A L M O N I D S 3 1.2.1 Genotype, Environment, and GenotypexEnvironment (Gx E) Interactions... 3 1.2.2 Genetic Mechanisms Leading, to Differentiation of Cultured and Wild Fish ....4 1.3 P H E N O T Y P I C C H A R A C T E R I S T I C S O F C U L T U R E D S A L M O N 5 1.3.1 Morphological and Biochemical Differences 10 1.3.2 Emergence Timing and Embryonic Growth 10 1.3.3 Growth 10 1.3.4 Metabolism and Physiology 11 1.3.5 Aggressive Behaviour 12 1.3.6 Predation Risk 13 1.3.7 Foraging Behaviour, Habitat Selection and Dispersal 13 1.3.8 Survival 14 1.3.9 Sexual Maturation and Spawning 15 1.4 G E N E T I C A N D E C O L O G I C A L I N T E R A C T I O N S B E T W E E N C U L T U R E D A N D W I L D FISH .... 16 1.4.1 Effects on Population Size and Inbreeding Depression 17 1.4.2 Importance of Maintaining Genetic Integrity of Wild Populations 18 1.4.3 Additive Genetic Effects 20 1.4.4 Heterosis 20 V 1.4.5 Outbreeding Depression 21 1.5 R E S E A R C H O B J E C T I V E S 2 2 7.5.1 Assessment of Genetically-Determined Differences in Phenotype 22 1.5.2 Role of Environment in Expression of Heterosis or Outbreeding Depression 23 1.5.3 Test for Selection on Endocrine Control of Growth Through Domestication. 23 1.6 R E F E R E N C E S 2 4 C H A P T E R II: G R O W T H A N D B E H A V I O U R A L C O N S E Q U E N C E S O F I N T R O G R E S S I O N O F A D O M E S T I C A T E D A Q U A C U L T U R E G E N O T Y P E INTO A N A T I V E S T R A I N O F C O H O S A L M O N (ONCORHYNCHUS KISUTCH) 40 2.1 I N T R O D U C T I O N 4 1 2 .2 M A T E R I A L S A N D M E T H O D S 4 3 2.2.1 Oncorhynchus kisutch Families 43 2.2.2 Growth Performance 46 2.2.3 Anti-predator Behaviour 48 2.2.4 Resistance to Bacterial Kidney Disease 48 2.2.5 Analysis of Line Means / Genetic Architecture 49 2.3 R E S U L T S 5 0 2.3.1 Growth Performance 50 2.3.2 Environmental Effects on Growth 53 2.3.3 Growth Comparisons with Different Wild Strains 56 2.3.4 Behavioural Differences 59 2.3.5 Resistance to Bacterial Kidney Disease 59 2.3.6 Analysis of Line Means 59 2 .4 D I S C U S S I O N 6 2 2 .5 A C K N O W L E D G M E N T S 6 7 2 .6 R E F E R E N C E S 6 8 C H A P T E R III: G R O W T H D I F F E R E N C E S A M O N G FIRST A N D S E C O N D G E N E R A T I O N H Y B R I D S O F D O M E S T I C A T E D A N D W I L D R A I N B O W T R O U T (ONCORHYNCHUS MYKISS) ... 76 3.1 I N T R O D U C T I O N 7 7 3.2 M A T E R I A L S A N D M E T H O D S 7 8 3.3 R E S U L T S 79 3.4 D I S C U S S I O N 83 3.5 A C K N O W L E D G M E N T S 85 3.6 R E F E R E N C E S 86 CHAPTER I V : OUTBREEDING DEPRESSION A F T E R T H R E E GENERATIONS OF INTROGRESSION OF FAST-GROWING C O H O SALMON INTO A NATIVE STRAIN 8 9 4.1 I N T R O D U C T I O N 90 4.2 M A T E R I A L S A N D M E T H O D S 93 4.3 R E S U L T S 96 4.4 D I S C U S S I O N 101 4.5 R E F E R E N C E S 104 CHAPTER V : GROWTH AND SURVIVAL TRADEOFFS LEADING TO OUTBREEDING DEPRESSION IN RAINBOW TROUT (ONCORHYNCHUS MYKISS) 1 1 0 5.1 I N T R O D U C T I O N I l l 5.2 M A T E R I A L S A N D M E T H O D S 113 5.2.1 Culture Conditions 114 5.2.2 Semi-natural Conditions under Competition 115 5.2.3 Semi-natural Habitat under Risk of Predation 116 5.3 R E S U L T S 119 5.3.1 Behaviour in a Culture Environment 119 5.3.2 Semi-natural Habitat under Competition 121 5.3.3 Semi-natural Habitat under Risk of Predation 121 5.3.4 Line Cross Analysis 128 5.3.5 Trade-off between Growth and Survival 131 5.3.6 Different Selection Pressures of Competition and Predation 131 5.4 D I S C U S S I O N 131 5.5 A C K N O W L E D G E M E N T S 138 5.6 R E F E R E N C E S 139 V l l C H A P T E R V I : E X P R E S S I O N O F H E T E R O S I S A N D O U T B R E E D I N G D E P R E S S I O N IN T H E W I L D A F T E R T H R E E G E N E R A T I O N S OF INTROGRESSION O F F A R M E D R A I N B O W T R O U T INTO A W I L D P O P U L A T I O N 146 6.1 INTRODUCTION 1 4 7 6 .2 M A T E R I A L S A N D M E T H O D S 1 4 9 6.3 R E S U L T S 1 5 3 6.4 DISCUSSION 1 5 7 6.5 R E F E R E N C E S 163 C H A P T E R VII : A L T E R A T I O N S IN C I R C U L A T I N G H O R M O N E L E V E L S A R I S I N G T H R O U G H D O M E S T I C A T I O N IN R A I N B O W T R O U T 166 7.1 INTRODUCTION 1 6 7 7.2 M A T E R I A L S A N D M E T H O D S 1 6 9 7.3 R E S U L T S 171 7 .4 DISCUSSION 1 7 5 7.5 R E F E R E N C E S 1 8 0 C H A P T E R VIII: G E N E R A L C O N C L U S I O N S 184 8.1 S U M M A R Y OF R E S E A R C H 1 8 5 8.1.1 Assessment of genetically-determined differences in phenotype 185 8.1.2 Role of the environment in expression of heterosis or outbreeding depression 7 5 5 8.1.3 Test for selection on endocrine control of growth through domestication.... 187 8.2 K N O W L E D G E G A P S A N D FURTHER R E S E A R C H 1 8 7 8.2.1 Assessing Genetic Differences 187 8.2.2 Assessing Consistency of Genetic Differences 188 8.2.3 Contribution of Phenotypic Differences to Fitness 189 8.2.4 Assessing Plasticity of Phenotypic Differences 189 8.3 R E L E V A N C E FOR C O N S E R V A T I O N OF PACIFIC S A L M O N 1 9 0 8.4 R E F E R E N C E S 1 9 3 A P P E N D I X A 195 V l l l L I S T O F T A B L E S Table 1.2 Summary of phenotypic differences between cultured and wi ld fish 6 Table 2.1 Cross design for 03 January 2002 producing the coho salmon lines used in this study 45 Table 2.2 Joint-scale analysis of individually reared families 60 Table 2.3 Joint-scale analysis of mixed groups reared under enriched or culture conditions 61 Table 4.1 Microsatellite primers used to determine the parentage of the surviving fry reared in the seminatural habitat 95 Table 4.2 Observed (Ho) and expected (HE) heterozygosities, probability of conformance to Hardy-Weinberg equilibrium (HWE) , and total number of alleles (NA) for each loci used in offspring identification 98 Table 5.1 Size of the predators and prey used in the semi-natural habitat under risk of predation 117 Table 5.2 Joint-scale analysis of behavior of the line crosses 129 Table 5.3 Joint-scale analysis of growth (in mass and length) and condition factor under culture or semi-natural conditions with competition or risk of predation 130 Table 6.1 Polymorphic microsatellite loci used to determine parentage of the rainbow trout offspring released into the natural lakes 151 Table 6.2 Observed (Ho) and expected (HE) heterozygosities, probability of conformance to Hardy-Weinberg equilibrium (HWE) , and total number of alleles (NA) for each loci used in offspring identification 154 Table 6.3 Size of the fry (and associated standard error) at release into the lakes on 29 July 2005, 50 days post-fertilization (dpf) and at the fall sampling 8 to 13 October 2005 (121 to 126 dpf) 158 Table 7.1 Size and growth data at sampling for the age-matched genotypes 172 ix L I S T OF F I G U R E S Figure 2.1 Size at age for individually reared coho salmon families 51 Figure 2.2 Relationship between the proportion of farm alleles in the genotype for the individually reared families and (a) SGRm, specific growth rate in mass, % bw day-1; (b) SGR1, specific growth rate in body length, % bl day-1; or (c) condition factor. 52 Figure 2.3 Mean size at age (and associated standard error) for mixed groups reared under culture and semi-natural (enriched) conditions 54 Figure 2.4 Relationship between the proportion of alleles of farm origin in the genotype and (a) SGRm, specific growth rate in mass, % bw day-1; (b) SGR1, specific growth rate in length, % bl day-1; or (c) condition factor 55 Figure 2.5 Mean mass (a) and mean length (b) at age for families representing four wild coho salmon populations examined in 2003 and a farm strain examined in both 2002 and 2003 57 Figure 2.6 Time required for recovery from predator attack and its relationship to (a) genotype; (b) mean mass; and (c) mean length 58 Figure 3.1 Growth performance of pure domestic (D), pure wild (W), F l hybrids (Fl), and F2 backcrosses to wild (Bw) and domestic (Bd) parents 80 Figure 3.2 Comparison of condition factors among the five genotypes as described in Figure 3.1 81 Figure 3.3 Growth frequency distribution of Bw (a, F l x wild) and Bd (b, F l x domestic) at 305 and 307 dpf, respectively 82 Figure 4.1 Specific growth rate in mass (SGRmass) and length (SGRlength) for pure farm fish (D, 100% domestic alleles), pure wild fish (W, 0%), and their hybrids: Fi (50%), BW1 (25%), and BW2 (12,5%) 99 Figure 4.2 Mean survival of the different genotypes from fertilization to first-feeding (A), and after 27 days of foraging under risk of predation (B), and estimated total survival (C) 100 X Figure 5.1 Relationship between the proportion of domestic alleles within the genotype and time required for the first food pellet to be consumed 120 Figure 5.2 Relationship between the proportion of domestic alleles in the genotype for families reared under culture conditions (closed circles) or semi-natural conditions with competition (open circles) and (a) SGR m , specific growth rate in mass, % bw day"1; (b) SGR, specific growth rate in body length, % bl day"1; or (c) condition factor 122 Figure 5.3 Relationship between the proportion of domestic alleles within the genotype and mean survival (and associated standard error) in a semi-natural habitat with competition among genotypes 123 Figure 5.4 Relationship between the proportion of domestic alleles within the genotype and mean survival of the families (and associated standard error) in a semi-natural habitat without risk of predation (no predator, closed circles) or under risk of predation (predator, open circles) 124 Figure 5.5 Relationship between the proportion of domestic alleles within the genotype and activity 126 Figure 5.6 Relationship between the proportion of domestic alleles within the genotype of families reared under semi-natural conditions without a predator (closed circles), or semi-natural conditions with a predator (open circles) and (a) SGR m , specific growth rate in mass, % bw day"1; (c) SGRi, specific growth rate in body length, % bl day"1; or (e) condition factor in Trial 1 and (b) SGR m ; (d) SGRi; or (f) condition factor in Trial 2 127 Figure 5.7 Correlations between activity in semi-natural habitats and fitness components measured as SGR m a ss, specific growth rate in mass, % bw day"1 (closed circles) and number of survivors in the habitat (open circles) 132 Figure 5.8 Correlations between growth measured as SGR m a ss in (a, b) and SGR | e n g th in (c, d) and survival (proportion) for fish reared under semi-natural conditions with competition (a, c) or under risk of predation (b, d) 133 Figure 6.1 Relationship between genotype and viability from fertilization to hatch 155 xi Figure 6.2 Estimated over-summer survival of the fry released into the two natural lakes, CPI and PPH 156 Figure 6.3 Growth in mass (SGRm ass) and length (SGR | e n gth) of the genotypes reared over summer in two natural lakes, CPI and PPH 159 Figure 7.1 Relationship between the proportion of domestic alleles within the genotype and (A) concentration of growth hormone (GH), (B) insulin-like growth factor (IGF-1), and (C) thyroid hormone (T3) in the plasma 173 Figure 7.2 Relationship between specific growth rate in mass (SGRmass) and the log of the concentration of GH, IGF-1, and T3 within the plasma 174 Figure 7.3 Concentration of circulating insulin-like growth factor I (A, IGF-1) and free T3 (B) in size-matched domestic and wild rainbow trout 176 X l l G L O S S A R Y Aquacultured - fish reared in culture throughout their entire life, usually for commercial purposes Cultured - broad term referring to all fish not entirely reared in the wild; often applied to strains that have been reared throughout their entire life history in aquaria, such as in captive brood programs Domestic - A formal definition of domestication was proposed by (Price 1984) as "that process by which a population of animals becomes adapted to man and to the captive environment by some combination of genetic changes occurring over generations and environmentally induced developmental events reoccurring during each generation"; domestication effects can be acquired very rapidly (within a single generation), and thus essentially all cultured fish may be at least partly domestic Farmed - synonymous with aquacultured Founder effect - chance change in the frequency of some genetic variants in populations as a result of a relatively small number of initial founders of the population Genetic drift - random change in allelic frequencies that results from the sampling of gametes from generation to generation Genotype x environment (G x E) interactions - Non-linear responses of phenotypes to environmental conditions observed among genotypes. Hatchery - fish reared for part of their life history in hatchery facilities; usually produced from artificially spawned adult fish captured after returning from the wild, and are released back into the wild at various life history stages (depending on the species) Heterosis - fitness of hybrids that exceeds the mean performance of the parental lines Hybrid vigour - synonymous with heterosis Inbreeding - non-random mating (with respect to genotype) where the mating individuals are more closely related than those drawn from the population by chance Inbreeding depression - decline in mean fitness with increasing homozygosity within populations Ne - effective population size; effective number of breeding individuals; number of individuals that would give rise to the calculated sampling variance i f they bred as an idealized population Norm of reaction - the phenotypic expression of a genotype under different environmental conditions Outbreeding - non-random mating (with respect to genotype) where the mating individuals are less closely related than those drawn from the population by chance Outbreeding depression - fitness of hybrids is below that of the parental lines Plasticity - the environmentally sensitive production of alternative phenotypes by given genotypes xiv A C K N O W L E D G E M E N T S There are so many people that I must thank for their guidance, assistance, and support throughout my doctoral research. The fact that I had the opportunity to do this research in the first place is thanks to my supervisor, Bob Devlin. I feel incredibly lucky to have spent these years working under Bob's supervision. His dedication and enthusiasm for science are inspirational. I am extremely grateful for the freedom I was allowed to explore, and for his kindness, trust, and words of advice that have made this degree an enjoyable, enriching, and motivating experience. Gratitude is also extended to my co-supervisor, Eric Taylor, and the members of his lab for helping me maintain my connection to the university. I am grateful for the interaction during lab meetings, and appreciate the constructive comments during practice seminars. Many thanks are due to my amazing thesis committee of Colin Brauner, Tom Grigliatti, and Mike Whitlock. Their input during committee meetings was very useful in developing and improving my research, and led me to approach problems in alternative ways. I remember Mike Whitlock's comment at my first committee meeting that I should be able to call myself a " ologist", and it was up to me to decide how to fill in that blank. Well, after four years, I think I've settled on "molecular ecologist", but I must say that has been one of the more difficult, yet useful, tasks set out for me! Anyone who has done laboratory work with live salmon will understand the immense amount of time and commitment required to obtain your data. Consequently, this research would not have been possible without the huge amount of help that I've had from current and past members of the Devlin lab. First, I would like to express my thanks to Carlo Biagi. He has spent so much time helping me with various aspects of my research, from obtaining fish to collecting samples. Besides the technical assistance, I would like to sincerely thank Carlo for his honesty, support, and encouragement - his friendship has been greatly appreciated. Another person who was extremely helpful when I first started was Mark D'Andrade. Mark took the time to share his knowledge of fish husbandry, for which I'm very grateful. In addition, Mark and Geordia Rigter provided help with feeding, cleaning, and sampling of fish. Other people who have provided a lot of support in the aquarium over the last four years include Geoff Harrison, Ki-Whan Eom, Morgan Williams, Nicole Hofs, Hanna Kim, Ryan Becker, and Laura Nendick. I hope that they all realize how important they have been to this research. When collecting gametes and adult fish to make the crosses for my research, I received considerable cooperation and assistance from the staff at various hatcheries across the province, in particular the Chehalis River Hatchery. Thank you also to Target Marine Farm and Justin Henry for generously allowing me to use your domestic fish in my studies, for your assistance during gamete collection, and for agreeing to be my industry partner for my NSERC Industrial Postgraduate Scholarship. Spring Valley Trout Farm and Ted Brown were always very accommodating with my requests for domestic rainbow trout adults and gametes. I am full of gratitude for the staff of the Freshwater Fisheries Society of British Columbia, including Tim Yesaki and the X V technicians at the Fraser Valley Trout Farm, for their valuable assistance at several points during my PhD. When it came time to move from the aquarium to the molecular lab, Dionne Sakhrani, Peter Raven, Mitch Uh, and Ben Goh helped me develop the skills required to analyse my samples. I owe a special thanks to Dionne for her excellent organization, dependability, and willingness to fit my requests in with all the other demands on her time. I don't think anybody could have completed a job more efficiently than Dionne and I when we worked together! In the radioimmunoassay lab, I must thank Jack Lloyd Smith for his perseverance in getting the assays up and running. I wish to acknowledge Fred and Mare Sundstrom, who are not only colleagues to me, but also friends. I have greatly enjoyed discussing all aspects of science with them, from specific research projects to career options to off-the-wall amusing facts. I will miss sharing a lab with them. Another important colleague of mine is Pete Biro, who generously shared his knowledge of the research lakes at Merritt. Without Pete's help and advice, I would not have been able to conduct the research of C h a p t e r V I I , which I think is one of the most important studies in this thesis. Special thanks are due to my parents, Doug and Edna Vandersteen, for never questioning my judgement and supporting any goal I strived for. Thank you for instilling me with strength, stubbornness, and the ability to enjoy hard work. My grandpa, Bert Vandersteen, who sadly is not here to see me finally reach the end of my formal education, I thank for nurturing my interest in nature, and teaching me how to be patient, and to sit quietly and observe. Those are lessons that have served me well. My grandmother, Lillian Vandersteen, is also owed gratitude for always believing in me, and making me believe that I can do anything I set my mind to. To my extended family, Ken and Klasien Tymchuk, and Kerri Tymchuk, I wish to express my appreciation for their love, support and friendship. It is not only the student that must make sacrifices to obtain an education, but also their family. I cannot express how much I appreciate the efforts of Toni Weatherburn, my sister, in making sure my nieces, Kaylee, Emily and Ally, and nephew Dylan, grow up with me in their life even though my visits back home are few and far between. My sister also deserves thanks for her caring words and gestures that have helped buffer the most stressful times. Last, but certainly not least, I wish to express my gratitude for my husband, Jeff Tymchuk. I know he thought I would be a student for ever, and I thank him for his continued support in spite of it. Jeff, thank you for keeping me grounded over these last few years, and thank you for reminding me every day what is really important in life. xvi D E D I C A T I O N I wish to dedicate this thesis to all of you who refused to ever let me take the easy way. XVII C O - A U T H O R S H I P S T A T E M E N T C h a p t e r I is based on a review manuscript that is currently in press. As the primary author of this review, I was responsible for the preparation of the manuscript, including the search, summary and compilation of published literature related to this topic. C h a p t e r I I and I I I have been published with myself as the primary author. For both of these manuscripts, I have designed the experiments with the assistance of my supervisor, performed the research, data analysis and interpretation, and prepared the manuscripts. C h a p t e r I V is currently under review, and C h a p t e r V , V I a n d V I I will be submitted for review. I will be the primary author on these papers as well, and therefore was the primary contributor to the design and completion of the experiments, data analysis, and manuscript preparation. 1 CHAPTER I1: A R E V I E W OF T H E R O L E OF G E N O T Y P E A N D E N V I R O N M E N T I N P H E N O T Y P I C D I F F E R E N T I A T I O N A M O N G W I L D A N D C U L T U R E D S A L M O N I D S 1 This chapter is a summary of a review paper in press. The manuscript is included as an appendix in its entirety . Tymchuk, W. E. , Devlin, R. H. , and Withler, R. E . The role of genotype and environment in phenotypic differentiation among wild and cultured salmonids. State of Knowledge Series, Fisheries & Oceans Canada. 2 1.1 I N T R O D U C T I O N Body size is one of the most important phenotypic characteristics of an animal, influencing fitness parameters, such as other phenotypic traits, life history, and population ecology. In many species, there is a positive correlation between body size and fitness. Since the growth rate of an animal links its size at age, growth rate is therefore an important factor in optimal life history strategy which maximizes age-specific survival and reproductive fitness (Roff 1992). Although growth rates are heritable, with existing inter- and intra-population variability, growth rates can remain submaximal relative to the population's potential for enhanced growth. Compensatory growth and selective breeding for enhanced growth, for example, provide evidence of the physiological capacity for higher growth rates in a population. Increase of growth rates with supplementation of exogenous hormones (Donaldson et al. 1979) suggests growth rates are constrained below physiological capacity partly by endocrine control. Submaximal growth rates seem an unlikely outcome of natural selection, suggesting that there must be costs to enhanced growth if these observations are to be reconciled with principles of evolutionary theory (Calow 1982; Arendt 1997). These observations lead to some important questions regarding variability in growth within and among populations. First, many phenotypic traits have been correlated with growth, but in many instances, it is not clear i f growth is the cause or effect. Second, there is evidence of genetic variability leading to altered growth rates in a population, but the actual genetic basis for these changes is not completely understood. Third, since growth rates are often submaximal, there must be fitness costs associated with enhanced growth. With the development of animal strains that have their physiological capacity for growth pushed beyond normal limits seen in nature, we are able to begin exploring even further the existing tradeoffs between growth and fitness. The overall goal of this research is to improve understanding of the genetic changes 3 involved in evolution of enhanced growth and the fitness-related costs of maintaining the enhanced growth, using cultured and wild strains of salmonids as a model. 1.2 D E V E L O P M E N T OF C U L T U R E D S A L M O N I D S Demand for food from a burgeoning human population, coupled with natural and anthropogenic environmental effects on wild fish stocks, is shifting focus on seafood supply from natural fisheries to aquaculture and hatchery programs. World trends continue to forecast growth of the aquaculture industry at rates greater than for other food-production sectors, with a focus on finfish. Salmon culture began in Europe in the last millennium and in North America in the middle of the 17th century, but intensive application of salmonid enhancement and aquaculture began in the 1960s. Since then, production of salmonids has expanded dramatically in many countries in North America, Europe, Australasia and South America, providing significant economic benefit to those regions. In Canada, the majority of cultured fisheries production arises from salmon farming on both coasts and in inland facilities, and from hatchery programs on the Pacific coast. 1.2.1 Genotype, Environment, and Genotype x Environment (Gx E) Interactions Phenotypic differences between farmed and wild salmonids may arise from a combination of genetic and environmental effects, but in most cases the origin of the difference is not well defined. To assess genetic effects, experiments must be performed by rearing fish of different origins in a common environment. Environmental effects (i.e. phenotypic plasticity) can be tested by rearing fish of a common genetic background in different environments. As will be discussed below, cultured fish may have an altered genotype in response to selection pressures from an artificial environment, leading to genetically-based phenotypic differences between the cultured and wild strains. Alternatively, since salmonids are so phenotypically plastic (Hutchings 2004), they may 4 have an altered phenotype in response to the altered environmental conditions in which they are reared. These environmentally-based phenotypic differences would not be passed along to offspring as they do not have a genetic basis. Thus, environmental effects are anticipated to have single generation effects arising directly from escaped fish, whereas genetic differences are those which have the potential to affect a species on a longer time frame. It is therefore critical to separate the influence of genotype and environment, and to understand genotype by environment interactions in order to fully predict the genetic effects of interaction between wild and cultured fish. 1.2.2 Genetic Mechanisms Leading to Differentiation of Cultured and Wild Fish The mechanisms causing genetic changes in cultured fish include inbreeding, genetic drift and selection. Selection can be further broken down into three categories: artificial selection, natural selection in captivity, and relaxation of natural selection. It is generally accepted that farming of salmon generates genetic change due to intentional and unintentional selection in culture (Reisenbichler and Mclntyre 1977; Fleming and Einum 1997). Studies are now examining differences in gene expression altered by domestication in salmon (Roberge et al. 2006). In many cases, the goal of selection is to produce a homogenous line that demonstrates constancy in desired traits such as enhanced growth (Gjedrem 2000). Sometimes diversity is desired, and in this case domestication effects would be lower. Genetic changes may also occur for traits other than those that are the focus of the selective breeding program. For example, fifth generation farmed Atlantic salmon differed significantly from wild populations in loci other than those chosen for the selective breeding program (Mj0lner0d et al. 1997). The characteristic of the genetic changes caused by indirect selection during domestication can vary according to the method of fish culture and the extent of time that fish spend in the artificial environment (Utter and Epifanio 2002). For example, characteristics of the culture environment such as rearing density and the source of food into the tank may affect the evolution of aggressive behaviour or increased motivation for surface-feeding. 5 1.3 P H E N O T Y P I C C H A R A C T E R I S T I C S OF C U L T U R E D S A L M O N Domestication can have a significant impact on life history traits in salmonids (Thorpe 2004); see Table 1.2 for a summary of documented phenotypic differences between domestic and wild fish. Many fitness-related traits, such as growth, competitive ability and anti-predator behaviour, have been found to have a genetic component. Domestication may select for traits related to improved growth rates, earlier age at maturity and spawning, greater survival, increased tolerance to high temperature and resistance to disease (Hynes et al. 1981). Differences between wild and cultured fish represent a continuum, ranging from differences among natural strains, through differences between wild and sea-ranched fish, to differences between wild and highly-selected and domestic cultured fish. Alterations in fitness-related traits in hatchery fish should be typical of differences expected in aquacultured salmon, although the latter may show a greater magnitude of change due to an increased length of time under intentional and due to unintentional selection, usually conducted in isolation from wild genetic pools. Transgenic fish, which can be viewed as an extreme form of selective breeding, are not considered in the present discussion except when examined as a model system for assessment of genotype/phenotype relationships (Devlin et al. 2001). The phenotypic differences found between cultured and wild strains may not only be due to genetic differentiation between the strains, but may also arise as a consequence of the different environments in which the fish are reared. There are few studies that have definitively assessed the relative contribution of genotype and environment. This complicates efforts to find trends in the growing body of literature on phenotypic differences between wild and farmed salmonids. Additionally, even when it is known that a trait is due to genetic differentiation between the strains, there is still a lack of knowledge regarding the expression of the different genotypes in different environments (i.e. genotype by environment interaction). 6 T A B L E 1.2 Summary of phenotypic differences between cultured and wild fish. The result is the observed characteristic of the farmed fish in relation to the wild fish. The cause is due to environment (E) or genotype (G) and was either found to exist (+), not exist (-) or was not tested (0). An environmental cause for phenotypic differences refers to the environment experienced by the fish prior to the experiment. A positive contribution for both E and G indicates either that the experimental design could not isolate the contribution of one factor, or that both factors affected phenotype. TRAIT RESULT ENVIRONMENT E G SPECIES SOURCE Growth Increased Wild + Increased Wild + Increased Culture 0 Wild Increased Culture 0 Decreased Semi-Natural 0 Increased Wild 0 Increased Culture 0 Increased Culture 0 No difference Semi-Natural 0 Increased Culture 0 Decreased Wild + Increased Wild + Increased Culture + Decreased Wild + Increased Culture 0 Increased Wild 0 Increased Wild 0 Increased Culture 0 Increased Culture 0 Increased Semi-natural 0 + Rainbow Ayles 1975 + Rainbow Ayles & Baker 1983 + Atlantic Einum & Fleming 1997 + Atlantic Fleming & Einum 1997 + Atlantic Fleming & Einum 1997 + Atlantic Fleming et al. 2000 + Atlantic Fleming et al. 2002 + Brook trout Flick & Webster 1964 - Brook trout Flick & Webster 1964 + Coho Hershberger et al. 1990 + Brook trout Keller & Plosila 1981 + Brook trout Lachance & Magnan 1990 + Brook trout Mason et al. 1967 + Brook trout Mason et al. 1967 + Coho McClelland et al. 2005 + Atlantic McGinnity et al. 1997 + Atlantic McGinnity et al. 2003 + Rainbow Tymchuk & Devlin 2005 + Coho Tymchuk et al. 2006 + Coho Tymchuk et al. 2006 7 TRAIT RESULT ENVIRONMENT E G SPECIES SOURCE Morphology Altered Culture + + Atlantic Fleming & Einum 1997 Altered Culture + 0 Atlantic Fleming et al. 1994 Altered Wild + + Coho Swain et al. 1991 Metabolism Altered Culture + + Brown trout Carline & Machung 2001 & Physiology No difference Culture 0 - Atlantic Dunmall & Schreer 2003 Altered Culture 0 + Atlantic Fleming et al. 2002 Altered Culture 0 + Atlantic Handeland et al. 2003 Altered Culture + + Atlantic Poppe et al. 2003 Altered Culture + + Atlantic Poppe et al. 1997 No difference Culture 0 - Brown trout Sanchez et al. 2001 Altered Culture 0 + Atlantic Thodesen et al. 1999 Increased Culture 0 + Atlantic Einum & Fleming 1997 Increased Culture 0 + Atlantic Fleming & Einum 1997 Decreased Semi-natural 0 + Atlantic Fleming & Einum 1997 Increased Wild 0 + Atlantic McGinnity et al. 2003 Increased Wild 0 + Atlantic McGinnity et al. 1997 No difference Culture + - Atlantic Mork et al. 1999 Heterogenous substrate Decreased Culture + + Atlantic Mork et al. 1999 Homogenous substrate 8 TRAIT RESULT ENVIRONMENT E G SPECIES SOURCE Anti-predator Decreased Culture 0 + Atlantic Einum & Fleming 1997 behaviour Decreased Culture 0 + Atlantic Fleming & Einum 1997 No difference Semi-natural 0 - Atlantic Fleming & Einum 1997 Decreased Culture 0 + Steelhead Johnsson & Abrahams 1991 Decreased Culture 0 + Atlantic Johnsson et al. 2001 Decreased Culture 0 + Coho Tymchuk et al. 2006 Foraging No difference Wild - - Atlantic Jacobsen & Hansen 2001 behaviour Habitat No difference Wild 0 - Atlantic Einum & Fleming 1997 selection & Dispersal Altered Culture + + Atlantic Morketal . 1999 Altered Semi-natural + + Masu Nagataet al. 1994 Survival Decreased Wild + + Brown trout Aerestrup et al. 2000 No difference Wild 0 - Atlantic Einum & Fleming 1997 Decreased Wild, summer 0 + Brook trout Flick & Webster 1964 No difference Wild, winter 0 - Brook trout Flick & Webster 1964 Variable Wild + + Brook trout Fraser 1981 Decreased Wild + + Brook trout Keller & Plosila 1981 Decreased Wild + + Brook trout Lachance & Magnan 1990 Increased Wild, winter + + Brook trout Mason et al. 1967 Decreased Wild, summer + + Brook trout Mason et al. 1967 Decreased Wild 0 + Atlantic McGinnity et al. 2003 Decreased Wild 0 + Atlantic McGinnity et al. 1997 9 TRAIT RESULT ENVIRONMENT E G SPECIES SOURCE Migration Altered Wild + + Atlantic Heggberget et al. 1993 Altered Wild + + Atlantic Webb et al. 1991 Reproductive Altered Wild + + Brook trout Lachance & Magnan 1990 Physiology No difference Culture 0 - Chinook Bryden et al. 2004 Spawning Altered Wild 0 + • Atlantic Fleming et al. 2000 behaviour Altered Semi-natural + + Atlantic Fleming et al. 1996 Altered Wild + + Atlantic Luraetal. 1993 Altered Wild + + Atlantic Okland et al. 1995 Altered Wild + + Atlantic Saegrovetal. 1997 Altered Semi-natural 0 + Atlantic Weir et al. 2005 Reproductive Decreased Wild + + Atlantic Clifford et al. 1998 success High density Increased Wild + + Atlantic Clifford et al. 1998 Low density Decreased Wild 0 + Atlantic Fleming et al. 2000 Decreased Semi-natural + + Atlantic Fleming et al. 1996 Decreased Wild + + Atlantic Saegrov et al. 1997 High density No difference Wild - - Atlantic Saegrov et al. 1997 Low density Decreased Semi-natural + + Atlantic Weir et al. 2005 10 1.3.1 Morphological and Biochemical Differences Morphological differences between cultured and wild salmon have been reported but in most cases environment rather than genetics was found to have the most influence on these characteristics (Kazakov and Semenova 1986; Swain et al. 1991; Fleming et al. 1994). Rearing fish in a culture environment can lead to environmentally-determined differences in morphology relative to fish reared in the wild. The extent of these differences will depend on the length of time spent within the artificial environment, and the intensity of the culture conditions (such as crowding, food supply, etc.). Multiple generations of strains kept within the culture environment could lead to genetically-based morphological differences arising from selection for traits affording fitness benefits in culture. 1.3.2 Emergence Timing and Embryonic Growth There are few studies on the genetic basis of differences in development rate between cultured and wild fish. Differences in development rate have been noted between hatchery and wild strains of rainbow trout (Ferguson et al. 1985) and it has been determined that development rate is controlled by at least one major locus in this species (Robison et al. 1999, 2001). The actual genes affecting this process in domestic fish are not yet known, although the expression of growth hormone genes in transgenic fish has been shown to significantly affect development rate with consequent effects on survival fitness (Devlin et al. 2004; Sundstrom et al. 2005). 1.3.3 Growth In general, domestic fish tend to grow faster than wild fish, although this is not universally the case (Einum and Fleming 2001; Bryden et al. 2004) and is greatly influenced by the rearing environment due to such factors as availability of food resources and temperature. However, there can also be large differences in growth 11 between cultured and wild strains that are due to genetic differences between the strains. The magnitude of the growth differences caused by genotype will be dependent on the purpose and history of the cultured strain. Aquacultured strains that have been intensely selected for enhanced-growth will show a larger shift in growth phenotype from the founding line relative to a strain that has not experienced directed selection. There is still limited knowledge on how environment will act on the inherent genetic differences (i.e. will environmental conditions affect different genotypes in distinct ways through genotype x environment interactions). 1.3.4 Metabolism and Physiology There is limited information on the physiological differences between farmed and wild strains. It is often thought that altered feed conversion efficiency underlies the ability of farmed strains of fish to growth faster. This has been supported by studies on Atlantic salmon (Thodesen et al. 1999) although no differences in conversion efficiency were found between wild and selected strains of brown trout (Sanchez et al. 2001). Altered endocrine regulation, in particular involving changes of the growth hormone / insulin-like growth factor axis, may also explain the enhanced growth of selected strains of fish (Fleming et al. 2002). Selection in a culture environment may also alter other traits directly related to physiology such as smoltification and saltwater tolerance (Shrimpton et al. 1994; Ugedal et al. 1998; Handeland et al. 2003; Leonard and McCormick 2001), swimbladder morphology (Poppe 1997), heart morphology (Poppe 2003) and critical thermal maxima (Carline and Machung 2001). These differences were suggested to be genetically-determined. Physiology may not always differ between farmed and wild strains. For example, no differences in swimming and cardiac performance were observed between farmed and wild Atlantic salmon (Dunmall and Schreer 2003). In general, cultured fish will tend to show physiological differences relative to wild fish. However, it is often not clear whether these differences are a cause, or a consequence of, other phenotypic differences between the strains (such as growth or behaviour differences). It is therefore difficult to clarify whether physiological 12 differences have a genetic basis per se, or if they are a product of the environment. Any observed differences are most likely due to a combination of these factors, unless cultured strains have been specifically selected for altered physiological characteristics (such as higher feed conversion efficiencies, for example). 1.3.5 Aggressive Behaviour Aggressive behaviour is often affected by artificial rearing. A study on the agonistic behaviour of domestic and wild steelhead trout suggested that four to seven generations of domestication resulted in behavioural divergence of the two populations (Berejikian et al. 1996), although Johnsson et al. (1996) found no difference in dominance between fish selected for enhanced growth and wild individuals. Studies on domestic farmed Atlantic salmon found genetically-based increases in aggression level relative to wild fish (Einum and Fleming 1997; Fleming et al. 2002; McGinnity et al. 1997, 2003). Interestingly, other research has found that selection for enhanced growth results in indirect selection for tameness, as opposed to aggression (Doyle and Talbot 1986). For example, wild Atlantic salmon were found to make more aggressive attacks toward farmed individuals than farmed made on wild (Mork et al. 1999). In an attempt to explain the discrepancy on the genetic link between aggressive behaviour and growth, (Ruzzante and Doyle 1991) concluded that agonistic behaviour will be inversely proportional to growth when selection occurs in an environment with forced social interaction and unlimited food resources. Cultured and wild fish do show differences in the level of aggression that they show to conspecifics, although there has not been a consistent trend as to whether aggression increases or decreases under culture. The common assertion is that aggression will decrease under culture when fish are reared in crowded conditions and do not have to fight for limited food resources. Knowledge is lacking on the expression of genetically-determined differences in aggression under varied environmental conditions, and the extent of behavioral plasticity that can be expressed by the different strains. 13 1.3.6 Predation Risk Several studies indicate that anti-predator behaviour has been altered through domestication so that domesticated fish are more willing to risk predation in order to feed and as a consequence, may have higher mortality rates. Farmed Atlantic salmon show increased risk-taking behaviour relative to wild fish (Einum and Fleming 1997; Fleming and Einum 1997; Fleming et al. 2002;) and a reduced response to predators, as measured by flight and heart rate response (Johnsson et al. 2001). Domestic/wild hybrid rainbow trout with enhanced growth rate have been observed to be more willing to risk predation in order to feed relative to wild individuals (Johnsson and Abrahams 1991). Similarly, brown trout selected for enhanced growth showed reduced antipredator behaviour relative to wild fish (Johnsson et al. 1996), and growth-enhanced transgenic salmonids have been observed in some but not all cases to expose themselves to greater predation risk and suffer higher levels of predation mortality (Abrahams and Sutterlin 1999; Sundstrom et al. 2003, 2004, 2005; Vandersteen Tymchuk et al. 2005). A genetically-determined reduced response to predators seems to be a consistent trend in domestic strains across several species. However, these studies have tested the predator response in an artificial environment with no real risk of being consumed by a predator. There are no indications of whether the domestic strains would respond to a real predator in the wild, or more importantly to overall fitness, i f their realized mortality would be any different from that experienced by the wild fish. 1.3.7 Foraging Behaviour, Habitat Selection and Dispersal In addition to being more willing to risk predation in order to feed, domestic fish may have altered foraging strategies relative to wild individuals. These foraging strategies may not be as suitable for the natural environment as they have evolved within the structure of the culture environment. Generally, cultured fish have demonstrated less effective or efficient foraging relative to wild conspecifics, particularly during the initial period after released from culture (Bachman 1984; Johnsen and Ugedal 1986; McKinnell 14 et al. 1997; Munakata et al. 2000; Jacobsen and Hansen 2001). Differences have also been observed in habitat selection (Dickson and MacCrimmon 1982; Mesa 1991; Mork et al. 1999) and dispersal (Joergensen and Berg 1991; Levings et al. 1986; Nagata et al. 1994). There is no evidence for a strong genetic basis for any observed differences in foraging strategy. Environment would likely have the strongest effect on this phenotypic characteristic, and would be expected to quickly overcome any genetic differences that had evolved within a culture environment. A genetic basis for this trait would probably be due to phenotypic expression of other phenotypic traits, such as growth or morphology, that would drive foraging behaviour characteristics. 1.3.8 Survival Escaped domestic salmon can survive in the wild, although recapture rates of farmed fish vary considerably. The ability of farmed fish to survive upon escape will depend on several factors including the timing of escape, escape location, and environmental conditions within the habitat (Flick and Webster 1964; Mason et al. 1967; Fraser 1981; Keller and Plosila 1981; Hansen and Jonsson 1994; Einum and Fleming 1997; Hansen et al. 1997; McGinnity et al. 1997, 2003). Importantly, not all studies have indicated that domestic fish will incur higher mortality rates in nature (Hvidsten and Lund 1988; Kennedy and Greer 1988). Survival is a trait that is influenced by all other phenotypic traits, and the environment in which they are expressed. Cultured fish, either through a plastic response to their environment or through an adaptive response to altered selection pressures, tend to express phenotypic characteristics best suited for the culture environment. Consequently, they tend to not have as high survival as wild fish in a natural environment. However, it is still not clear i f cultured fish that experience a natural environment throughout their life history will still show decreased survival relative to the wild fish. It is not known how strong the genetic basis of survival is, and whether the cultured fish still have the ability to show a phenotypically plastic response to the environment that will maximize their ability to survive. 15 1.3.9 Sexual Maturation and Spawning Genetic impact of farmed fish on wild populations will depend in part on the reproductive behaviour of farmed fish in the wild. Evidence suggests that farmed fish have the ability to successfully breed in the wild, although contradicting results do occur. There are generally significant differences in breeding potential (Table 1.12) of cultured and wild fish (Fleming and Gross 1992, 1993; Fleming et al. 1996), although other studies have found similar reproductive success for hatchery and native fish in the wild (Palm et al. 2003; Dannewitz et al. 2004). Morphology and life history traits related to reproductive behaviour respond evolutionarily to altered selection regime in the hatchery environment (Fleming and Gross 1989; Fleming 1994). The genetic impact of aquacultured salmon on wild populations will depend not only on the size of the wild population, but also on variation in breeding success (Fleming and Petersson 2001). Cultured fish typically have reduced spawning success (from both environmental and genetic causes) which is frequently more severe in males (Fleming and Gross 1993; Berejikian et al. 1998; Fleming et al. 1996, 2000; Berejikian et al. 2001; Fleming and Petersson 2001; Weir et al. 2004) although this is not always the case (Garant et al. 2003) . Consistent trends have been observed in that cultured fish often have the physiological ability to spawn (Peck 1988; Lachance and Magnan 1990; Lura and Saegrov 1993; Berejikian et al. 1997; Lacroix et al. 1997; Clifford et al. 1998; Bessey et al. 2004; Bryden et al. 2004), but altered spawning behaviour limits their success (Hansen et al. 1987; Heggberget et al. 1993; 0kland et al. 1995; Fleming et al. 1996; Weir et al. 2004) . Even though the reproductive success of farmed fish may be low, the potential for significant gene flow still exists since the population of farmed fish often outnumbers the population of resident wild fish (at least in the case for Atlantic salmon), at times by as much as 3:1 (Lund et al. 1994; Lura and 0kland 1994). We do not have any data on the ability of farmed and wild Pacific salmon to spawn in nature, but comparisons between hatchery and wild coho salmon indicate that trends observed for Atlantic salmon may be typical of the phenotypic changes expected during domestication. 16 1.4 G E N E T I C A N D E C O L O G I C A L I N T E R A C T I O N S B E T W E E N C U L T U R E D A N D W I L D FISH In addition to increasing understanding of the genetic and physiological basis for altered rates of growth and the relationship between growth and fitness, comparisons of cultured and wild fish have an important role in the conservation of wild fish populations. Cultured fish may enter the natural environment by purposeful release for conservation and enhancement purposes, or they may escape from aquaculture facilities. From as early as 20 years ago, fisheries biologists have noticed that escaped farmed salmon were present in wild populations (Hansen et al. 1987; Gudjonsson 1991; Lura and Sa^grov 1991; Jacobsen et al. 1992; Crozier 1993; Carr et al. 1998; McKinnell et al. 1997; Clifford et al. 1998; Volpe et al. 2000; Morton and Volpe 2002) and concern has emerged regarding the concurrent genetic and ecological consequences of this interaction arising from differences in phenotype and genotype between cultured and wild fish populations (Crossman 1991; Hindar et al. 1991; Fleming et al. 1996; Hansen et al. 1997). Early reviews of the literature on the outcome of interaction between introduced and native fish populations suggested that introductions have usually been harmful to the native fish populations (Allendorf 1991; Hindar et al. 1991), but see (Peterson 1999). Genetic effects of domestic fish may be direct or indirect. Direct genetic effects include the alteration of the wild genome (introgression) due to interbreeding between wild and domesticated fish, or the production of sterile hybrids. Indirect effects include the impact of reduced effective population size or altered selection pressure arising from competition or the introduction of pathogens (Skaala et al. 1990; Krueger and May 1991; Waples 1991). Interbreeding between cultured and native populations has been generally found to be disadvantageous when the genetic effects impact fitness-related traits (Hindar et al. 1991), but genetic effects of hybridization between farmed and wild salmon are somewhat unpredictable and may differ between populations. Most studies have focused on the fitness of the Fi generation when exploring the impacts of interbreeding between domestic and wild strains. Such hybrids may in fact have enhanced fitness due to hybrid 17 vigour, or heterosis. A clearer picture would be provided by following the fitness of the F2 and later generations, as this in when outbreeding depression will begin to be expressed (see below). There is also limited information on the ability of introgressed genotypes to show an evolutionary response to environmental pressures. It is not known whether hybrid populations that phenotypically appear to be "wild" would evolve in a manner similar to true wild fish populations. The genetic impact of escaped cultured fish on wild populations will depend on the demographics of the wild population, the magnitude and frequency of the escape, and the extent of introgression of aquacultured genotypes into the wild population (Hutchings 1991). The phenotype of wild and farmed hybrids may vary depending on the source of the wild population (for example see Einum and Fleming 1997). 1.4.1 Effects on Population Size and Inbreeding Depression Escaped cultured fish (possessing low genetic variability) may reduce the effective population size (N e) of wild populations (Ryman 1997) since breeding within a culture environment would increase the variance in family size, leading to a possible population bottleneck and a concurrent founders effect. However, escapees could also increase effective population size if novel alleles are being added to a breeding group from distant populations. Analysis of microsatellite loci in wild and hatchery-reared brown trout used for enhancement purposes indicated reduced N e and a loss of genetic variability (Hansen et al. 2000). A potential effect of reduced population size may be that native fish may need to mate with genetically similar fish, leading to inbreeding depression. Inbreeding depression can occur in fish species as demonstrated in a study by Gjerde et al. (1983) where rainbow trout showed inbreeding depression for survival. Empirical studies (Wang et al. 2002) have indicated that inbreeding has a negative impact on fitness by causing a shift in fitness-related phenotypic traits. Another impact of increased difficulty in finding mates is the positive correlation between hybridization rates (for example between Atlantic salmon and brown trout) and the proportion of escaped farmed fish (Hindar and Balstad 1994). 18 1.4.2 Importance of Maintaining Genetic Integrity of Wild Populations Genotype, in addition to environment, determines the adaptive phenotypic characteristics of salmonids, and as such it is likely that disruption of this genetic structure may have short-term and long-term effects on individual fitness as well as the future resilience of populations to natural and anthropogenic pressures. Significant evidence now exists demonstrating the unique genetic character of populations of several salmonid species. Native salmonid populations can be considered as genetically distinct stocks that have evolved adaptations to maximize fitness under selection regime of their local environments. Genetic differentiation among salmonid populations is well documented (King et al. 2001; Stahl 1987) and in some cases can be correlated with adaptive phenotypes (Beacham and Murray 1987; Beacham et al. 1988; Groot and Margolis 1991; Taylor 1991; Clarke et al. 1995; Quinn 2005). Clear examples of genetically-determined life-history traits of adaptive significance are known, for example that of age of smoltification in Chinook salmon (Clarke et al. 1994) where inland populations tend to migrate to sea after spending at least one year in streams (stream type) whereas coastal populations migrate to the ocean in their first year (ocean type). Another good example of a life-history adaptation is the nonanadromous form of sockeye salmon, kokanee. The two forms of sockeye can be found together in the same population, but they seem to maintain their genetic identity through assortative mating (Foote et al. 1989) and reduced fitness of hybrids (Wood and Foote 1996) Genetic differentiation of populations at the molecular level has been detected using allozymes and D N A markers in the mitochondrial and nuclear genomes, and although such genetic differentiation among populations does not necessarily arise from selection, in many cases it does. For Atlantic salmon in Europe, genetic diversity both within and among populations has been correlated with performance characteristics relevant to survival and recruitment (Bourke et al. 1997; Verspoor 1997). Similarly, an 19 examination of the genetic structure of wild and hatchery brown trout indicated that more than 60% of the genetic variability in the wild populations was due to differences between the populations (Garcia-Marin et al. 1991). In contrast, only 3% of the genetic variability for the hatchery fish was explained by differences between populations, demonstrating that the hatchery fish represent a much more homogenous population with a narrower genetic base. Several species of Pacific salmon in British Columbia show very clear evidence of genetic differentiation down to the subpopulation level (Beacham and Withler 1985; Beacham and Murray 1987; Beacham et al. 2002) and in some cases this differentiation is correlated with evolutionary significant units (ESUs) for the species (Wood and Foote 1996). The homing behaviour of salmon may facilitate genetic differentiation (which may occur by genetic forces in addition to selection), and indeed populations forced into genetic isolation (e.g. chinook salmon introduced into New Zealand from California approximately 90 years ago) already display evidence of phenotypic divergence between populations (Kinnison et al. 1998). Thus, natural populations carry unique reservoirs of genes, and gene combinations, which are specially suited for providing adaptive phenotypes in the environment in which they evolved. Consequently, interactions between strains of fish with different genetic backgrounds are anticipated in some cases to genetically alter local native populations and reduce their viability by altering local phenotypic adaptations (Bams 1976; Verspoor 1988) and threatening intraspecific genetic diversity (Ryman 1997). In addition to providing a cushion against extinction from environmental change, genetic variability in fish populations can hide deleterious recessive genes. Under non-equilibrium conditions (e.g. increased homozygosity due to inbreeding), as the frequency of heterozygotes in the population decreases, there is a loss of alleles and an increase in the expression of recessive alleles that may reduce survival when expressed. Indeed, heterozygosity, a measure of genetic variability, has been observed to be positively correlated with measures of fitness in several instances (Wang et al. 2002). For example, a study on six strains of hatchery-reared rainbow trout showed that higher heterozygosity 20 related to faster development rate (as measured by hatching time) and egg size (Danzmann et al. 1985, 1986; Ferguson et al. 1985). In Chinook salmon (Arkush et al. 2002), and rainbow trout (Ferguson and Drahushchak 1990), higher heterozygosity was associated with greater disease resistance. 1.4.3 Additive Genetic Effects There is a growing body of literature indicating that phenotypic differences between domestic and wild fish are largely a result of additive genetic differences. Therefore, the phenotypic effects of domestication tend to be diluted with repeated backcrossing into wild populations. Interbreeding between farmed and native Atlantic salmon generally resulted in hybrids with intermediate expression of traits such as aggression, growth and anti-predator behaviour (Einum and Fleming 1997). McGinnity et al. (1997, 2003) found that first-generation and second-generation backcross hybrids were intermediate between wild and farm Atlantic salmon in growth, survival and parr maturity rates. For coho salmon, there is a strong correlation between the proportion of domestic genes within the genotype and measures of growth (McClelland et al. 2005). Introgression of domesticated and wild rainbow trout (Ayles and Baker 1983) and Chinook salmon (Bryden et al. 2004) show similar trends for growth and other measures of physiology (such as disease resistance). 1.4.4 Heterosis Crossing different strains with low genetic variability may re-establish lost alleles and allelic combinations, which could lead to hybrid vigour, or heterosis, where the Fi generation would have increased fitness relative to the parental stocks. Heterosis is most likely to occur if the parental stocks are inbred and not highly genetically divergent. However, i f the parent stocks came from different habitats, the resulting progeny may not be well adapted for either habitat. An allele that is advantageous in one genome or environment may be disadvantageous to overall fitness in another genome or 21 environment. Cross-breeding experiments on five Norweigian strains of Atlantic salmon did not find significant heterosis for survival or body weight (Gjerde and Refstie 1984). Further experiments support a lack of heterosis for growth rate in Atlantic salmon fry (Friars et al. 1979). Heterosis for competitive ability may have been detected in hybrids of Atlantic and farm salmon in one instance, but the results are not clear (Einum and Fleming 1997). However, heterosis for growth and survival was detected when two different strains of brook trout were crossed (Webster and Flick 1981). Other studies on rainbow trout (including crosses among inbred lines) have detected evidence of heterosis for body weight (Ayles and Baker 1983; Gjerde et al. 1983; Gjerde 1988; Wangila and Dick 1996). The introduction of low numbers of genetically novel fish into large populations may in some cases be beneficial by providing otherwise unavailable natural variation to the population which can be acted on by natural selection. 1.4.5 Outbreeding Depression Outbreeding depression could cause a negative impact on fitness by disrupting co-adapted allele complexes, which would be indicated by the F 2 generation (Lynch and Walsh 1998). The route of genetic introgression is likely to be through hybridization rather than by pure farm stock displacing pure wild populations, and thus outbreeding depression may not be apparent until recombination has separated co-adapted genotypes. Studies indicate that outbreeding depression can occur in fish populations. After following hybrids of genetically isolated pink salmon, the Fi generation had increased genetic variability relative to the control fish, followed in the F 2 generation by very low survival and increased bilateral asymmetry indicating that outbreeding depression had occurred (Gharrett et al. 1991). More recent studies have confirmed the previous observation that outbreeding depression is possible in populations of pink salmon (Gharrett et al. 1999; Gilk et al. 2004). However, a study by McGinnity et al. (1997) undertaken to examine the survival and growth of farmed, wild and hybrid Atlantic salmon in a natural environment, found no evidence for outbreeding depression as the fitness of the hybrid genotypes was intermediate between the farmed and wild genotypes. 22 No evidence of outbreeding depression was found in crosses of wild coho salmon from different populations (Smoker et al. 2004), and no studies have yet tested for the presence of outbreeding depression during introgression of farmed Pacific salmon into wild populations. The difficulty in obtaining high statistical power in these experiments for estimates of outbreeding depression suggest that these negative results should be interpreted with caution until further data has been developed. 1.5 R E S E A R C H O B J E C T I V E S Through comparisons of phenotype, genotype and fitness of fast-growing domestic and slow-growing wild coho salmon, rainbow trout, and their hybrids, I wish to extend the knowledge summarized above regarding the fitness tradeoffs in the wild of adaptation to a culture environment. This information will indicate the strongest selection pressures on body size and growth of an organism in the wild, and will allow test of the theoretical mode of gene action leading to the phenotypic changes that arise through the process of domestication and selection for faster growth. The main goal of my research can be broken down in the three objectives listed below. For each objective 1 and 3, both coho salmon and rainbow trout species were used. This provides an interesting comparison as the rainbow trout have been domesticated for a much longer time (over 100 years) relative to the coho salmon, which have been reared in the culture environment for the past five to seven generations. Comparisons of the two species may indicate i f the genetic and phenotypic changes observed among the strains are conserved in the process of domestication. 1.5.1 Assessment of Genetically-Determined Differences in Phenotype C h a p t e r II summarizes the observed differences in growth and behaviour of wild and domestic coho salmon and their first- and second-generation hybrids. In C h a p t e r III I have summarized the preliminary growth differences among the strains of rainbow trout 23 (domestic, wild, Fi hybrids, and Fi hybrids backcrossed to both the wild and domestic parents). 1.5.2 Role of Environment in Expression of Heterosis or Outbreeding Depression For this objective, I have taken the novel step of extending the line crosses to a third-generation of hybridization, representing second-generation backcrosses to the wild parental strain. This design simulates the scenario of having domestic fish escape into the wild and interbreeding with the wild population, and allows testing of delayed fitness affects that may not arise except through outbreeding depression. C h a p t e r I V tests the relative fitness of the coho salmon strains when reared in semi-natural mesocosms under competition and risk of predation. With the use of the rainbow trout strains in C h a p t e r V , I have attempted to isolate the relative contribution of risk of predation and competition to selection pressures limiting the growth of fish in the natural environment. C h a p t e r V I extends this examination by combining risk of predation and competition in a field experiment conducted in whole-lake experiment, with replication. 1.5.3 Test for Selection on Endocrine Control of Growth Through Domestication Differences in growth and behaviour among wild and cultured salmonids are often thought to arise from differences in growth hormone (GH) and / or insulin-like growth factor (IGF-I). In C h a p t e r V I I , I have measured hormone profiles (GH, IGF-I, and thyroid hormone T 3 ) of the rainbow trout strains used in Ojective 1 to see i f circulating hormone concentrations may be related to the observed differences in phenotype (specifically growth and behaviour). 24 i.6 R E F E R E N C E S Aarestrup, K. , C. Nielsen, and S.S. Madsen. 2000. Relationship between gill Na+, K+-ATpase activity and downstream movement in domesticated and first-generation offspring of wild anadromous brown trout (Salmo trutta). Can. J. Fish. Aquat. Sci. 57: 2086-2095. Abrahams, M . V. , and A. Sutterlin. 1999. The foraging and anti-predator behaviour of growth-enhanced transgenic Atlantic salmon. Anim. Behav. 58:933-942. Allendorf, F. W. 1991. Ecological and genetic effects of fish introductions: synthesis and recommendations. Can. J. Fish. Aquat. Sci. 48:178-181. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Q. Rev. Biol. 72:149-177. Arkush, K. D., A . R. Giese, H. L. Mendonca, A. M . McBride, G. D. Marty, and P. W. Hedrick. 2002. Resistance to three pathogens in the endangered winter-run chinook salmon (Oncorhynchus tshawytscha): effects of inbreeding and major histocompatibility complex genotypes. Can. J. Fish. Aquat. Sci. 59:966-975. Ayles, G. B. 1975. Influence of genotype and environment on growth and survival of rainbow trout (Salmo gairdneri) in central Canadian aquaculture lakes. Aquaculture. 6: 181. Ayles, G. B., and R. F. Baker. 1983. Genetic differences in growth and survival between strains and hybrids of rainbow trout (Salmo gairdneri) stocked in aquaculture lakes in the Canadian prairies. Aquaculture 33:269-280. Bachman, R. A . 1984. Foraging Behavior of Free-Ranging Wild and Hatchery Brown Trout in a Stream. Trans. Am. Fish. Soc. 113:1-32. Bams, R. A . 1976. Survival and propensity for homing as affected by presence or absence of locally adapted paternal genes in two transplanted populations of pink salmon (Oncorhynchus gorbuscha). J. Fish. Res. Board Can. 33:2716-2725. 25 Beacham, T. D., and C. B. Murray. 1987. Adaptive variation in body size age, morphology, egg size and developmental biology of chum salmon (Oncorhynchus keta) in British Columbia. Can. J. Fish. Aquat. Sci. 44:244-261. Beacham, T. D., K. J. Supernault, M . Wetklo, B. Deagle, K. Labaree, J. R. Irvine, J. R. Candy, K. M . Miller, R. J. Nelson, and R. E. Withler. 2002. The geographic basis for population structure in Fraser river chinook salmon (Oncorhynchus tshawytscha). Fish. Bull. 101:229-242. Beacham, T. D., and R. E. Withler. 1985. Heterozygosity and morphological variability of chum salmon (Oncorhynchus keta) in southern British Columbia. Heredity 54:313-322. Beacham, T. D., R. E. Withler, C. B. Murray, and L. W. Barner. 1988. Variation in body size, morphology, egg size, and biochemical genetics of pink salmon in British Columbia. Trans. Am. Fish. Soc. 117:109-126. Berejikian, B. A. , S. B. Mathews, and T. P. Quinn. 1996. Effects of hatchery and wild ancestry and rearing environments on the development of agonistic behaviour in steelhead trout (Oncorhynchus mykiss) fry. Can. J. Fish. Aquat. Sci. 53:2004-2014. Berejikian, B. A. , E. P. Tezak, L. Park, E. LaHood, S. L . Schroder, and E. Beall. 2001. Male competition and breeding success in captively reared and wild coho salmon (Oncorhynchus kisutch). Can. J. Fish. Aquat. Sci. 58:804-810. Berejikian, B. A. , E. P. Tezak, S. L. Schroder, C. M . Knudsen, and J. J. Hard. 1998. Reproductive behavioral interactions between wild and captively reared coho salmon (Oncorhynchus kisutch). ICES J. Mar. Sci. 54:1040-1050. Bessey, C , R. H. Devlin, N . R. Liley, and C. A . Biagi. 2004. Reproductive Performance of Growth-Enhanced Transgenic Coho Salmon. Trans. Am. Fish. Soc. 133:1205-1220. Bourke, E. A. , J. Coughlan, H. Jansson, P. Galvin, and T. F. Cross. 1997. Allozyme variation in populations of Atlantic salmon located throughout Europe: diversity 26 that could be compromised by introductions of reared fish? ICES J. Mar. Sci. 54:974-985. Bryden, C. A. , J. W. Heath, and D. D. Heath. 2004. Performance and heterosis in farmed and wild Chinook salmon (Oncorhynchus tshawytscha) hybrid and purebred crosses. Aquaculture 235:249-261. Calow, P. 1982. Homeostasis and fitness. Am. Nat. 120:416-419. Carline, R. F., and J. F. Machung. 2001. Critical thermal maxima of wild and domestic strains of trout. Trans. Am. Fish. Soc. 130:1211-1216. Carr, J. W., J. M . Anderson, F. G. Whoriskey, and T. Dilworth. 1998. The occurence and spawning of cultured Atlantic salmon (Salmo solar) in a Canadian river. ICES J. Mar. Sci. 54:1064-1073. Clarke, W. C , C. Groot, and L. Margolis. 1995. Physiological Ecology of Pacific Salmon. U B C Press Clarke, W. C , R. E. Withler, and J. E. Shelbourn. 1994. Inheritance of smoking phenotypes in backcrosses of hybrid stream-type x ocean-type Chinook salmon (Oncorhynchus tshawytscha). Estuaries 17:13-25. Clifford, S. L. , P. McGinnity, and A. Ferguson. 1998. Genetic changes in Atlantic salmon (Salmo salar) populations of northwest Irish rivers resulting from escapes of adult farm salmon. Can. J. Fish. Aquat. Sci. 55:358-363. Crossman, E. J. 1991. Introduced Freshwater Fishes: A Review of the North American Perspective with Emphasis on Canada. Can. Tech. Rep. Fish. Aquat. Sci. 48. Crozier, W. W. 1993. Evidence of genetic interaction between escaped farmed salmon and wild Atlantic salmon (Salmo salar L.) in a northern Irish river. Aquaculture 113:19-29. Dannewitz, J., E. Petersson, J. Dahl, T. Prestegaard, A . -C . Lof, and T. Jarvi. 2004. Reproductive success of hatchery-produced and wild-born brown trout in an experimental stream. J. Appl. Ecol. 41:355-364. 27 Danzmann, R. G., M . M . Ferguson, and F. W. Allendorf. 1985. Does enzyme heterozygosity influence developmental rate in rainbow trout? Heredity 56:417-425. Danzmann, R. G., M . M . Ferguson, F. W. Allendorf, and K. L. Knudsen. 1986. Heterozygosity and developmental rate in a strain of rainbow trout (Salmo gairdneri). Evolution 40:86-93. Devlin, R. H. , C. A . Biagi, and T. Y . Yesaki. 2004. Growth, viability and genetic characteristics of G H transgenic coho salmon strains. Aquaculture 236:607-632. Devlin, R. H. , C. A . Biagi, T. Y . Yesaki, D. E. Smailus, and J. C. Byatt. 2001. Growth of domesticated transgenic fish. Nature 409:781-782. Dickson, T. A. , and H. R. MacCrimmon. 1982. Influence of hatchery experience on growth and behavior of juvenile Atlantic salmon (Salmo salar) within allopatric and sympatric stream populations. Can. J. Fish. Aquat. Sci. 39:1453-1458. Donaldson, E. M . , U . H . Fagerlund, D. A . Higgs, and J. R. McBride. 1979. Hormonal enhancement of growth in fish. Pp. 455-597 in W. S. Hoar, J. D. Randall and J. R. Brett, eds. Fish Physiology. Vol VIII. Bioenergetics and Growth. Academic Press, New York. Doyle, R. W., and A . J. Talbot. 1986. Artificial selection on growth and correlated selection on competitive behaviour in fish. Can. J. Fish. Aquat. Sci. 43:1059-1064. Dunmall, K. M . , and J. F. Schreer. 2003. A comparison of the swimming and cardiac performance of farmed and wild Atlantic salmon, Salmo salar, before and after gamete stripping. Aquaculture 220:869-882. Einum, S., and I. A . Fleming. 1997. Genetic divergence and interactions in the wild among native, farmed and hybrid Atlantic salmon. J. Fish Biol. 50:634-651. Einum, S., and I. A . Fleming. 2001. Implications of stocking: ecological interactions between wild and released salmonids. Nord. J. Freshwat. Res. 75:56-70. 28 Ferguson, M . M . , R. G. Danzmann, and F. W. Allendorf. 1985. Developmental divergence among hatchery strains of rainbow trout (Salmo gairdneri), part 2. Hybrids. Can. J. Genet. Cytol. 27:298-307. Ferguson, M . M , and L. R. Drahushchak. 1990. Disease resistance and enzyme heterozygosity in rainbow trout. Heredity 64:413-417. Fleming, I. A . 1994. Captive breeding and the conservation of wild salmon populations. Conserv. Biol. 8:886-888. Fleming, I. A. , T. Agustsson, B. Finstad, J. I. Johnsson, and B. T. Bjornsson. 2002. Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 59:1323-1330. Fleming, I. A. , and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES J. Mar. Sci. 54:1051-1063. Fleming, I. A. , and M . R. Gross. 1989. Evolution of adult female life history and morphology in a pacific salmon (coho: Oncorhynchus kisutch). Evolution 43:141-157. Fleming, I. A. , and M . R. Gross. 1992. Reproductive behavior of hatchery and wild coho salmon (Oncorhynchus kisutch): Does it differ? Aquaculture 103:101-121. Fleming, I. A. , and M . R. Gross. 1993. Breeding success of hatchery and wild coho salmon (Oncorhynchus kisutch) in competition. Ecol. Appl. 3:230-245. Fleming, I. A. , K. Hindar, I. B. Mjoelneroed, B. Jonsson, T. Balstad, and A. Lamberg. 2000. Lifetime success and interactions of farm salmon invading a native population. P. Roy. Soc. Lond. B Bio. 267:1517-1523. Fleming, I. A. , B. Jonsson, and M . R. Gross. 1994. Phenotypic divergence of sea-ranched, farmed, and wild salmon. Can. J. Fish. Aquat. Sci. 51:2808-2824. Fleming, I. A. , B. Jonsson, M . R. Gross, and A. Lamberg. 1996. An experimental study of the reproductive behaviour and success of farmed and wild Atlantic salmon (Salmo salar). J. Appl. Ecol. 33:893-905. 29 Fleming, I. A. , and E. Petersson. 2001. The ability of released, hatchery salmonids to breed and contribute to the natural productivity of wild populations. Nord. J. Freshwat. Res. 75:71-98. Flick, W. A. , and D. A. Webster. 1964. Comparative first year survival and production in wild and domestic strains of brook trout, Salvelinus fontinalis. Trans. Am. Fish. Soc. 93:58-69. Flick, W. A. , and D. A. Webster. 1976. Production of wild, domestic, and interstrain hybrids of brook trout (Salvelinus fontinalis) in natural ponds. J. Fish. Res. Bd. Can. 33:1525-1539. Foote, C. J., C. C. Wood, and R. E. Withler. 1989. Biochemical genetic comparison of sockeye salmon and kokanee, the anadromous and nonanadromous forms of Oncorhynchus nerka. Can. J. Fish. Aquat. Sci. 46:149-158. Fraser, J . M . 1981. Comparative survival and growth of planted wild, hybrid, and domestic strains of brook trout (Salvelinus fontinalis) in Ontario lakes. Can. J. Fish. Aquat. Sci. 38:1672-1684. Friars, G.W., J.K. Bailey, and R.L. Saunders. 1979. Considerations of a method of analyzing diallel crosses of Atlantic salmon. Canadian Journal of Genetics and Cytology 21: 121-128. Garant, D., J. J. Dodson, and L. Bernatchez. 2003. Differential reproductive success and heritability of alternative reproductive tactics in wild Atlantic salmon (Salmo salar L.). Evolution 57:1133-1141. Garcia-Marin, J. L. , P. E. Jorde, N . Ryman, F. M . Utter, and C. Pla. 1991. Management implications of genetic differentiation between native and hatchery populations of brown trout (Salmo trutta) in Spain. Aquaculture 95:235-249. Gharrett, A.J. , and W.W. Smoker. 1991. Two generations of hybrids between even- and odd-year pink salmon (Oncorhynchus gorbuscha). Can. J. Fish. Aquat. Sci. 48: 1744-1749. Gilk, S.E., L A . Wang, C.L. Hoover, W.W. Smoker, S.G. Taylor, A . K . Gray, and A.J . Gharrett. 2004. Outbreeding depression in hybrids between spatially separated 30 pink salmon, Oncorhynchus gorbuscha, populations: Marine survival, homing ability, and variability in family size. Environ. Biol. Fish. 69: 287-297. Gjedrem, T. 2000. Genetic improvement of cold-water fish species. Aquacult. Res. 31:25-33. Gjerde, B. 1988. Complete diallele cross between six inbred groups of rainbow trout, Salmo gairdneri. Aquaculture 75: 71-87. Gjerde, B., K. Gunnes, and T. Gjedrem. 1983. Effect of inbreeding on survival and growth in rainbow trout. Aquaculture 34:327-332. Gjerde, B., and T. Refstie. 1984. Complete diallel cross between five strains of Atlnatic salmon. Livestock Production Science 11:207-226. Groot, C , and L. Margolis. 1991. Pacific Salmon Life Histories. U B C Press, Vancouver, BC. Gudjonsson, S. 1991. Occurrence of reared salmon in natural salmon rivers in Iceland. Aquaculture 98:133-142. Handeland, S. O., B. T. Bjornsson, A. M . Arnesen, and S. O. Stefansson. 2003. Seawater adaptation and growth of post-smolt Atlantic salmon (Salmo salar) of wild and farmed strains. Aquaculture 220:367-384. Hansen, L. P., J. A . Jacobsen, and R. A . Lund. 1997. The incidence of escaped farmed Atlantic salmon, Salmo salar L. , in the Faroese fishery and estimates of catches of wild salmon. ICES, Copenhagen, Denmark. Hansen, L. P., and B. Jonsson. 1994. Development of sea ranching of Atlantic salmon, Salmo salar L. , towards a sustainable aquaculture strategy. Aquacult. Fish. Manag. 25:199-214. Hansen, M . M . , E. D. Ruzzante, E. E. Nielsen, and D. K. Mensberg. 2000. Microsatellite and mitochondrial D N A polymorphism reveals life-history dependent interbreeding between hatchery and wild brown trout (Salmo trutta L.). Mol . Ecol. 9:583-594. 31 Heggberget, T. G., F. 0kland, and O. Ugedal. 1993. Distribution and migratory behaviour of adult wild and farmed Atlantic salmon (Salmo salar) during return migration. Aquaculture 118:73-83. Hershberger, W. K. , J. M . Myers, R. N . Iwamoto, W. C. Macauley, and A . M . Saxton. 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85:187-197. Hindar, K. , and T. Balstad. 1994. Salmonid culture and interspecific hybridization. Conserv. Biol. 8:881-882. Hindar, K. , N . Ryman, and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Distribution and migratory behaviour of adult wild and farmed Atlantic salmon (Salmo salar) during return migration. 48:945-957. Hutchings, J. A . 1991. The threat of extinction to native populations experiencing spawning intrusions by cultured Atlantic salmon. Aquaculture 98:119-132. Hutchings, J. A . 2004. Norms of reaction and phenotypic plasticity in salmonid life histories. Pp. 510 in A . P. Hendry and S. C. Stearns, eds. Evolution Illuminated: salmon and their relatives. Oxford University Press, Inc., New York. Hvidsten, N . A. , and R. A . Lund. 1988. Predation on hatchery-reared and wild smolts of Atlantic salmon, Salmo salar L. , in the estuary of River Orkla, Norway. J. Fish Biol. 33:121-126. Hynes, J. D., E. H. Brown, J. H . Helle, N . Ryman, and D. A . Webster. 1981. Guidelines for the culture of fish stocks for resource management. Can. J. Fish. Aquat. Sci. 38:1867-1876. Jacobsen, J. A. , and L. P. Hansen. 2001. Feeding habits of wild and escaped farmed Atlantic salmon, Salmo salar L., in the Northeast Atlantic. ICES J. Mar. Sci. 58:916-933. Jacobsen, J. A. , L. P. Hansen, and R. A . Lund. 1992. Occurrence of farmed salmon in the Norwegian Sea. Ices, Copenhagen (Denmark). 32 Joergensen, J., and S. Berg. 1991. Stocking experiments with 0+ and 1+ trout parr, Salmo trutta L. , of wild and hatchery origin: 2. Post-stocking movements. Journal of Fish Biology 39:171-180. Johnsen, B. O., and O. Ugedal. 1986, Feeding by hatchery-reared and wild brown trout, Salmo trutta L. , in a Norwegian stream. Aquacult. Fish. Manage. 17:281-287. Johnsson, J. I., and M . V . Abrahams. 1991. Interbreeding with domestic strain increases foraging under threat of predation in juvenile steelhead trout (Oncorhynchus mykiss): An experimental study. Can. J. Fish. Aquat. Sci. 48:243-247. Johnsson, J. I., and B. T. Bjornsson. 1994. Growth hormone increases growth rate, appetite and dominance in juvenile rainbow trout, Oncorhynchus kisutch. Anim. Behav. 48:177-186. Johnsson, J. I., J. Hojesjo, and I. A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can. J. Fish. Aquat. Sci. 58:788-794. Johnsson, J. I., E. Petersson, E. Jonsson, B. T. Bjornsson, and T. Jarvi. 1996. Domestication and growth hormone alter antipredator behaviour and growth patterns in juvenile brown trout, Salmo trutta. Can. J. Fish. Aquat. Sci. 53:1546-1554. Kallio-Nyberg, I., and M . L. Koljonen. 1997. The genetic consequence of hatchery-rearing on life-history traits of the Atlantic salmon (Salmo salar L.): A comparative analysis of sea-ranched salmon with wild and reared parents. Aquaculture 153:207-224. Kazakov, R. V. , and O. V . Semenova. 1986. Morphological characteristics of hatchery-reared and wild young Atlantic salmon Salmo salar L. Proceedings of the Zoological Institute, Russian Academy of Science. Keller, W. T., and D. S. Plosila. 1981. Comparison of domestic, hybrid and wild strains of brook trout in a pond fishery. N . Y . Fish Game J. 28:123-137. 33 Kennedy, G. J. A. , and J. E. Greer. 1988. Predation by cormorants, Phalacrocorax carbo (L.), on the salmonid populations of an Irish river. Aquacult. Fish. Manage. 19:159-170. King, T. L., S. T. Kalinowski, W. B. Schill, A . P. Spidle, and B. A . Lubinski. 2001. Population structure of Atlantic salmon (Salmo salar L.): A range-wide perspective from microsatellite D N A variation. Mol . Ecol. 10:807-821. Kinnison, M . , M . Unwin, N . Boustead, and T. Quinn. 1998. Population-specific variation in body dimensions of adult chinook salmon (Oncorhynchus tshawytscha) from New Zealand and their source population, 90 years after introduction. Can. J. Fish. Aquat. Sci. 55:554-563. Krueger, C. C , and B. May. 1991. Ecological and genetic effects of salmonid introductions in North America. Can. J. Fish. Aquat. Sci. 48: 66-77. Lachance, S., and P. Magnan. 1990. Performance of domestic, hybrid, and wild strains of brook trout, Salvelinus fontinalis, after stocking: The impact of intra- and interspecific competition. Can. J. Fish. Aquat. Sci. 47:2278-2284. Lacroix, G. L. , B. J. Galloway, D. Knox, and D. MacLatchy. 1997. Absence of seasonal changes in reproductive function of cultured Atlantic salmon migrating into a Canadian river. ICES J. Mar. Sci. 54:1086-1091. Leonard, J. B. K., and S. D. McCormick. 2001. Metabolic enzyme activity during smolting in stream- and hatchery-reared Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 58:1585-1593. Levings, C. D., C. D. McAllister, and B. D. Change. 1986. Differential use of the Campbell River Estuary, British Columbia, by wild and hatchery-reared juvenile chinook salmon (Oncorhynchus tshawytscha). Can. J. Fish. Aquat. Sci. 43:1386-1397. Lund, R. A. , L. P. Hansen, and F. 0kland. 1994. Escaped farmed salmon and geographical zones established for wild fish protection. NINA Oppdragsmelding 303:15. 34 Lura, LL, and F. 0kland. 1994. Content of synthetic astaxanthin in escaped farmed Atlantic salmon, Salmo salar, L. ascending Norwegian rivers. Fisheries Manag. Ecol. 1:205-216. Lura, H. , and H. Saegrov. 1991. A method of separating offspring from farmed and wild Atlantic salmon (Salmo salar) based on different ratios of optical isomers of astaxanthin. Can. J. Fish. Aquat. Sci. 48:429-433. Lura, H. , and H. Saegrov. 1993. Timing of spawning in cultured and wild Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) in the River Vosso, Norway. Ecol. Freshwat. Fish 2:167-172. Lynch, M . , and J. B. Walsh. 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, M A Mason, J. W., O. M . Brynilds, and P. E. Degurse. 1967. Comparative survival of wild and domestic strains of brook trout in streams. Trans. Am. Fish. Soc. 96:313-&. McClelland, E. K. , J. M . Myers, J. J. Hard, L . K. Park, and K. A. Naish. 2005. Two generations of outbreeding in coho salmon (Oncorhynchus kisutch): effects on size and growth. Can. J. Fish. Aquat. Sci. 62:2538-2547. McGinnity, P., P. Prodohl, A. Ferguson, R. Hynes, N . 6 Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. P. Roy. Soc. Lond. B 270:2443-2450. McGinnity, P., C. Stone, J. B. Taggart, D. Cooke, D. Cotter, R. Hynes, C. McCamley, T. Cross, and A. Ferguson. 1997. Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: use of D N A profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54:998-1008. McKinnell, S., A . J. Thomson, E. A . Black, B. L. Wing, C. M . Guthrie, III, J. F. Koerner, and J. H. Helle. 1997. Atlantic salmon in the North Pacific. Aquacult. Res. 28:145-157. 35 Mesa, M . G. 1991. Variation in feeding, aggression, and position choice between hatchery and wild cutthroat trout in an artificial stream. Trans. Am. Fish. Soc. 120:723-727. Mj0lner0d, I. B., U . H. Refseth, E. Karlsen, T. Balstad, K. S. Jakobsen, and K. Hindar. 1997. Genetic differences between two wild and one farmed population of Atlantic salmon (Salmo salar) revealed by three classes of genetic markers. Hereditas 127:239-248. Mork, O. I., B. Bjerkeng, and M . Rye. 1999. Aggressive interactions in pure and mixed groups of juvenile farmed and hatchery-reared wild Atlantic salmon Salmo salar L. in relation to tank substrate. Aquacult. Res. 30:571-578. Morton, A. , and J. P. Volpe. 2002. A description of escaped farmed Atlantic salmon Salmo salar captures and their characteristics in one Pacific salmon fishery area in British Columbia, Canada, in 2000. Alaska Fishery Research Bulletin 9:102-110. Munakata, A. , B. T. Bjoernsson, E. Joensson, M . Amano, K. Ikuta, S. Kitamura, T. Kurokawa, and K. Aida. 2000. Post-release adaptation processes of hatchery-reared honmasu salmon parr. J. Fish Biol. 56:163-172. Nagata, M . , M . Nakajima, and M . Fujiwara. 1994. Dispersal of wild and domestic masu salmon fry (Oncorhynchus masou) in an artificial channel. J. Fish Biol 45:99-110. Okland, F., T. G. Heggberget, and B. Jonsson. 1995. Migratory behaviour of wild and farmed Atlantic salmon (Salmo salar) during spawning. J. Fish Biol. 46:1-7. Palm, S., J. Dannewitz, T. Jarvi, E. Petersson, T. Prestegaard, and N . Ryman. 2003. Lack of molecular genetic divergence between sea-ranched and wild sea trout (Salmo trutta). Mol . Ecol. 12:2057-2071. Peck, J. W. 1988. Fecundity of hatchery and wild lake trout in Lake Superior. J. Great Lakes Res. 14:9-13. Peterson, R. G. 1999. Potential genetic interaction between wild and farm salmon of the same species. Pp. 23. The Office of the Commissioner for Aquaculture Development, Fisheries and Oceans Canada. 36 Poppe, T. H. , D. Griffiths, H. Meldal. 1997. Swimbladder abnormality in farmed Atlantic salmon Salmo salar. Dis. Aquat. Organ. 30:73-76. Poppe, T. J., G. Gunnes, B. Torud. 2003. Heart morphology in wild and farmed Atlantic salmon Salmo salar and rainbow trout Oncorhynchus mykiss. Dis. Aquat. Organ. 57:103-108. Quinn, T. 2005. The Behavior and Ecology of Pacific Salmon and Trout. University of Washington Press, Seattle. Reisenbichler, R. R., and J. D. Mclntyre. 1977. Genetic differences in growth and survival of juvenile hatchery and wild steelhaed trout, Salmo gairdneri. J. Fish. Res. Board Can. 34:123-128. Roberge, C , S. Einum, H. Guderley, and L. Bernatchez. 2006. Rapid evolutionary changes of gene transcription profiles in farmed Atlantic salmon. Mol . Ecol. 15: 9-20. Robison, B. D., P. A . Wheeler, K. Sundin, P. Sikka, and G. H. Thorgaard. 2001. Composite interval mapping reveals a major locus influencing embryonic development rate in rainbow trout (Oncorhynchus mykiss). J. Hered. 92:16-22. Robison, B. D., P. A . Wheeler, and G. H. Thorgaard. 1999. Variation in development rate among clonal lines of rainbow trout (Oncorhynchus mykiss). Aquaculture 173:131-141. Roff, D. A . 1992. The evolution of life histories: theory and analysis. Chapman and Hall, New York. Ruzzante, D. E., and R. W. Doyle. 1991. Rapid behavioural changes in medaka (Oryzias latipes) caused by selection fro competitive and noncompetitive growth. Evolution 45:1936-1946. Ryman, N . 1997. Minimizing adverse effects of fish culture: understanding the genetics of populations with overlapping generations. ICES J. Mar. Sci. 54:1149-1159. Sanchez, M . P., B. Chevassus, L. Labbe, E. Quillet, and M . Mambrini. 2001. Selection for growth of brown trout (Salmo trutta) affects feed intake but not feed efficiency. Aquat. Liv. Res. 14:41-48. 37 Saegrov, H. , K. Hindar, S. Kaalaas, and H. Lura. 1997. Escaped farmed Atlantic salmon replace the original salmon stock in the River Vosso, western Norway. Academic Press, London (UK). Shrimpton, J. M . , N . J. Bernier, and D. J. Randall. 1994. Cortisol dynamics and smolting in hatchery and wild coho salmon. Fish Physiology Association, Vancouver, Canada. Skaala, 0., G. Dahle, K. E. J0rstad, and G. Na^vdal. 1990. Interactions between natural and farmed fish populations: Information from genetic markers. J. Fish Biol . 36:449-460. Smoker, W. W., I. A . Wang, A. J. Gharrett, and J. J. Hard. 2004. Embryo survival and smolt to adult survival in second-generation outbred coho salmon. Journal of Fish Biology 65:254-262. Stahl, G. 1987. Genetic population structure of Atlantic salmon. Pp. 121-140 in N . Ryman and F. Utter, eds. Population genetics and fishery management. University of Washington Press, Seattle. Sundstrom, F. L. , R. H. Devlin, J. I. Johnsson, and C. A . Biagi. 2003. Vertical position reflects increased feeding motivation in growth hormone transgenic coho salmon (Oncorhynchus kisutch). Ethology 109:701 -712. Sundstrom, L. F., M . Lohmus, and R. H. Devlin. 2005. Selection on increased intrinsic growth rates in coho salmon, Oncorhynchus kisutch. Evolution 59:1560-1569. Sundstrom, L. F., M . L5hmus , R. H. Devlin, J. I. Johnsson, C. A . Biagi, and T. Bohlin. 2004. Feeding on profitable and unprofitable prey: comparing behaviour of growth-enhanced transgenic and normal coho salmon (Oncorhynchus kisutch). Ethology 110:381-396. Swain, D. P., and B. E. Riddell. 1990. Variation in agonistic behavior between newly emerged juveniles from hatchery and wild populations of coho salmon, Oncorhynchus kisutch. Can. J. Fish. Aquat. Sci. 47:566-571. Taylor, E. B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98:185-207. 38 Thodesen, J., B. Grisdale-Helland, S. J. Helleand, and B. Gjerde. 1999. Feed intake, growth and feed utilization of offspring from wild and selected Atlantic salmon {Salmo salar). Aquaculture 180:237-246. Thorpe, J. E. 2004. Life history response of fishes to culture. J. Fish Biol. 65:263-285. Tymchuk, W. E., C. A . Biagi, R. E. Withler, and R. H. Devlin. 2006. Growth and behavioural consequences of introgression of a domesticate aquaculture genotype into a native strain of coho salmon (Oncorhynchus kisutch). Trans. Am. Fish. Soc. 135:442-445. Tymchuk, W. E., and R. H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245:295-300. Ugedal, O., B. Finstad, B. Damsgard, and A . Mortensen. 1998. Seawater tolerance and downstream migration in hatchery-reared and wild brown trout. Aquaculture 168:395-405. Utter, F., and J. Epifanio. 2002. Marine aquaculture: Genetical potentialities and pitfalls. Rev. Fish Biol. Fisher. 12:59-77. Vandersteen Tymchuk, W. E., M . V . Abrahams, and R. H. Devlin. 2005. Competitive ability and mortality of growth-enhanced transgenic coho salmon fry and pan-when foraging for food. Trans. Am. Fish. Soc. 134:381-389. Verspoor, E. 1988. Reduced genetic variability in first-generation hatchery populations of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci.45:1686-1690. Verspoor, E. 1997. Genetic diversity among Atlantic salmon (Salmo salar L.) populations. Proceedings of an ICES/NASCO Symposium held in Bath, England, 18-22 April 1997, Academic Press, London (UK), Dec 1997. Vol . 54. Verspoor, E., and C. G. De Leaniz. 1997. Stocking success of Scottish Atlantic salmon in two Spanish rivers. J. Fish Biol. 51:1265-1269. Volpe, J. P., E. B. Taylor, D. W. Rimmer, and B. W. Glickman. 2000. Evidence of natural reproduction of aquaculture-escaped Atlantic salmon in a coastal British Columbia river. Conservation Biology 14:899-903. 39 Wang, S., J. J. Hard, and F. Utter. 2002. Genetic variation and fitness in salmonids. Conserv. Genet. 3:321-333. Wangila, B.C.C., and T.A. Dick. 1996. Genetic effects and growth performance in pure and hybrid strains of rainbow trout, Oncorhynchus mykiss (Walbaum) (order: Salmoniformes, family: Salmonidae). Aquae. Res. 27: 35-41. Waples, R. S. 1991. Genetic interactions between hatchery and wild salmonids: Lessons from the Pacific Northwest. Can. J. Fish. Aquat. Sci. 48:124-133. Webb, J.H., D.W. Hay, P.D. Cunningham, and A.F. Youngson. 1991. The spawning behaviour of escaped farmed and wild adult Atlantic salmon (Salmo salar 1.) in a northern Scottish river. Aquaculture. 98: 97-110. Webster, D.A., and W.A. Flick. 1981. Performance of indigenous, exotic, and hybrid strains of brook trout (Salvelinus fontinalis) in waters of the Adirondack mountains, New York. Can. J. Fish. Aquat. Sci. 38: 1701-1707. Weir, L . K. , J. A . Hutchings, I. A . Fleming, and S. Einum. 2005. Spawning behaviour and success of mature male Atlantic salmon (Salmo salar) parr of farmed and wild origin. Can. J. Fish. Aquat. Sci. 62: 1153-1160. Weir, L. K., J. A . Hutchings, I. A . Fleming, and S. Einum. 2004. Dominance relationships and behavioural correlates of spawning success in farmed and wild male Atlantic salmon, Salmo salar. J. Anim. Ecol. 73:1069-1079. Wood, C. C , and C. J. Foote. 1996. Evidence for sympatric genetic divergence of anadromous and nonanadromous morphs of sockeye salmon (Oncorhynchus nerka). Evolution 50:1265-1279. 40 CHAPTER II2: G R O W T H A N D B E H A V I O U R A L C O N S E Q U E N C E S O F I N T R O G R E S S I O N O F A D O M E S T I C A T E D A Q U A C U L T U R E G E N O T Y P E INTO A N A T I V E S T R A I N O F C O H O S A L M O N (ONCORHYNCHUS KISUTCH) 2 A version of this chapter has been accepted for publication. Tymchuk, W . E . , C. Biagi, R. Withler, and R. H . Devlin. 2006. Growth and behavioural consequences of introgression of a domesticated aquaculture genotype into a native strain of coho salmon. Transactions of the American Fisheries Society 135: 442-455. 41 2.1 I N T R O D U C T I O N Little is known about the genetic consequences of reproductive interaction between farmed strains of Pacific salmon and their wild counterparts. Differences in phenotype between wild and aquacultured salmon are caused both by an environmental effect due to different culture conditions and genetic effects arising from indirect and direct selection pressures in the culture environment. It is the genetic differences between the strains that will determine the long-term consequences of introgression into a native population. The characteristic of the genetic changes caused by indirect selection during domestication can vary according to the method of fish culture and the extent of time that fish spend in an artificial environment (Utter and Epifanio 2002). Depending on the culture program, the artificial rearing component could range from release at the fry or smolt stage (such as for hatchery enhancement) to spending the entire life-history within the culture environment (such as for the farmed fish used in this study). Thus, domestication involves adaptation to a culture environment (Price 1984), and aquaculture strains may have reduced fitness in the wild relative to native populations which have presumably maximized overall fitness (Pyke et al. 1977; Roff 1984). In addition to a domestication effect, farmed fish usually possess genetic differences arising from directed selection (mass selection, family selection, etc.) for enhanced growth. Growth rates in many species have been shown to be highly heritable (Gjedrem 1976), and inter- and intra-population variability exists in nature which could allow natural selection for increased growth. Indeed, fast-growing strains of commercially important salmonids have been developed through the use of selective breeding programs (e.g. Gjedrem 1983; Gjerde 1986; Hershberger et al. 1990). Estimates of genetic variability for performance traits have indicated that there is high potential for enhanced growth in coho salmon through selective breeding (Iwamoto et al. 1982, 1990; Myers et al. 2001). After only four generations under a selective breeding program, coho salmon demonstrated a 60% increase in weight with consistently high levels of 42 heritability (Hershberger et al. 1990). Selective breeding for enhanced growth typically selects the largest mature individuals to breed for the future generation, and is usually applied without knowledge of the underlying genetic and physiological mechanisms that are being selected for. Evolution of quantitative characters is generally assumed to be caused by a large number of genes, each with a small effect. However, this assumption is often not tested, and there is evidence that major genes may be involved in adaptation in some instances (Orr and Coyne 1992). Understanding the genetic basis of phenotypic variation among strains, the response to selection and the inheritance of hybrid phenotypes is of value to aquaculture in developing marker-assisted selection programs for desired traits. Further, this information is necessary to predict and minimize impact of interaction between fast-growing domesticated and natural populations (Utter et al. 1993) through understanding of tradeoffs between growth, survival and reproduction as well as the underlying genetic mechanisms responsible for the growth differences. Body size is one of the most important phenotypic characteristics of an animal, influencing fitness parameters, life history, and population ecology. Since the growth rate of an animal links its size at age, growth rate is therefore an important factor in optimal life history strategy which maximizes age-specific survival and reproductive fitness (Roff 1984). However, in many cases growth rates remain submaximal in the wild relative to a species' upper physiological potential for growth as demonstrated by compensatory growth responses, selective breeding, and hormone supplementation studies (Donaldson 1970; Donaldson et al. 1979; Metcalfe and Monaghan 2001). As many fitness traits are positively correlated with size, submaximal growth rates seem an unlikely outcome of natural selection unless there are major counteracting fitness costs associated with fast growth, as suggested by principles of evolutionary theory (Calow 1982; Arendt 1997) and by empirical studies on the costs of compensatory growth after growth retardation (Ali et al. 2003). It is expected that directed selection for enhanced growth will accentuate the behavioural, morphological and physiological differences between farmed and native fish, and further reduce the fitness of farmed fish in the natural environment beyond that occurring through domestication. For Atlantic salmon, farmed and farmed x 43 wild hybrids were able to survive in the wild (McGinnity et al. 1997; Fleming et al. 2000) and successfully produced F 2 generation offspring (McGinnity et al. 2003), but the presence of farmed genotypes lowered the overall fitness of the population. The present study examines the phenotypic effect of introgression of growth-enhanced domesticated Pacific salmon into a native genetic background over three generations under controlled laboratory conditions. As there is significant variation in growth among populations of coho salmon (Withler and Beacham 1994), growth of three additional natural populations of this species was also examined under the same conditions to assess whether initial differences between the farmed and wild strains are reflective of natural variability among populations or arise from changes caused by domestication and selection for enhanced growth. Further, the contribution of additive and dominance genetic effects to observed differences in growth and behaviour among the parental lines (farmed and native) and hybrid crosses (Fi and F 2 hybrids and F 2 backcross progeny) have been examined. 2.2 M A T E R I A L S A N D M E T H O D S 2.2.1 Oncorhynchus kisutch Families Crosses between farmed (D) and native Chehalis River (Ch) coho salmon were made December 1998 at the West Vancouver Laboratory (WVL), BC, producing pure D, pure Ch, and Fi offspring. Gametes from farmed parents were obtained from Target Marine Hatchery, Sechelt, BC. This growth-enhanced domesticated strain of coho was originally derived from Kitimat River and Robertson Creek populations (BC) and has been selected for enhanced growth for six generations. Substantial anecdotal information from aquaculture and laboratory observations, in addition to some published data (Devlin et al. 2001), have previously suggested that this strain is faster growing than native coho salmon populations in British Columbia. Gametes from Chehalis River parents were obtained from a mix of hatchery-produced and naturally-spawned adult fish returning to 4 4 the Chehalis River Hatchery. Eggs from four females were pooled, subdivided and fertilized separately with sperm from four males to produce the pure D and F i strains. For the pure Ch strain, eggs from nine females were pooled, subdivided and fertilized separately with sperm collected from five different males. Offspring from these fertilizations were reared at West Vancouver Lab until maturity. Using these mature adults, the following crosses were performed at W V L on 03 January 2002 (Table 1): pure farm (D; two families), pure Chehalis (Ch; eight families), farm x Chehalis and Chehalis x farm (Fi; six families), F i x F i (F2; eight families), and F i x Chehalis and Chehalis x F i ( F 2 backcrosses, Bch.; eight families each). To assess the variability among populations of coho salmon found in nature, the following additional fertilizations were completed: 13 November 2002, Kitimat (five families); 20 November 2002, Robertson Creek (five families) and Big Qualicum (four families); 27 November 2002, pure domestic Target Marine (five families). Crosses of farmed Target Marine fish were fertilized again at this time to provide a reference point to the previous set of fertilizations in January 2002. Incubation temperatures were adjusted as needed so that the eggs hatched on the same day. 45 T A B L E 2.1 Cross design for 03 January 2002 producing the coho salmon lines used in this study. The farm (D), native Chehalis River (Ch) and farm x native hybrid (Fi) females (represented by numbers) and males (represented by letters) were crossed to produce parental strains in addition to Fi x wild backcrosses (Bch) and F2 hybrids. The shaded cells indicate the families that were reared as both individual and mixed family groups. Growth of these families was tracked until maturity. A l l other families were terminated October 2002. D(l) D(2) Ch(l) Ch(2) Ch(3) Ch(4) F, (1) F, (2) Fi (3) F, (4) D (A) F, D(B) D Fi D (C) D Fi D (D) F, Ch(A) Ch Ch B C h B C h Ch(B) F, Ch Ch ' B C h " B C h Ch (C) Fi • Ch Ch B C h B C h Ch(D) Ch Ch B C h B C h Fi(A) B C h B C h F 2 F 2 Fi(B) B C h B C h . . " F 2 F 2 . Fi (C) -Bch ' B C h F 2 F 2 Fi(D) B C h B C h F 2 F 2 46 2.2.2 Growth Performance On 06 M a y 2002 the alevins were transferred to rearing tanks and feeding was initiated. Families were reared both separately (individual groups consisted of 50 siblings from the same family) and together. The mixed groups consisted of 15 individuals selected from 12 families from the D , Ch , F i and Bch crosses. The families represented in the mixed groups are indicated in Table 1. The individual groups were reared in 200 L tanks under culture conditions (well water at a flow no less than 1 L min" 1 kg" 1, temperature 11°C; supplemental aeration) and were fed commercial feed (Skretting, Vancouver, Canada; composition of feed varied over time to meet the requirements of different life stages) to satiation several times per day. In October 2002, due to space requirements of the growing fish, all the individually reared families were terminated with the exception of those families that were also represented in the mixed groups (see Table 1). Two mixed groups were reared under culture conditions, and two were reared under semi-natural conditions in l m x 3m stream troughs with enriched environments (creek water, gravel substrate, branches and rocks) and limited food resources (< 50% of the amount fed to the culture-reared groups). A l l fish used in the mixed groups were marked at the first-feeding fry stage by injection of different colours of visible elastomer (VIE, Northwest Marine Technology, Inc.) in different body locations so that they could be identified to family. A t 280 days post-fertilization (dpf), a Passive Integrated Transponder (PIT) tag was inserted into each fish so that individual growth rates could be tracked. A t this time, an adipose fin was collected from the mixed group fish as the V I E tags did not provide completely reliable identification for two of the groups (red vs orange). Consequently, microsatellite analysis conducted on genomic D N A extracted from the fin tissue was utilized to allow family identification when the V I E tag discrimination was not clear. 47 Genotyping of five polymorphic loci in the parental fish enabled genotyping and unambiguous family assignment of progeny that could not be identified by tag colour. The loci used were Ogo2 (Olsen et a l , 1998), Omy325 (McConnell et al., 1997), Onelll (Olsen et al., 2000), Otsg253b (Williamson et al., 2002) and Ssa407 (Cairney et al., 2000). Amplification products were size fractionated (96 samples per gel), on 4.5% denaturing polyacrylamide gels. The gels were run at 3,000 V for 2.5 hr at a gel temperature of 51 °C on an ABI 377 automated D N A sequencer. Allele sizes were determined with Genescan 3.1 and Genotyper 2.5 software (PE Biosystems, Foster City, C A , U.S.A.). Mass and length measurements were collected on May 08, 2002 and then approximately every two months from August 2002 until February 2003. This provided a total of five time points in their growth trajectory at 126, 220, 280, 349, and 430 days post-fertilization. The last measurement marked the end of the freshwater phase for the fish, after which they were reared in sea water. In May 2002, sub-samples of 25 individuals from each family were measured, whereas all individuals in each family were measured in subsequent sampling periods. Since specific growth rate may be correlated with size of the fish at the beginning of the experiment, a standardized estimate of specific growth rate in mass (SGR m ) was used: (2.1) S G R m = M * ~ M ' A x l 0 0 b x time where b is the allometric mass exponent for the relation between growth rate and mass, and M\ and M 2 represent mass at t\ (124 dpi) and h (430 dpf). Estimates of b have not been estimated directly for coho salmon, but similarities between estimates for brown trout (0.308) and Atlantic salmon (0.310) suggest that b = 0.31 may be an acceptable estimate for salmonids (Elliott et al. 1995; Elliott and Hurley 1997; Quinn et al. 2004) and thus was used in the present study. Specific growth rate in length (SGRi ) was calculated as (2.2) SGR = l n ( Z 2 ) - l n ( Z l ) x l 0 0 time 48 where L\ and £2 represent length at t\ (124 dpf) and ?2 (430 dpf). Condition factor (K) was calculated as: (2.3) K = fFiLr Jxl00 Fixed effects A N O V A (and Kruskal-Wallis rank analysis when data were not normal) was used to determine statistical relationships between groups, followed by Tukey's pairwise comparison to distinguish groups (SIGMASTAT, Systat Software, Point Richmond, CA). 2.2.3 Anti-predator Behaviour Anti-predator behaviour trials were conducted in September 2002 on each individually-reared family. To test for strain differences in response to predation, a predator attack was simulated with a great blue heron model. A similar technique has been used to measure predator response in rainbow trout (Jonsson et al. 1996; Hojesjo et al. 1999) and Atlantic salmon (Johnsson et al. 2001). After the predator attack, a food pellet was offered to the group of fish every 10 s until a pellet was consumed. The time required until consumption of one pellet was then recorded. These trials took place before first feeding in the morning (fish had not eaten for 18 hours) and were repeated four times for each family. 2.2.4 Resistance to Bacterial Kidney Disease After the sampling period in February 2003, the individually reared fish were transferred from their 200 L tanks and grouped together into two 2 m circular tanks. After this transfer, the fish began to suffer natural mortality from bacterial kidney disease (BKD). The mortality of fish due to B K D was recorded until May 2003 at which point all fish were treated with erythromycin. 49 2.2.5 Analysis of Line Means / Genetic Architecture Joint-scaling regression technique was used to test for additivity and dominance and their contribution to the phenotypic divergence between slow- and fast-growing salmon, following the procedure outlined in Lynch and Walsh (1998). The joint-scaling test uses least-squares regression to estimate the model parameters and then compares the observed line means with predicted means (Cavalli 1952; Hayman 1960). This analysis was carried out for mass, length, S G R m , S G R i , condition factor, and response to predator attack. The following model was fit to the observed line means, (2.4) *>:=M,i// 0 +M , . 2 a ,+M, 3 £,+£ ; . where z, is the /th line mean, uo is the mean of all line means, o.t is the additive genetic effect, 5, is the dominance effect, and s, is the residual error. The error term incorporates deviation of the observed and predicted line means in addition to any epistatic effects which could not be estimated directly due to a limited number of degrees of freedom. With M as the matrix of coefficients, the linear model becomes, (2.5) z = M a + e where a is a vector containing estimates of the model parameters, mean (ju), additive effects (a) and dominance effects (S). The weighted least-squares solution is ( 2 6 ) a = ( M ^ M V " 1 M T V 'z with V as a diagonal matrix of squared standard errors of the means, and z a vector of line means. The sampling variances and covariances of the parameter estimates are given by the covariance matrix (2.7) C = ( M ^ - ' M ) - 1 The additive and additive-dominance genetic models were tested by sequential model fitting, beginning with the simpler additive model. The additive model was tested with a goodness-of-fit test statistic 50 (2.8) Y a r ^ with degrees of freedom equal to the number of lines (four) minus the number of estimated parameters. For the additive model, df = 2. The additive-dominance model was then tested, with df = 1. The likelihood-ratio test was used to see if the additive-dominance model (AD) was a significant improvement over the additive model (A) (2.9) withdf= 1. A — ZA ZAD 2.3 R E S U L T S 2.3.1 Growth Performance In the present study, all fish were produced and reared under similar conditions within a controlled laboratory environment to minimize environmental effects such that observed differences in growth and behaviour could be primarily attributed to genetic effects. No significant differences were observed when the same families were reared in replicate tanks, supporting the assertion that environmental effects were being effectively minimized. For the individually-reared families of farm (D), Chehalis (Ch), Fi hybrid, F 2 hybrid, and F 2 backcross (Fi x native; Bch) genealogy, genetic background had an effect on final size at age, for both mass (Fig. 2.1a; F 4 ; i 3 = 5.983, p = 0.012) and length (Fig. 2.1b; Fzi.n = 6.833, p = 0.008). Strong and significant correlations exist between mass and genotype (r2 = 0.705, p < 0.001) and between length and genotype (r2 = 0.743, p < 0.001). Pair-wise comparisons indicated that D families were larger than Ch or Bch at final size. As early as the first weight and length measurement at 126 dpf, the rank of mean mass and length of the crosses was as follows: D > F 2 > Fi > Bch > Ch. This trend continued for the duration of the growth trial. There were no differences in C V for mass (F 4 , i 3 = 0.782, p = 0.564) among the families (D = 22.9 ± 2.7%; F, = 18.6 ± 2.3%; F 2 =28.0 ± 10.4%; B C h = 22.9 ± 3.5%; Ch = 18.0 ± 0.8%), 100 150 200 250 300 350 400 450 Days Post-Fertilization Figure la. 0 4 , , , , , , 1 100 150 200 250 300 350 400 450 Days Post-Fertilization Figure lb. FIGURE 2.1 Size at age for individually reared coho salmon families. Points represent genotype means and the associated standard error for (a) mass; and (b) length. Groups are as follows: D, farm; Ch, Chehalis; F i , farm x Ch, and Ch x farm; F 2 , Fi x F i ; Bch, Fi x C h a n d C h x F i . 52 ( T l *Ch 0 25 50 75 100 % Farm Alleles Figure 2a. 0.44 -4 0 25 50 75 100 % Farm Alleles Figure 2b. [F2 T " 1 *Ch i Fl 0 25 50 75 100 % Farm Alleles Figure 2c. F I G U R E 2.2 Relationship between the proportion of farm alleles in the genotype for the individually reared families and (a) SGRm, specific growth rate in mass, % bw day-1; (b) SGR1, specific growth rate in body length, % bl day-1; or (c) condition factor. Points represent genotype means and associated standard errors. Groups are as described in Figure 2.1, where D = 100%, F, and F 2 = 50%, B C h = 25% and Ch = 0%. The F i and F 2 values are offset by 1 % for clarity only. The regression line connecting the D and W means indicates a priori expectation of additive genetic effects; character means for all hybrids should fall on this line if divergence is due to genes with only additive effects. 53 There was a genotype effect on specific growth rate from 126 to 430 dpf for both mass (SGRm, Fig. 2.2a; A N O V A F 3 = 7.656, p = 0.006) and length (SGRi, Fig. 2.2b; A N O V A F3 = 4.904, p = 0.024), with significant correlations between genotype and SGR m (r 2= 0.675, p < 0.001) and between genotype and SGR, (r2 = 0.554, p = 0.002). The rank of SGR for both mass and length was the same as that for final mass and length. Condition factor did not vary significantly among the families (Fig. 2.2c; F 4 1 3 = 1.570, p = 0.263). 2.3.2 Environmental Effects on Growth Growth of mixed groups of D, Ch, F i and Bch reared under culture and semi-natural conditions was tracked over time. There was a genotype effect on growth in mass (Fig. 2.3a,b; F 3 , , 6 = 19.664, p < 0.001) and length (Fig. 2.3c,d; F 3 , i 6 = 20.295, p < 0.001), and the rank of size at age was the same as that found for the individually reared groups with D > Fi > Bch > Ch. As for individually-reared groups, there were strong and significant correlations between mass and genotype (culture, r = 0.730, p < 0.001; semi-natural, r 2 = 0.687, p < 0.001) and length and genotype (culture, r 2 = 0.777, p < 0.001; semi-natural, r 2 = 0.750, p < 0.001). Pair-wise comparisons indicated that all genotypes were different in both mass and length, except for Ch and Bch- There was not a detectable genotype x environment interaction (mass, F 3 j i 6 = 0.782, p = 0.521; length, F 3 , i 6 = 0.777, p = 0.524). C V for mass was higher when families were reared as mixed groups (for both culture and enriched conditions) relative to individually-reared families (F2,24 = 4.608, p = 0.020). There was no genotype effect on C V for mass ( F 3 ; 2 4 = 2.925, p = 0.054) nor a genotype x environment interaction (F6,24 = 1 -495, p = 0.222). There was a significant genotype effect on specific growth rate in mass (SGR m , Fig. 2.4a; F 3 ; i 6 = 13.897, p < 0.001) and specific growth rate in length (SGR, Fig. 2.4b; F 3 , i 6 - 3.705, p = 0.034) with no environment effect detected between the culture and enriched habitats (SGR m , F U 6 = 0.394, p = 0.539; S G R i , F i , i 6 = 0.391, p = 0.541) and no genotype x environment interaction (SGR m , F 3 > i6 = 0.833, p = 0.495; S G R i ; F 3 j i 6 = 0.786, p = 0.519). Pair-wise comparisons indicated that the pure domestics had significantly 54 30 io H 30 10 Enriched 100 150 200 250 300 350 400 450 Days Post-Fertilization 100 150 200 250 300 350 400 450 Days Post-Fertilization Figure 3a. Figure 3b. 0 0 100 150 200 250 300 350 400 450 100 150 200 250 300 350 400 450 Days Post-Fertilization Days Post-Fertilization Figure 3c. Figure 3d. F I G U R E 2.3 Mean size at age (and associated standard error) for mixed groups reared under culture and semi-natural (enriched) conditions. Graphs are: (a) mass over time under culture conditions; (b) mass over time in an enriched environment; (c) length over time under culture conditions; (d) length over time in an enriched environment. Groups are as decribed in Fig. 2 .1 . 55 2.0 A O 0 25 50 75 % Farm Alleles Figure 4a. 0.4 O 0 25 50 75 % Farm Alleles Figure 4b. • Culture O Enriched '-a 0.0115 o U Ch B, Fl 0 25 50 75 % Farm Alleles Figure 4c. F I G U R E 2 .4 Relationship between the proportion of alleles of farm origin in the genotype and (a) SGRm, specific growth rate in mass, % bw day-1; (b) SGRI, specific growth rate in length, % bl day-1; or (c) condition factor, determined in each case for the mixed groups reared under culture (solid circle) and enriched (open circles) environments. Points represent genotype means and associated standard errors. Groups are as described in Figure 2.1 and 2 . 2 . 56 higher SGR m relative to F i , Bch, and Ch and significantly higher SGR| relative to Bch and Ch. Significant correlations were found between SGR m and genotype for both the culture (r2 = 0.738, p < 0.001) and enriched (r2 = 0.661, p = 0.001) environments. Genotype did not have a significant effect on condition factor among the mixed-family groups (Figure 4c; F 3 1 6 = 2.144, p = 0.121) and there were no significant genotype x environment interactions (F6,i6 = 1.311, p = 0.290). Environment (culture or enriched) did have a significant impact on condition factor (F2,i6 = 22.851, p < 0.001) with higher condition factors observed for the culture environment. 2.3.3 Growth Comparisons with Different Wild Strains Crosses representative of different natural coho populations were reared in the hatchery and their growth tracked over time. At smoltification (430 dpf, 2002 for Chehalis and 470 dfp, 2003 for all other populations) there was significant variability among the populations in both mass (Fig. 2.5a; F 5 J 5 = 14.148, p < 0.001) and length (Fig. 2.5b; F 5 ; i 5 = 14.825, p < 0.001) with the rank size as follows: Domestic > Robertson > Kitimat > Chehalis (430 dpf, 2002) > Big Qualicum. Pair-wise comparisons detected no significant difference between the domestic crosses reared in consecutive years (2002 vs 2003), indicating that rearing conditions were not significantly different between the two rearing periods. The farm families were significantly larger in mass than all populations except for Robertson; for length, Chehalis and Big Qualicum populations were significantly smaller than all other populations. Until 227 dpf, each family was reared in replicate. No tank effects were detected between replicate groups (paired t-test, t.36 = -0.106, p = 0.916) indicating that variation due to environment was being successfully minimized with consistent satiation feeding of all families and periodic tank rotations. 57 30 100 150 200 250 300 350 400 450 500 Days Post-Fertilization Figure 5a. 2 -\ , , , , , n , 1 100 150 200 250 300 350 400 450 500 Days Post-Fertilization Figure 5b. FIGURE 2.5 Mean mass (a) and mean length (b) at age for families representing four wild coho salmon populations examined in 2003 and a farm strain examined in both 2002 and 2003. fe 1 fe o 1 0 25 50 75 100 % Farm Alleles Figure 6a. 12 H fe 2 4 I P r = 0.66 Figure 6b. 0 4 8 12 Mean Mass (rank) 16 Figure 6c 0 , 4 8 12 16 Mean Length (rank) FIGURE 2.6 Time required for recovery from predator attack and its relationship to (a) genotype; (b) mean mass; and (c) mean length. The groups are as described in Figure and 2.2. 59 2.3.4 Behavioural Differences Fish in individually reared families were attacked with a model of a great blue heron (three strikes per tank), and then offered a pellet every 10 seconds until a pellet was consumed. There was a significant genotype effect on the rank time to first feeding (Fig. 2.6a; F 4 1 3 = 4.036, p = 0.038), with farm families taking significantly less time (mean 10.2 ± 0.2 s) to begin feeding relative to wild families (22.5 ± 5.8 s). The rank of mean time to first feeding for the five genotypes was as follows: D < F 2 < Fi < Bch < Ch. There was a significant negative correlation between the rank mean mass of the family and the rank time to first feeding (Fig. 2.6b; y = 13.52 - 0.80x, r 2 = 0.66). Rank mean length was also negatively correlated with rank time to first feeding, although not as strong as for mass (Fig. 2.6c; y = 11.80 - 0.57x, r 2 .= 0.33). 2.3.5 Resistance to Bacterial Kidney Disease Mortality in sea water post-smoltification due to natural B K D exposure was monitored in duplicate tanks. There was no difference in mortality between the two tanks ( F U 8 = 0.139, p = 0.713). The differences in mortality between D (25 ± 9%), Fi (15 ± 6%), F 2 (6 ± 9%), B C h (14 ± 9%), and Ch (16 ± 9%) were not significant (F 4 , i 8 = 0.574, p = 0.685), and there was no genotype x tank interaction (F^is = 0.179, p = 0.946). 2.3.6 Analysis of Line Means The lines in Figure 2.2 (growth of individually reared groups), Figure 2.4 (growth of group-reared families) and Figure 2.6a (response to risk of predation) represent an a priori expectation that the character means for all hybrids should fall on this line if the fast- and slow-growing strains have diverged primarily from variation at genetic loci with additive effects. Dominance effects would cause displacement of hybrids above or below the line, with Fi displacement twice that of F 2 and Bch- The size of the displacement is anticipated to be proportional to the degree of dominance. T A B L E 2.2 Joint-scale analysis of individually reared families. Values represent X1 statistic measuring goodness-of-fit of line means to genetic models, where a larg value indicates a poor fit. Underline indicates adequate fit of model to data. Additive Additive - Dominance degrees of freedom 3 2 anti-predator response 0.98 mass 22.54 20.13 length (L48 S G R m H i SGRi 2JA T A B L E 2.3 Joint-scale analysis of mixed groups reared under enriched or culture conditions. Values represent x 2 statistic measuring goodness-of-fit of line means to genetic models, where a large value indicates a poor fit. Underline indicates adequate fit of model to data. Enriched Culture Additive Additive - Dominance Additive Additive - Dominance degrees of freedom mass length SGR m SGR, 2 14.71 0.84 0.52 1.79 1 0.93 2 20 .15 1.95 0.61 5.31 1 1.38 62 In all cases, a model of no effect was rejected, indicating genetic divergence in all characters tested (mass, length, SGR m , SGRi, condition factor, and anti-predator behaviour). For the individually-reared families (Table 2.2), all traits except mass were not significantly different from the additive model. Results for the group-reared families (Table 2.3) were similar, with all traits except mass fitting an additive genetics model. The additive-dominance model provided a significant improvement in fit for regressions of mass, but had no effect for any of the other traits. 2.4 D I S C U S S I O N The present study was undertaken to examine the phenotypic consequences arising from introgression of a domesticated aquaculture genotype into a wild strain of Pacific salmon. We have found that growth differences between strains of coho salmon are largely a result of additive genetic differences, and that phenotypic effects of domesticated genotypes are largely diminished within two generations of backcrossing to wild salmon. These observations agree with research on the introgression of domesticated and wild rainbow trout (Oncorhynchus mykiss) genotypes and their corresponding effects on growth (Tymchuk and Devlin 2005). McGinnity et al. (2003) found similar results in that Fi and backcross hybrids were intermediate between wild and farm Atlantic salmon (Salmo salar) in growth, survival and parr maturity rates. Bryden et al. (2004) also concluded that differences between farmed and wild Chinook salmon were due mainly to additive genetic variation. The farm strain of coho salmon used in the present research was originally derived from hybrids of Kitimat River and Robertson Creek coho salmon populations. Pure crosses made from natural Kitimat River and Robertson Creek populations showed faster growth relative to the other natural populations tested in this study (Big Qualicum and Chehalis Rivers). This inherent difference among strains may explain some of the growth differences observed between the farm and Chehalis coho salmon strains in the 63 present study, but the farm strain had growth rates higher than the populations from which is was derived, indicating that size-selection has also occurred. Body mass at smoltification did not fit a simple additive genetic model among pure and hybrid strains of coho salmon, although specific growth rate for mass did. For the individually-reared families, a lack of fit with the additive-dominance model would suggest significant epistasis for final mass, although this term would also contain any effects due to common environment (one family per tank) as there were not enough line crosses to estimate epistasis directly. The additive-dominance model was an adequate fit for the mixed groups reared under either culture or enriched conditions. Artificial propagation is known to cause genetic-based differences in behaviour (Reisenbichler and Rubin 1999; Huntingford 2004). Optimal foraging theory assumes that the fitness associated with an animal's foraging behaviour has been maximized by natural selection, thereby maximizing net rate of energy intake (Pyke et al. 1977). Animals will generally have to make trade-offs between rate of energy intake and risk of predation in order to maximize fitness. Indeed, coho salmon are able to adjust their behaviour in response to varying risks of predation by reducing their foraging behaviour, therefore making trade-offs between risk of predation and food intake that are dependent on hunger level of the fish (Dill and Fraser 1984; Abrahams and Dil l 1989). Altered selection regimes may have increased the motivation to eat in faster-growing strains of fish, resulting in adoption of a more "high-gain/high-risk" foraging strategy (Johnsson 1993). Fish with high growth rates due to hormone supplement (by injection or transgenesis) do have higher appetites and are more willing to risk increased predation in order to forage (Higgs et al. 1975; Johnsson et al. 1996; Devlin et al. 1999; Dunham 1999; Abrahams and Sutterlin 1999; Abrahams and Pratt 2000; Sundstrom et al. 2003; Sundstrom et al. 2004a). The genetic divergence of anti-predator response between the fast- and slow-growing strains of fish observed in the present study agrees withprevious studies on cultured salmonids (Johnsson and Abrahams 1991; Berejikian 1995; Sundstrom et al. 2004b), however it remains unclear what physiological mechanisms are driving these 64 differences. Response to risk of predation was correlated with the size of the fish, which could mean that the larger fish perceive a lower risk of being consumed by the predator or that the larger fish are more motivated to eat (due to either higher metabolic requirements or hormonal production leading to increased appetite and feeding behaviour). If the differences in behavioural response to risk of predation are due to feeding motivation, it is still not clear if the fish are driven to eat more to meet metabolic needs or by desire to eat more (due to hormonal control of appetite). Since line means for anti-predator response fit an additive model, behavioural changes are likely the combined action of many physiological characteristics. Fleming et al. (2002) have found evidence of a connection between selection for growth in domestic Atlantic salmon and corresponding changes in growth endocrinology and it will be informative to determine whether such effects are also contributing to phenotypic traits in the present study. The stronger correlation seen for mass (relative to length in Fig. 6a and b) may suggest that fat body size (arising from enhanced feeding motivation) is more important than growth (since length is generally a better indicator of growth). Results of this study indicate that the farm coho salmon studied are able to maintain their growth advantage under altered environmental conditions. This observation differs somewhat with Fleming and Einum (1997) who found that interpopulation competition had a negative impact on the growth of domesticated Atlantic salmon, under both culture and semi-natural conditions, with the domestic fish growing slower than the wild fish under semi-natural conditions. In contrast, domesticated farm coho salmon families grew the fastest under all conditions examined. A similar trend was found by McGinnity et al. (1997 and 2003) where farmed Atlantic salmon grew faster than wild conspecifics in the natural environment. In the present study, significant differences in CV in mass among the three rearing environments (individually-reared, mixed groups under culture conditions, mixed groups under enriched conditions) indicate that social interactions influenced the growth of the fish by increasing variability in growth when reared as mixed groups. Simulated habitats with more food-restricted environments, increased competition, or the addition of predators 65 may have resulted in different observations. Indeed, domestic strains of brook trout (Salvelinus fontinalis) also have been found to have reduced survival and slower rates of growth relative to wild or domestic x wild hybrids when released into natural lakes (Fraser 1981; Webster and Flick 1981). Wild steelhead trout (Oncorhynchus mykiss) had higher survival than hatchery and hatchery x wild fish when reared in natural streams, whereas the hatchery fish had the highest survival and growth under culture conditions (Reisenbichler and Mclntyre 1977). This study did not show the presence of either hybrid vigour in the F\ crosses, or outbreeding depression in the F 2 crosses for measures of growth or survival. A similar study with rainbow trout (Tymchuk and Devlin 2005) also found no evidence of heterosis in the Fi crosses, and Smoker et al. (2004) did not find evidence of outbreeding depression in crosses of wild coho salmon from different wild populations. Low statistical power in our experiment and in Smoker et al. (2004) for estimates of outbreeding depression in survival suggest that these negative results should be interpreted with caution. Furthermore, heterosis or outbreeding effects may only be detectable under certain environmental conditions, and the full complexity of the natural environment cannot be replicated with the lab. F 2 crosses between spatially separated populations have provided evidence of epistatic outbreeding depression in other studies suggesting that disruption of co-adapted gene complexes of epistatic genes can occur (Gharrett & Smoker 1991; Gharrett et al. 1999; Gilk et al. 2004). Knowledge of the genetic changes responsible for altered growth rates in fish is crucial information needed to increase our ability to predict the consequences of introgression between fast and slow-growing strains of fish. Farm salmonids have a well-documented escape record (Hansen et al. 1987; Gausen and Moen 1991; Gudjonsson 1991; McKinnell et al. 1997; Grozier 1998; Hansen et al. 1999), with evidence of successful interbreeding and introgression observed between domestic and wild Atlantic salmon (Crozier 1993; Carr et al. 1999), brown trout (Salmo trutta) (Barbat-Leterrier et al. 1989; Cagigas et al. 1999; Berrebi et al. 2000; Fritzner et al. 2001; Hansen et al. 2001), and rainbow trout (Campton and Johnston 1985). Experimental interactions 66 between farmed and wild Atlantic salmon resulted in a reduction in fitness (McGinnity et al. 2003), consistent with other research indicating that genetic interactions between wild and cultured fish have impacted natural populations (summarized in Utter and Epifanio 2002). In general, the genetic consequences of cultured fish interbreeding with wild populations tend to be unpredictable and disadvantageous to the locally- adapted population (Hindar et al. 1991). The long term consequences of such effects are not yet known. The present data suggest that, for at least one species of Pacific salmon, selection for enhanced growth has arisen mainly due to additive gene action. The effects of interaction between such strains and wild fish will depend on the frequency and magnitude of escapes, their ability to interbreed with conspecifics, the size of the receiving population, and, most critically, the fitness of the domesticated genotypes in nature. The present data provides a framework for estimating the number of generations that may be required to completely dilute any observable growth or behavioural differences caused by interaction between fast-growing strains of coho salmon used in aquaculture and wild populations. The effect of a single small introgression of farm alleles into a wild population, based on the present data, would be anticipated to be diluted with repeated backcrosses in the absence of selection, and little phenotypic effect would be detectable after just two or more generations. In contrast, escapes of large numbers of farmed individuals into small populations, or repeated escapes of moderate numbers over several generations, would be anticipated to have an impact on the phenotype of the receiving population. The critical question, as yet unanswered, is whether natural selection can restore wild population genetic structures from such introgressed populations. 67 2 .5 A C K N O W L E D G M E N T S We gratefully acknowledge funding from an NSERC Industrial Post-Graduate Scholarship, in partnership with Target Marine Hatcheries, to WET, and from the Canadian Biotechnology Strategy to RHD. We thank Hanna Kim for injecting fry with the elastomer tags, and Mark D'Andrade for setting up the semi-natural stream tanks. Discussion of the line-cross analysis with M . Whitlock was much appreciated. Comments from three anonymous reviewers have greatly improved the clarity of the manuscript. 68 2.6 R E F E R E N C E S Abrahams, M . V. , and L. M . Di l l . 1989. A determination of the energetic equivalence of the risk of predation. Ecology. 70:999-1007. Abrahams, M . V. , and T. C. Pratt. 2000. Hormonal manipulations of growth rate and its influence on predator avoidance - foraging trade-offs. Can. J. Zool. 78:121-127. Abrahams, M . V. , and A. Sutterlin. 1999. The foraging and antipredator behaviour of growth-enhanced transgenic Atlantic salmon. Anim. Behav. 5:933-942. A l i , M . , A . Nicieza, and R. J. Wootton. 2003. Compensatory growth in fishes: a response to growth depression. Fish Fish. 4:147-190. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Q. Rev. Biol. 72: 149-177. Barbat-Leterrier, A. , R. Guyomard, and F. Krieg. 1989. Introgression between introduced domesticated strains and Mediterranean native populations of brown trout (Salmo trutta L.). Aquat. Liv. Res. 2:215-223. Berejikian, B.A. 1995. The effects of hatchery and wild ancestry and experience on the relative ability of steelhead trout fry (Oncorhynchus mykiss) to avoid a benthic predator. Can. J. Fish. Aquat. Sci. 52:2476-2482. Berrebi, P., C. Poteaux, M . Fissier, and G. Cattaneo-Berrebi. 2000. Stocking impact and allozyme diversity in brown trout from Mediterranean southern France. J. Fish. Biol. 56:949-960. Bryden, C. A. , J. W. Heath, and D. D. Heath. 2004. Performance and heterosis in farmed and wild Chinook salmon (Oncorhynchus tshawytsha) hybrid and purebred crosses. Aquaculture 235:249-261. Cagigas, M . E., E. Vazquez, G. Blanco, and J. A . Sanchez. 1999. Genetic effects of introduced hatchery stocks on indigenous brown trout (Salmo trutta L.) populations in Spain. Ecol. Freshw. Fish 8:141-150. 69 Cairney, M . , J. B. Taggart, and B. Hoyheim. 2000. Characterization of microsatellite and minisatellite loci in Atlantic salmon (Salmo salar L.) and cross-species amplification in other salmonids. Mol. Ecol. 9:2175-2178. Calow, P. 1982. Homeostasis and fitness. Am. Nat. 120:416-419. Campton, D. E., and J. M . Johnston. 1985. Electrophoretic evidence for a genetic admixture of native and non-native rainbow trout in the Yakima River, Washington. Trans. Am. Fish. Soc. 114:782-793. Carr, J. W., J. M . Anderson, F. G. Whoriskey, and T. Dilworth. 1999. The occurrence and spawning of cultured Atlantic salmon (Salmo salar) in a Canadian River. ICES J. Mar. Sci. 54:1064-1073. Cavalli, L. L. 1952. A n analysis of linkage in quantitative inheritance. In Quantitative inheritance. Edited by E.C.R. Reeve, and C H . Waddington. Her Majesty's Stationery Office, London, pp. 135-144. Crozier, W. W. 1993. Evidence of genetic interaction between escaped farmed salmon and wild Atlantic salmon (Salmo salar L.) in a northern Irish river. Aquaculture 113:19-29. Crozier, W. W. 1998. Incidence of escaped farmed salmon, Salmo salar L. , in commercial salmon catches and fresh water in Northern Ireland. Fisheries Manage. Ecol. 5:23-29. Devlin, R. H , C. A . Biagi, T. Y . Yesaki, D. E. Smailus, and J. C. Byatt. 2001. Growth of domesticated transgenic fish. Nature 409:781-782. Devlin, R.H., J. I. Johnsson, D. E. Smailus, C. A . Biagi, E. Jonsson, and B. Bjornsson. 1999. Increased ability to compete for food by growth hormone-transgenic coho salmon Oncorhynchus kisutch (Walbaum). Aquae. Resource 30:479-482. Di l l , L . M . , and A. H. G. Fraser. 1984. Risk of predation and the feeding behavior of juvenile coho salmon (Oncorhynchus kisutch). Behav. Ecol. Sociobiol. 16:65-71. Donaldson, E. M . , U . H . M . Fagerlund, D. A . Higgs, and J. R. McBride. 1979. Hormonal enhancement of growth. In Fish Physiology. Edited by W.S. Hoar, D.J. Randall, 70 and J.R. Brett. Academic Press, New York, N .Y. , USA; London, England, pp. 455-597. Donaldson, L. R. 1970. Selective breeding of salmonid fishes. In Marine Aquaculture. Edited by W.J. McNeil. Oregon State University Press, Corvallis. pp. 65-74. Dunham, R. A . 1999. Utilization of Transgenic Fish in Developing Countries: Potential Benefits and Risks. J. World Aquacult. Soc. 30:1-11. Elliott, J. M . , M . A, Hurley, and R. J. Fryer. 1995. A new, improved growth model for brown trout, Salmo trutta. Funct. Ecol. 9:290-298. Elliott, J. M . , and M . A . Hurely. 1997. A funtional model for maximum growth of Atlantic salmon parr, Salmo salar, from two populations in Northwest England. Funct. Ecol. 11:592-603. Fleming, I. A. , T. Agustsson, B. Finstad, J. I. Johnsson, and B. T. Bjornsson. 2002. Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 59:1323-1330. Fleming, I. A. , and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES J. Mar. Sci. 54:1051-1063. Fraser, J. M . 1981. Comparative survival and growth of planted wild, hybrid, and domestic strains of brook trout (Salvelinus fontinalis) in Ontario lakes. Can. J. Fish. Aquat. Sci. 38:1672-1684. Fritzner, N . G., M . M . Hansen, S. S. Madsen, and K. Kristiansen. 2001. Use of microsatellite markers for identification of indigenous brown trout in a geographical region heavily influenced by stocked domesticated trout. J. Fish Biol. 58:1197-1210. Gausen, D., and V. Moen. 1991. Large-scale escapes of farmed Atlantic salmon (Salmo salar) into Norwegian rivers threaten natural populations. Can. J. Fish. Aquat. Sci. 48:426-428. Gharrett, A . J., and W. W. Smoker. 1991. Two generations of hybrids between even- and odd-year pink salmon (Oncorhynchus gorbuscha): A test for outbreeding depression? Can. J. Fish. Aquat. Sci. 48:1744-1749. 71 Gharrett, A . J., W. W. Smoker, R. R. Reisenbichler, and S. G. Taylor. 1999. Outbreeding depression in hybrids between odd- and even-broodyear pink salmon. Aquaculture 173:117-130. Gilk, S. E., I. A . Wang, C. L . Hoover, W. W. Smoker, and S. G. Taylor. 2004. Outbreeding depression in hybrids between spatially separated pink salmon, Oncorhynchus gorbuscha, populations: marine survival, homing ability, and variability in family size. Environ. Biol. Fish. 69:287-297. Gjedrem, T. 1976. Possibilities for genetic improvements in salmonids. J. Fish. Res. Bd. Can. 33:1094-1099. Gjedrem, T. 1983. Genetic variation in quantitative traits and selective breeding in fish and shellfish. Aquaculture 33:51-72. Gjerde, B. 1986. Growth and reproduction in fish and shellfish. Aquaculture 57:37-56. Gudjonsson, S. 1991. Occurrence of reared salmon in natural salmon rivers in Iceland. Aquaculture 98:133-142. Hansen, L. P., K. B. Doving, and B. Jonsson. 1987. Migration of farmed adult Atlantic salmon with and without olfactory sense, released on the Norwegian coast. J. Fish Biol. 30:713-721. Hansen, L. P., J. A . Jacobsen, and R. A . Lund. 1999. The incidence of escaped farmed Atlantic salmon, Salmo salar L., in the Faroese fishery and estimates of catches of wild salmon. ICES J. Mar. Sci. 56:200-206 Hansen, M . M . , E. E. Nielsen, D. Bekkevold, and K. L. Mensberg. 2001. Admixture analysis and stocking impact assessment in brown trout (Salmo trutta), estimated with incomplete baseline data. Can. J. Fish. Aquat. Sci. 58:1853-1860. Hayman, B. I. 1960. The separation of epistatic from additive and dominance variation in generation means. Genetica 31:133-146. Hershberger, W. K. , J. M . Myers, R. N . Iwamoto, W. C. Macauley, and A. M . Saxton. 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85:187-197. 72 Higgs, D. A. , E. M . Donaldson, H. M . Dye, and J. R. McBride. 1975. A preliminary investigation of the effect of bovine growth hormone on growth and muscle composition of coho salmon Oncorhynchus kisutch. Gen. Com. Endocr. 27:240-253. Hindar, K. , N . Ryman, and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Can. J. Fish. Aquat. Sci. 48:945-957. Hojesjo, J., J. I. Johnsson, and M . Axelsson. 1999. Behavioural and heart rate responses to food limitation and predation risk: an experimental study on rainbow trout. J. Fish Biol. 55:1009-1019. Huntingford, F. A . 2004. Implications of domestication and rearing conditions for the behaviour of cultivated fishes. J. Fish Biol. 65(Suppl. A): 122-142. Iwamoto, R. N . , J. M . Myers, and W. K. Hershberger. 1990. Heritability and genetic correlations for flesh coloration in pen-reared coho salmon. Aquaculture 86:181-190. Iwamoto, R. N . , A . M . Saxton, and W. K. Hershberger. 1982. Genetic estimates for length and weight of coho salmon during fresh water fearing. J. Hered. 73:187-191. Johnsson, J. I. 1993. Big and brave: size selection affects foraging risk under risk of predation in juvenile rainbow trout, Oncorhynchus mykiss. Anim. Behav. 45:1219-1225. Johnsson, J. I., and M . V . Abrahams. 1991. Interbreeding with domestic strain increases foraging under threat of predation in juvenile steelhead trout Oncorhynchus mykiss: an experimental study. Can. J. Fish. Aquat. Sci. 48:243-247. Johnson, J. I., J. Hojesjo, and I. A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can. J. Fish. Aquat. Sci. 58:788-794. Johnsson, J. I., E. Petersson, E. Jonsson, B. T. Bjornsson, and T. Jarvi. 1996. Domestication and growth hormone alter antipredator behaviour and growth 73 patterns in juvenile brown trout, Salmo trutta. Can. J. Fish. Aquat. Sci. 53:1546-1554. Jonsson, E., J. I. Johnsson, and B. T. Bjornsson. 1996. Growth hormone increases predation exposure of rainbow trout. P. Roy. Soc. Lond. B. 263:647-651. Lynch, M . , and J. B. Walsh. 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, M A . McConnell, S. K., D. E. Ruzzante, P. O'Reilly, and L. Hamilton. 1997. Microsatellite loci reveal highly significant genetic differentiation among Atlantic salmon (Salmo salar L.) stocks from the east coast of Canada. Mol . Ecol. 6:1075-1089. McGinnity, P., P. Prodohl, A. Ferguson, R. Hynes, N . 6 Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. P. Roy. Soc. Lond. B 270:2443-2450. McGinnity, P., C. Stone, J. B. Taggart, D. Cooke, D. Cotter, R. Hynes, C. MaCamley, T. Cross, and A. Ferguson. 1997. Genetic impact of escaped farmed Atlantic salmon on native populations: use of D N A profiling to assess freshwater performance of wild, farmed and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54:998-1008. McKinnell, S., A . J. Thomson, E. A . Black, B. L. Wing, C. M . Guthrie III, J. F. Koerner, and J. H. Helle. 1997. Atlantic salmon in the North Pacific. Aquae. Res. 28:145-157. Metcalfe, N . B., and P. Monaghan. 2001. Compensation for a bad start: grow now, pay later? TREE 16:254-260. Myers, J. M . , W. K. Hershberger, A. M . Saxton, and R. N . Iwamoto. 2001. Estimates of genetic and phenotypic parameters for length and weight of marine net-pen reared coho salmon (Oncorhynchus kisutch Walbaum). Aquae. Res. 32:277-285. Olsen, J. B., P. Bentzen, and J. E. Seeb. 1998. Characterization of seven microsatellite loci derived from pink salmon. Mol. Ecol. 7:1083-1090. 74 Olsen, J. B., S. L . Wilson, E. J. Kretschmer, K. C. Jones, and J. E. Seeb. 2000 Characterization of 14 tetranucleotide microsatellite loci derived from Atlantic salmon. Mol. Eco. 9:2155-2234. Orr, H. A. , and J. A . Coyne. 1992. The genetics of adaptation: A reassessment. Am. Nat. 140:725-742. Price, E. O. 1984. Behavioral aspects of animal domestication. Q. Rev. Biol. 59:1-32. Pyke, G. H., H . R. Pulliam, and E. L. Charnov. 1977. Optimal foraging: a selective review of theory and tests. Q. Rev. Biol. 52:137-154. Quinn, T. P., L. A . Vollestad, J. Peterson, and V. Gallucci. 2004. Influences of freshwater and marine growth on the egg size-egg number tradeoff in coho and chinook salmon. Trans. Am. Fish. Soc. 133:55-65. Reisenbichler, R. R., and S. P. Rubin. 1999. Genetic changes from artificial propagation of Pacific salmon affect the productivity and viability of supplemented populations. ICES J. Mar. Sci. 56:459-466. Reisenbichler, R. R., and J. D. Mclntyre. 1977. Genetic differences in growth and survival of juvenile hatchery and wild steelhead trout, Salmo gairdneri. J. Fish. Res. Bd. Can. 34:123-128: Roff, D. A . 1984. The evolution of life history parameters in teleosts. Can. J. Fish. Aquat. Sci. 41:989-1000. Smoker, W. W., I. A . Wang, A. J. Gharrett, and J. J. Hard. 2004. Embryo survival and smolt to adult survival in second-generation outbred coho salmon. J. Fish Biol. 65(Suppl. A):254-262. Sundstrom, F. L. , R. H. Devlin, J. I. Johnsson, and C. A . Biagi. 2003. Vertical position reflects increased feeding motivation in growth hormone transgenic coho salmon (Oncorhynchus kisutch). Ethology 109: 701-712. Sundstrom, L. F., M . Lohmus, J. I. Johnsson, and R. H. Devlin. 2004a. Growth hormone transgenic salmon pay for growth potential with increased predation mortality. P. Roy. Soc. Lond. B. 271:350-352. 75 Sundstrom, L. F., E. Petersson, J. Hojesjo, J. I. Johnsson, and T. Jarvi. 2004b. Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behav. Ecol. 15:192-198. Tymchuk, W.E. and R. H . Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus kisutch). Aquaculture. 245:295-300. Utter, F., and J. Epifanio. 2002. Marine aquaculture: Genetic potentialities and pitfalls. Rev. Fish Biol. Fisher. 12:59-77. Utter, F., K. Hindar, and N . Ryman. 1993. Genetic effects of Aquaculture on natural salmonid populations. Pages 144-165 in K. Heen, R. L . Monahan, and F. Utter, editors. Salmon Aquaculture. Fishing News Books, Oxford, U K . Williamson, K. S., J. F. Cordes, and B. P. May. 2002. Characterization of microsatellite loci in chinook salmon (Oncorhynchus tshawytscha) and cross-species amplification in other salmonids. Mol . Ecol. Notes 2:17-19. Webster, D. A. , and W. A. Flick. 1981. Performance of indigenous, exotic, and hybrid strains of brook trout (Salvelinus fontinalis) in waters of the Adirondack Mountains, New York. Can. J. Fish. Aquat. Sci. 38:1701-1707. Withler, R. E., and T. D. Beacham. 1994. Genetic variation in body weight and flesh colour of coho salmon (Oncorhynchus kisutch) in British Columbia. Aquaculture 119:135-148. 76 C H A P T E R III 3: G R O W T H D I F F E R E N C E S A M O N G FIRST A N D S E C O N D G E N E R A T I O N H Y B R I D S O F D O M E S T I C A T E D A N D W I L D R A I N B O W T R O U T (ONCORHYNCHUS MYKISS) 3 A version of this chapter has been published. Tymchuk, W . E. and R. H . Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245: 295-300. 77 3.1 I N T R O D U C T I O N Use of selective breeding programs to improve a trait of interest, such as growth, requires phenotypic variation among individuals for the trait. This variation must be at least partly determined by additive genetic variation for selection to cause genetic changes leading to the desired phenotypic effect. Salmonids (salmon and trout and their relatives) demonstrate high phenotypic variability in growth rate, and a high proportion of this variability (h2 > 0.50) can be explained by additive genetic variation (Iwamoto et al. 1982; Myers et al. 2001; Martyniuk et al. 2003). Therefore, selection practices may be applied to salmonid populations leading to strains that have enhanced growth rates relative to the wild stocks from which they were derived (Hershberger et al. 1990). Within domesticated strains of rainbow trout, Oncorhynchus mykiss, size at age, and therefore growth rate, has a strong genetic basis (Gjedrem 1976; Gall and Huang 1988; Su et al. 1996), which has allowed effective selection for enhanced growth in strains of this species for use in aquaculture (Ayles and Baker 1983). Selective breeding for enhanced growth typically selects the largest mature individuals to breed for the future generation, utilizing either family or individual selection. Selection is therefore being applied to growth, with little knowledge of the underlying genetic and physiological mechanisms that are being selected for. Other intentional (and unintentional) differences in behaviour, morphology and physiology arise throughout the selection, or domestication, process (Einum and Fleming 2001), such as earlier maturity when selecting for enhanced growth in rainbow trout (Crandell and Gall 1993). Faster growth is often associated with fitness tradeoffs (such as an increased risk of predation; Dil l and Fraser 1984; Biro et al. 2004; Sundstrom et al. 2004) which have resulted in wild-strain fish expressing less than maximal growth rates in nature (Arendt 1997). This study tracks the growth over time of pure and hybrid strains of domestic (fast-growing) and wild (slow-growing) rainbow trout. Assessing the phenotypic expression of introgression of the domestic genotype into the wild genetic background is 78 a first step in addressing two important goals: 1) determining the genetic and physiological changes that have occurred during domestication to cause enhanced growth; and 2) providing a genetic basis for risk-assessment of the interaction between domestic and wild fish populations which may occur in nature. 3.2 M A T E R I A L S A N D M E T H O D S Strains of O. mykiss, ranging from fast-growing pure domestic (D) to slow-growing pure wild (W), were reared at the West Vancouver Laboratory (WVL), BC. Gametes from domestic parents were obtained from Spring Valley Trout Farm, Langley, B C , and gametes from wild parents were collected from nature from Pennask Lake (south-central interior of BC). Early maturing individuals from the domestic strain were crossed with late maturing individuals from the wild strain on 5 July 1999. Gametes from Fi (domestic x wild hybrids) were obtained from these hybrid strains cultured at W V L , and the following crosses were fertilized at W V L on 2 6 June 2002 : pure wild (W, four families) and Fi x wild backcross (B w , four families). On 28 October 2 0 0 2 , additional crosses were fertilized: wild x domestic (Fi, four families); F] x domestic backcross (Bd, two families); and pure domestic (D, four families). It was necessary to perform crosses at two different times since the wild and domestic females largely mature in the spring and fall, respectively, and in 2 0 0 2 (unlike in 1999) no individuals of both strains matured at the same time. Consequently, cryopreserved W and Fi sperm, collected 2 6 June 2 0 0 2 from wild and hybrid fish, was used on 28 October 2 0 0 2 in crosses with domesticated females (following the method outlined in Wheeler and Thorgaard 1991). The June 2 0 0 2 and the October 2 0 0 2 set of crosses will be referred to as the spring and fall sets. At first feeding, 3 0 fry were sampled from each family and placed into 2 0 0 L tanks. Families were reared individually under culture conditions (well water at a flow no less than 1 L min"1 kg"1, constant temperature at 11°C; supplemental aeration) and were fed commercial salmonid feed (Skretting, Vancouver, Canada; composition of feed varied over time to meet the requirements of different life stages) to satiation several 79 times per day. The spring set of families were individually measured for mass and length on 18 October 2002, 29 April 2003, 29 August 2003, 30 September 2003, and 24 March 2004 corresponding to 114, 307, 429, 461 and 637 days post-fertilization (dpf). The fall set of families was measured on 16 May 2002, 29 August 2003, 30 September 2003, and 24 March 2004 corresponding to 200, 305, 337 and 513 dpf. On 29 August 2003, a PIT (Passive Integrated Transponder) tag was inserted into each individual. This allowed specific growth rates (SGR = (lnW 2 - lnWi)/growth interval (days) x 100) for individual fish also to be tracked from 29 August 2003 to 24 March 2004. Condition factors were calculated as CF = weight (grams)/length (cm)3 x 100. 3.3 R E S U L T S Mass (Fig. 3.1a) and length (Fig. 3.1b) measurements over time indicate that the rank of the genotypes (from fastest to slowest growth) was as follows: D > Bd > Fi > B w > W. At two time points where sampling yielded age-matched fish (305 and 307 dpf, for the spring and fall set of crosses, respectively) there was a significant and strong correlation between mass and genotype (Fig. 3.1c; r 2 = 0.902, F i ^ = 147.731, p < 0.001) and between length and genotype (Fig. 3.Id; r 2 = 0.919, F u 6 = 181.989, p < 0.001). There were significant differences in mass (F 4 ji3 = 73.740, p < 0.001) and length (F 4 >i3 = 105.152, p < 0.001) among genotypes, with significant pair-wise comparisons as indicated in Figure 3.1c and 3.Id. There were also significant correlations between genotype and specific growth rates in mass (Fig. 3.1e; r 2 = 0.854, F i ; i6 = 93.609, p < 0.001) and length (Fig. 3.If; r 2 - 0.858, F i ; 1 6 = 96.557, p < 0.001). Significant differences were found among the different genotypes for S G R m a s s (F 4 ji3 = 22.198, p < 0.001) and SGRi e n gth (F 4 , i 3 = 22.198, p < 0.001). Rank of condition factor (Fig. 3.2) followed the same trends as body size and specific growth rate with D > Bd > Fi > B w > W. Differences in condition factor were significant when comparing both age-matched (at 305 and 307 dpf, Fig. 3.2; F 4 ] n = 80 400 100 200 300 400 500 600 700 (a) Days Post-Fertilization 30 o-l , , , , , 1 100 200 3O0 400 500 600 700 (b) Days Post-Fertilization F I G U R E 3.1 Growth performance of pure domestic (D), pure wild (W), F l hybrids (Fl), and F2 backcrosses to wild (Bw) and domestic (Bd) parents. These are labeled as %D in figure legend. Growth performance was measured as mass (a) and length (b) over time. Significant differences in mass (c) and length (d) for age-matched families, and SGRmass (e) and SGR | e n gth (f) are labeled on the graphs with different lower case letters. 81 1.6 • 1.4 • 1-2 H # age-matched O season-matched 5 i o 25 50 75 100 % Domestic Genotype F I G U R E 3.2 Comparison of condition factors among the five genotypes as described in Figure 3.1. Condition factors were compared for both age-matched (at 305 and 307 days post-fertilization) and season-matched (24 March 2004) groups. Size data for age-matched fish were collected at different times of the year. Data for season-matched fish were collected on the same date although the fish are not age-matched. Significant differences for each set are as indicated on the graph. 82 5- io Mass (grams) a ) B w 20 40 Mass (grams) F I G U R E 3.3 Growth frequency distribution of Bw (a, F l x wild) and Bd (b, F l x domestic) at 305 and 307 dpf, respectively. Note difference in weight scales between Figures a and b. 83 44.188, p < 0.001) and season-matched (calculated with 24 March 2004 size data, Figure 3.2; F 4 , i 3 = 16.315, p < 0.001) families. The size frequency distributions of the B W (Fig. 3.3a) and BH (Fig. 3.3b) crosses show a lack of a strong bimodal size distribution, with fewer individuals expressing faster growth than would be expected i f growth was controlled by one major locus. Comparison of the two distributions illustrates again the magnitude of the growth difference between the domestic and wild fish. 3.4 DISCUSSION It is clear that there have been genetic changes during the domestication of rainbow trout that have led to significant improvements in growth performance under culture conditions. In addition to enhanced growth, the domestic fish have higher condition factors than wild fish which indicate that the domestic fish have a proportionally larger increase in growth in mass relative to growth in length. This may be due to stronger selective pressure for increased mass, although that cannot be determined from this study. Presumably, a slender body form provides a fitness advantage in nature, whereas under culture conditions, growth enhancement can be achieved through selection for both overall growth and by changes in weight-length relationships. Growth enhancement in itself does not necessarily result in changes in body shape since growth-enhanced transgenic trout made in a wild strain retain their slender body form (Devlin et al. 2001). Previous selection of rainbow trout during domestication may not have been isolated to growth enhancement, but may also include performance under culture conditions, suggesting that the relative growth rates among strains may differ according to rearing environment. From the current results, it appears that additive genetic effects can explain a large amount of the variation in growth differences between the wild and domesticated strains examined. The Fi hybrids have phenotypic values very close to the mean of the parental values, as would be expected if the domesticated fish differ from the slow-growing 84 strains primarily in genes with additive effects. There is no strong evidence for heterosis, as has been found for previous hybrid crosses between wild Pennask and other domesticated trout populations (Ayles and Baker 1983). Analysis of the size-frequency distribution of the F2 backcrosses supports the additive model of more than one gene involved in growth enhancement of the domesticated strain relative to wild fish. If faster growth was controlled by one major locus, a clear bimodal size-frequency distribution would be expected for these backcrossed lines. Although both distributions are not normal (Kolmogorow-Smirnov; B d = 6.100, p = 0.16; B w = 0.131, p < 0.001), the distributions do suggest that at least more than one locus is involved in the observed phenotypic differences in growth between the strains. Rainbow trout have been extensively used as a model fish species, and as a consequence much is known about the genetic make-up of this species (Sakamoto et al. 2000; Thorgaard et al. 2002). By combining current genetic knowledge (Sakamoto et al. 2000) with the expression of phenotype of hybrid strains such as are documented here, candidate genes may be identified which display strong correlations with growth rate. Such techniques applied to rainbow trout have recently discovered evidence for quantitative trait loci within domesticated strains for growth (Martyniuk et al. 2003) and other traits, including upper thermal tolerance (Jackson et al. 1998; Perry et al. 2001), maturation (Sakamoto et al. 1999; Martyniuk et al. 2003), embryonic development (Robison et al. 2001), and disease resistance (Palti et al. 1999). It will be of interest to determine whether such loci also played a major role in the transformation of wild fish to the domesticated growth phenotype. Development of multiple generation hybrid strains between populations of wild slow-growing and domesticated fast-growing rainbow trout provides an opportunity to explore the genetic architecture of growth and the changes which have occurred during the domestication process. It will also provide valuable insight into the consequences of interaction (either ecological or genetic) between domestic and wild trout populations, as has been undertaken with studies of cultured and wild Atlantic salmon (Salmo salar). Data from Fi and F2 crosses of domestic and wild Atlantic salmon indicated additive 85 genetic variation for survival in the wild (McGinnity et al. 1997, 2003), which suggests that implications of introgression of the domestic genotype into wild fish populations can be diluted over time. 3.5 A C K N O W L E D G M E N T S This research was funded by an NSERC Industrial Post-Graduate Scholarship, in partnership with Target Marine Hatcheries, to WET, and by the Canadian Biotechnology Strategy to RHD. We thank Tim Yesaki and staff at the Freshwater Fisheries Society of British Columbia Summerland Trout Hatchery for facilitating collection of wild trout. 86 3.6 R E F E R E N C E S Arendt, J.D., 1997. Adaptive intrinsic growth rates: an integration across taxa. Q. Rev. Biol. 72:149-177. Ayles, G.B. and Baker, R.F., 1983. Genetic differences in growth and survival between strains and hybrids of rainbow trout (Salmo gairdneri) stocked in aquaculture lakes in the Canadian prairies. Aquaculture 33:269-280. Biro, P.A., Abrahams, M.V. , Post, J.R., Parkinson, E.A., 2004. Predators select against high growth rates and risk-taking behaviour in domestic trout populations. Proc. R. Soc. Lond. B 271:2233-2237. Crandell, P.A., Gall, G.A.E., 1993. The genetics of body weight and its effect on early maturity based on individually tagged rainbow trout (Oncorhynchus mykiss). Aquaculture 117:77-93. Devlin, R.H., Biagi, C.A., Yesaki, T.Y., Smailus, D.E., Byatt, J .C , 2001. Growth of domesticated transgenic fish. Nature 409:781-782. Di l l , L. M . , and A. H. G. Fraser. 1984. Risk of predation and the feeding behavior of juvenile coho salmon (Oncorhynchus kisutch). Behav. Ecol. Sociobiol. 16:65-71. Einum, S., and Fleming, L A . , 2001. Implications of Stocking: Ecological Interactions Between Wild and Released Salmonids. Nord. J. Fresh. Res. 75:56-70. Gall, G.A.E., and Huang, N . , 1988. Heritability and selection schemes for rainbow trout body weight. Aquaculture 73:43-56. Gjedrem, T., 1976. Possibilities for genetic improvements in salmonids. J. Fish. Res. Board Can. 33:1094-1099. Hershberger, W.K., Myers, J .M., Iwamoto, R.N., Macauley, W.C., and Saxton, A . M . . 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85:187-197. Iwamoto, R.N., Saxton, A . M . , and Hershberger, W.K., 1982. Genetic estimates for length and weight of coho salmon during freshwater rearing. J. Hered. 73:187-191. 87 Jackson, T.R., Ferguson, M . M . , Danzmann, R.G., Fishback, A . G . , Ihssen, P.E., O'Connel, P.E., 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. Martyniuk, C.J., Perry, G.M.L. , Mogahadam, H.K., Ferguson, M . M . , and Danzmann, R.G., 2003. The genetic architecture of correlations among growth-related traits and male age at maturation in rainbow trout. J. Fish Biol. 63:746-764. McGinnity, P., Prodohl, P., Ferguson, A. , Hynes, R., 6 Maoileidigh, N . , Baker, N . , Cotter, D., O'Hea, B., Cooke, D., Rogan, G., Taggart, J., Cross, T., 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proc. R. Soc. B. 270:2443-2450. McGinnity, P., Stone, C , Taggart, J.B., Cooke, D., Cotter, D., Hynes, R., McCamley, C , Cross, T., Ferguson, A. , 1997. Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: use of D N A profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54:998-1008. Myers, J.M., Hershberger, W.K., Saxton, A . M . , and Iwamoto, R.N., 2001. Estimates of genetic and phenotypic parameters for length and weight of marine net-pen reared coho salmon (Oncorhynchus kisutch Walbaum). Aquae. Res. 32:277-285. Palti, Y . , Parsons, J.E., and Thorgaard, G.H., 1999. Identification of candidate D N A markers associated with IHN virus resistance in backcross of rainbow (Oncorhynchus mykiss) and cutthroat (O. clarki). Aquaculture 173:81-94. 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. Robison, B.D., Wheeler, P. A. , Sundin, K., Sikka, P., and Thorgaard, G.H., 2001. Composite interval mapping reveals a major locus influencing embryonic development rate in rainbow trout (Oncorhynchus mykiss). J. Hered. 92:16-22. 88 Sakamoto, T., Danzmann, R.G., Gharbi, K. , Howard, P., Ozaki, A . , Khoo, S.K., Woram, R.A., Okamoto, N . , Ferguson, M . M . , Holm, L.-E. , Guyomard, R., and Hoyheim, B., 2000. A microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) characterized by large sex-specific differences in recombination rates. Genetics 155:1331-1345. Sakamoto, T., Danzmann, R.G., Okamoto, N . , Ferguson, M . M . , and Ihssen, P.E., 1999. Linkage analysis of quantitative trail loci associated with spawning time in rainbow trout (Oncorhynchus mykiss). Aquaculture 173:33-43. Su, G.-S., Liljedahl, L.-E. , and Gall, G.A.E., 1996. Genetic and environmental variation of body weight in rainbow trout (Oncorhynchus mykiss). Aquaculture 144:71-80. Sundstrom, L. F., E. Petersson, J. Hojesjo, J. I. Johnsson, and T. Jarvi. 2004. Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behav. Ecol. 15:192-198. Thorgaard, G.H., Bailey, G.S., Williams, D., Buhler, D.R., Kaattari, S.L., Ristow, S.S., Hansen, J.D., Winton, J.R., Bartholomew, J.L., Nagler, J.J., Walsh, P.J., Vijayan, M . M . , Devlin, R.H., Hardy, R.W., Overturf, K.E . , Young, W.P., Robison, B.D., Rexroad, C., and Palti, Y . , 2002. Status and opportunities for genomics research with rainbow trout. Comp. Biochem. Phys. B 133:609-646. Wheeler, P.A., and Thorgaard, G.H. 1991. Cryopreservation of rainbow trout semen in large straws. Aquaculture 93:95-100. 89 CHAPTER I V 4 : O U T B R E E D I N G D E P R E S S I O N A F T E R T H R E E G E N E R A T I O N S OF I N T R O G R E S S I O N O F F A S T - G R O W I N G C O H O S A L M O N INTO A N A T I V E S T R A I N 4 A version of this manuscript is being prepared for submission for publication. Tymchuk, W. E. and R. H. Devlin. Outbreeding depression after three-generations of introgression of fast-growing coho salmon into a native population. 90 4.1 I N T R O D U C T I O N Body size is often positively correlated with fitness components (Roff 1992; Stearns 1992) such as increased fecundity (Shine 1988; Honek 1993) or egg size (van den Berghe and Gross 1989) and increased survival during early life stages (Sogard 1997). Since growth rates have not evolved to a maximum in nature (restricted only by physiological or phylogenetic constraints) but instead show local adaptation (Arendt 1997), selection for faster growth is being counterbalanced by opposing selective pressures (Dill and Fraser 1984; Travis 1989; Schluter et al. 1991). Faster growing individuals have been found to suffer increased mortality due to predation (Gotthard 2000; Munch and Conover 2003; Sundstrom et al. 2004, 2005) and individuals frequently make tradeoffs in foraging and growth under risk of predation (Lima and Di l l 1990). Strains of salmon reared within aquaculture environments have been unintentionally and intentionally selected for faster growth rates (e.g. Gjedrem 1976, 2000; Hershberger et al. 1990). For domesticated coho salmon, the basis of this enhancement of growth has been shown to be genetically-determined mainly through additive genetic variance (McClelland et al. 2005; Tymchuk et al. 2006), and the rates of growth are higher than those expressed by native coho populations under both culture and semi-natural conditions (Tymchuk et al. 2006). Further, their response to a simulated predation attack indicated that this strain of coho may be more susceptible to risk of predation during foraging. A decreased response to risk of predation is a common observation when comparing domestic farm fish with wild conspecifics (Johnsson and Abrahams 1991; Johnsson et al. 1996; Einum & Fleming 1997; Fleming & Einum 1997; Johnsson et al. 2001; Fleming et al. 2002). Selection for faster growth and adaptation to a culture environment (domestication) is known to incur fitness tradeoffs such as altered foraging behaviour and reduced anti-predator behaviour (Dill and Fraser 1984; Johnsson and Abrahams 1991; Biro et al. 2004; Sundstrom et al. 2004a, 2004b; Tymchuk et al. 2006), and may therefore have a significant effect on survival of fast-growing domestic individuals in a natural environment. 91 Cultured fish may enter the natural environment by purposeful release for conservation and enhancement purposes, or they may escape from aquaculture facilities (Hansen et al. 1987; Gudjonsson 1991; Lura and Saegrov 1991; Jacobsen et al. 1992; Crozier 1993; Carr et al. 1997; McKinnell et al. 1997; Clifford et al. 1998; Volpe et al. 2000; Morton and Volpe 2002). Concern has emerged regarding the concurrent genetic and ecological consequences of this interaction arising from differences in phenotype and genotype between cultured and wild fish populations (Crossman 1991; Hindar et al. 1991; Fleming et al. 1996; Hansen et al. 1997). Early reviews of the literature on the outcome of interaction between introduced and native fish populations suggested that introductions have usually been harmful to the native fish populations (Allendorf 1991; Hindar et al. 1991), but see (Peterson 1999). There is a growing body of literature indicating that phenotypic differences between farm and wild fish (due to indirect and direct selection) are largely (although not completely) a result of additive genetic differences. Most studies have examined phenotypes of pure domestic and wild strains and their Fi hybrids (Ayles and Baker 1983; Einum and Fleming 1997; McGinnity et al. 1997, 2003; Bryden et al. 2004; McClelland et al. 2005;Tymchuk and Devlin 2005; Tymchuk et al. 2006), with the latter showing an intermediate phenotype between the two parental types. However, only recently has examination of the mutigenerational effects of longer term introgression been examined (McGinnity et al. 2003; McClelland et al. 2005; Tymchuk & Devlin 2005; Tymchuk et al. 2006) which is required to understand the genetic mechanisms involved in controlling phenotype. These studies have indicated that following the introduction of domesticated genes into a wild strain, the phenotypic effects of introgression will be diluted out with repeated backcrossing into wild populations, or i f sustained introduction of domestication alleles is occurring, a phenotypic shift in the wild population may be anticipated. Thus, although additive gene action is shown to have a substantial effect on the expression of phenotype in hybrids, it is still not clear how these phenotypes are influenced by environment, and/or non-additive genetic effects. The fitness of hybrids may be altered by endogenous selection (such as physiological 92 abnormalities that are expressed regardless of environment) or exogenous selection (environment-specific fitness difference) or a combination of both, so that the magnitude of heterosis or outbreeding depression may be dependent on the environment (Burke and Arnold 2001). The fitness consequences of introgression through interpopulation hybridization are difficult to predict and assess. Interbreeding among distinct populations that have low genetic variability may re-establish lost alleles and allelic combinations, which could lead to hybrid vigour, or heterosis, with the Fi generation having increased fitness relative to the parental lines (Lynch and Walsh 1998). Heterosis is most likely to occur if the parental lines are inbred and not highly genetically divergent. However, i f the parent populations come from different habitats to which they are well adapted, the resulting hybrid progeny may not be well adapted for either habitat (Allendorf and Waples 1996). An allele that is advantageous in one environment may be disadvantageous to overall fitness in another environment. If such alleles act in a co-dominant or additive fashion, this may result in outbreeding depression due to disruption of local adaptation. Outbreeding depression could also cause a negative impact on fitness by disrupting coadapted allele complexes, which would not become apparent until the F 2 generation, or beyond if the coadapted allele complexes consist of tightly linked assemblages of genes that require multiple generations of recombination to break apart. The purpose of the present study was to examine the fitness effects of three generations of introgression of coho salmon populations adapted to two different environments (culture vs wild). There have been some studies testing for outbreeding depression in fish populations after two generations of introgression (Gharrett and Smoker 1991; Gharrett et al. 1999; Smoker et al. 2004; McClelland et al. 2005; Tymchuk & Devlin 2005; Tymchuk et al. 2006), but to our knowledge this is the first study that has followed the impact of introgression for three generations (which will help determine i f there are delayed effects of outbreeding depression). By making use of genotyping at multiple microsatellite loci to determine parentage, we were able to release first-feeding fry as mixed genotype groups into semi-natural mesocosms with minimal handling, 93 thereby replicating more closely a natural environment and improving our ability to accurately predict the consequences of interaction between farm and wild Pacific salmon. 4.2 M A T E R I A L S A N D M E T H O D S Crosses were made using pure and hybrid farm and native coho salmon (Oncorhynchus kisutch) adults from strains that have been previously shown to have altered rates of growth (Tymchuk et al. 2006). Four females were collected from Chehalis River Hatchery, British Columbia (BC) on 6 January 2005. The eggs from each female were split into four aliquots and fertilized on the same day with milt from a pure domestic (D), pure wild (W), domestic x wild hybrid (Fi), and F i x wild backcross (B w ) male which had been reared at the Fisheries and Oceans Laboratory in West Vancouver, B C , Canada (DFO/UBC Centre for Aquaculture and Environmental Research). Four males of each genotype were used in the fertilizations, with the exception of the F i male which was used to fertilize all four females (since only one male of this genotype was producing milt). Thus, in total 16 crosses were generated. Eyed eggs of pure domestic strain were also obtained from Target Marine Farms and reared at W V L with the rest of the crosses, so that five different genotypes were used in the following experiment (D, F i , Bwi, Bw2 and W) which contained a range of domesticated alleles of 100, 50, 25, 12.5 and 0%, repectively. At the eyed stage, all non-fertilized eggs were counted and removed from the egg boxes. At hatch, the total number of viable alevin and the number of eyed eggs that did not hatch was also recorded so that viability of the different genotypes could be tested. On 17 April 2005 a total of 16 fry of each of the five genotypes (four fry from each of four families except for the Fi and B W 2 crosses in which the 16 fry derived from only two families due to low viability in two families) were pooled into 10 mesocosms (5 x 1 x 0.4 m, 80 fry per tank) enriched with gravel bottom. A slow water current (10 L/min) was maintained within the mesocosm by supplying creek water at one end, and allowing it to flow out the other end. Two predators (coho salmon parr, mean mass 15.5 ±3 .1 g, mean length 11.4 ± 0.8 cm) were added to each tank on the same day. Fry were fed with live 94 Artemia larvae two times per day, initiated on the day they were placed into the semi-natural habitats. To enable comparisons of growth and survival between the culture and seminatural environments, fish were also reared under culture conditions. For each family, 50 fry were placed into 200 L tanks with a steady supply of fresh creek water and were fed to satiation several times per day with commercial fish food (Skretting). On 14 May 2005 all surviving fry were removed from the tanks and blood from each individual was collected and placed into 100 ul of 0.01% NaOH and frozen for microsatellite analysis to determine the genotype of the survivors. Prior to microsatellite analysis, the blood was thawed, heated for 5 min at 95 C, and briefly centrifuged. Genotypes (D, F i , B w i , Bw2 and W) of the survivors were determined by identification of their parents (of known genotype) based on five polymorphic microsatellite D N A loci [OMMJ231, OMM1270 (Rexroad III and Palti 2003), Ogo2 (Olsen et al. 1998), and Ssa407 (Cairney et al. 2000), with the forward primer in each pair labeled (see Table 4.1)]. Microsatellite amplification via polymerase chain reaction (PCR) was performed in 10 ul reactions containing 1 x reaction buffer [20 m M Tris-HCl (pH 8.4), 50 mM KC1], 1.5 mM M g C l 2 , 0.2 mM each dNTP, 0.55 uM each primer, 0.25 units Taq polymerase (Invitrogen), and ~ 10 ng of template D N A (0.5 uL of NaOH solution). Samples were amplified using an Applied Biosystems Inc. (ABI) GeneAmp® PCR System 2700 thermocycler using touchdown PCR: one cycle of 95°C for 3 min; 10 cycles of 95°C for 30 s, 63°C for 1 min (- 0.5°C per cycle), and 72°C for 1 min; 20 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 30 s; one cycle of 72°C for 20 min. Microsatellites were size fractionated using an ABI 3130x1 automated D N A sequencer. Electrophoretic (allele size) data were analyzed using GeneMapper® software version 3.7 (ABI). The computer program P R O B M A X (Danzmann 1997) was used to assign parentage of each individual via exclusion technique (since the genotype of every parent was known). Briefly, the program systematically excludes parent-pairs of known genotype which contain loci incompatible with the genotype of the offspring, until one remaining parent-pair remains. The offspring are then classified as either D, F i , Bwi, Bw2, or W according to the genotype of the parents. To minimize the sensitivity of this T A B L E 4.1 Microsatellite primers used to determine the parentage of the surviving fry reared in the seminatural habitat. Loci Unit ~T T Primer F Primer R GenBank Acc. # QMM1231 OMM1270 Ogo2 Ssa407 (TCTA) 1 2 (AC) 2 7 (GA) 2 4 ( G A C A ) 3 7 AF470011 T C C A C C T G C T C T G A C C T C T A C T C A G C A G C C A G A G A A C A G T A A G C A T G T AF470035 G C C A T T T G G G A A T C A G A G A G T T C T A C A C G G A A A C C C T G A C A G G A A T A C T G A G AF009794 AJ402724 A C A T C G C A C A C C A T A A G C A T TGTGTAGGCAGGTGTGGAC CGACTGTTTCCTCTGTGTTGAG CACTGCTGTTACTTTGGTGATTC 96 method to genotyping errors, all allele sizes were verified by.hand using the GeneMapper® software. GENEPOP software (http://wbiomed.curtin.edu.au/genepop/index.html) was used to calculate observed and expected heterozygosities and to test their conformance to Hardy-Weinberg equilibrium. Specific growth rate in mass and length was calculated at SGR = (\r1W2 - \nWi)/( t2 - (2) x 100 where Wi and W2 are the mass or length at time ^ and t2, respectively. Fixed effect one-way A N O V A (genotype as the factor) was used to test for differences in length, mass, and specific growth rate in length and mass between groups under culture conditions, followed by Tukey's pairwise comparison to distinguish groups (SIGMASTAT, Systat Software, Point Richmond, CA). Growth within the semi-natural habitats was analyzed as a randomized block design with genotype as the factor and trough as the block. Fixed factor A N O V A was used to test for differences among genotypes in viability from the eyed egg stage to first-feeding. To test for an increase or decrease in survival under risk of predation relative to the wild genotype, survival of the domestic and hybrid genotypes was compared to the wild genotype with G-tests incorporating Williams' correction (McGinnity et al. 1997, 2003; Sokal and Rohlf 1995). The estimated total survival was obtained by multiplying the survival rates from fertilization to first-feeding and from first-feeding to 27 days after first-feeding, for each genotype. G-tests were used to test for differences in survival relative to the wild genotype. 4.3 R E S U L T S At the first-feeding stage when the yolk had been completely absorbed, there were no differences in mass (F^n = 0.788, p = 0.556) or length (F^n = 1.342, p = 0.315) among the genotypes. The rate of growth for the duration of the experiment (27 days) varied according to genotype under culture conditions (Fig. 4.1), for mass (F^n = 4.898, p = 0.016) but not length (F^n = 2.702, p = 0.086). There is evidence of dominance for growth in the Fi line cross, however this may be an artifact of using only one male for all the fertilizations. There was an effect of environment on growth, with all genotypes having reduced specific growth rate of mass and length in the semi-natural stream tanks 9 7 (but still all with positive growth). The D and Fi salmon outgrew all other genotypes (Bwi, Bw2, and W) within the culture environment, but under semi-natural conditions there were no differences in rates of growth among the different genotypes. From the microsatellite analysis to determine genotypes of the surviving offspring, 3 % of the offspring could not be assigned to a parent-pair and were excluded from the analysis. Table 4 .2 summarized the observed and expected heterozygosities for the four loci. Ssa407 was not significantly different from Hardy-Weinberg equilibrium, but OMM1231 had higher heterozygosity while OMM1270 and ogo2 had lower than expected heterozygosity. The number of alleles at each locus ranged from 6 to 18 (Table 4 .2) . There was an effect of genotype on the proportion of viable alevin that hatched from fertilized eggs (Fig. 4 .2 , F 4 = 5 .628 , p = 0 . 0 0 6 ) with the rank order of viability as D > W > Bwi > Fi > B w 2 - After foraging for 2 7 days under risk of predation in ten seminatural mesocosms, there was a mean of 19 ± 3 survivors remaining from an initial population of 80 . There were no differences in the number of fry surviving in the ten tanks (G-test, p = 0 . 9 7 8 ) indicating that the level of mortality and food restriction was relatively consistent among the habitats. However, there was unequal representation of the genotypes after foraging under risk of predation (Fig. 4 .2 , x 2 4 = 15 .29 , p = 0 . 0 0 4 ) with the rank order of survival at this stage being the same as for survival to hatch. Pair-wise comparisons detected no difference between the pure parental strains. The B w 2 hybrids had lower survival than W, and all hybrids had lower survival than D (all p < 0 .05) . Negligible mortality ( < 5 % for any family) of the fry reared under culture conditions indicates that mortality was due to the effect of the natural environment (i.e. predation or food availability) and not due to poor endogenous viability at this stage. The estimated total survival from fertilization to 2 7 days after first-feeding followed the same rank as observed for both early life history stages above. These data are consistent with outbreeding depression effects on survival occurring by hybridization between wild and domesticated coho salmon strains, but such effects are only manifested under semi-natural conditions with low food availability and risk or predation. 98 T A B L E 4.2 Observed (Ho) and expected (HE) heterozygosities, probability of conformance to Hardy-Weinberg equilibrium (HWE), and total number of alleles ( N A ) for each loci used in offspring identification. Trait OMM1231 . OMM1270 Ogo2 Ssa407 H0 0.882 0.841 0.784 0.724 HE 0.845 0.950 0.805 0.641 HWE 0.000 0.000 0.000 0.088 N A 10 18 9 6 99 3.5 2.5 a? (23 1.5 0.5 1.0 0.8 f , 0.6 c 01 o r tO 0.4 0.2 • Control O Semi-Natural I 2 2 0 25 50 75 100 % Domestic Alleles • Control O Semi-Natural 5 £ o 0.0 0 25 50 75 100 % Domestic Alleles F I G U R E 4.1 Specific growth rate in mass (SGRmass) and length (SGRlength) for pure farm fish (D, 100% domestic alleles), pure wild fish (W, 0%), and their hybrids: Fi (50%), BWI (25%), and BW2 (12.5%). Fish were reared either in an aquaculture environment (culture, solid circles) or in semi-natural mesocosms (semi-natural, empty circles). Points represent family means and the associated standard error. 100 A) B) C) i in 2 0.15 13 iY3 0.00 0 12.5 25 50 100 % Domest ic Alleles 0 12.5 25 50 100 % Domest ic Alleles 0 12.5 25 50 100 % Domest ic Alleles F I G U R E 4.2 Mean survival of the different genotypes from fertilization to first-feeding (A), and after 27 days of foraging under risk of predation (B), and estimated total survival (C). Error bars represent the associated standard errors, and letters indicate significant differences among genotypes. The estimated total survival (C) was obtained by multiplying the survival rates of the egg and fry stage, for each genotype. 101 4.4 D I S C U S S I O N The present study has investigated the effects on survival of fry produced by three-generations of introgression of farmed fish into wild populations. After three generations of introgression of farmed alleles into a native population, there was evidence of outbreeding depression (within a simulated natural environment) indicated by decreased viability from egg fertilization to first-feeding, and decreased survival under risk of predation for 27 days after the first-feeding stage. Studies on salmonids support observations that hybridization will have mainly negative effects on salmon populations, either due to introgression of maladapted genotypes arising from natural local adaptation (Gharrett and Smoker 1991; Gharrett et al. 1999; Gilk et al. 2004; Smoker et al. 2004) or from changes arising through domestication (McGinnity et al. 1997, 2003; Fleming et al. 2000; McClelland et al. 2005; Tymchuk et al. 2006). In the present study, the hybrid strains had rates of survival less than the parental strains. High survival of the pure domestic strain suggests that a reduced response to risk of predation seen in the strain (Tymchuk et al. 2006; Chapter 2) does not, in and of itself, always lead to increased mortality of domestic fish under predation risk (albeit in semi-natural mesocosms). Although farm coho salmon tend to outgrow the wild genotypes under culture conditions and in seminatural environments provided with artificial food (Tymchuk et al. 2006), the present study has shown that they were not able to maintain their potential for faster growth under the semi-natural conditions tested in this experiment, suggesting that the use of natural prey items provided in limited quantities within this mesocosm strongly represses the expression of the domestic growth phenotype in pure and hybrid strains. However, even though the different genotypes did not show significantly different rates of growth, there were still significant differences in survival during the fry stage, indicating an inability to balance the tradeoff between obtaining food and avoiding predators. One possibility is that the hybrids may have a behavioral phenotype similar to that of the wild-type and show caution in foraging under risk of predation, but due to a metabolism influenced by the domestic alleles (which would be altered to allow faster growth) they then have a greater risk of succumbing to 102 starvation. Further tests on the alteration or separation of the relationship between behavior and physiology may yield interesting insight to this question. This study tested for outbreeding depression during only a small portion of the life history of coho salmon (from fertilization to one month following emergence). Although mortality during this time period can be severe and therefore would act as a strong selection pressure against less fit phenotypes (Elliott 1994; Einum and Fleming 2000), to understand the full fitness potential of specific strains, it is crucial to also know the relative net fitness over the entire life history as well as the reproductive fitness of the different genotypes. For example, i f domestic fish that had escaped into the wild were not able to successfully breed, there would not be any concern for introgression of domestic alleles into the wild population (although this does not rule out the potential for single generational impacts due to direct negative ecological interactions with the native population). It is also necessary to know the relative reproductive and net fitness of the hybrids to determine the rate at which the domestic alleles continue to be passed throughout the population. It will also be important to assess whether the outbreeding depression observed in this study is due to differences between populations that inherently existed between the strains before domestication (the farm line was derived from Kitimat populations of coho, while the native strain was of Chehalis origin), or if the differences in fitness arise from changes that have occurred due to selection in a culture environment. A previous study (Tymchuk et al. 2006) indicated that phenotypic differences (growth and behavior) between the farm and wild Kitimat strains are not as great in magnitude as the differences between farm and wild Chehalis strains, indicating that there have been phenotypic alterations due to domestication. In conclusion, this experiment provides further evidence indicating that interpopulation hybridization can have a detrimental effect on native populations of salmon. In particular, farm and wild hybrids showed decreased viability from fertilization to hatch with the Bw2 strain having the lowest viability. The Bw2 strain also showed reduced survival relative to the farm and wild parental strains after foraging under risk of predation. The low survival of the Fi hybrids relative to the backcross lines could possibly be due to the fact that only one male was used for all the Fi fertilizations. 103 When the survival estimates for the egg and fry stage were combined, there was a strong indication of outbreeding depression after three generations of introgression between the farm and wild strains. It is clear that the studies assessing risk of interaction between domestic and wild fish populations must incorporate a range of environmental characteristics as habitat will strongly influence the expression of heterosis or outbreeding depression, and their survival and reproductive fitness consequences. 104 4.5 R E F E R E N C E S Allendorf, F. W. 1991. Ecological and genetic effects of fish introductions: synthesis and recommendations. Can. J. Fish. Aquat. Sci. 48:178-181. Allendorf, F. W., and R. S. Waples. 1996. Conservation and genetics of salmonid fishes. Pages 238-280 in J. C. Avise, and J. L . Hamrick, editors. Conservation Genetics: Case Histories from Nature. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Q. Rev. Biol. 72:149-177. Ayles, G. B., and R. F. Baker. 1983. Genetic differences in growth and survival between strains and hybrids of rainbow trout (Salmo gairdneri) stocked in aquaculture lakes in the Canadian prairies. Aquaculture 33:269-280. Biro, P. A. , M . V . Abrahams, J. R. Post, and E. A . Parkinson. 2004. Predators select against high growth rates and risk-taking behaviour in domestic trout populations. P. Roy. Soc. Lond. B 271:2233-2237. Bryden, C. A. , J. W. Heath, and D. D. Heath. 2004. Performance and heterosis in farmed and wild Chinook salmon (Oncorhynchus tshawytscha) hybrid and purebred crosses. Aquaculture 235:249-261. Burke, J. M . , and M . L. Arnold. 2001. Genetics and the fitness of hybrids. Annu. Rev. Genet. 35:31-52. Cairney, M . , J. B. Taggart, and B. Hoyheim. 2000. Characterization of microsatellite and minisatellite loci in Atlantic salmon (Salmo salar L.) and cross-species amplication in other salmonids. Mol. Ecol. 9:2155-2234. Carr, J. W., J. M . Anderson, F. G. Whoriskey, and T. Dilworth. 1998. The occurence and spawning of cultured Atlantic salmon (Salmo salar) in a Canadian river. ICES J. Mar. Sci. 54:1064-1073. Clifford, S. L. , P. McGinnity, and A . Ferguson. 1998. Genetic changes in Atlantic salmon (Salmo salar) populations of northwest Irish rivers resulting from escapes of adult farm salmon. Can. J. Fish. Aquat. Sci. 55:358-363. Crossman, E. J. 1991. Introduced freshwater fishes: A review of the North American perspective with emphasis on Canada. Can. Tech. Rep. Fish. Aquat. Sci. 48. 105 Crozier, W. W. 1993. Evidence of genetic interaction between escaped farmed salmon and wild Atlantic salmon (Salmo salar L.) in a northern Irish river. Aquaculture 113:19-29. Danzmann, R. G. 1997. P R O B M A X : A computer program for assigning unknown parentage in pedigree analysis from known genotypic pools of parents and progeny. J. Hered. 88:333. Dil l , L. M . , and A . H. G. Fraser. 1984. Risk of predation and the feeding behavior of juvenile coho salmon (Oncorhynchus kisutch). Behav. Ecol. Sociobiol. 16:65-71. Einum, S., and I. A . Fleming. 1997. Genetic divergence and interactions in the wild among native, farmed and hybrid Atlantic salmon. J. Fish Biol. 50:634-651. Einum, S., and I. A . Fleming. 2000. Selection against late emergence and small offspring in Atlantic salmon (Salmo salar). Evolution 54:628-639. Elliott, J. M . 1994. Quantitative Ecology and the Brown Trout. Oxford University Press, USA. Fleming, I. A. , T. Agustsson, B. Finstad, J. I. Johnsson, and B. T. Bjornsson. 2002. Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 59:1323-1330. Fleming, I. A. , and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES J. Mar. Sci. 54:1051-1063. Fleming, I. A. , K. Hindar, I. B. Mjoelneroed, B. Jonsson, T. Balstad, and A . Lamberg. 2000. Lifetime success and interactions of farm salmon invading a native population. P. Roy. Soc. Lond. B 267:1517-1523. Fleming, I. A. , B. Jonsson, M . R. Gross, and A. Lamberg. 1996. An experimental study of the reproductive behaviour and success of farmed and wild Atlantic salmon (Salmo salar). J. Appl. Ecol. 33:893-905. Gharrett, A . J., and W. W. Smoker. 1991. Two generations of hybrids between even- and odd-year pink salmon (Oncorhynchus gorbuscha): A test for outbreeding depression? Can. J. Fish. Aquat. Sci. 48:1744-1749. Gharrett, A . J., W. W. Smoker, R. R. Reisenbichler, and S. G. Taylor. 1999. Outbreeding depression in hybrids between odd- and even-broodyear pink salmon. Aquaculture 173:117-130. 106 Gilk, S. E., I. A . Wang, C. L. Hoover, W. W. Smoker, S. G. Taylor, A . K. Gray, and A. J. Gharrett. 2004. Outbreeding depression in hybrids between spatially separated pink salmon, Oncorhynchus gorbuscha, populations: marine survival, homing ability, and variability in family size. Environ. Biol. Fish. 69:287-297. Gjedrem, T. 1976. Possibilities for genetic improvements in salmonids. J. Fish. Res. Bd. Can. 33:1094-1099. Gjedrem, T. 2000. Genetic improvement of cold-water fish species. Aquaculture Research 31:25-33. Gotthard, K. 2000. Increased risk of predation as a cost of high growth rate: an experimental test in the speckled wood butterfly, Pararge aegeria. J. Anim. Ecol. 69:896-902. Gudjonsson, S. 1991. Occurrence of reared salmon in natural salmon rivers in Iceland. Aquaculture 98:133-142. Hansen, L. P., K. B. Doving, and B. Jonsson. 1987. Migration of farmed adult Atlantic salmon with and without olfactory sense, released on the Norwegian coast. J. Fish Biol. 30:713-721. Hansen, L. P., J. A . Jacobsen, and R. A . Lund. 1997. The incidence of escaped farmed Atlantic salmon, Salmo salar L., in the Faroese fishery and estimates of catches of wild salmon. ICES, Copenhagen, Denmark. Hershberger, W. K. , J. M . Myers, R. N . Iwamoto, W. C. Macauley, and A. M . Saxton. 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85:187-197. Hindar, K. , N . Ryman, and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Can. J. Fish. Aquat. Sci. 48:945-957. Honek, A . 1993. Intraspecific variation in body size and fecundity in insects: a general relationship. Oikos 66:483-492. Jacobsen, J. A. , L . P. Hansen, and R. A . Lund 1992. Occurrence of farmed salmon in the Norwegian Sea. Ices, Copenhagen, Denmark. Johnsson, J. I., and M . V . Abrahams. 1991. Interbreeding with domestic strain increases foraging under threat of predation in juvenile steelhead trout (Oncorhynchus mykiss): An experimental study. Can. J. Fish. Aquat. Sci. 48:243-247. 107 Johnsson, J. I., J. Hojesjo, and I. A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can. J. Fish. Aquat. Sci. 58:788-794. Johnsson, J. I., E. Petersson, E. Joensson, B. T. Bjoernsson, and T. Jaervi. 1996. Domestication and growth hormone alter antipredator behaviour and growth patterns in juvenile brown trout, Salmo trutta. Can. J. Fish. Aquat. Sci. 53:1546-1554. Lima, S. L. , and L. M . Di l l . 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68:619-640. Lura, H. , and H. Saegrov. 1991. A method of separating offspring from farmed and wild Atlantic salmon (Salmo salar) based on different ratios of optical isomers of astaxanthin. Can. J. Fish. Aquat. Sci. 48:429-433. Lynch, M . , and B. Walsh 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, M A . McClelland, E. K. , J. M . Myers, J. J. Hard, L . K. Park, and K. A . Naish. 2005. Two generations of outbreeding in coho salmon (Oncorhynchus kisutch): effects on size and growth. Can. J. Fish. Aquat. Sci. 62:2538-2547. McGinnity, P., P. Prodohl, A. Ferguson, R. Hynes, N . 6 Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. P. Roy. Soc. Lond. B 270:2443-2450. McGinnity, P., C. Stone, J. B. Taggart, D. Cooke, D. Cotter, R. Hynes, C. McCamley, T. Cross, and A. Ferguson. 1997. Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: use of D N A profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54:998-1008. McKinnell, S., A . J. Thomson, E. A . Black, B. L . Wing, C. M . Guthrie, III, J. F. Koerner, and J. H. Helle. 1997. Atlantic salmon in the North Pacific. Aquacult. Res. 28:145-157. 108 Morton, A. , and J. P. Volpe. 2002. A description of escaped farmed Atlantic salmon Salmo salar captures and their characteristics in one Pacific salmon fishery area in British Columbia, Canada, in 2000. Alaska Fishery Research Bulletin 9:102-110. Munch, S. B., and D. O. Conover. 2003. Rapid growth results in increased susceptibility to predation in Menidia menidia. Evolution 57:2119-2127. Olsen, J. B., P. Bentzen, and J. E. Seeb. 1998. Characterization of eight microsatellite loci derived from pink salmon. Mol . Ecol. 7:1087-1089. Peterson, R. G. 1999. Potential genetic interaction between wild and farm salmon of the same species. Pp. 23. The Office of the Commissioner for Aquaculture Development, Fisheries and Oceans Canada. Rexroad III, C. E., and Y. Palti. 2003. Development of Ninety-Seven Polymorphic Microsatellite Markers for Rainbow Trout. Trans. Am. Fish. Soc. 132:1214-1221. Roff, D. A . 1992. The evolution of life histories: theory and analysis. Chapman and Hall, New York. Schluter, D., T. D. Price, and L. Rowe. 1991. Conflicting selection pressures and life history trade-offs. P. Roy. Soc. Lond. B 246:11-27. Shine, R. 1988. The evolution of large body size in females: a critique of Darwin's fecundity advantage model. Am. Nat. 131:124-131. Smoker, W. W., I. A . Wang, A . J. Gharrett, and J. J. Hard. 2004. Embryo survival and smolt to adult survival in second-generation outbred coho salmon. J. Fish Biol. 65:254-262. Sogard, S. M . 1997. Size-selective mortality in the juvenile stage of teleost fishes: a review. Bull. Mar. Sci. 60:1129-1157. Sokal, R. R., and F. J. Rohlf 1995. Biometry. Freeman, New York. Stearns, S. C. 1992. The evolution of life histories. Oxford University Press, New York. Sundstrom, L. F., M . Lohmus, and R. H. Devlin. 2005. Selection on increased intrinsic growth rates in coho salmon, Oncorhynchus kisutch. Evolution 59:1560-1569. Sundstrom, L . F., M . Lohmus, R. H. Devlin, J. I. Johnsson, C. A . Biagi, and T. Bohlin. 2004a. Feeding on profitable and unprofitable prey: comparing behaviour of growth-enhanced transgenic and normal coho salmon (Oncorhynchus kisutch). Ethology 110:381-396. 109 Sundstrom, L. F., M . Lohmus, J. I. Johnsson, and R. H. Devlin. 2004. Growth hormone transgenic salmon pay for growth potential with increased predation mortality. P. Roy. Soc. Lond. B. 27LS350-S352. Sundstrom, L. F., E. Petersson, J. Hojesjo, J. I. Johnsson, and T. Jarvi. 2004b. Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behav. Ecol. 15:192-198. Travis, J. 1989. The role of optimizing selection in natural populations. Annu. Rev. Ecol. Syst. 20:279-296. Tymchuk, W. E., C. A . Biagi, R. E. Withler, and R. H. Devlin. 2006. Growth and behavioural consequences of introgression of a domesticate aquaculture genotype into a native strain of coho salmon (Oncorhynchus kisutch). Trans. Am. Fish. Soc. 135:442-445. Tymchuk, W. E., and R. H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245:295-300. van den Berghe, E. P., and M . R. Gross. 1989. Natural selection resulting from female breeding competition in a pacific salmon (coho: Oncorhynchus kisutch). Evolution 43:125-140. Volpe, J. P., E. B. Taylor, D. W. Rimmer, and B. W. Glickman. 2000. Evidence of natural reproduction of aquaculture-escaped Atlantic salmon in a coastal British Columbia river. Conserv. Biol. 14:899-903. Wolf, D. E., N . Takebayashi, and L. H. Rieseberg. 2001. Predicting the risk of extinction through hybridization. Conserv. Biol. 15:1039-1053. 110 CHAPTER V S : G R O W T H A N D S U R V I V A L T R A D E O F F S L E A D I N G T O O U T B R E E D I N G D E P R E S S I O N I N R A I N B O W T R O U T (ONCORHYNCHUS MYKISS) 5 A version of this manuscript is currently under review for publication. Tymchuk, W. E., Devlin, R. H. and Sundstrom, L. F . Growth and Survival Tradeoffs Leading to Outbreeding Depression in Rainbow Trout (Oncorhynchus mykiss). Evolution. I l l 5.1 I N T R O D U C T I O N The fitness consequences of introgression through interpopulation hybridization are difficult to predict and assess. Interbreeding among distinct populations that have low genetic variability (e.g. from inbreeding or bottlenecks) has the potential to re-establish lost alleles and allelic combinations, or mask deleterious recessive alleles, which could lead to hybrid vigour, or heterosis (Lynch and Walsh 1998). However, i f the parent populations come from different habitats, the resulting progeny may not be well adapted for either habitat (Allendorf and Waples 1996). A n allele that is advantageous in one environment may be disadvantageous to overall fitness in another environment. This may result in outbreeding depression due to disruption of local adaptation via a loss of additive x additive epistasis. The observed heterosis of an Fi line will be determined by a tradeoff between positive dominance effects acting within loci, and negative epistatic effects acting among loci (Lynch and Walsh 1998). Outbreeding depression may also arise from disruption of the linkage arrangement of co-adapted allele complexes, which would not become apparent until the F i generation, or beyond if the coadapted allele complexes consist of very tightly linked genes that require many generations of recombination to break apart. Fast-growing strains of commercially important salmonids have been developed through the use of selective breeding programs (Gjedrem 1983; Gjerde 1986; Hershberger et al. 1990). Interbreeding between wild and cultured strains offish provides a useful model to understand the genetic mechanisms underlying fitness effects caused by introgression, and the evolutionary constraints of enhanced growth in a natural environment. Faster growth is often strongly correlated with higher survival, particularly during early life stages (Sogard 1997) and large adults tend to have higher reproductive success than smaller adults (Roff 1992; Stearns 1992). This would suggest that a species should evolve to grow as fast as possible within their physiological constraints, but this is not what is observed in nature (Arendt 1997). In order to realize their fitness potential at reproductive maturity, an organism must obtain energy to grow and survive in its natural environment. For many organisms, there can be a trade-off between energetic gain (leading to growth) and mortality since individuals must assess the risk of predation each time they engage in foraging activity to 112 achieve a high rate of growth (Lima and Dil l 1990). This trade-off is often controlled through behavioural responses of the individual to its environment. Selection for faster growth and adaptation to a culture environment (domestication) is known to incur fitness tradeoffs such as altered foraging behaviour and reduced anti-predator behaviour (Biro et al. 2004; Di l l and Fraser 1984; Sundstrom et al. 2004a; Sundstrom et al. 2004b; Tymchuk et al. 2006), and may therefore have a significant effect on survival of fast-growing domestic individuals in a natural environment. There is a growing body of literature indicating that phenotypic differences between domestic and wild fish (due to indirect and direct selection) are largely a result of additive genetic differences. Therefore, the phenotypic effects of introgression tend to be diluted with repeated backcrossing into wild populations (Ayles and Baker 1983; Bryden et al. 2004; Einum and Fleming 1997; McClelland et al. 2005; McGinnity et al. 2003; McGinnity et al. 1997; Tymchuk et al. 2006; Tymchuk and Devlin 2005). Although additive gene action is shown to have a substantial effect on the expression of phenotype in hybrids, it is still not clear how these phenotypes are influenced by environment, and by non-additive genetic effects. Fitness of hybrids may be altered by endogenous (intrinsic) selection (such as physiological abnormalities that are expressed regardless of environment), exogenous (extrinsic) selection (environment-specific fitness difference), or by a combination of both, resulting in the magnitude of heterosis or outbreeding depression being dependent on the environment (Burke and Arnold 2001). Most studies that have tested hybrid fitness have done so only within the laboratory environment, and have therefore measured intrinsic mechanisms of outbreeding depression. Even if reciprocal transplant studies are conducted with fitness of the parents and hybrids assessed in both parental environments (Emms and Arnold 1997; Hatfield and Schluter 1999; Nagy 1997) it is still difficult to differentiate between the extrinsic and intrinsic mechanisms leading to outbreeding depression (Rundle and Whitlock 2001). The purpose of this study was to test, using a rainbow trout (Oncorhynchus mykiss) model system, the dependence of hybrid fitness on environment (exogenous selection). Three questions were asked to address this objective: 1) Does relative fitness of parental and hybrid lines differ according to environment; 2) How do environmental 113 characteristics alter relative fitness?; and 3) Is there delayed expression of outbreeding depression after three generations of introgression? Multiple fitness components were measured both under culture and simulated natural environmental conditions. First, i f relative fitness of parental and hybrid lines is dependent on environment, we would expect to see differences in the relationship between genotype and fitness among the different environments. Second, we Have attempted to isolate the influence of competitive interactions and risk of predation on the relative growth and survival of the wild and backcrossed lines to assist in identifying critical environmental variables influencing the magnitude of heterosis and/or outbreeding depression. To address the third question, we have tracked the introgression of an artificially-selected genotype into a wild genetic background over three generations. To our knowledge, this is the first study reporting on the impact of introgression in fish populations over three generations. This information has allowed examination of whether there are delayed fitness effects of outbreeding depression which would appear as reduced fitness of the second-generation backcross line relative to the parental strains. Results of these experiments are relevant to improving our evolutionary understanding of the interaction among genomes during hybridization events, as well to assessing the risk of introgression of domesticated fish into wild populations. 5.2 M A T E R I A L S A N D M E T H O D S Strains of rainbow trout (Oncorhynchus mykiss) expressing different rates of growth have been utilized in the present study to address our research objectives. The genetic basis of growth differences between these strains has been established in an earlier common-garden experiment (Tymchuk and Devlin 2005). Gametes were collected from domestic farmed (D; Campbell Lake Trout Farm, Little Fort, British Columbia (BC), wild (W; Pennask Lake, BC, Canada), domestic x wild hybrid (Fi), and Fi x wild backcross (Bi) trout in June 2004 (all hybrid strains were reared their entire life at the DFO/UBC Centre for Aquaculture and Environmental Research; see Tymchuk & Devlin 2005 for history of these parents). On 2 June 2004, a total of five pure domestic (D) full-sib crosses were made, utilizing five females and five males. On 9 June 2004, 114 gametes were collected from five wild caught Pennask Lake females. Eggs from each female were divided into five aliquots so that each female was crossed with one male of each of the genotypes used for the fertilizations (D, W, F i , and Bi). Some families were terminated due to poor viability, leaving the following line crosses: pure domestic (D, four families), pure wild (W, four families), domestic x wild hybrids (Fi, two families), Fi x wild backcrosses (Bi, four families) and B\ x wild backcrosses (B2, four families). A l l eggs were reared in Heath trays with natural creek water. By 30 July 2004, the alevin had utilized most of their yolk mass and were transferred to rearing tanks so that feeding could be initiated. The fry were reared under three different environmental conditions: 1) regular culture conditions; 2) semi-natural conditions with competitive interactions among genotypes; and 3) semi-natural conditions with risk of predation. 5.2.1 Culture Conditions Under culture conditions, families were reared separately and the individual groups consisted of 50 siblings from the same family. The individual groups were reared in 200 L tanks (four families of each genotype except for the Fi line which had two families; 18 tanks in total) with natural creek water at a flow no less than 1 L min"1 kg"1, temperature 5 - 16°C, supplemental aeration, and were fed commercial feed (Skretting, Vancouver, Canada; composition of feed varied over time to meet the requirements of different life stages) to satiation several times per day. Individual families were not reared in replicate since a previous study (on fish reared in the same tanks and conditions as used in this experiment) indicated that fish grown under these conditions did not differ between replicates (Tymchuk et al. 2006; Tymchuk and Devlin 2005). Instead, we decided to maximize the number of different families used to represent each genotype. A model of a great blue heron was used to assess the relative response to risk of predation of the different genotypes under culture conditions. A similar technique has been used to measure predator response in rainbow trout (Hojesjo et al. 1999; Jonsson et al. 1996), brown trout (Sundstrom et al..2005), Atlantic salmon (Johnsson et al. 2001), and coho salmon (Tymchuk et al. 2006). These antipredator trials were conducted between 31 January 2005 and 9 February 2005. In total, six behavioral observations were 115 made: three control and three simulated attack trials. There was at least one day of regular feeding between each observation. For the simulated predator attack trials, prior to addition of food, the heron model was dunked into the tank three times in rapid succession. The protocol for the behavioral observations was to record the time until a pellet of food was consumed. Food pellets were dropped into the tank, with a new pellet added as soon as the previous reached the bottom of the tank. Timing was started when the first pellet was dropped, and was stopped when a fish ate the pellet. For analysis, the mean of the three control and three simulated attack observations for each family was used. Behavior (time to consumption of first food pellet) was analysed with a two-way A N O V A to test for effect of genotype and predator (control or simulated attack). The dependent variable for the behavior test was the family mean of the three control and three simulated attack observations. Specific growth rate for mass (SGRmass) and length (SGRiength) was calculated as SGRmass or length = [(hi(M2 or Li) - \n{M\ or L\j) I time] * 100 where M\ (or L\) is the mass (or length) at 22 October 2004 and 1M2 (or Li) is mass (or length) at 23 May 2005. Time represents the number of days between the initial and final measurements, which in this case is 213 days. Condition factor was calculated as K = M2L2'3 x 100. Differences in growth and condition were tested with a one-way A N O V A with family means as the dependent variable and genotype as the factor. A l l statistical analyses were performed with SIGMASTAT (Systat Software, Point Richmond, CA) or SYSTAT® Version 10.2. For clarity, only significant p-values are reported in the results section. 5.2.2 Semi-natural Conditions under Competition Under semi-natural conditions with competitive interactions among genotypes, unfed fry were combined into mixed genotype groups consisting of 40 fish of each of five genotypes (D, W, Fj, B i , and B2). The mixed groups were replicated and reared in two large circular tanks (6 m diameter) enriched with gravel bottom, rocks, logs, branches, and water current (approximately 1 m s"1 generated with two pumps per tank). The water source and temperature was the same as for the fish reared under culture conditions (i.e. 116 natural creek water). These fry received only natural sources of food consisting of brine shrimp (Artemia sp), bloodworms (Chironomidae), amphipods (Gammarus sp), fruit fly adults (Drosophila melanogaster), and any other natural food that arrived through the creek water line or naturally fell into or grew within the tank (such as ants, wasps, and mosquito larvae). The fish placed into the semi-natural habitats were fin-clipped according to genotype through a combination of adipose and left or right pectoral fin clips. Genotypes were clipped differently for each of the two habitats. Fish were removed from the habitats 29 May 2005 (219 days after.they were placed into the system). Specific growth rate for mass (SGRmass) and length (SGRi e n gth) and condition factor was calculated as described for the culture environment experiment above. Here, Time represents the number of days between the initial (22 October 2004) and final (29 May 2005) measurements, which in this case is 219 days. Differences in growth and condition were tested with a one-way A N O V A with family means as the dependent variable and genotype as the factor. Survival of the genotypes was compared with a G-test using the pure wild crosses as a reference with the null hypothesis that there is no difference in survival among the genotypes. Mortality of the corresponding families under culture conditions was minimal, so no corrections were made to survival measured under semi-natural conditions with competition. 5.2.3 Semi-natural Habitat under Risk of Predation The semi-natural habitats for the predation experiment consisted of 200 L tanks with gravel bottoms and a large hollow cinder block to provide cover. For each genotype, 20 fry (pooled from four different families) were placed into each of four tanks (20 tanks in total). Half of the tanks also had one predator (coho salmon, Oncorhynchus kisutch, parr). This experiment was conducted twice: the first trial started on August 5 t h, 2004 and ended August 19 t h, 2004; the second trial started August 20 t h, 2004 and ended September 16 th, 2004. See Table 5.1 for a summary of the mean size of the predators and prey placed into these two trials. The number of trout prey visible, both before and after feeding, was scored from behind an opaque barrier to assess their willingness to expose 117 T A B L E 5.1 Size of the predators and prey used in the semi-natural habitat under risk of predation. Numbers in brackets are associated standard errors. Mass (g) Length (cm) Trial 1 Predators 47.8 (3.9) 15.6(0.3) D 0.12(0.03) 0.74 (0.12) W 0.14(0.03) 0.48 (0.11) F i 0.12(0.03) 0.76 (0.14) B, 0.12(0.04) 0.58 (0.13) B 2 0.13 (0.03) 0.50(0.13) Trial 2 Predators 67.2 (6.7) 17.3 (0.6) D 2.4 (0.1) 3.9 (0.2) W 2.6(0.1) 3.7 (0.2) Fi 2.5 (0.1) 4.0 (0.2) B, 2.5 (0.1) 3.8(0.2) B 2 2.6(0.1) 3.7(0.3) 118 themselves to predator attack. The fish were fed live artemia twice per day at 08:00 and 14:00. Growth rates were calculated as described above with initial date being 5 August 2004 and final date 19 August 2004 in trial 1 and 20 August 2004 and 16 September 2004, initial and final dates, respectively, in trial 2. Consequently, Time was 14 days for Trial 1 and 27 days for Trial 2. Condition factor was calculated as described above. Growth rates of the fry in the semi-natural habitats under risk of predation were tested with a two-way A N O V A on ranks with SGRmass/iength (tank mean) as the dependent variable and genotype and predator (present or absent) as the two factors. Survival under risk of predation was adjusted according to survival of the fish in the control tanks to ensure that any differences in mortality among the genotypes were due to the presence of a predator, and not genetically-determined physiological effects on viability. Survival was arcsin square root transformed and analysed with a two-way A N O V A using genotype and predator (present vs absent) as the fixed factors. Behavior (proportion of fish visible) was arcsin square root transformed and analysed with M A N O V A with activity before and after feeding as dependent variables and genotype and predator (present or absent) as the two factors. 5.2.4 Line Cross Analysis Joint-scaling regression technique was used to test for additive and additive-dominance models of gene action and their contribution to the phenotypic divergence between the wild and domestic trout, following the procedure outlined in (Lynch and Walsh 1998; Tymchuk et al. 2006). The joint-scaling test uses least-squares regression to estimate the model parameters and then compares the observed line means with the predicted means (Cavalli 1952; Hayman 1960). This analysis was carried out on family means from the D, W, F i , B i and B 2 crosses for time to hatch, SGR m , SGRi, behavior, and survival for all three environments (culture, competitive interactions, and risk of predation). 119 5.3 R E S U L T S 5.3.1 Behaviour in a Culture Environment Within the culture environment, all individual families of trout responded to the simulated predator attack (Fig. 5.1) by increasing the amount of time before consuming their first pellet (F ] i 2 6 = 17.876, p < 0.001). Genotype had an influence on the time it took to consume the first offered food pellet (F 4 ; 2 6 = 3.189, p = 0.029) with the domestic fish showing a faster response than all other groups. The genotypes deviated in size with the domestic families being larger in mass (F4 >i3 = 5.629, p = 0.007), length (F 4j3 = 3.200, p = 0.049) and condition factor (F4,i3 = 5.745, p = 0.008) relative to the wild and B 2 families (determined by Tukey's pairwise comparisons). 120 0 J , . • . 1 0 25 50 75 100 % Domestic Alleles F I G U R E 5.1 Relationship between the proportion of domestic alleles within the genotype and time required for the first food pellet to be consumed (family means and associated standard errors) under culture conditions without a simulated predator attack (control, solid circles) and after a simulated predator attack with a model of a blue heron (predator attack, open circles). Groups are as follows: pure domestic (D, 100%), pure wild (W, 0%), domestic x wild hybrids (Fi, 50%), Fi x wild backcrosses (Bi, 25%) and Bi x wild backcrosses (B2, 12.5%). The line connecting the D and W means indicates a priori expectation of additive genetic effects; character means for all hybrids should fall on this line if divergence is due to genes with only additive effects. 121 5.3.2 Semi-natural Habitat under Competition At the start of the experiment, there were no differences in mass, length, or condition factor among the different genotypes placed into the semi-natural system at the fry stage. At the final measuring period (using only the survivors), there was a strong effect of genotype on SGR m a S s(Fig. 5.2A; F 4 ; 5 = 15.315, p = 0.005) and SGR| e n gth (Fig. 5.2B; F 4 ] 5 = 19.324, p = 0.003). Final condition factor did not vary according to genotype (Fig. 5.2C) Survival proportions of the different genotypes from the two replicates were grouped as the different tanks did not have a significant effect on relative survival (paired t-test, t4 = -0.688, p = 0.529). There was an effect of genotype on survival of the fry (Fig. 5.3; X 4 = 71.490, p < 0.001) with domestic, Fi and Bi families having higher survival relative to B2 and pure wild fry. While obtaining a mass and length measurement on the survivors, the stomachs of randomly selected fish were cut open to look for the presence of any identifiable food items. In addition to ants and wasps, several of the largest fish had smaller conspecifics inside their stomach, providing evidence of cannibalism within the system. 5.3.3 Semi-natural Habitat under Risk of Predation To examine the effect of domestic alleles on actual predation mortality, experiments were undertaken in semi-natural habitats supplied with natural food sources. Including both trials, the overall survival of the trout fry was 92 ± 4% in the control (non-predator) habitats, and 71 ± 5% in the habitats with a predator (Fig. 5.4). Data from the two trials was pooled since there was no difference between the replicates for survival in the control (ti 8 = 1.213, p = 0.241) or predator ( t n = -1.244, p = 0.235) habitats. During the sampling, a few fish were found in the screen trap collecting water from a block of five tanks. Since the genotype (source tank) of these fish could not be verified with certainty, five predator tanks (one of each genotype) were excluded from this analysis. 122 • Culture o Semi-natural 0 25 50 75 100 % Domestic Alleles 0.5 • Culture o Semi-natural 0 25 50 75 100 % Domestic Alleles 1.6 1.4 1.2 i.o H 0.8 0 25 50 75 100 % Domestic Alleles F I G U R E 5.2 Relationship between the proportion of domestic alleles in the genotype for families reared under culture conditions (closed circles) or semi-natural conditions with competition (open circles) and (a) SGR m , specific growth rate in mass, %bw day"';(b) SGR|, specific growth rate in body length, % bl day"1; or (c) condition factor. Points represent genotype means and associated standard errors. Groups are as described in Figure 1, where D = 100%, F, = 50%, B 2 = 25%, B i = 12.5% and W = 0%. The line connecting the D and W means indicates a priori expectation of additive genetic effects. 123 1.0 -o.o -I I I I I 0 25 50 75 100 % Domestic Alleles F I G U R E 5.3 Relationship between the proportion of domestic alleles within the genotype and mean survival (and associated standard error) in a semi-natural habitat with competition among genotypes. Genotype groups are as described in Figure 1. The line connecting the D and W means indicates a priori expectation of additive genetic effects. 124 1.0 0.8 1 "£ 0.6 13 if) 0.4 0.2 • i • no predator O predator 0 25 50 75 % Domestic Alleles 100 F I G U R E 5.4 Relationship between the proportion of domestic alleles within the genotype and mean survival of the families (and associated standard error) in a semi-natural habitat without risk of predation (no predator, closed circles) or under risk of predation (predator, open circles). Genotype groups are as described in Figure 1. The line connecting the D and W means under risk of predation indicates a priori expectation of additive genetic effects. 125 Presence of a predator decreased overall survival (Fi^s - 10.770, p = 0.003). There was also an effect of genotype on survival (F^s = 2.793, p = 0.048), but no detectable interaction between genotype and predator (F^s = 1.340, p = 0.283). Activity (number of fry visible) in all tanks was recorded both before and after food was added to the system for five days prior to the addition of predators into half the tanks. Genotype had an effect on the activity level of the fish (Fig. 5.5 A ; Wilks A-g.sg = 0.400, p < 0.001) with univariate tests indicating that the effect of genotype was present in the activity of fish after feeding, but not before feeding. After the predators were added to half the tanks, the activity before and after feeding was recorded each day until the end of the experiment (Fig. 5.5B and 5.5C). There was a strong genotype effect on the activity level of the fry (Wilks' A,8,58 = 0.3 86, p < 0.001) both before and after feeding, with the D and Fj fry showing significantly higher levels of activity relative to the other genotypes (similar to that seen above in Fig. 5.5A). The presence of the predator caused a significant reduction in the activity of the fry relative to the non-predator habitats, both before and after feeding (Wilks' %2,29 = 0.432, p < 0.001). At the start of trial 1, the fry did not differ in mass according to genotype (F 4 J 95 = 2.206, p = 0.074) although the pure domestic fry were smaller in length relative to all other genotypes (F 4 J 95 = 8.688, p < 0.001). By the beginning of trial 2, this relationship was reversed, and the domestic and F i families were significantly larger in mass (F 4 J 95 = 20.716, p < 0.001) and length (F 4 , 95 = 9.137, p < 0.001). Growth rates differed between the two trials for both SGRmass ( t 3 g = 5.908, p < 0001) and S G R i e n g t h (t 38 = 4.178, p < 0.001) so the trials were analyzed separately. For both trials, there were no differences in growth rates or condition among the different genotypes in the control and predator habitats (Fig. 5.6). The presence of a predator in the semi-natural habitats did not alter growth rates or condition factor in trial 1 nor trial 2. 1.0 i 0.8 -0.6 -s 0.4 -0.2 -0.0 -Prior to Addition of Predators • P r io r t o f e e d i n g O A f t e r f e e d i n g 0 25 50 75 100 % Domestic Alleles B 1.0 0.8 0.6 0.4 0.2 0.0 Control o P r i o r t o f e e d i n g A f t e r f e e d i n g 0 25 50 75 100 % Domestic Alleles Predator • P r i o r t o f e e d i n g O A f t e r f e e d i n g 0 25 50 75 100 % Domestic Alleles F I G U R E 5.5 Relationship between the proportion of domestic alleles within the genotype and activity (measured as the proportion of fry visible within the tanks) prior to the addition of live artemia (prior to feeding, closed circles) and after the addition of artemia (after feeding, open circles) for (a) the first week prior to the addition of predators; (b) the control habitats that did not receive a predator; and (c) the habitats that had a predator. Points represent genotype means and associated standard errors. Genotype groups are as described in Figure 1. The line connecting the D and W means indicates a priori expectation of additive genetic effects. to O N 127 F I G U R E 5.6 Relationship between the proportion of domestic alleles within the genotype of families reared under semi-natural conditions without a predator (closed circles), or semi-natural conditions with a predator (open circles) and (a) SGR m , specific growth rate in mass, % bw day"1; (c) SGR|, specific growth rate in body length, % bl day"1; or (e) condition factor in Trial 1 and (b) SGR m ; (d) SGRi; or (f) condition factor in Trial 2. 128 5.3.4 Line Cross Analysis Additive genetic effects adequately explained the behavioural phenotypes (Table 5.2) of the line crosses measured as response to food in a culture environment (Fig. 5.1), and activity of the fry (number visible) in a semi-natural habitat before feeding (Fig. 5.5A-C) both in the presence and absence of a predator. However, an additive-dominance model of gene action provided a better fit to the means of the line crosses for the activity of the fry after feeding in the semi-natural habitats with risk of predation (Fig. 5.5A, B). In these habitats, the Fi crosses showed heterosis for activity with levels close to the mean of the domestic lines. Growth in mass and length of the fry reared under both culture and semi-natural conditions (Fig. 5.2A, B) did not fit an additive or dominance model, indicating epistasis (Table 5.3). Under culture conditions, the Fi and B2 families showed higher rates of growth than would be expected by an additive model, whereas under semi-natural conditions the Bi crosses showed higher than expected rates of growth. Condition factor in both environments (Fig. 5.2C) adequately fit a dominance model, with the backcrosses having condition factors higher than the mean of the Fi and wild lines under culture conditions, and heterosis being indicated by the Fi lines in the semi-natural habitat. During Trial 1 of the predation experiment (Table 5.3), growth in mass, growth in length, and condition factor (Fig. 5.6) did not differ from an additive model. In the tanks without a predator in Trial 2 (Table 5.3), the line crosses did not fit an additive or dominance model for S G R m a s s , SGRi e n grh, or condition (Fig. 5.6). Under these conditions, the Fi families had rates of growth lower than the mid-parent means, and the B i families had higher rates of growth, and condition factor, than expected. Under risk of predation, the line crosses fit a dominance model for SGRm ass, SGR|ength, and condition (Fig. 5.6) with the Fi and Bi crosses showing lower growth rates than predicted from the parental means. Survival under competition did not differ from an additive model (p = 0.681) whereas under risk of predation, the line means did not fit either an additive (p = 0.0002) 129 T A B L E 5.2 Joint-scale analysis of behavior of the line crosses. Values represent %2 statistic measuring goodness-of-fit of line means to genetic models, with the associated p-value in brackets. Trait Habitat Treatment Additivity Dominance Best Fit Model Response to food culture control 0.287 (0962) Additivity simulated predator 0.010 (0.997) Additivity Activity before feeding semi-natural control 2.492 (0.477) Additivity no predator 2.037 (0.565) Additivity predator3.636 (0.305) Additivity Activity after feeding semi-natural control 6.949 (0.074) 1.586 (0.452) Dominance no predator5.263 (0.154) 1.626 (0.444) Dominance predator2.828 (0.419) Additivity 130 T A B L E 5.3 Joint-scale analysis of growth (in mass and length) and condition factor under culture or semi-natural conditions with competition or risk of predation. Values represent X 2 statistic measuring goodness-of-fit of line means to genetic models, with the associated p-value in brackets. Trait Habitat Treatment Additivity Dominance Best Fit Model SGR m a s s culture 10.228 (0.017) 8.827 (0.012) Neither semi-natural competition 11.540 (0.009) 8.035 (0.018) Neither no predator (Trial 1) 2.175 (0.537) Additivity no predator (Trial 2) 19.373 (O.001) 19.261 (O.001) Neither risk of predation (Trial 1) 3.110 (0.375) Additivity risk of predation (Trial 2) 6.642 (0.084) 0.361 (0.835) Dominance SGR|ength culture 13.119(0.004) 12.596 (0.002) Neither semi-natural competition37.256 (<0.001) 17.319 (<0.001) Neither no predator (Trial 1) 0.763 (0.858) Additivity no predator (Trial 2) 11.907 (0.008) 10.338 (0.006) Neither risk of predation (Trial 1) 1.122 (0.772) Additivity risk of predation (Trial 2) 7.017 (0.071) 0.879 (0.644) Dominance Condition culture 7.522 (0.057) 3.080 (0.214) Dominance semi-natural competition 20.685 (<0.001) 0.819(0.664) Dominance no predator (Trial 1) 2.374 (0.498) Additivity no predator (Trial 2)64.686 (<0.001) 63.004 (O.001) Neither risk of predation (Trial 1) 2.761 (0.430) Additivity risk of predation (Trial 2) 7.537 (0.057) 4.803 (0.091) Dominance 131 or dominance model (p = 0.0001), suggesting epistasis. Outbreeding depression was indicated by the B 2 line means (Fig. 5.4). 5.3.5 Trade-off between Growth and Survival Within the semi-natural habitats under risk of predation, there was a negative correlation between the survival of fry and their activity both before and after feeding (Figs. 5.7A, C). The correlation was reversed for specific growth rate (in mass) and activity (both before and after feeding, Fig. 5.7A, C). Within the semi-natural habitats that did not have a predator (Figs. 5.7B, D ) , there were no correlations between activity and estimates of fitness (growth or survival) with the exception of a negative correlation between activity before feeding and growth rate in mass (Fig. 5.7B). 5.3.6 Different Selection Pressures of Competition and Predation There were strong positive correlations between growth in both mass and length and survival within the semi-natural habitats with competition (Figs. 5.8A, C). Under risk of predation, the correlations between growth and survival were,negative (Figs. 5.8B, D ) . 5.4 D I S C U S S I O N It is clear from this study that the ability to detect heterosis and outbreeding depression will be influenced by the environment in which the hybrids are studied. A reduction in fitness of hybrids may be due to disruption of local adaptation (environment), disruption of co-adapted gene complexes (physiology), or a combination of both. Results from this experiment indicate that disruption of local adaptation likely plays a prominent role in the reduced fitness experienced by progeny produced after three generations of introgression of domestic alleles into a wild population, which supports 132 C Risk of Predation D No Predators R z = 0.266, p = 0.049 R 2 = 0.433, p = 0.008 0.2 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 Activity After Feeding ^ Activity After Feeding • SGRmass O Survival F I G U R E 5.7 Correlations between activity in semi-natural habitats and fitness components measured as S G R m a s s , specific growth rate in mass, % bw day"1 (closed circles) and number of survivors in the habitat (open circles). Correlations were tested for activity before feeding (a) under risk of predation; and (b) with no predators; and for activity after feeding (c) under risk of predation; and (d) with no predators. Points represent the values obtained for each habitat (tank) within the experiment. 133 Competitor! B Predation mass 0.15 0.20 0.25 0.30 0.35 0.40 S G R l e n g t h 1.0 -0.8 -- •• • •• 0.6 - • • • 0.4 - • 0.2 -• • 0.0 - R2 = 0.276, p =' 0.045 D SGR, mass 1.0 - • • 0.8 -0.6 - • • 0.4 -0.2 -o.o - R2 '= 0.341, p = 0.022 8 10 0.0 0.4 0.8 1.2 1.6 2.0 2.4 SGR l e n g t h FIGURE 5.8 Correlations between growth measured as S G R m a s s in (a, b) and SGRlength in (c, d) and survival (proportion) for fish reared under semi-natural conditions with competition (a, c) or under risk of predation (b, d). Points represent the values obtained for each habitat (tank). 134 theory that adaptation to local conditions will be the primary cause of outbreeding depression in species such as salmon that are locally adapted on a small geographic scale (Templeton 1986). However, the nonlinearity of survival under risk of predation suggests that intrinsic mechanisms (such as genetic incompatibilities) are interacting with the ecological selection pressures on the hybrid strains. Under culture conditions, the Fi families tended to show heterosis for growth (in mass and length) in the direction of the domestic parental line, and for condition factor in the direction of the wild line. These trends were also observed under semi-natural conditions under competition, but were altered when reared with a risk of predation (in Trial 2). Heterosis was also observed in the activity of the fish after feeding in a semi-natural habitat without risk of predation where the activity of the Fi fry was not different from the activity of the domestic fry. Cross-breeding experiments with Atlantic salmon did not find significant heterosis for survival or body weight (Friars et al. 1979; Gjerde and Refstie 1984) although heterosis has been detected in brook trout (Webster and Flick 1981), rainbow trout (Ayles and Baker 1983; Gjerde 1988; Gjerde et al. 1983; Wangila and Dick 1996) and coho salmon (McClelland et al. 2005). Thus, heterosis effects seem to be dependant on both the species and strains used as well as the specific environmental conditions and possibly developmental stages examined. The route of genetic introgression of domestic strains into wild populations is likely to be through hybridization rather than by pure farm stock displacing pure wild populations (Clifford et al. 1998; Fleming et al. 2000), and thus outbreeding depression may not be apparent until recombination has separated co-adapted genotypes. In the present experiments, evidence of outbreeding depression for survival under risk of predation was not apparent until three generations of introgression of domestic alleles into a wild background (with B 2 hybrids showing rates of survival lower than either parental strain), suggesting that closelyrlinked loci may contain co-adapted alleles in these strains. While outbreeding depression may be first anticipated in F 2 and first-generation backcross progeny, no evidence of depression was observed in the present 135 studies for estimates of survival in the semi-natural habitat with competition, nor for growth in culture or semi-natural conditions with either competition or risk of predation. Similarly, no outbreeding depression effects were detected in F 2 or first-generation backcross progeny among domesticated and wild coho strains (Tymchuk et al. 2006), rainbow trout (Tymchuk and Devlin 2005), Atlantic salmon (McGinnity et al. 1997) nor among hybrids of different wild strains of coho salmon (Smoker et al. 2004). However, evidence of outbreeding depression for survival of F 2 hybrids between odd- and even-year pink salmon has been documented, and may arise due to epistatic gene action resulting from disruption of co-adapted gene complexes (Gharrett and Smoker 1991; Gharrett et al. 1999; Gilk et al. 2004). The difficulty in obtaining high statistical power in these experiments for estimates of outbreeding depression suggest that negative results should be interpreted with caution until further data has been developed. Due to the complexity of the environment in which salmon live, there is an intricacy inherent in the relationship between their phenotype, the environment, and their consequent fitness. Using rainbow trout as a model species, the present study shows that distinct environments can indeed establish different tradeoffs in behaviour, growth and mortality. In semi-natural environments where different strains competed for food resources, there was a positive correlation between growth and survival with faster-growing fish being able to monopolize limited food resources, and ultimately being able to prey upon slower-growing cohorts. The introduction of predators into these experiments, however, profoundly altered this relationship such that in semi-natural conditions under risk of predation, there was a negative correlation between growth and survival. In the habitats with predators, there was also a positive correlation between activity and growth, and a negative correlation between activity and survival. Although these opposing correlations suggest a behaviorally mediated tradeoff between growth and survival, this effect is probably not mediated by a plastic behavioural response to the predators since growth rates were no faster in the habitats without predators. In fact, the highest rates of growth were found in habitats with the highest rates of predation where there was a significant negative correlation between survival and growth, providing 136 support that food availability was the limiting factor for growth. In non-natural systems (such as with growth-enhanced transgenic salmonids), growth is highly dependant on food availability, and cannibalism can play a significant role in influencing survival under competitive conditions (Devlin et al. 2004). In the present experiments under culture and semi-natural conditions with competition, there was a positive correlation between the proportion of alleles from the fast-growing domestic strain within the genotype and the growth and survival of the fry. Fast-growing trout were able to maintain their growth advantage when reared under semi-natural conditions with competition, although growth of all families was not as large as when they were reared under culture conditions. For both mass and length, the ratios of SGR between D and W are approximately the same for the two environments (Fig. 3) suggesting they have a similar growth advantage in both conditions, even when growth is suppressed. This is supported by the line-cross analysis that found similar relationships among the genotypes in both the culture and semi-natural environment with competition. In semi-natural conditions under risk of predation, surviving fry with more of a domestic genetic constitution did not show faster rates of growth, perhaps because the fastest growing fry were more susceptible to predation, or possibly due to the influence of life stage (early fry stage). The fry reared under competition remained within the system for a year, whereas the fry in the predation experiments remained within the habitats for only three weeks, suggesting that first-feeding fry may have a different initial response to the environment relative to trout at later stages of growth. For example, it is possible that early fry expend less energy in the presence of a predator by shifting their activity to a more cryptic behaviour and allowing acquired energy to be directed more for growth. Such an effect may not occur in more domestic strains if they are less reactive to predation threat, as suggested by observations in the present study as well as other previous studies (Einum and Fleming 1997; Fleming and Einum 1997; Johnsson et al. 1996; Tymchuk et al. 2006). The present study has revealed that there are genetically-determined behavioural differences among the trout strains examined, and these are expressed both under culture 137 and semi-natural conditions. Relative behavioral differences among the genotypes showed more consistency among the three different environments compared with relative growth (see Table 2 vs Table 3). Under culture conditions, there was a negative correlation between the proportion of domestic alleles within the genotype and the time it took for the fish to begin feeding after food was made available in their habitats, with the pure domestic crosses taking the shortest amount of time to begin feeding. This trend was observed both before and after a simulated predator attack, even though all genotypes responded to the perceived risk of predation by increasing the time it took to resume feeding. In the semi-natural habitats under risk of predation, the D and Fi crosses tended to be more active both before and after feeding, in both the control and predator habitats. The D and Fi crosses also showed a stronger response to feeding compared to the other crosses. Other research has documented evidence of reduced antipredator behaviour in other fast-growing strains when compared with wild fish (Abrahams and Sutterlin 1999; Devlin et al. 1999; Einum and Fleming 1997; Fleming et al. 2002; Fleming and Einum 1997; Johnsson et al. 2001; Sundstrom et al. 2003; Sundstrom et al. 2004a; Tymchuk et al. 2006), although it has not been shown if these observations would hold across different environments. This research further supports the theory that a reduced response to risk of predation may be a conserved phenotypic result of domestication that will be expressed even in a natural environment. This does not mean, however, that realized survival will necessarily differ among the domestic and wild strains. It will be interesting, and important, to extend this research and track growth and survival of pure and hybrid crosses in an environment incorporating both competitive interactions and risk of predation. Depending on the intensity of the predation and the level of food within the habitat, a positive correlation between growth and survival in a competitive environment may be altered (strengthened or weakened), with theory and empirical data indicating that the effect of the introgressed domestic alleles may have less of an impact under competition and risk of predation if, for example, predation makes additional resources available that causes a reduction in the intensity of competitive 138 interactions (Chase et al. 2002). Alternatively, a more stressful environment may expose a greater magnitude of outbreeding depression due to genetic effects that are essentially neutral under less stressful condition (Kondrashov and Houle 1994). Even if the differences in growth and behaviour caused by introgression of farmed alleles are minimized under natural conditions, introgression may still impact the genetic integrity of wild populations. Significant evidence now exists demonstrating the unique genetic character of populations of several salmonid species. Complex adaptive differences among populations are likely due to genotypes at many loci (Orr 1998), and native salmonid populations can be considered as genetically distinct stocks that have evolved adaptations to maximize fitness under selection regimes of their local environments. Genetic differentiation among salmonid populations is well documented (Beacham et al. 2001; Beacham et al. 2002; King et al. 2001; Stahl 1987) and in many cases can be correlated with adaptive phenotypes (Beacham and Murray 1987; Beacham et al. 1988; Clarke et al. 1995; Clarke et al. 1994; Groot and Margolis 1991; Quinn 2005; Taylor 1991). Indeed, minimizing the phenotypic expression of the genetic differences between strains may lower the intensity of natural selection working to eliminate the introgression of the new alleles. Faster-growing fish that are now escaping into the environment may not fully show their potential for faster growth, but they will be altering the genetic background of the fish populations nonetheless. There is still limited information regarding the impact of introgression on the ability of a population to evolve adaptations to changing environments, and to cope with further introgression events. 5.5 A C K N O W L E D G E M E N T S We thank M . Lohmus, C. Biagi, G. Harrison, M . Williams, G. Rigter, and N . Hofs for their invaluable assistance with fertilization of the crosses, data collection and rearing of the fish. Funding for these experiments was provided by the Canadian Regulatory System for Biotechnology to RHD. 139 5.6 R E F E R E N C E S Abrahams, M . V. , and A . Sutterlin. 1999. The foraging and antipredator behaviour of growth-enhanced transgenic Atlantic salmon. Animal Behaviour 5:933-942. Allendorf, F. W., and R. S. Waples. 1996. Conservation and genetics of salmonid fishes. Pp. 238-280 in J. C. Avise and J. L. Hamrick, eds. Conservation Genetics: Case Histories from Nature. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. The Quarterly Review of Biology 72:149-177. Ayles, G. B., and R. F. Baker. 1983. Genetic differences in growth and survival between strains and hybrids of rainbow trout (Salmo gairdneri) stocked in aquaculture lakes in the Canadian prairies. Aquaculture 33:269-280. Beacham, T. D., J. R. Candy, K. J. Supernault, T. Ming, A. Schulze, D. Tuck, K. H. Kaukinen, J. R. Irvine, K. M . Miller, and R. E. Withler. 2001. Evaluation and application of microsatellite and major histocampatibiity complex variation for stock identification of coho salmon in British Columbia. Transactions of the American Fisheries Society 130:1116-1149. Beacham, T. D., and C. B. Murray. 1987. Adaptive variation in body size age, morphology, egg size and developmental biology of chum salmon (Oncorhynchus keta) in British Columbia. Can. J. Fish. Aquat. Sci. 44:244-261. Beacham, T. D., K. J. Supernault, M . Wetklo, B. Deagle, K. Labaree, J. R. Irvine, J. R. Candy, K. M . Miller, R. J. Nelson, and R. E. Withler. 2002. The geographic basis for population structure in Fraser river chinook salmon (Oncorhynchus tshawytscha). Fisheries Bulletin 101:229-242. Beacham, T. D., R. E. Withler, C. B. Murray, and L. W. Barner. 1988. Variation in body size, morphology, egg size, and biochemical genetics of pink salmon in British Columbia. Transactions of the American Fisheries Society 117:109-126. Biro, P. A. , M . V . Abrahams, J. R. Post, and E. A . Parkinson. 2004. Predators select against high growth rates and risk-taking behaviour in domestic trout populations. Proceedings of the Royal Society of London B 271:2233-2237. 140 Bryden, C. A. , J. W. Heath, and D. D. Heath. 2004. Performance and heterosis in farmed and wild Chinook salmon (Oncorhynchus tshawytscha) hybrid and purebred crosses. Aquaculture 235:249-261. Burke, J. M . , and M . L. Arnold. 2001. Genetics and the fitness of hybrids. Annual Review of Genetics 35:31-52. Cavalli, L. L. 1952. An analyis of linkage in quantitative inheritance. Pp. 135-144 in E. C. R. Reeve and C. H. Waddington, eds. Quantitative inheritance. Her Majesty's Stationery Office, London. Chase, J. M . , P. A . Abrams, J. P. Grover, S. Diehl, P. Chesson, R. D. Holt, S. A . . Richards, R. M . Nisbet, and T. J. Case. 2002. The interaction between predation and competition: a review and synthesis. Ecology Letters 5:302-315. Clarke, W. C , C. Groot, and L. Margolis. 1995. Physiological Ecology of Pacific Salmon. U B C Press Clarke, W. C , R. E. Withler, and J. E. Shelbourn. 1994. Inheritance of smolting phenotypes in backcrosses of hybrid stream-type x ocean-type Chinook salmon (Oncorhynchus tshawytscha). Estuaries 17:13-25. Clifford, S. L. , P. McGinnity, and A. Ferguson. 1998. Genetic changes in Atlantic salmon (Salmo salar) populations of northwest Irish rivers resulting from escapes of adult farm salmon. Canadian Journal of Fisheries and Aquatic Sciences/Journal Canadien des Sciences Halieutiques et Aquatiques. Ottawa [Can. J. Fish. Aquat. Sci./J. Can. Sci. Halieut. Aquat.] 55:358-363. Devlin, R. H , C. A . Biagi, and T. Y . Yesaki. 2004. Growth, viability and genetic characteristics of G H transgenic coho salmon strains. Aquaculture 236:607-632. Devlin, R. H , J. I. Johnsson, D. E. Smailus, C. A. Biagi, E. Jonsson, and B. T. Bjornsson. 1999. Increased ability to compete for food by growth hormone-transgenic coho salmon Oncorhynchus kisutch (Walbaum). Aquacult. Res. 30:479-482. Dil l , L. M . , and A. H. G. Fraser. 1984. Risk of predation and the feeding behavior of juvenile coho salmon (Oncorhynchus kisutch). Behavioral Ecology and Sociobiology 16:65-71. 141 Einum, S., and I. A . Fleming. 1997. Genetic divergence and interactions in the wild among native, farmed and hybrid Atlantic salmon. Journal of Fish Biology [J. Fish Biol ] 50:634-651. Emms, S. K. , and M ; L. Arnold. 1997. The effect of habitat on parental and hybrid fitness: transplant experiments with Louisiana irises. Evolution 51:1112-1119. Fleming, I. A. , T. Agustsson, B. Finstad, J. I. Johnsson, and B. T. Bjornsson. 2002. Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Sciences 59:1323-1330. Fleming, I. A. , and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES Journal of Marine Science [ICES J. Mar. Sci.] 54:1051-1063. Fleming, I. A. , K. Hindar, I. B. Mjoelneroed, B. Jonsson, T. Balstad, and A. Lamberg. 2000. Lifetime success and interactions of farm salmon invading a native population. Proceedings of the Royal Society of London, Series B: Biological Sciences 267:1517-1523. Friars, G. W., J. K. Bailey, and R. L. Saunders. 1979. Considerations of a method of analyzing diallel crosses of Atlantic salmon. Canadian Journal of Genetics and Cytology 21:121-128. Gharrett, A . J., and W. W. Smoker. 1991. Two generations of hybrids between even-and odd-year pink salmon (Oncorhynchus gorbuscha): A test for outbreeding depression? Canadian Journal of Fisheries and Aquatic Sciences 48:1744-1749. Gharrett, A . J., W. W. Smoker, R. R. Reisenbichler, and S. G. Taylor. 1999. Outbreeding depression in hybrids between odd- and even-broodyear pink salmon. Aquaculture 173:117-130. Gilk, S. E., I. A . Wang, C. L. Hoover, W. W. Smoker, S. G. Taylor, A . K. Gray, and A . J. Gharrett. 2004. Outbreeding depression in hybrids between spatially separated pink salmon, Oncorhynchus gorbuscha, populations: marine survival, homing 142 ability, and variability in family size. Environmental Biology of Fishes 69:287-297. Gjedrem, T. 1983. Genetic variation in quantitative traits and selective breeding in fish and shellfish. Aquaculture 33:51-72. Gjerde, B. 1986. Growth and reproduction in fish and shellfish. Aquaculture 57:37-56. Gjerde, B. 1988. Complete diallel cross between six inbred groups of rainbow trout, Salmo gairnderi. Aquaculture 75:71-87. Gjerde, B., K. Gunnes, and T. Gjedrem. 1983. Effect of inbreeding on survival and growth in rainbow trout. Aquaculture 34:327-332. Gjerde, B., and T. Refstie. 1984. Complete diallel cross between five strains of Atlnatic salmon. Livestock Production Science 11:207-226. Groot, C , and L. Margolis. 1991. Pacific Salmon Life Histories.564. Hatfield, T., and D. Schluter. 1999. Ecological speciation in sticklebacks: environment-dependent hybrid fitness. Evolution 53:866-873. Hayman, B. I. 1960. The separation of epistatic from additive and dominance variation in generation means. Genetica 31:133-146. Hershberger, W. K., J. M . Myers, R. N . Iwamoto, W. C. Macauley, and A. M . Saxton. 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85:187-197. Hojesjo, J., J. I. Johnsson, and M . Axelsson. 1999. Behavioural and heart rate responses to food limitation and predation risk: an experimental study on rainbow trout. Journal of Fish Biology 55:1009-1019. Johnsson, J. I., J. Hojesjo, and I. A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Canadian journal of fisheries and aquatic sciences/Journal canadien des sciences halieutiques et aquatiques. Ottawa ON 58:788-794. Johnsson, J. I., E. Petersson, E. Jonsson, B. T. Bjornsson, and T. Jarvi. 1996. Domestication and growth hormone alter antipredator behaviour and growth 143 patterns in juvenile brown trout, Salmo trutta. Canadian Journal of Fisheries and Aquatic Sciences 53:1546-1554. Jonsson, E., J. I. Johnsson, and B. T. Bjornsson. 1996. Growth hormone increases predation exposure of rainbow trout. Proceedings of the Royal Society of London B:647-651. King, T. L., S. T. Kalinowski, W. B. Schill, A . P. Spidle, and B. A . Lubinski. 2001. Population structure of Atlantic salmon (Salmo salar L.): A range-wide perspective from microsatellite D N A variation. Molecular Ecology 10:807-821. Kondrashov, A. S., and D. Houle. 1994. Genotype-environment interactions and the estimation of the genomic mutation rate in Drosophila melanogaster. Proceedings of the Royal Society of London 258:221-227. Lima, S. L. , and L. M . Di l l . 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology Lynch, M . , and B. Walsh. 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, M A . McClelland, E. K. , J. M . Myers, J. J. Hard, L . K. Park, and K. A. Naish. 2005. Two generations of outbreeding in coho salmon (Oncorhynchus kisutch): effects on size and growth. Canadian Journal of Fisheries and Aquatic Science 62:2538-2547. McGinnity, P., P. Prodohl, A. Ferguson, R. Hynes, N . 6 Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proceedings of the Royal Society of London B 270:2443-2450. McGinnity, P., C. Stone, J. B. Taggart, D. Cooke, D. Cotter, R. Hynes, C. McCamley, T. Cross, and A. Ferguson. 1997. Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: use of D N A profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54:998-1008. 144 Nagy, E. S. 1997. Selection for native characters in hybrids between two locally adapted plant subspecies. Evolution 51:1469-1480. Orr, H. A . 1998. The population genetics of adaptation: the distribution of factors fixed during adaptive evolution. Evolution 52:935-949. Quinn, T. 2005. The Behavior and Ecology of Pacific Salmon and Trout. University of Washington Press Roff, D. A . 1992. The evolution of life histories: theory and analysis. Chapman and Hall, New York. Rundle, H. D., and M . C. Whitlock. 2001. A genetic interpretation of ecologically dependent isolation. Evolution 55:198-201. Smoker, W. W., I. A . Wang, A . J. Gharrett, and J. J. Hard. 2004. Embryo survival and smolt to adult survival in second-generation outbred coho salmon. Journal of Fish Biology 65:254-262. Sogard, S. M . 1997. Size-selective mortality in the juvenile stage of teleost fishes: a review. Bulletin of Marine Science 60:1129-1157. Stahl, G. 1987. Genetic population structure of Atlantic salmon. Pp. 121-140 in N . Ryman and F. Utter, eds. Population genetics and fishery management. University of Washington Press, Seattle. Stearns, S. C. 1992. The evolution of life histories. Oxford University Press, New York. Sundstrom, L. F., R. H. Devlin, J. I. Johnsson, and C. A . Biagi. 2003. Vertical position relects increased feeding motivation in growth hormone transgenic coho salmon (Oncorhynchus kisutch). Ethology 109:701-712. Sundstrom, L. F., M . Lohmus, R. H. Devlin, J. I. Johnsson, C. A . Biagi, and T. Bohlin. 2004a. Feeding on profitable and unprofitable prey: comparing behaviour of growth-enhanced transgenic and normal coho salmon (Oncorhynchus kisutch). Ethology 110:381-396. Sundstrom, L. F., E. Petersson, J. Hojesjo, J. I. Johnsson, and T. Jarvi. 2004b. Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behavioral Ecology 15:192-198. 145 Sundstrom, L. F., E. Petersson, J. I. Johnsson, J. Dannewitz, J. Hojesjo, and T. Jarvi. 2005. Heart rate responses to predation risk in Salmo trutta are affected by the rearing environment. Journal of Fish Biology 67:1280-1286. Taylor, E. B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98:185-207. Templeton, A . R. 1986. Coadaptation and outbreeding depression. Pp. 105-116 in M . E. Soule, ed. Conservation biology: the science of scarcity and diversity. Sinauer, Sunderland, M A . Tymchuk, W. E., C. A . Biagi, R. E. Withler, and R. H. Devlin. 2006. Growth and behavioural consequences of introgression of a domesticate aquaculture genotype into a native strain of coho salmon (Oncorhynchus kisutch). Transactions of the American Fisheries Society 135:442-445. Tymchuk, W. E., and R. H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245:295-300. Wangila, B. C. C , and T. A . Dick. 1996. Genetic effects and growth performance in pure and hybrid strains of rainbow trout, Oncorhynchus mykiss (Walbaum) (order: Salmoniformes, family: Salmonidae). Aquae. Res. 27:35-41. Webster, D. A. , and W. A. Flick. 1981. Performance of indigenous, exotic, and hybrid strains of brook trout (Salvelinus fontinalis ) in waters of the Adirondack Mountains, New York. Canadian Journal of Fisheries and Aquatic Sciences 38:1701-1707. 146 CHAPTER V I 6 : E X P R E S S I O N O F H E T E R O S I S A N D O U T B R E E D I N G D E P R E S S I O N I N T H E W I L D A F T E R T H R E E G E N E R A T I O N S O F I N T R O G R E S S I O N O F F A R M E D R A I N B O W T R O U T I N T O A W I L D P O P U L A T I O N A version of this chapter is being submitted for publication. Tymchuk, W. E., Biro, P. A., and Devlin, R. H . Expression of heterosis and outbreeding depression in the wild after three generations of introgression of farmed rainbow trout into a wild population. 147 6.1 I N T R O D U C T I O N According to optimal foraging theory, the ability to obtain food resources from the environment is an important component of fitness upon which selection will act. Often, however, an individual cannot maximize extraction of food resources from the environment without facing a concurrent increase in risk of predation (Werner et al. 1983; Werner and Gilliam 1984; Lima and Dil l 1990; Werner and Anholt 1993). For example, juvenile salmonids may choose to forage in habitats with abundant food resources, but at a greater risk of predation, or may spend more time foraging which takes time away from hiding from or watching for predators (and thereby increase risk of being consumed by a predator). Prey are known to respond behaviourally to risk of predation (Sih 1987; Lima and Dil l 1990) by for example, increasing their cryptic behaviour or reducing foraging effort. Domestication through adaptation to a culture environment has been shown to alter behavioural responses to risk of predation with domestic fish showing a reduced response to predators (Johnsson and Abrahams 1991; Fleming and Einum 1997; Johnsson et al. 2001; Tymchuk et al. 2006). In fish populations, faster growing fish are often at an advantage when under competition for food resources, since larger fish have a competitive advantage over smaller fish when foraging, and a larger size often corresponds with a lower risk of predation (Werner and Gilliam 1984). In some cases fish may grow so fast, and possess such an advantage that they begin to cannibalize their intraspecific cohorts (Tymchuk, personal observation; Devlin et al. 2004). However, additional investment of time acquiring food resources required by fast-growing fish to maintain their faster growth may also put them at a disadvantage by increasing their exposure to, and mortality from, predators. Predation and competition are two important characteristics of the natural environment that shape optimal life history characteristics of a species. However, which selection pressure is stronger can vary depending on circumstances. For example, there may be a fitness advantage to growing fast and out-competing conspecifics when 148 predation threat is low, and/or future resources will be limited, even though there may be an initial cost of increased mortality due to predation. Alternatively, under high risk of predation and/or very low food availability, slow growth may provide increased fitness by reducing costs associated with foraging, such energy expenditure and predation risk. It is not fully known how the two selection pressures interact to structure the ecology of a fish population which vary in the expression of these traits. For example, when two populations that express two extremes of the growth / predation risk tradeoff encounter each other, how would risk of predation and competitive interactions combine and ultimately resolve over the course of several generations? Studies on larval anuran indicate that risk of predation can alter competitive interactions among prey species through non-lethal mechanisms (Werner 1991). If domestic fish have become adapted to a culture environment, they may respond less to a real risk of predation in the wild, and those which survive would therefore maintain high foraging activity and faster growth. In contrast, wild fish tend to show lower rates of growth as risk of predation increases, spending more time either hiding or foraging in less risky habitats with poor food resources. Consequently, differences in risk of predation and foraging tradeoffs between the domestic and wild fish may increase the size difference between them upon interaction, providing domestic fish with an enhanced competitive advantage. However, this benefit from competition may not always payoff, for example i f risk of predation is so high that few domestic fish actually survive to benefit from enhanced growth and experience improved reproductive fitness. In the present study, we test the fitness consequences of altered rates of growth under selection from competition for food and risk of predation. Specifically, we have utilized strains of rainbow trout (Oncorhynchus mykiss) known to express different rates of growth within both culture and semi-natural environments (Tymchuk and Devlin 2005). This experimental model makes use of pure wild (slow growing) and domestic (fast growing) strains, as well as three backcross generations of intraspecific hybrids between fast-growing (domestic) and slow-growing (wild) trout which were released into two experimental research lakes in nature. Previous work using these lakes found that 149 pure strains of fast-growing domestic fish suffer increased mortality relative to wild fish when risk of predation is high, but not when risk of predation is low (Biro et al. 2004) . This research extends previous work by testing for fitness costs of intermediate rates of growth, as expressed by the domestic x wild hybrids. This information is of relevance for assessing the multigenerational risk of escaped or released domestic fish should they successfully interbreed with wild populations. 6.2 M A T E R I A L S A N D M E T H O D S Strains of rainbow trout expressing different rates of growth have been utilized in the present study to address our research objectives. The genetic basis of growth differences has been established in earlier laboratory-based common-garden experiments, where fast growth among domestic and hybrid trout compared to wild strains was found to be largely controlled by additive genetic variance (Tymchuk and Devlin 2005) . The domestic strain (D) used was the Spring Valley Trout Farm (Langley, BC) strain: B i and B2 backcross progeny (see below) used in this study utilized domestic stock directly from Spring Valley, whereas pure domestic and Fi hybrids used the Spring valley stock collected from Campbell Lake Trout Farm, Little Fort, British Columbia (BC), Canada. The wild strain (W) of trout used was collected from nature from Pennask lake, BC, Canada, for all generations. On 7 June 2005 , five single-pair pure domestic (D) families were fertilized, utilizing five females and five males. On 9 June 2005 , sperm was collected from domestic (D), wild (W), domestic x wild hybrid (Fi), and Fi x wild backcross (Bw) males (six different males for each genotype) reared at Fisheries and Oceans Canada Laboratory (UBC/DFO Centre for Aquaculture and Environmental Research) in West Vancouver, BC (Tymchuk and Devlin 2005) . Eggs were collected from 240 wild-caught Pennask Lake females and fertilized in single-pair crosses, each female being crossed with one of the four possible male genotypes (D, W, F i , Bw). These fertilizations provided family lines representing pure domestic (D, n = 5), pure 150 wild (W, n = 49), domestic x wild hybrids (F\, n = 68), Fj x wild backcrosses (Bi, n = 67) and Bw x wild backcrosses (B2, n = 56). Fertilized eggs were incubated at the laboratory until they had absorbed their yolk at which time they were measured for weight and length and numbers within groups determined. Subsequently, fry were transported to the experimental lake area in bags containing oxygen, within coolers containing melting ice. On 28 July 2005 trout were measured and released the following day as first-feeding fry into two small experimental lakes (PHH and CPI) with similar characteristics (3.3 and 4 ha respectively, maximum depth for both lakes is 18 m) located adjacent to each other in south-central British Columbia, Canada (49°50'- 49°56' N , 120°33' - 120°34* W). Both lakes were assessed on 11 June 2005 to verify that they contained predators. Gil l netting revealed the presence of yearling and older rainbow trout (present from previous year's research) as well as insect predators such as dragon fly nymphs. A more comprehensive estimate of predator density and size was completed during the fall sampling, and these data are included in the results section. Each genotype was stocked at a density of 1000 fry/ha except for the domestic fry which were stocked at 200 fry/ha (due to low availability of eggs when crosses were performed). Fry were left undisturbed within the these lakes to experience natural selective and competitive forces until 8 to 13 October 2005, at which time lethal gillnet sampling for five consecutive nights with a standardized effort was used to estimate the population size and growth of the fry that survived over the summer (Post et al. 1999; Biro et al. 2003a). Since vulnerability to gillnetting increases with body size, the total catch was adjusted for size dependence in recapture probability by an established size-specific vulnerability function determined from previous mark-recapture experiments in the lakes, providing an estimate of the number of survivors for each genotype (Post et al. 1999; Biro et al. 2003a,b; Biro et al. 2004). A tissue sample (adipose fin) was retained from each fish and stored in ethanol. D N A was extracted from fins using DNeasy 96 Tissue Kit (Qiagen) as described by the manufacturer. Genotypes of the survivors were determined by identification of parents based on five polymorphic microsatellite D N A loci [Table 6.1, OMM1234 and T A B L E 6.1 Polymorphic microsatellite loci used to determine parentage of the rainbow trout offspring released into the natural lakes. Loci Unit GenBank Acc. # Primer F Primer R OMM1234 (GATA) 2 2 AF470014 CAGCAGCCACTCAATGTTAG GCCTTGTCTCTCAGATCGTT OMM1270 (AC) 2 7 AF470035 GCCATTTGGGAATCAGAGAGGTTCTACA CGGAAACCCTGACAGGAATACTGAG Ssa407 (GACA) 3 7 AJ402724 TGTGTAGGCAGGTGTGGAC CACTGCTGTTACTTTGGTGATTC Ssa408 (GACA) 2 7 AJ402725 AATGGATTACGGGTACGTTAGACA CTCTTGTGCAGGTTCTTTCATCTGT Ssa417 (GATA) 1 1 4 AJ402734 AGACAGGTCCAGACAAGCACTCA ATCAAATCCACTGGGGTTATACTG 152 OMM1270 (Rexroad III and Palti 2003), Ssa407, Ssa408, and Ssa417 (Caimey et al. 2000)]. Microsatellite amplifications via polymerase chain reactions (PCRs) were performed in 10 ul reactions containing 1 x reaction buffer [20 m M Tris-HCl (pH 8.4), 50 m M KC1], 1.5 m M M g C l 2 , 0.2 m M each dNTP, 0.55 uM each primer, 0.25 units Taq polymerase (Invitrogen), and ~ 10 ng of template DNA. Samples were amplified using an Applied Biosystems Inc. (ABI) GeneAmp® PCR System 2700 thermocycler using touchdown PCR: one cycle of 95°C for 3 min; 10 cycles of 95°C for 30 s, 63°C for 1 min (- 0.5°C per cycle), and 72°C for 1 min; 20 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 30 s; one cycle of 72°C for 20 min. Microsatellites were size fractionated using an ABI 3130x1 automated D N A sequencer. Electrophoretic data were analyzed using GeneMapper® software version 3.7 (ABI). The computer program P R O B M A X (Danzmann 1997) was used to assign parentage of each individual via exclusion technique. Although this technique is more sensitive to genotyping error and mutation, it was chosen since the genotype of every parent was known, and there was potential for some fish captured from the lakes to not have been from our spring release. GENEPOP software (http://wbiomed.curtin.edu.au/Renepop/index.html) was used to calculate observed and expected heterozygosities and to test their conformance to Hardy-Weinberg equilibrium. Groups of trout representative of each genotype were also grown within the laboratory in 10°C well water and fed daily to satiation as described (Tymchuk and Devlin, 2005). Specific growth rate for mass (SGRm ass) and length (SGRi e n gth) was calculated as SGR m a S s or length = [(ln(M2 or Z 2 ) - ln(Mi or L\)) I ( 2^ - ^i)] * 100 where M\ (or L\) and M\ (or L\) represent mass (or length) at t\ and h. Growth rates were calculated using vulnerability-corrected mean mass for genotype (Post et al. 1999; Biro et al. 2003a,b; Biro et al. 2004). Condition factor was calculated as K = WtL,~3 x 100. A l l statistical analyses were performed with SIGMASTAT (Systat Software, Point Richmond, CA). One-way A N O V A with genotype as a fixed factor was used to test for differences among the line crosses in size and growth, using family means as the response variable. Tukey's test was used for unplanned pairwise comparisons between 153 genotypes. Paired student's t-test was used to test for differences in growth between the two lakes. Survival of the genotypes was compared with a G-test using the pure wild crosses as a reference with the null hypothesis that there is no difference in survival among the genotypes. Mortality of the corresponding families raised under culture conditions was minimal, so no corrections were made to survival measured in the wild. 6.3 R E S U L T S Viability from fertilization to hatch varied with genotype (Fig. 6.1, all p < 0.001) with the Fi crosses showing lower viability relative to the W, Bj and B2 crosses, and domestic crosses being intermediate to and not different from other groups. After release into the natural lakes, a total of 75 (0.5% of total released) and 191 (1.1% of total) fry were captured in the fall from CPI and PPH, respectively. For CPI , 61 individuals (81%) were identified according to family, and 158 individuals from PPH (83%>) were likewise genotyped. Individuals that could not be assigned to a parent-pair, or that were assigned to more than one parent-pair, were excluded from the analysis. The number of alleles detected at each loci ranged from 9 to 18 (Table 6.2). For all loci except Ssa408 the observed heterozygosity was significantly different from that expected from Hardy-Weinberg equilibrium. Ssa407, Ssa417, and OMM1234 had lower heterozygosity than expected, while observed heterozygosity of OMM1270 was higher than expected. Survival of the fry in the experimental lakes varied according to genotype (Fig. 6.2). In both lakes, D fry had the highest rates of survival, followed by the W. The hybrid crosses (Fi, Bi and B2) all showed survival rates lower than both D and W (G-test, all p < 0.001). Survival varied according to lake with all genotypes experiencing lower survival in CPI relative to PPH (all p < 0.001). The estimated number of predators within CPI (n = 555, mean length = 26.3 ±8 .1 cm) was approximately three times higher than the number in PPH (n = 190, mean length = 27.4 ±8.1 cm) as determined by gill-netting during the fall sampling period. 1 5 4 T A B L E 6 .2 Observed (Ho) and expected (H£) heterozygosities, probability of conformance to Hardy-Weinberg equilibrium (HWE), and total number of alleles ( N A ) for each loci used in offspring identification. Trait OMM1234 OMM1270 Ssa407 Ssa408 Ssa417 Ho 0 . 8 7 6 0 . 9 2 5 0 . 7 1 5 0 . 8 0 9 0 . 5 3 7 HE 0 . 8 7 6 0 . 8 9 8 0 . 7 3 5 0 . 7 9 9 0 . 7 4 3 HWE 0 . 0 1 2 0 . 0 1 7 0 . 0 0 3 0 . 6 3 5 0 . 0 0 0 N A 18 18 13 1 0 9 155 1.0 7 0.8 -0.0 - L - ^ - . • • . • 1 0 25 50 75 100 % Domestic Alleles F I G U R E 6.1 Relationship between genotype and viability from fertilization to hatch. Points represent the family means and their associated standard error, and different letters indicate significant differences among the means. 156 • CPI D O 0.3 - O PPH ival 0.2 -ival Surv 0.1 -A o • d 0.0 -B • o a • b C o • b B 8 c i 0 1 25 50 75 1 100 % Domestic Alleles F I G U R E 6.2 Estimated over-summer survival of the fry released into the two natural lakes, CPI and PPH. Points represent the proportion of fish surviving from those released 2 0 July 2005 . Letters indicate significant differences in survival among the genotypes (G-test, all p < 0.001). 157 At release into the lakes at 50 days post-fertilization (dpf), only the D fry differed in size from the other genotypes (Table 6.3; reduced mass, F 4 J 4 5 = 11.826, p < 0.001; shorter length, F 4 j 4 5 = 65.728, p < 0.001). At the end of the experiment (121 to 126 dpf), none of the groups differed in mass, length, or condition factor (all p > 0.200). However, since the D fry started at a smaller size, they showed faster rates of growth (Fig. 6.3) in both mass and length (both p < 0.001) relative to all other genotypes. There were no differences in size, condition, or growth of the sampled fry between the two lakes. 6.4 D I S C U S S I O N Results of this experiment indicate that under certain conditions, domestic fish that have been selected within a culture environment for faster rates of growth may be able to survive at least as well as wild fish within the same environment. However, offspring generated from introgression of domestic alleles into a wild population over three generations showed survival rates in the field lower than either parental strain. This indicates that even though the pure domestic fish can successfully avoid mortality from predation or starvation, domestic x wild hybrids and backcrosses may be at a disadvantage. Since the source of mortality cannot be known for certain in the wild, these hybrid offspring may not necessarily have suffered an increase risk of predation, but may have also died due to starvation. The mortality of the different crosses within the culture environment was negligible from first-feeding onward, so there were no intrinsic differences in viability at this stage that may have caused the trend in mortality observed in the field. A previous study comparing growth and mortality between domestic and wild rainbow trout in these same lakes found that the domestic fish showed high survival advantage under no risk of predation, a small survival advantage under an intermediate risk of predation, and lower survival when predation was high (Biro et al. 2004). In the present study, no differences in the relationship among genotypes were observed from 158 T A B L E 6.3 Size of the fry (and associated standard error) at release into the lakes on 2 9 July 2005 , 50 days post-fertilization (dpf) and at the fall sampling 8 to 13 October 2005 (121 to 126 dpf). D fry were smaller than all other genotypes at release. Size of the fry during the fall sampling differed between the two lakes ( P P H vs C P H ) , but there were no significant differences among the genotypes. Genotype D B, B 2 W Mass (g) Length (cm) Initial Fall P P H Fall C P H Initial Fall P P H Fall C P H 0.057 6.43 7.93 2 .09 7.57 5.47 (0.017) (0.45) (0 .21) (0 .03) (0-14) (0-31) 0 .112 7.78 8.42 2 .57 7.44 4 .84 (0.018) (0 .81) (0.29) (0.07) (0 .13) (0.22) 0.094 5.10 7.50 2 .50 7.40 4.88 (0.017) (0.74) (0.35) (0.08) (0 .15) (0.29) 0.093 6.53 7.81 2 .50 6.69 3.54 (0.023) (0.85) (0.35) (0.05) (0.17) (0.27) 0.098 6.37 7.97 2.51 7.11 4 .32 (0.019) (0.57) (0 .21) (0.12) (0.14) (0.30) 159 11.0 10.5 10.0 J 0 25 50 75 100 % Domestic Alleles 0 25 50 75 100 % Domestic Alleles F I G U R E 6.3 Growth in mass (SGRmaSs) a n d length (SGR|ength) of the genotypes reared over summer in two natural lakes, CPI and PPH. Points represent the family means and associated standard error, and letters indicate significant differences among genotypes. 160 lakes experiencing different levels of predation. There are three possible reasons for the observed differences between these two studies. First, the fry used in this study were released at first-feeding stage, whereas Biro et al. (2004) reared the fry in culture conditions until they were approximately 15 cm in length before release, and thus differences in developmental stage and rearing experience exist between the studies. Second, the strains of wild and domestic fish used in the two studies may not have been of identical lineage, and may therefore express different relative phenotypes. Third, the difference in results between these two studies may have arisen from different levels of overall mortality. In this study the overall survival of the fry was only 4% for the high predator lake, and 10% for the low predator lake, whereas in Biro et al. (2004), the overall survival under risk of predation was 38%. In similar studies with Atlantic salmon, lower survival of hybrid and backcross offspring between farm (domestic) and wild fish relative to wild fish has been observed (McGinnity et al. 2003), although in these studies the pure farm parental strain also had reduced survival relative to the wild fish (differing from the present results). Another study found no difference in fitness between farm and wild strain fish in the natural environment (Fleming et al. 2000), suggesting that strain differences and environmental fluctuations are likely playing important roles in influencing the fitness outcomes of introgression between wild and domestic strains of salmonids. It is often argued that if faster rates of growth were advantageous in terms of overall fitness, this characteristic would have evolved within the natural environment (Arendt 1997). Evidence of the ability of fish to express periods of faster growth provides further support that growth rates in nature are restricted to levels below the physiological maximum (Ali et al. 2003). This argument assumes however that any increase in magnitude of growth rate would offer the same advantage or disadvantage to an individual. There may in fact be other optimal growth rates that would incur similar levels of fitness, but perhaps the intermediate steps arriving at that rate actually cause lower fitness. Depending on the shape of the tradeoff curve, it is possible to have, more than one optimal solution (Partridge et al. 1991). For example, i f there is a tradeoff 161 between a fitness advantage due to increased competitive ability and a fitness disadvantage due to increased risk of predation, a fish with a small increase in growth rate may have a higher increase in risk of predation than the benefit they receive by being better competitors. There is perhaps a minimum growth shift that must occur before this tradeoff is balanced. It has been well established for the fish used in this experiment that there is a linear relationship between the proportion of domestic alleles within the genotype and growth (Tymchuk and Devlin 2005) under conditions without competition or predation. Since the hybrid and backcross families tend to show growth rates intermediate to the hybrid strains, perhaps they are no longer able to balance predation risk with competitive advantage, and therefore show reduced survival relative to the parental strains. It would be interesting to determine whether domestic fish express altered phenotypes to the extent that they would occupy a different niche from the wild fish so that they would actually coexist within a natural environment without ecological interaction. Models predict that species which share the same food resources and predators may be able to coexist if, for example, one species is better at acquiring food while the other minimizes risk of predation. Altered balance of the predation risk / foraging gain tradeoff has been implicated in coexistence of fish assemblages (Werner & Hall 1977; McPeek 1996; McPeek et al. 2001). In conclusion, this research shows, for the first time in nature, the relative fitness consequences of introgression beyond first-generation hybrids of domestics alleles into a wild population of rainbow trout. Through the use of microsatellite analysis to identify individuals according to strain, we were able to estimate fitness in nature directly from the fry stage (first-feeding stage) without having to rear fish in culture environments first in order to have them at a size large enough to mark. Since selection due to mortality at this early life-history stage can be quite strong, it is important to remove the influence of the culture environment which this study also has done for the first time. It is clear that the pure domestic strains need not have reduced survival relative to the wild population for introgression to have a negative impact on fitness of the hybrid offspring. Thus, even 162 though the domestic strain showed faster growth and higher survival relative to the wild strain, the hybrids ( F i , B i and B 2 ) showed a reduction in over-summer survival relative to both parental strains. Thus, assuming equal reproductive fitness among the strains, we predict that over the long term, sustained introgression could threaten the conservation of wild populations. The extent of the decline of the wild population will depend on the size and frequency of the release or escape of domestic fish. If there is unequal reproductive fitness among the strains, the genetic threat to the wild population may be reduced, particularly if the domestic fish cannot successfully interbreed with the wild population in the first place, or if the hybrids have such a great reduction in fitness that they are unable to survive to maturity. 163 6.5 R E F E R E N C E S A l i , M . , A . G. Nicieza, and R. J. Wooton. 2003. Compensatory growth in fishes: a response to growth depression. Fish Fish. 4:147-190. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Q. Rev. Biol. 72:149-177. Biro, P. A. , M . V . Abrahams, J. R. Post, and E. A . Parkinson. 2004. Predators select against high growth rates and risk-taking behaviour in domestic trout populations. P. Roy. Soc. Lond. B 271:2233-2237. Biro, P. A. , J. R. Post, and E. A . Parkinson. 2003. From individuals to populations: prey fish risk-taking mediates mortality in whole-system experiments. Ecology 84:2419-2431. Cairney, M . , J. B. Taggart, and B. Hoyheim. 2000. Characterization of microsatellite and minisatellite loci in Atlantic salmon (Salmo salar L.) and cross-species amplication in other salmonids. Mol . Ecol. 9:2155-2234. Danzmann, R. G. 1997. P R O B M A X : A computer program for assigning unknown parentage in pedigree analysis from known genotypic pools of parents and progeny. J. Hered. 88:333. Fleming, I. A. , and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES J. Mar. Sci. 54:1051-1063. Fleming, I. A. , K. Hindar, I. B. Mjoelneroed, B. Jonsson, T. Balstad, and A . Lamberg. 2000. Lifetime success and interactions of farm salmon invading a native population. P. Roy. Soc. Lond. B . 267:1517-1523. Johnsson, J. I., J. Hojesjo, and I. A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can. J. Fish. Aquat. Sci. 58:788-794. Lima, S. L. , and L. M . Di l l . 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68:619-640. 164 McGinnity, P., P. Prodohl, A. Ferguson, R. Hynes, N . 6 Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. P. Roy. Soc. Lond. B 270:2443-2450. McPeek, M . A . 1996. Tradeoffs, food web structure, and the coexistence of habitat specialists and generalists. Am. Nat. 145:S124-S138. McPeek, M . A. , M . Grace, and J. M . L. Richardson. 2001. Physiological and behavioral responses to predators shape the growth/predation rsik trade-off in damselflies. Ecology 82:1535-1545. Nelson, R. J., and T. D. Beacham. 1999. Isolation and cross species amplication of microsatellite loci useful for study of Pacific salmon. Anim. Genet. 30:225-244. Olsen, J. B., P. Bentzen, and J. E. Seeb. 1998. Characterization of eight microsatellite loci derived from pink salmon. Mol. Ecol. 7:1087-1089. Partridge, L. , R. Sibly, R. J. H . Beverton, and W. G. Hil l . 1991. Constraints in the evolution of life histories. Philos. T. Roy. Soc. B 332:3-13. Post, J. R., E. A . Parkinson, and N . T. Johnston. 1999. Density-dependent processes in structured fish populations: assessment of interaction strengths in whole-lake experiments. Ecol. Mono. 69:155-175. Rexroad III, C. E., and Y. Palti. 2003. Development of Ninety-Seven Polymorphic Microsatellite Markers for Rainbow Trout. Trans. Am. Fish. Soc. 132:1214-1221. Sih, A . 1987. Predators and prey lifestyles: an evolutionary and ecological overview in W. C. Kerfoot, and A. Sih, editors. Predation: direct and indirect impacts on aquatic communities. University of New England Press, Hanover, New Hampshire. Tymchuk, W. E., C. A . Biagi, R. E. Withler, and R. H. Devlin. 2006. Growth and behavioural consequences of introgression of a domesticate aquaculture genotype into a native strain of coho salmon (Oncorhynchus kisutch). Trans. Am. Fish. Soc. 135:442-445. 165 Tymchuk, W. E., and R. H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245:295-300. Werner, E. E. 1991. Nonlethal effects of a predator on competitive interactions between two anuran larvae. Ecology 72:1709-1720. Werner, E. E., and B. R. Anholt. 1993. Ecological consequences of the trade-off between growth and mortality rates mediated by foraging activity. Am. Nat. 142:242-272. Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in size-structured populations. Annu. Rev. Ecol. Syst. 15:393-425. Werner, E. E., J. F. Gilliam, D. J. Hall, and G. G. Mittelbach. 1983. An experimental test of the effects of predation risk on habitat use in fish. Ecology 64:1540-1548. Werner, E. E., and D. J. Hall. 1977. Competition and habitat shift in two sunfishes (Centrarchidae). Ecology 58:869-876. 166 CHAPTER V I I 7 : A L T E R A T I O N S I N C I R C U L A T I N G H O R M O N E L E V E L S A R I S I N G T H R O U G H D O M E S T I C A T I O N I N R A I N B O W T R O U T 7 A version of this chapter is being prepared for submission. Tymchuk, W. E. and Devlin, R. H. Alterations in circulating hormone levels arising through domestication in rainbow trout. 167 7.1 INTRODUCTION Growth of an organism involves an increase in cell number (hyperplasia) and / or cell size (hypertrophy). The control of growth is still not completely understood and involves a multifaceted system of regulation influenced by genotype and environment, utilizing cellular controls which are modulated by endocrine signals that act within the body to direct hyperplasia and hypertrophy. The ability of pituitary extracts to increase body growth was first observed in rats over 85 years ago (Evans and Long 1921). Since this time, much research has been conducted in mammals and other animals including fish to elucidate the mechanism by which hormones regulate growth, and how they in turn are regulated (Donaldson et al. 1979). The growth hormone (GH) and insulin-like growth factor I (IGF-I) axis has been highly conserved through evolution of teleost fish and plays a crucial role in their growth (Bjornsson 1997; Duan 1997). Originally, it was thought that G H induces growth indirectly by stimulating the production of IGF-I (Daughaday et al. 1972), but more recent research indicated that G H can also influence growth directly by stimulating cell differentiation, while IGF-I causes cell multiplication (Green et al. 1985); this has be termed the "dual effector theory of action". In fish, the effect of G H on growth has been thoroughly demonstrated through experiments with exogenous supplementation of the hormone (Higgs et al. 1977; Donaldson et al. 1979; Devlin et al. 1994; Johnsson et al. 1996; McLean et al.1996; Bjornsson 1997; Johnsson et al. 1999) and through synthesis of GH-transgenic fish (Zhu et al. 1986; Dunham et al. 1991; Du et al. 1992; Devlin et al. 1994; Martinez et al. 1995; Rhaman etal. 1998; Nam et al. 2001). Exogenous supplementation of G H has been found to increase levels of circulating IGF-I (Moriyama 1995), and positive correlations have been found between growth rate and concentration of plasma IGF-I in Chinook salmon (Beckman et al. 1998). G H and IGF-I are not the only hormones involved in the control of growth, and do not act in isolation from other endocrine pathways. In particular, thyroid hormones have also been implicated to have an important role in the regulation of GH, IGF-I, and 168 growth (Donaldson et al. 1979; Higgs et al. 1982). 3,5,3'-triiodothyronine (T3) is the presumed active form of thyroid hormone (de Luze & Leloup 1984; MacLatchy & Eales 1990). In mammals, T3 is required for mRNA expression of G H in the pituitary (Koenig et al. 1987), and inhibition of T3 with a competitive inhibitor 6-n-propyl-2-thiouracil (PTU) reduces the growth-promoting effects of G H (Kang and Devlin 2004). Further, G H was found to increase plasma T3 in the eel (Anguilla anguilla) (de Luze and Leloup 1984) and in rainbow trout through increased thyroxine (T4) deiodinase activity (MacLatchy and Eales 1990). Thyroid hormone receptors have been found on the liver of rainbow trout (Bres and Eales 1986) and T3 has been found to stimulate hepatic production of IGF-I in tilapia (Schmid et al. 2003). Thus, a clear linkage between thyroid and growth hormones also exists in fish. Domestic rainbow trout (Oncorhynchus mykiss) show increased growth and reduced antipredator response relative to native wild strains (Johnson and Abrahams 1991; Tymchuk and Devlin 2005). Although these traits have been shown to be controlled by a strong genetic component, it is still not clear what physiological mechanisms are mediating these observed differences in phenotype. The purpose of the present experiments were to determine if there have been genetically-based alterations in the G H / IGF-I /thyroid hormone axes that can explain the differences in growth and behaviour observed among the domestic and wild strains of rainbow trout. To increase understanding of how selection may act on endocrine systems, we have also examined the hormone profiles of first- and second-generation hybrids between wild and domestic strains which provide animals which represent intermediate steps in the genetic differentiation of domesticaed strains from wild. This study is also unique because it includes comparisons of both size- and age-matched fish. It is often difficult to test for endocrine differences which influence growth because of consequent body size differences which arise among strains which are correlated with plasma hormone levels independent of inherent strain-specific growth effects. To isolate the influence of genetically-determined hormone differences, plasma was sampled from fish that were the same age (but different size), and from fish that were size-matched (but consequently the 169 slow-growing wi ld fish were older than the domestic fish). Comparisons of these samples allow determination of the effect of size on measured concentration of hormone within the plasma. 7.2 M A T E R I A L S A N D M E T H O D S Strains of O. mykiss, ranging from fast-growing pure domestic (D) to slow-growing pure wi ld (W), were reared at the Fisheries and Oceans Canada and University of British Columbia aquatic laboratory (Centre for Aquaculture and Environmental Research; C A E R ) in West Vancouver, B C . Gametes from domestic parents were obtained from Spring Valley Trout Farm, Langley, B C , and gametes from wi ld parents were collected from nature from Pennask Lake (south-central interior of B C ) . Early maturing individuals from the domestic strain were crossed with late maturing individuals from the wi ld strain on 5 July 1999. Gametes from F i (domestic x wi ld hybrids) were obtained from these hybrid strains cultured at C A E R , and the following crosses were fertilized at C A E R on 26 June 2002: pure wi ld (W, four families) and F i x wi ld backcross ( B w , four families). On 28 October 2002, additional crosses were fertilized: wi ld x domestic (F i , four families); F i x domestic backcross (Ba, two families); and pure domestic (D, four families). It was necessary to perform crosses at two different times since the wi ld and domestic females largely mature in the spring and fall, respectively, and in 2002 (unlike in 1999) no individuals of both strains matured at the same time. Consequently, cryopreserved W and F i sperm, collected 26 June 2002 from wild and hybrid fish, was used on 28 October 2002 in crosses with domesticated females following the method outlined in Wheeler and Thorgaard (1991). To enable comparison of size-matched domestic and wi ld fish, an additional five single-pair crosses of pure domestic strain were fertilized 2 June 2004. A t first feeding, 30 fry were sampled from each family and placed into 200 L tanks. Families were reared individually under culture conditions (well water at a flow no less than 1 L min" 1 kg" 1, constant temperature at 11°C; supplemental aeration) and 170 were fed commercial salmonid feed (Skretting, Vancouver, Canada; composition of feed varied over time to meet the requirements of different life stages) to satiation several times per day. The age-matched fish were sampled on 26 September 2003 and the size-matched fish were sampled on 8 August 2005. Immediately after the fish were terminated (by over-anesthetization in MS222), they were measured in length and mass and a blood sample was taken from the caudal vessels via transection of the caudal peduncle. Blood was kept on ice for < 30 min from collection and then centrifuged at 4°C at 3000 x g for 5 min so that plasma could be collected and stored at -80°C until analysis. Concentration of plasma G H was measured by radioimmunoassay following the protocol outlined by Swanson (1994) and using the chloramine-T method of iodination of the growth hormone. Growth hormone and the required antibodies (salmon/trout growth hormone, salmon/trout growth hormone antiserum, and goat anti-rabbit IgG antiserum) were obtained from GroPep (Adelaide, Australia). IGF-I was assayed with the use of an IGF-I radioimmunoassay kit obtained from GroPep following their recommended protocol. IGF-binding proteins were removed from the serum prior to assay using an acid ethanol extraction protocol. A limited amount of radioactive antigen, together with a known amount of corresponding antibody, was combined with the assay sample. The radiolabeled and unlabelled antigens compete for the same antibody binding sites. A second antiserum precipitates the antibody-bound antigen, separating it from the unbound antigen which remains in solution. After removal of the supernatant, the precipitate was counted in a gamma-counter. The amount of radiolabel present in the precipitate is directly related to the amount of unlabelled antigen present in the original sample. The IGF-I concentration in the samples is calculated from a calibration curve generated using the standards provided. Free triiodothyronine was measured with a competitive enzyme immunoassay obtained from Alpco Diagnostics (25-BC-1006), following the protocol contained within the assay kit. For each hormone, all samples from each collection (age-matched and size-matched) were completed in one assay so that there was no interassay variation in measurements of hormone concentration. 171 7.3 R E S U L T S 7.3.1 Age-matched Fish Size of the age-matched fish varied according to genotype for both mass (F 4 j gs = 43.567, p < 0.001) and length (F 4 , 8 5 = 31.123, p < 0.001) with genotypes composed of more domestic alleles showing larger size (Table 7.1). The growth of the fish showed a similar positive linear relationship with the proportion of domestic alleles within the genotype for both specific growth rate in mass (F 4 >85 - 38.748, p < 0.001) and length (F 4 ; 85 = 1 1.343, p < 0.001). Condition also varied according to genotype (F 4,g5 = 50.203, p < 0.001) with pure domestic fish having the highest condition, and pure wild the lowest. Circulating G H was influenced by genotype (F 4 ; gs = 3.63, p = 0.01), although there was not a linear trend between genotype and hormone concentration. The W, hybrid, and backcross genotypes did not have different concentrations of G H , but the D fish showed significantly higher levels. Concentration of IGF-1 in the plasma varied according to genotype (Fig. 7.1; F 4 > g s = 14.575, p < 0.001) and showed a similar trend as that seen for growth (with pure D having the highest levels, pure W the lowest, and the hybrids and backcrosses showing intermediate concentrations). Similar results were obtained for T 3 (F 4 ] 85 = 3.4, p = 0.021), showing a positive correlation between hormone level and growth rate associated with degree of domestication. There were significant correlations between growth and circulating levels of all three hormones. The correlations were strongest for SGR m a S s and hormone concentration (Fig. 7.2; IGF-I, r 2 = 0.529, p < 0.001; T 3 , r 2 - 0.465, p < 0.001; GH, r 2 = 0.080, p = 0.022) although the correlations with SGRie n gth were also significant (IGF-I, r1 = 0.184, p < 0.001; T 3 , r 2 = 0.247, p < 0.003; GH, r 2 = 0.070, p = 0.031). 172 T A B L E 7.1 Size and growth data at sampling for the age-matched genotypes. Numbers represent the means for each genotype with the associated standard deviation in brackets. Each genotype was represented by five individuals from four different families, with the exception of Bd which had five individuals from two families. Genotype Mass (g) Length (cm) SGRmass SGRlength Condition D B D 103.2 19.8 2 .09 0.68 1.316 (24.2) 54.9 (1.5) 16.1 (0.17) 1.45 (0.05) 0.48 (0 .080) 1.215 (26.9) 48.8 (2.5) 15.4 (0.43) 1.75 (0 .13) 0.53 (0.087) 1.234 r\ B W W (23.6) 36.2 (2-4) 14.3 (0 .31) 1.18 (0 .13) 0 .39 (0 .111) 1.084 (22.6) 18.7 (2.9) 12.2 (0.39) 1.07 (0.12) 0.48 (0 .081) 0.974 (7.4) (1.8) (0 .21) (0.22) (0 .056) 173 2.0 -, 1.5 -1.0 -X 0.5 -0.0 25 50 75 100 % Domestic Alleles CTl CD 25 50 75 % Domestic Alleles 25 50 75 % Domestic Alleles FIGURE 7.1 Relationship between the proportion of domestic alleles within the genotype and (A) concentration of growth hormone (GH), (B) insulin-like growth factor (IGF-1), and (C) thyroid hormone (T3) in the plasma. Points represent the mean and associated standard errors for each genotype which are as follows: wild (W, 0% domestic), Fi x wild (B W ) 25% domestic), wild x domestic (Fi, 50% domestic), Fi x domestic (Bd 75% domestic), and domestic (D, 100%) domestic). 174 o • A 0.0 0.5 1.0 1.5 2.0 2.5 3.0 SGR _ O 0.0 0.5 1.0 1.5 2.0 2.5 3.0 SGR„,„ W Bw F l Bd D W Bw F l Bd D b 81 • w O Bw T F l A Bd • D 0.0 0.5 1.0 1.5 2.0 2.5 3.0 SGRm=„ FIGURE 7.2 Relationship between specific growth rate in mass (SGRmass) and the log of the concentration of GH, IGF-1, and T 3 within the plasma. Groups are as described in Figure 7.1. The regression line includes data from all genotypes. 175 7.3.2 Size-matched Fish For the size-matched fish, there were no differences in length ( D = 24.7 ± 1.0 cm, W = 25.3 ± 2.4 cm, F i i 3 8 = 1.341, p = 0.254), although the domestic fish were larger in mass ( D = 198.1 ± 28.6 g, W = 162.2 ± 50.6 g, F U 8 = 7.635, p = 0.009) due to the different body morphology which exists between the strains (domestic fish did have a higher mean condition relative to the wild fish: 1.315 ± 0.088 for D , 0.972 ±0.012 for W; Fpg = 227.197, p < 0.001). Pure domestic fish were found to have higher levels of circulating IGF-1 (F i > 3 8 = 4.412, p = 0.042) and T3 ( F U 8 = 48.2 52, p < 0.001) than that found in size-matched wild trout (Fig. 7.3). Relative to the previous measurements of hormone levels in plasma from age-matched fish, within strain comparisons of the concentration of IGF-I estimated for the wild fish differed from the previous estimate (t3s = 2.906, p = 0.006), although levels remained the same between sampling groups for the domestic fish (t3g = -0.804, p = 0.426). Estimated circulating levels of T 3 did not vary between age- and size-matched wild (t3g = 1.531, p = 0.139) or domestic fish (t3s = -1.724, p = 0.093). There were no correlations between size (mass and length) and hormone concentration (for both T 3 and IGF-I). 7.4 D I S C U S S I O N Results of this experiment show that increased growth due to presence of domestic alleles within the genome is correlated with increased circulating levels of IGF-I and T 3 . G H was also increased in the pure domestic fish, but the hybrid crosses did not show intermediate levels of hormone that would be predicted by their intermediate rates of growth. The differences in hormone concentration between domestic and wild fish were still present even when the fish were of the same length (although mass and condition was still greater for the domestic fish), indicating that genetic effects rather 176 20 15 cn c i o H A Domestic Wild Domestic Wild F I G U R E 7.3 Concentration of circulating insulin-like growth factor I (A, IGF-I) and free T3 (B) in size-matched domestic and wild rainbow trout. The domestic fish had significantly higher levels of both hormones relative to the wild fish. Bars represent the genotype means and associated standard errors. 177 •.-»**' than body-size effects were largely responsible for these differences. For IGF-I, the wild fish showed higher levels in the plasma when they were sampled at an older age (and therefore large size), although there were no differences between the two sampling periods for the domestic fish even though they were of different sizes. It is possible that the relationship between IGF-I and size or growth is stronger at smaller sizes, and then tends to level off as the fish gets larger. Although supplementation with growth hormone causes increased growth in fish, there are conflicting data reported among correlations between growth rate and growth hormone levels within the plasma. There is some evidence indicating that this genetically-based variability can be related to growth differences among different strains. For example, in Atlantic salmon (Salmo salar), plasma and pituitary G H has been found to vary significantly among strains with a positive correlation with growth rates (Fleming et al. 2002; Nielsen et al. 2001). Although there was not a strong correlation between circulating G H and growth rate in the present experiments, there were differences in plasma concentration of GH between the wild and domestic strains which show the maximum difference in growth rate in this experiment. In contrast, a study comparing circulating G H of domestic strains of rainbow trout with different growth rates did not find differences in concentration of G H between the strains examined, although the difference in growth between these strains was much less than for the strains examined in the present study (Valente et al. 2003). Measuring plasma levels of hormones provides only a snapshot of the hormone profile within the fish. The amount of G H measured in the plasma at the time of sampling is influenced by hormone release (which in rainbow trout is known to be episodic; Gomez et al. 1996) and the clearance rate of hormones from the plasma (which can be influenced by such factors as the number of receptors or hormone binding proteins within the plasma). G H also has a short half-life in circulation of between 20 to 45 minutes. Therefore, it is difficult to get a reliable estimate of G H production and, for example, a low level of G H in the pituitary may indicate a high secretion rate or a low level of GH-synthesis, and a high plasma level may reflect high synthesis or low 178 turnover. Measuring just one aspect of the hormone profile (such as plasma concentration) may not provide a good indication of the whether the hormone is being up-regulated or down-regulated. Therefore, further research that would add valuable information to this study is to measure the expression of GH, growth hormone receptor (GH-R), and IGF-I mRNA within various tissues of different strains of fish. The endocrine system provides a connection between the external environment and an animal's internal physiological response to environmental conditions. As such, hormone action is altered by many factors including temperature, food supply, competition, and life-history stage (Bjornsson 1997; Duan 1997; Power et al. 2001) and the biological action of hormones is crucial for normal animal growth and development. It is likely that hormone levels expressed within an individual are adaptive and therefore play an important role in determining their overall fitness. Consequently, altered endocrine pathways arising from artificial selection within an aquaculture environment may be anticipated to alter fitness in the wild. Hormones not only alter the physiology of an organism, but can also have a profound influence on behaviour, with the relationship between hormones and behaviour being quite complex (Nelson 2000). It is clear that G H supplementation and transgenesis increases appetite and reduces antipredator behaviour in rainbow trout and coho salmon (Johnsson and Bjornsson 1994; Jonsson et al. 1996; Devlin et al. 1999; Sundstrom et al. 2003, 2005; Vandersteen Tymchuk et al. 2005), and is therefore hypothesized to potentially decrease survival within the natural environment where predation pressure is present. However, experiments testing these theories in the wild by supplementing brown trout with exogenous sources of G H have not detected any fitness costs of the higher levels of G H (Johnsson et al. 1999, 2000). An increase in G H may also decrease allocation of energy to lipid reserves (Weatherley and Gi l l 1987; O'Connor et al. 1993) so that fish expressing higher levels of G H may have decreased survival during any period when food resources are low and fish must depend on stored lipid reserves to maintain their metabolic requirements. This expectation may not hold true however if the increase in appetite with increased G H levels (which would lead to an increase in lipid levels) overrides the lipolytic effect of GH. 179 It is clear that endocrine control of growth and behaviour occurs through a complex network of interaction among different hormones, including GH, IGF-I, and T 3 as indicated in this study. Intentional and unintentional selection within a culture environment (domestication) has altered the hormone profiles of rainbow trout when compared with wild conspecifics. Introgression of domestic alleles into a wild background produces fish with hormone profiles that tend to be intermediate to those expressed by the parental strains. These altered hormone profiles may restrict the hybrid individuals from producing physiological and behavioural responses appropriate for their experienced environment, and may provide some explanation for the reduced fitness observed for the hybrid lines when reared in the natural environment (Chapter VI). Further research will be necessary to determine if the altered hormone profiles are the cause, or consequence of, physiological and behavioural differences leading to variability in fitness among the different phenotypes. 180 7.5 R E F E R E N C E S Beckman, B. R., D. A . Larsen, S. Moriyama, B. Lee-Pawlak, and W. D. Dickhoff. 1998. Insulin-like growth factor-1 and environmental modulation of growth during smoltification of spring chinook salmon (Oncorhynchus tshawytscha). Gen. Comp. Endocr. 109:325-335. Bjornsson, B. T. 1997. The biology of salmon growth hormone: from daylight to dominance. Fish Physiol. Biochem. 17:9-24. Bres, 0., and J. G. Eales. 1986. Thyroid hormone binding to isolated trout (Salmo gairdneri) liver nuclei in vitro: binding affinity, capacity, and chemical specificity. Gen. Comp. Endocr. 61:29-39. Daughaday, W. H., K. Hall, M . S. Raben, W. D. Salmon, Jr, J. L . Van den Brande, and J. J. Van Wyk. 1972. Somatomedin: proposed designation for sulphation factor. Nature 235:107. de Luze, A. , and J. Leloup. 1984. Fish growth hormone enhances peripheral conversion of thyroxine to triidothyronine in the eel (Angutila anguilla L.). Gen. Comp. Endocr. 56:308-312. Devlin, R. H , C. A . Biagi, and T. Y . Yesaki. 2004. Growth, viability and genetic characteristics of G H transgenic coho salmon strains. Aquaculture 236:607-632. Devlin, R. H , C. A . Biagi, T. Y . Yesaki, D. E. Smailus, and J. C. Byatt. 2001. Growth of domesticated transgenic fish. Nature 409:781-782. Devlin, R. H , T. Y . Yesaki, C. A . Biagi, E. M . Donaldson, P. Swanson, and W.-K. Chan. 1999. Extraordinary salmon growth. Nature 371:209-210. Devlin, R. H , T. Y. Yesaki, E. M . Donaldson, S. J. Du, and C.-L. Hew. 1995. Production of germline transgenic Pacific salmonids with dramatically increased growth performance. Can. J. Fish. Aquat. Sci. 52:1376-1384. Donaldson, E. M . , U . H. Fagerlund, D. A . Higgs, and J. R. McBride. 1979. Hormonal enhancement of growth in fish. Pages 455-597 in W. S. Hoar, J. D. Randall, and J. 181 R. Brett, editors. Fish Physiology. Vol VIII. Bioenergetics and Growth. Academic Press, New York. Duan, C. 1997. The insulin-like growth factor system and its biological actions in fish. Am. Zool. 37:491-503. Evans, H. M . , and J. A . Long. 1921. The effect of anterior lobe administered intraperitoneally upon growth, maturity and oestrous cycles of the rat. Anat. Rec. 21:62. Gomez, J. M . , T. Boujard, A. Fostier, and P. Y . Le Bail. 1996. Characterization of growth hormone nycthermeral plasma profiles in catheterized rainbow trout (Oncorhynchus mykiss). J. Exp. Zool. 274:171-180. Green, H., M . Morikawa, and T. Nixon. 1985. A dual effector theory of growth-hormone action. Differentiation 29:195-198. Higgs, D. A. , U . H. Fagerlund, J. G. Eales, and A. M . McBride. 1982. Application of thyroid and steroid hormones as anabolic agents in fish culture. Comp. Biochem. Physiol. B. 73:143-176. Johnsson, J. I., and B. T. Bjornsson. 1994. Growth hormone increases growth rate, appetite and dominance in juvenile rainbow trout, Oncorhynchus kisutch. Anim. Behav. 48:177-186. Johnsson, J. I., E.. Joensson, E. Petersson, T. Jaervi, and B. T. Bjoernsson. 2000: Fitness-related effects of growth investment in brown trout under field and hatchery conditions. J. Fish Biol. 57:326-336. Johnsson, J. I., E. Petersson, E. Joensson, B. T. Bjoernsson, and T. Jaervi. 1996. Domestication and growth hormone alter antipredator behaviour and growth patterns in juvenile brown trout, Salmo trutta. Can. J. Fish. Aquat. Sci. 53:1546-1554. Johnsson, J. I., E. Petersson, E. Joensson, T. Jaervi, and B. T. Bjoernsson. 1999. Growth hormone-induced effects on mortality, energy status and growth: a field study on brown trout (Salmo trutta). Funct. Ecol. 13:514-522. 182 Jonsson, E., J. I. Johnsson, and B. T. Bjornsson. 1996. Growth hormone increases predation exposure of rainbow trout. P. Roy. Soc. Lond. B 263:647-651. Kang, D.-Y., and R. H. Devlin. 2004. Effects of 3,5,3'-triiodo-L-thyronine (T3) and 6-n-propyl-2-thiouracil (PTU) on growth of GH-transgenic coho salmon, Oncorhynchus kisutch. Fish Physiol. Biochem. 0:1-9. Koenig, R. J., G. A . Brent, R. L. Warne, P. R. Larsen, and D. D. Moore. 1987. Thyroid hormone receptor binds to a site in the rat growth hormone promoter required for induction by thyroid hormone. P. Natl. Acad. Sci. 84:5670-5674. MacLatchy, D. L. , and J. G. Eales. 1990. Growth hormone stimulates hepatic thyroxine 5'-monodeiodinase activity and 3,5,3'-triiodothyronine levels in rainbow trout, Salmo gairdneri. Gen. Comp. Endocrinol. 78:164-172. Moriyama, S. 1995. Incrased plasma insulin-like growth factor-1 (IGF-1) following oral and intraperitoneal administration of growth hormone to rainbow trout, Oncorhynchus mykiss. Growth Regulat. 5:164-167. Nelson, R. J. 2000. An Introduction to Behavioural Endocrinology, 2 n d ed. Sinauer Associates Inc., Sunderland, Massachusetts, U.S.A. Nielsen, C., G. Holdensgaard, H. C. Petersen, B. T. Bjornsson, and S. S. Madsen. 2001. Genetic differences in physiology, growth hormone levels and migratory behaviour of Atlantic salmon smolts. J. Fish Biol. 59:28-44. O'Connor, P. K. , B. Reich, and M . A . Sheridan. 1993. Growth hormone stimulates hepatic lipid mobilization in rainbow trout, Oncorhynchus mykiss. J. Comp. Phys. 163:427-431. Power, D. M . , L. Llewellyn, M . Faustino, M . A . Nowell, B. T. Bjornsson, I. E. Einarsdottir, A . V . M . Canario, and G. E. Sweeney. 2001. Thyroid hormones in growth and development of fish. Comp. Biochem. Phys. C 130:447-459. Schmid, A. C , I. Lutz, W. Kloas, and M . Reinecke. 2003. Thyroid hormone stimulates hepatic IGF-I mRNA expression in a bony fish, the tilapia Oreochromis mossambicus, in vitro and in vivo. Gen. Comp. Endocr. 130:129-134. 183 Sundstrom, F. L. , R. H. Devlin, J. I. Johnsson, and C. A . Biagi. 2003. Vertical position reflects increased feeding motivation in growth hormone transgenic coho salmon (Oncorhynchus kisutch). Ethology 109:701-712. Sundstrom, L . F., M . Lohmus, and R. H. Devlin. 2005. Selection on increased intrinsic growth rates in coho salmon, Oncorhynchus kisutch. Evolution 59:1560-1569. Swanson, P. 1994. Radioimmunoassay of fish growth hormone, prolactin, and somatolactin. Pages 545-556 in Hochachka, and Mommsen, editors. Biochemistry and Molecular Biology of Fishes. Tymchuk, W. E., and R. H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245:295-300. Valente, L. M . P., P.-Y. Le Bail, E. F. S. Gomes, and B. Fauconneau. 2003. Hormone profile of fast- and slow-growing strains of rainbow trout (Oncorhynchus mykiss) in response to nutritional state. Aquaculture 219:829-839. Vandersteen Tymchuk, W. E., M . V . Abrahams, and R. H. Devlin. 2005. Competitive ability and mortality of growth-enhanced transgenic coho salmon fry and pan-when foraging for food. Trans. Am. Fish. Soc. 134:381-389. Weatherley, A. H. , and H. S. Gi l l 1987. The biology of fish growth. Academic Press, London. Wheeler, P. A. , and G. H . Thorgaard. 1991. Cryopreservation of rainbow trout semen in large straws. Aquaculture 93:95-100. CHAPTER V I I I : G E N E R A L C O N C L U S I O N S 185 8.1 S U M M A R Y O F R E S E A R C H 8.1.1 Assessment of genetically-determined differences in phenotype In Chapter II and Chapter III, I have shown that there are strong linear relationships between phenotype (specifically growth and behaviour) and genotype (proportion of domestic alleles within the genome) for both coho salmon and rainbow trout. It is interesting that both species show these trends, even though they differ in the extent of time in which they have been reared in the culture environment. The more recent domestication of coho salmon is obvious in the reduced magnitude of difference in phenotype between pure domestic and wild strains, relative to that seen between domestic and wild rainbow trout. Results of these experiments suggest that the observed phenotypic differences seen arising through domestication, or selection for enhanced growth, likely occur through small changes of many alleles that work together additively to alter phenotype. The significance of this is that domestic fish that escape or are released into the wild and successfully interbreed with wild fish will tend to have their phenotype diluted over time with further backcrossing to the wild population. However, this assumption is based only on data collected to this point in the laboratory environment, and also would not hold true if there were repeated introgressions of domestic fish into the wild population.. 8.1.2 Role of the environment in expression of heterosis or outbreeding depression The results of these experiments extend the research conducted in the previous objective by trying to simulate more realistically the natural environment to determine i f the linear relationship between phenotype and genotype holds under all environmental conditions. Another important aspect of the experiments in this objective is that the line crosses were extended to a third-generation of hybridization, representing second-generation backcrosses to the wild parental strain. This provides an opportunity to test i f 186 a lack of outbreeding depression observed in other studies holds true even i f the lines are introgressed even further, allowing for more tightly-linked gene complexes to break apart. In Chapter IV the relative fitness of the coho salmon strains was tested when they were reared in semi-natural mesocosms under competition and risk of predation. With the use of microsatellite analysis to identify the offspring to genotype, the fish were able to go into the mesocosms immediately at first-feeding to minimize any impact of the culture environment. It is clear from this experiment that as environmental conditions are more representative of nature, there is much more variability in the phenotype within genotypes, and growth and survival start to deviate from the linear correlations predicted in earlier experiments. There was also evidence of outbreeding depression with the backcross lines showing reduced survival relative to the parental strains. These trends were further supported with the use of the rainbow trout strains in Chapter V. In these experiments, I have attempted to isolate the relative contribution of risk of predation and competition to selection pressures limiting the growth of fish in the natural environment. Under competition, the fastest growing fish had a fitness advantage and were able to survive and grow better (even to the extent of cannibalizing on the slower growing fish). In this environment, there was still a linear relationship between the proportion of domestic genes within the genome and growth and survival. However, when reared under risk of predation, there were signs of outbreeding depression again with the backcross lines showing low levels of survival relative to the parental strains. This led to the question of how competition and risk of predation might interact together on the fitness of the different genotypes, and was the motivation for Chapter VI which presents the results of an experiment conducted in two whole-lakes. Contrary to common assumption, the fast-growing domestic trout did not suffer increased mortality within the natural environment tested, and in fact showed higher rates of survival relative to the wild fish. A l l hybrid strains (Fi, Bwi, and Bw2) showed survival rates lower than both parental strains, indicating some fitness cost of expressing intermediate phenotypes. 187 8.1.3 Test for selection on endocrine control of growth through domestication Differences in growth and behaviour among wild and cultured salmonids are often thought to arise from differences in growth hormone (GH) and / or insulin-like growth factor (IGF-I). In Chapter VII, hormone profiles (GH, IGF-I, and thyroid hormone T 3 ) of the rainbow trout strains used in previous studies correlate with measures of growth, indicating that the faster-growing domestic fish tend to have higher levels of circulating IGF-I, T 3 , and to a lesser extent, GH. Even when domestic and wild strains were size-matched for length, there were still significant differences in the levels of circulating IGF-I and T3 (GH was not measured in these fish). Hybrid strains with rates of growth intermediate to the parental strains were found to have intermediate concentrations of circulating IGF-I and T 3 , although G H levels were not different from the wild parental line. A disruption of an adaptive balance of interacting hormones within an individual may underly their inability to achieve an adaptive tradeoff between growth and survival under natural conditions. For example, one hypothesis generated from this research is that the backcross lines showing reduced fitness relative to the pure wild and domestic strains may have hormone levels that support an increase in metabolism (allowing for faster growth) but do not promote the behavioural phenotype required to meet these increased metabolic demands (such as taking increased risks of predation in order to forage) and thereby suffer increased mortality due to starvation. 8.2 K N O W L E D G E G A P S A N D F U R T H E R R E S E A R C H 8.2.1 Assessing Genetic Differences It is clear that phenotypic differences (particularly growth) between cultured and wild fish are at least partly due to altered genotype. What is not yet understood are the specific genetic changes that have occurred to cause these phenotypic differences. For example, traits that are controlled by many alleles of small effect will present different 188 risks to wild populations, and will require different management strategies, relative to traits that are caused by a small number of alleles of large effect. A better understanding of the genetic changes underlying desired traits will also aid in the development of custom aquaculture strains through the use of marker-assisted selection. Such genetic information may be obtained from 1) further breeding studies (e.g. assessing heteritabilities for critical traits in wild and cultured populations under culture and natural conditions, and the scale to which outbreeding depression and/or heterosis are at play among populations), 2) from experiments mapping and identifying genes responsible for , specific phenotypes, and 3) from gene expression studies identifying candidate genes involved in fitness related processes. 8.2.2 Assessing Consistency of Genetic Differences As an extension of the previous point, it will be interesting to assess if the genetic changes arising through the process of domestication are a conservative process. There has been little comparison among strains and species of cultured fish to determine if the genetic alterations leading to phenotypic differences occur in predictable patterns, or i f each strain is developed through a unique set of alleles. This information will determine if a general risk management strategy could be applied to broad area, or if plans must be developed on a population-by-population basis. Another knowledge gap within this concept is the fact that it is still not clear which genetic changes observed in these and other studies are due to the process of domestication per se, or are due to inherent differences among populations. The only way to test this definitively is to compare a domestic strain against the wild strain from which they were originally derived. Problems arise though with strains that have been selected over long periods of time, such as the rainbow trout, because it is difficult to actually determine the original population. Even if the original wild population is known, it likely would not have remained genetically constant throughout several generations, or may have even become extinct. The experiments described above have tried to address 189 this issue by comparing different wild populations in addition to the domestic population. Although there is significant phenotypic variability among populations, the variability is surpassed in magnitude by the differences between the domestic and wild fish. These comparisons provide some insight into the genetic and phenotypic changes that have arisen due to adaptation to a culture environment. 8.2.3 Contribution of Phenotypic Differences to Fitness Altering the expression of a phenotypic trait can alter overall fitness. Different phenotypic traits will interact in a complex manner to determine the fitness of an individual. While there is an abundant amount of literature on discrete phenotypic differences among cultured and wild strains, there is a need for more complex analyses of how these differences interact during the life-history of the fish and consequently impact their ability to survive and reproduce. It is clear from the experiments described above that focusing on only one trait, such as growth, would not provide a good estimate of the fitness of a population in the wild. Evidence for this is the observation that although in semi-natural mesocosms and in the wild all genotypic groups showed similar rates of growth, their survival rates were significantly different. There is an uncoupling of any relationship between growth and fitness. A critical need is to extend laboratory studies and assess identified genetic differences such that true determinations of their influence on fitness in nature can be determined. It will also be important to examine the ability and the rate that populations may be able to revert back to naturally-selected genotypes and phenotypes following introgression events. 8.2.4 Assessing Plasticity of Phenotypic Differences Theoretical population genetic models indicate that wild-type phenotypes under stabilizing selection are expected to be more canalized with individuals or the populations 190 showing a flat reaction norm that is buffered against genetic variation (Wagner et al. 1997; Stearns and Kawecki 1994). It is often argued that sensitivity of a character to genetic and environmental perturbations is generally negatively correlated with its influence on fitness (Sterns et al. 1995), although this may not always hold true. The ability of domestic fish to grow faster within a culture environment may have resulted from a de-canalization of a wild-type phenotype. Perhaps the selected individuals have increased their phenotypic plasticity so that they are able to do well in a culture environment, yet still maintain the ability to adopt the phenotype most fit for a natural environment (such as reduced feeding under high risk of predation or low food resources). This hypothesis is only beginning to be tested. For example, due to the difficulty of behavioural observations in natural environments, there are few studies that test whether behavioural differences among strains observed in an artificial environment are an accurate predictor of the behavioural characteristics that will be displayed in the natural environment. There is a need for more rigorous assessments of the behavioural plasticity of cultured and wild strains to assess whether domestication has altered the mean and/or the breadth of the reaction norm. Determination of any alterations in phenotypic plasticity caused by domestication would have important consequences to studies examining the fitness impacts of introgression of native and non-native populations. There is limited information on the conquences of genetic interaction between populations that show different levels of phenotypic plasticity. Assessing the reaction norms of hybrid strains will enhance understanding of the expression of different levels of plasticity, and will help to determine the costs and limitations to selection of plastic responses. 8.3 R E L E V A N C E F O R C O N S E R V A T I O N O F P A C I F I C S A L M O N Escaped farmed salmon have made their presence known in the wild, and although their numbers may have a large variation, in general there is a positive correlation between the number of escaped farmed fish in the population and the size of 191 the local aquaculture industry. Farmed fish demonstrate fitness-related differences relative to wild fish including less responsiveness to risk of predation and altered foraging strategies that tend to incur higher metabolic costs. For these reasons, and other considerations as discussed above, farmed salmon tend to have lower survival than wild fish. However, some farmed fish do survive, and evidence indicates that these fish may be capable of successfully reproducing in the wild. As a consequence, the farmed fish may successfully pass along their genetic material with a general trend of lower genetic variability in addition to altered fitness-related traits. A model such as that developed by (Emlen 1991) may be useful to predict the impact of interbreeding between a particular escaped strain and the local wild population. If the hybrids are expected to demonstrate high levels of outbreeding depression, an alternate strain of domestic fish should be considered for the farming operation in order to reduce this impact. For example, by developing and using a life-history model, Byrne et al. (1992) were able to gain some insight on the potential consequences of stocking of hatchery fry and smolts on the native steelhead. The factors influencing the outcome of the supplementation were the number of spawners, the number of stocked fish, the number and fitness of progeny from the stocked fish, and the extent of interbreeding between the two strains. If the fitness of native, stocked and hybrid fish was assumed to be equal, stocking of fish led to replacement of the native fish population, as opposed to enhancement. The outcome of genetic interaction between farmed and wild populations is difficult to predict as our understanding of genetic dynamics is poorly developed and understood for age-structured populations with overlapping generations, such as shown by salmonid populations. Consequently, conservative approaches have been recommended when assessing genetic impact risks (Ryman 1997; Waples 1991). It is clear that an important first step is to minimize escape of cultured fish into the wild as they may pose a genetic risk to wild populations (Altukhov and Salmenkhova 1990; Krueger and May 1991). Effort should also be directed at developing molecular techniques to better identify and monitor introgression of cultured strains into wild 192 populations. The use of triploid fish in aquaculture may eliminate the genetic impact, and reduce the ecological impact, of escaped farmed fish on wild stocks (Cotter et al. 2000; Devlin and Donaldson 1992). The genetic effects on interaction between cultured and wild fish populations will depend on the frequency and magnitude of escapes, their ability to interbreed with conspecifics, the size of the receiving population, and, most critically, the fitness of the domesticated genotypes in nature. The effect of a small introgression of farm alleles into a wild population, based on the present studies, would be anticipated to be diluted with repeated backcrosses in the absence of selection, and little phenotypic effect would be detectable after just two or more generations. In contrast, escapes of large numbers of farmed individuals into small populations, or repeated escapes of moderate numbers over several generations, would be anticipated to have an impact on the phenotype of the receiving population. The critical question, as yet unanswered, is whether natural selection can restore into such introgressed populations genotypes yielding phenotypes and fitness well adapted for current and future environmental conditions. 193 8.4 R E F E R E N C E S Altukhov, Y . P., and E. A . Salmenkhova. 1990. Introductions of distinct stocks of chum salmon, Oncorhynchus keta (Walbaum) into natural populations of the species. J. Fish Biol. 37:25-33. Byrne, A. , T. C. Bjornn, and J. D. Mclntyre. 1992. Modeling the response of native steelhead to hatchery supplementation programs in an Idaho river. North American Journal of Fisheries Management 12:62-78. Cotter, D., V . O'Donovan, N . O'Maoileidigh, G. Rogan, N . Roche, and N . P. Wilkins. 2000. An evaluation of the use of triploid Atlantic salmon (Salmo salar L.) in minimising the impact of escaped farmed salmon on wild populations. Aquaculture 186:61-75. Devlin, R. H. , and E. M . Donaldson. 1992. Containment of genetically altered fish with emphasis on salmonids. In C. L. Hew and G. L. Fletcher, eds. Transgenic Fish. World Scientific, Singapore. Emlen, J. M . 1991. Heterosis and outbreeding depression: a multilocus model and an application to salmon production. Fish. Res. 12:187-212. Krueger, C. C , and B. May. 1991. Ecological and genetic effects of salmonid introductions in North America. Can. J. Fish. Aquat. Sci. 48:66-77. Ryman, N . 1997. Minimizing adverse effects of fish culture: understanding the genetics of populations with overlapping generations. ICES J. Mar. Sci. 54:1149-1159. Stearns, S. C , M . Kaiser, and T. J. Kawecki. 1995. The differential genetic and environmental canalization of fitness components in Drosophila melanogaster, J. Evol. Biol. 8:539-557. Stearns, S. C. and T. J. Kawecki. 1994. Fitness sensitivity and the canalization of life history traits. Evolution. 48:1438-1450. Wagner, G. P., G. Booth, and H. Bagheri-Chaichian. 1997. A population genetic theory of canalization. Evolution. 51:329-347. 194 Waples, R. S. 1991. Genetic interactions between hatchery and wild salmonids: lessons from the Pacific Northwest. Canadian Journal of Fisheries and Aquatic Sciences 48:124-133. 195 A P P E N D I X A The role of genotype and environment in phenotypic differentiation among wild and cultured salmonids Wendy E. Tymchuk1'2, Robert H. Devlin 1 ' 2 Ruth E. Withler3 Centre for Aquaculture and Environmental Research, Fisheries & Oceans Canada, 4160 Marine Drive, West Vancouver, BC, Canada V 7 V 1N6 department of Zoology, University of British Columbia, Vancouver, B C , Canada V6T 1Z4 fisheries & Oceans Canada, 4160 Hammond Bay Road, Nanaimo, BC, Canada V9T 6N7 196 Executive Summary This paper reviews the existing literature examining genetic influences on and consequences of the interaction between cultured and wild salmonids. The paper identifies major phenotypic changes that have occurred in domestic strains (e.g. morphology, physiology, and behavior), and examines whether these changes have effects on fitness in laboratory and natural environments. Long-term effects of interactions between domestic and wild strains will primarily arise from genetic effects, but the phenotype of domestic strains relative to wild strains arises from both genetic and environmental forces. Separating these causal components ofphenotype is required to understand the potential effects of introgression events, yet achieving this goal remains a difficult task. Studies in the wild are required to fully determine the fitness of domestic and wild strains and thus examine potential long-term consequences arising from their interaction. Genotype, in addition to environment, determines the adaptive phenotypic characteristics (i.e. reproductive capabilities and ongoing survival) of salmonids, and, as such, it is likely that disruption of this genetic structure may have short-term and long-term effects on individual fitness as well as the future resilience of populations to natural and anthropogenic pressures. Domestication has been noted to have a significant effect on life history traits in salmonids (Thorpe 2004). Domestication may select for many different traits, including improved growth rates, earlier age at maturity and spawning, greater survival, increased tolerance to high temperature and resistance to disease (Hynes et al. 1981). Differences between wild and cultured fish represent a phenotypic continuum, ranging from differences among natural strains, to differences between wild and sea-ranched fish, to differences between wild and highly selected domestic cultured fish (Figure 1). Alterations in fitness-related traits in hatchery fish should be typical of differences expected in aquacultured salmon, although the latter may show a greater magnitude of change due to an increased length of time under intentional and indirect selection, which is usually conducted in isolation from wild genetic pools. Accumulated evidence now indicates that some fitness-related traits affected by domestication, such as growth, competitive ability, and anti-predator behavior, are in part genetically controlled. Transgenic fish, which can be viewed as an extreme form of domestication, are not considered in the present discussion except when examined as a model system for assessment of genotype/phenotype relationships (Devlin et al. 2001). 197 Figure 1. Representation of the relationships among phenotypic states for one hypothetical set of wild, hatchery, domestic, and transgenic strains. The ranges of phenotype observed among different wild populations are anticipated to overlap considerably with each other and with hatchery strains derived from them. Domestic strains that have undergone directed and unintentional selection are expected to possess phenotypes different from wild and hatchery populations, often in a common direction away from the wild phenotype, and in some cases to an extent seen for transgenic strains. Transgenic strains can possess a wide range of phenotypic transformations for novel traits, and for existing traits previously possessed by the host strain, or may have no change in phenotype from the host strain. Phenotypic differences between cultured and wild fish Rearing fish in a culture environment can lead to environmentally determined differences in morphology relative to those reared in the wild. The extent of these differences depends on the type and the length of time spent within the artificial environment, and the intensity of the culture conditions such as crowding, food supply, etc. Multiple generations of strains kept within the culture environment may lead to genetically based morphological differences arising from selection for traits affording fitness benefits in culture. Domestication has also been shown to alter the physiology offish. Environmental factors such as availability offood resources and temperature will of course have an effect on growth of fish. However, there can also be large differences in growth between cultured and wild strains as a result of genetic differences between the strains. The magnitude of the growth differences caused by genotype will be dependent on the purpose and history of the cultured strain. Aquacultured strains that have been intensely selected for enhanced growth show a larger shift in growth phenotype from the founding line compared with those strains that have not 198 experienced directed selection. It is important to note that it is often not clear whether physiological differences are a cause or consequence of other phenotypic differences between the strains (such as growth or behavior differences). It is therefore difficult to clarify whether physiological differences have a genetic basis per se, or if they are a product of the environment. Behavioral differences commonly arise during domestication. Cultured and wild fish do show differences in the level of aggression displayed towards conspecifics, although there has not been a consistent trend as to whether aggression increases or decreases under culture. A common assertion is that aggression will decrease under culture when fish are reared in crowded conditions and do not have to fight for limited food resources. A genetically determined reduced response to predators seems to be a consistent trend in domestic strains across several species. In contrast, little research has been conducted to reveal genetic control of foraging strategy, habitat selection, and dispersal. A genetic basis for altered foraging strategy could arise from phenotypic expression of other genetically influenced traits such as growth or morphology, which would drive foraging behavior characteristics. Expression of physiological and behavioral phenotypes will ultimately determine survival. Survival is influenced by most other phenotypic traits, and the environment in which they are expressed. Cultured fish, either through a plastic response to their environment or through an adaptive response to altered selection pressures, tend to express phenotypic characteristics best suited for the culture environment. Consequently, they tend to have a lower survival than wild fish in a natural environment. However, few studies have examined whether cultured fish that experience a natural environment throughout their life history will still show decreased survival relative to the wild fish. Furthermore, the strength of the genetic basis of survival is not known, nor is it clear whether cultured fish still have the ability to show a phenotypically plastic response to the environment that will maximize their ability to survive. Reproductive capabilities of domesticated fish are often affected. The literature consistently observes that cultured fish often have the physiological ability to spawn, but that altered spawning behavior limits their success. While the reproductive success offarmed fish may be low, the potential for significant gene flow still exists because the population offarmed fish often outnumbers the population of resident wild fish (at least in the case for Atlantic salmon), at times by as much as 3:1 (Lund et al. 1994; Lura and 0kland 1994). There are no data comparing the ability of farmed and wild Pacific salmon to spawn in nature, but comparisons between hatchery and wild coho salmon, and studies examining cultured wild strains indicate that trends observed for Atlantic salmon may be typical of the phenotypic changes expected during domestication. Genetic effects offarmed fish on wild populations would depend in part on the reproductive behavior offarmed fish in the wild. Evidence suggests that farmed fish have the ability to breed successfully in the wild, although contradicting results exist. There are generally significant differences in breeding potential between cultured and wild fish (Fleming and Gross 1992, 1993; Fleming et al. 1996; Berejikian et al. 1997; Bessey et al. 2004), although other studies have found similar reproductive success for hatchery and native fish in the wild (Dannewitz et al. 2004; Palm et al. 2003). Morphology and life history traits related to reproductive behavior respond evolutionarily to altered selection regime in the hatchery environment (Fleming 1994; Fleming and Gross 1989). The genetic effect of aquacultured salmon on wild populations will depend not only on the size of the wild population, but also on variation in breeding success (Fleming and Petersson 2001). 199 Cause of phenotypic differentiation between cultured and wild strains Phenotypic differences between farmed and wild salmonids may arise from a combination of genetic and environmental effects, but in most cases, the origin of the difference is not well defined. Environmentally based phenotypic differences would not be passed to offspring as they do not have a genetic basis, and are thus anticipated to have single generation effects arising directly from escaped fish. In contrast, genetic differences have the potential to affect the wild populations of a species over a longer time frame. Thus, it is therefore critical to separate the influence of genotype and environment. To assess genetic effects, experiments must be performed by rearing fish of different origins in a common environment (i.e. common-garden experiments.). Such experiments can help determine whether cultured fish have an altered genotype that has arisen in response to selection pressures from an artificial environment. Environmental effects (i.e. phenotypic plasticity) can be tested by rearing fish of a common genetic background in different environments, revealing whether phenotypic plasticity (Hutchings 2004) may have altered phenotype in response to the environmental conditions. Currently, there is still limited knowledge on how the environment will act on inherent genetic differences among strains (i.e. will environmental conditions affect different genotypes in distinct ways through genotype x environment interactions). For example, fast-growing domestic fish may have a greater growth advantage relative to wild fish under culture conditions than they do in nature. An understanding of genotype by environment interactions remains one of the most critical components influencing phenotype and fitness. Research in this area is required to improve prediction of genetic effects arising from interaction between wild and cultured fish. Mechanism of genetic interaction Genetic effects of domestic fish may be direct or indirect. Direct genetic effects include the alteration of the wild genome (introgression) as a result of interbreeding between wild and domesticated fish, or the production of sterile hybrids. Indirect effects include the effect of reduced effective population size or altered selection pressure arising from competition or the introduction of pathogens (Krueger and May 1991; Skaala et al. 1990; Waples 1991). Genetic effects of hybridization between farmed and wild salmon are somewhat unpredictable and may differ between populations, but most interactions have been generally found to be disadvantageous when the genetic effects alter fitness-related traits (Hindar et al. 1991). Most studies have focused on the fitness of the Fi generation when exploring the effects of interbreeding between domestic and wild strains. While such hybrids may have enhanced fitness due to hybrid vigor, the negative effects of outbreeding depression are not manifested until F2 and later generations, and thus simple first-generation hybrid studies have limited predictive value. The genetic effect of escaped cultured fish on wild populations will also depend on the demographic of the wild population, the magnitude and frequency of the escape, and the extent of introgression of aquacultured genotypes into the wild population (Hutchings 1991). The phenotype of wild and farmed hybrids may vary depending on the source and genetic structure of the wild population (e.g., see Einum and Fleming 1997). Anadromous populations of salmonids may be somewhat resistant to introgression due to aspects of their complex life histories such as overlapping maturation age classes and straying among distinct populations (Utter and Epifanio 2002). Furthermore, genetic distance between the two populations does not seem to be a reliable indicator of the potential effects of introgression (Utter and Epifanio 2002). 200 KNOWLEDGE GAPS AND RECOMMENDATIONS • More fully define the genetic basis of domestic traits and the mechanisms by which they alter phenotype. It is clear that phenotypic differences (particularly growth) between cultured and wild fish are due in part to altered genotype. However, the specific genetic changes that have occurred to cause these phenotypic differences are not yet understood. For example, traits that are controlled by many alleles of small effect will present different risks to wild populations and will require different management strategies relative to traits that are caused by a small number of alleles of large effect. A better understanding of the genetic changes underlying desired traits will also aid in the development of custom aquaculture strains through the use of marker-assisted selection. Such genetic information may be obtained from: 1) additional breeding studies (e.g. assessing heteritabilities for critical traits in wild and cultured populations under culture and natural conditions, and the scale to which outbreeding depression and/or heterosis are at play among populations); 2) experiments mapping and identifying genes and alleles responsible for specific phenotypes; and, 3) gene expression studies identifying candidate genes involved in fitness-related processes. • Determine whether conserved genetic and physiological pathways are employed among domestic strains to achieve alteration of specific trait. Further to the above, it will be crucial to assess whether genetic changes arising through the process of domestication are a conservative process. There has been little comparison among strains and species of cultured fish to determine if the genetic alterations leading to phenotypic differences occur in predictable patterns, or i f each strain is developed through a unique set of alleles. This information will determine whether a general risk management strategy could be generalized, or i f plans must be developed on a case-by-case basis. • Extensive research is required to determine which environmental variables play controlling roles in influencing the magnitudes of phenotypic differences among wild and between wild and domestic strains (i.e. improve our knowledge of phenotypic plasticity and genotype x environment interactions). Because of the difficulty of making observations in natural environments, there are few studies that test whether differences among strains observed in an artificial environment are an accurate predictor of the characteristics that will be displayed in the natural environment. Thus, there is a need for more rigorous assessments of the plasticity of cultured and wild strains to assess whether domestic genotypes have response to environmental conditions which differ from wild type in non-parallel ways (i.e. genotype x environment interactions). This area of research is critical. 201 • Undertake experiments to evaluate the contribution of phenotypic differences between domestic and wild strains to survival and reproductive fitness. Altering the expression of a phenotypic trait can alter overall fitness. Different phenotypic traits will interact in a complex manner to determine the fitness of an individual. While there is much literature on discrete phenotypic differences among cultured and wild strains, there is a need for more complex analyses of how these differences interact during the life history of the fish and consequently influence their ability to survive and reproduce. • Fitness evaluations must be undertaken in nature to provide information to reliably predict net fitness and consequences of domestic genotypes introgressed into wild populations. Without data from nature, laboratory experiments may reveal forces causing phenotypic and fitness differences, but their true magnitudes cannot be known with certainty. It is critical to extend laboratory studies and assess identified genetic differences such that true determinations of their influence on fitness in nature can be determined. It will also be important to examine the ability and the rate that populations may be able to revert to naturally selected genotypes and phenotypes following introgression events. • Given current uncertainty in our ability to a priori predict consequences of introgression, research directed to monitoring and minimizing interactions should be supported. The outcome of genetic interaction between farmed and wild populations is difficult to predict as our understanding of genetic dynamics is poorly developed for age-structured populations with overlapping generations such as those shown by salmonid populations. Consequently, conservative approaches have been recommended when assessing genetic effect risks (Ryman 1997; Waples 1991). Clearly, an important first step is to minimize escape of cultured fish into the wild (Altukhov and Salmenkhova 1990; Krueger and May 1991). Effort should also be directed at developing molecular techniques to better identify and monitor introgression of cultured strains into wild populations, particularly for reproductively mature stages and consequent early stages of their progeny. The use of triploid fish or other containment techniques in aquaculture may eliminate genetic effects, and reduce the ecological consequences of escaped farmed fish on wild stocks (Cotter et al. 2000; Devlin and Donaldson 1992). • Develop models that make use of the emerging understanding of the relationship between genotype, phenotype and fitness to allow prediction of the consequences of introgression of domestic and wild strain. Recent research has revealed that many phenotypic traits that differ between wild and domestic strains are controlled by additive genetic variation (Tymchuk et al. 2006, McGinnity et al. 1997, 2003, Fleming et al. 2000). These observations could now allow estimation of the effects of introgression on the genotype of wild populations, assuming neutral fitness. Further, modeling exercises can allow sensitivity analysis to estimate risk arising from different genotypes under various introgression scenarios, and, coupled with studies of natural fitness among genotypes, may be used in the future to predict consequences in the wild. 202 INTRODUCTION Demand for food from a burgeoning human population, coupled with natural and anthropogenic environmental effects on wild fish stocks, is shifting focus on seafood supply from natural fisheries to aquaculture and hatchery programs. This trend is global. In 2001 aquaculture accounted for approximately 37.4% (59.7 million metric tons) of the world's food supply from aquatic sources (FAO 2002). In Canada in 2002, aquaculture accounted for 14.2% (171 thousand metric tons) of the total volume and 22.3% ($628 M) of the value of aquatic food production (DFO 2002) indicating the importance of this sector for food production. World trends continue to forecast growth of the aquaculture industry, with a focus on finfish, at rates greater than for other food-production sectors. In Canada the majority of cultured fisheries production arises from salmon farming on both coasts and in inland facilities, and from hatchery programs on the Pacific coast. HISTORY OF WILD AND CULTURED SALMONID INTERACTIONS Salmon culture began in Europe in the last millennium and in North America in the middle of the 17th century, but intensive application of salmonid enhancement and aquaculture only began in the 1960s. Since then, production of salmonids has expanded dramatically in many countries in North America, Europe, Australasia, and South America, providing significant economic benefit to those regions. A large portion of salmon culture is conducted in net pens in marine environments, but cultured fish are also reared in hatcheries. Cultured fish may enter the natural environment by purposeful release for conservation and enhancement purposes, or they may escape from aquaculture facilities. Over the past 20 years, fisheries biologists have noticed that escaped farmed salmon were present in wild populations (Carr et al. 1997a; Clifford et al. 1998; Crozier 1993; Gudjonsson 1991; Hansen et al. 1987; Jacobsen et al. 1992; Lura and Saegrov 1991; McKinnell et al. 1997; Morton and Volpe 2002; Volpe et al. 2000). Concern has emerged regarding the concurrent genetic and ecological consequences of this interaction arising from differences in phenotype and genotype between cultured and wild fish populations (Crossman 1991; Fleming et al. 1996; Hansen et al. 1997; Hindar et al. 1991). Early reviews of the literature on the outcome of interaction between introduced and native fish populations suggested that introductions have usually been harmful to the native fish populations (Allendorf 1991; Hindar et al. 1991), but see (Peterson 1999). PHENOTYPIC CHARACTERISTICS OF CULTURED SALMON Scope of the issue Domestication can have a significant influence on life history traits in salmonids (Thorpe 2004); see Table 1 for a summary of documented phenotypic differences between domestic and wild fish. Many fitness-related traits such as growth, competitive ability, and anti-predator behavior have been found to have a genetic component. Domestication may select for traits 203 related to improved growth rates, earlier age at maturity and spawning, greater survival, increased tolerance to high temperature, and resistance to disease (Hynes et al. 1981). Differences between wild and cultured fish represent a continuum, ranging from differences among natural strains, to differences between wild and sea-ranched fish, to differences between wild and highly selected and domesticated cultured fish. Alterations in fitness-related traits in hatchery fish should be typical of differences expected in aquacultured salmon, although the latter may show a greater magnitude of change due to an increased length of time under intentional and indirect selection, which is usually conducted in isolation from wild genetic pools. Transgenic fish, which can be viewed as an extreme form of domestication, are not considered in the present discussion except when examined as a model system for assessment of genotype/phenotype relationships (Devlin et al. 2001). The phenotypic differences found between cultured and wild strains may not only be due to genetic differentiation between the strains. They may also occur as a consequence of the different environments in which the fish are reared. Few studies have definitively assessed the relative contribution of genotype and environment. This complicates efforts to find trends in the growing body of literature on phenotypic differences between wild and farmed salmonids. Additionally, even when it is known that a trait is due to genetic differentiation between the strains, there is still a lack of knowledge regarding the expression of the different genotypes in different environments (i.e. genotype by environment interaction). Table 1. Summary of phenotypic differences between cultured and wild fish. T R A I T RESULT ENVIRONMENT E G S P E C I E S S O U R C E Morphology Altered Culture + + Atlantic salmon Fie ming & Einum 1997 Altered Culture + 0 Atlantic salmon Fie ming et al. 1994 Altered Wild + + Coho salmon Swain et al. 1991 Growth Increased Wild + + Rainbow trout Ayles 1975 Increased Wild + + Rainbow trout Ayles & Baker 1983 Increased Culture 0 + Atlantic salmon Einu m & Fleming 1997 Wild Increased Culture 0 + Atlantic salmon Fie ming & Einum 1997 Decreased Semi-Natural 0 + Atlantic salmon Fie ming & Einum 1997 Increased Wild 0 + Atlantic salmon Fie ming et al. 2000 Increased Culture 0 + Atlantic salmon Fie ming et al. 2002 Increased Culture 0 + Brook trout Flick & Webster 1964 No difference Semi-Natural 0 - Brook trout Flick & Webster 1964 Increased Culture 0 + Coho salmon Hershberger et al. 1990 Decreased Wild + + Brook trout Keller & Plosila 1981 Increased Wild + + Brook trout Lachance & Magnan 1990 Increased Culture + + Brook trout Mason et al. 1967 Decreased Wild + + Brook trout Mason et al. 1967 Increased Culture 0 + Coho salmon McClelland et al. 2005 Increased Wild 0 + Atlantic salmon McGin nity et al. 1997 Increased Wild 0 + Atlantic salmon McGin nity et al. 2003 Increased Culture 0 + Rainbow trout Tymchuk & Devlin 2005 Increased Culture 0 + Coho salmon Tymchuk et al. in press Increased Semi-natural 0 + Coho salmon Tymchuk et al. in press Metabolism & Altered Culture + + Brown trout Carline & Machung 2001 Physiology Rainbow trout No difference Culture 0 - Atlantic salmon Du nmall & Schreer 2003 204 TRAIT RESULT ENVIRONMENT EG SPECIES S O U R C E Altered Culture 0 + Atlantic salmon Fie ming et al. 2002 Altered Culture 0 + Atlantic salmon Handeland et al. 2003 Altered Culture + + Atlantic salmon Pop pe et al. 2003 Altered Culture + + Atlantic salmon Pop pe et al. 1997 No difference Culture 0 - Brown trout Sanchez et al. 2001 Altered Culture 0 + Atlantic salmon Thodesen et al. 199 9 Aggression Increased Culture 0 + Atlantic salmon Einu m & Fleming 1997 Increased Culture 0 + Atlantic salmon Fie ming & Einum 1997 Decreased Semi-natural 0 + Atlantic salmon Fie ming & Einum 1997 Increased Wild 0 + Atlantic salmon McGin nity et al. 2003 Increased Wild 0 + Atlantic salmon McGin nity et al. 1997 No difference Culture + - Atlantic salmon Mork et al. 1999 Heterogenous substrate Decreased Culture + + Atlantic salmon Mork et al. 1999 Homogenous substrate Anti-predator Decreased Culture 0 + Atlantic salmon Einu m & Fleming 1997 behavior Decreased Culture 0 + Atlantic salmon Fie ming & Einum 1997 No difference Semi-natural 0 - Atlantic salmon Fie ming & Einum 1997 Decreased Culture 0 + Steelhead trout Johnsson & Abrahams 1991 Decreased Culture 0 + Atlantic salmon Johnsso n et al. 2001 Decreased Culture 0 + Coho salmon Tymchuk et al. in press Foraging behavior No difference Wild - - Atlantic salmon Jacobsen & Han sen 2001 Habitat selection & No difference Wild 0 - Atlantic salmon Einu m & Fleming 1997 Dispersal Altered Culture + + Atlantic salmon Mork et al. 1999 Altered Semi-natural + + Masu salmon Nagata et al. 1994 Survival Decreased Wild + + Brown trout Aerestrup et al. 2000 No difference Wild 0 - Atlantic salmon Einu m & Fleming 1997 Decreased Wild, summer 0 + Brook trout Flick & Webster 1964 No difference Wild, winter 0 - Brook trout Flick & Webster 1964 Variable Wild + + Brook trout Fraser 1981 Decreased Wild + + Brook trout Keller & Plosila 1981 Decreased •Wild + + Brook trout Lachance & Magnan 1990 Increased Wild, winter + + Brook trout Mason et al. 1967 Decreased Wild, summer + + Brook trout Mason et al. 1967 Decreased Wild 0 + Atlantic salmon McGin nity et al. 2003 Decreased Wild 0 + Atlantic salmon McGin nity et al. 1997 Migration Altered Wild + + Atlantic salmon Heg gberget et al. 1993 Altered Wild + + Atlantic salmon Webb et al. 1991 205 TRAIT RESULT ENVIRONMENT EG SPECIES SOURCE. Reproductive Altered Wild ' + + Brook trout Lachance & Magnan Physiology 1990 No difference Culture 0 Chinook salmon Bryden et al. 2004 Spawning behav iour Altered Wild 0 + Atlantic salmon Fie ming et al. 2000 Altered Semi-natural + +- Atlantic salmon Fie ming et al. 1996 Altered Wild + + Atlantic salmon L uraetal. 1993 Altered Wild + + Atlantic salmon Okland etal. 1995 Altered Wild + + Atlantic salmon SEE grov et al. 1997 Altered Semi-natural 0 + Atlantic salmon Weir etal. 2004 Reproductive Decreased Wild + + Atlantic salmon Cli ffordetal. 1998a,b success High density Increased Wild + + Atlantic salmon Cli ffordetal. 1998a,b Low density Decreased Wild 0 + Atlantic salmon Fie ming et al. 2000 Decreased Semi-natural + + Atlantic salmon Fie ming et al. 1996 Decreased Wild + + Atlantic salmon Sas grov et al. 1997 High density No difference Wild - - Atlantic salmon Sas grov et al. 1997 Low density Decreased Semi-natural + + Atlantic salmon Weir et al. 20 04 The result is the observed characteristic of the farmed fish in relation to the wild fish. The cause is due to environment (E) or genotype (G) and was either found to exist (+), not exist (-) or was not tested (0). An environmental cause for phenotypic differences refers to the environment experienced by the fish prior to the experiment. A positive contribution for both E and G indicates either that the experimental design could not isolate the contribution of one factor, or that both factors affected phenotype. Morphological and biochemical differences Morphological differences between cultured and wild salmon have been reported (Table 2) but in most cases, environment rather than genetics was found to have the most influence on these characteristics (Fleming et al. 1994; Kazakov and Semenova 1986; Swain et al. 1991). Fleming et al. (1994) and Einum and Fleming (2001) have noted that some morphological features associated with juvenile hatchery fish persist into the marine phase, whereas others do not. Fleming and Einum (1997) did find morphological differences between farmed and wild Atlantic salmon (farmed individuals had a more robust body and smaller rayed fins) that were caused by genetic differences since all fish were reared in the same environment for the experiment (common-garden experiment). 206 Table 2. Morphological differences between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon F more robust with smaller rayed fins relativeFleming & Einum 1997 to W F had smaller heads and fins and narrower Fleming et al. 1994 caudal peduncles; F were more robust with smaller rayed fins F = farmed, D = domestic, Hy = hybrid, W = wild Thus, rearing fish in a culture environment can lead to environmentally determined differences in morphology relative to fish reared in the wild. The extent of these differences will depend on the length of time spent within the artificial environment, and the intensity of the culture conditions (such as crowding, food supply, etc.). Multiple generations of strains kept within the culture environment could lead to genetically based morphological differences arising from selection for traits affording fitness benefits in culture. Emergence timing and embryonic growth There are few studies on the genetic basis of differences in development rate between cultured and wild fish (Table 3). Differences in development rate have been noted between hatchery and wild strains of rainbow trout (Ferguson et al. 1985) and it has been determined that development rate is controlled by at least one major locus in this species (Robison et al. 2001; Robison et al. 1999). The actual genes affecting this process in domesticated fish are not yet known, although the expression of growth hormone genes in transgenic fish has been shown to significantly affect development rate with consequent effects on survival fitness (Devlin et al. 2004; Sundstrom et al. 2005). Table 3. Differences in development between cultured and wild salmonids. SPECIES RESULT SOURCE Rainbow trout Controlled by at least one major locus Robison etal. 199, 2001 We need to increase our understanding of the influence of domestication on embryonic development rates because timing of emergence of fry from redds can have a significant effect on fry survival (through altered predation risk or competition advantage) and is often adapted for the local environment. Growth Fast-growing strains of commercially important salmonids have been developed through the use of selective breeding programs (Gjedrem 1983; Gjerde 1986; Hershberger et al. 1990). 207 Both mass and family selection methods have been utilized, with the typical levels of growth enhancement achieved being approximately 10-15% per generation. Heritabilities for growth enhancement remain relatively high (> 0.2), indicating that sufficient genetic variability remains within many aquacultured strains to allow for further enhancement of this trait (Gjedrem 2000). In general, domestic fish tend to grow faster than wild fish, although this is not universally the case (Bryden et al. 2004; Einum and Fleming 2001). Growth is greatly influenced by the rearing environment (Table 4). The following sections summarize growth enhancements which have been achieved by domestication among different salmonid species. Table 4. Differences in growth between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon F had higher growth rate in the hatchery and wild Einum & Fleming 1997 F outgrew W in the hatchery, W outgrew F in a semi-natural habitat Hy outgrew W in the natural environment Fleming & Einum 1997 Fleming et al. 2000 F outgrew W Fleming et al. 2002 Brook trout F outgrew W D maintained size advantage relative to W when released into semi-natural ponds and allowed to over-winter F had lower growth relative to W in the wild McGinnity et al. 1997, 2003, 1997 Flick & Webster 1964 Keller & Plosila 1981 Pacific salmon and trout D grew faster than Hy and W in hatchery but could not Mason et al. 1967 maintain this growth advantage in the wild D and Hy maintained size advantage over W when released Lachance & Magnan 1990 into small lakes (over two year period) F outgrew W Ayles et al. 1979; Ayles & Baker 1983 F outgrew W Lachance & Magnan 1990 F demonstrated 60% increase in weight Hershberger et al. 1990 F outgrew W McClelland et al. 2005 F outgrew W Tymchuk et al. in press; Tymchuk & Devlin 2005 F = farmed, D = domestic, Hy = hybrid, W = wild Atlantic salmon and sea trout Altered selection has been observed to cause genetic changes leading to faster growth of farmed Atlantic salmon (Fleming et al. 2000, 2002; Fleming and Einum 1997; Gjedrem 1979; Johnsson et al. 1996; Kallio-Nyberg and Koljonen 1997; McGinnity et al. 1997, 2003; McGinnity et al. 1997; Petersson and Jarvi 1995). An additional study found that farmed strains 208 of both Atlantic salmon and sea trout were larger at smolt, but the study could not differentiate between environmental or genetic influence (Petersson et al. 1996). Brook trout Domestic brook trout (propagated for >30 years) have been observed to grow faster than both wild and hybrid fish in the hatchery, but were not able to maintain their growth advantage in the wild (Mason et al. 1967). Other studies found that faster-growing domestic strains of brook trout were able to maintain their growth advantage relative to wild strains when reared in semi-natural ponds (Flick and Webster 1976; Lachance and Magnan 1990). Pacific salmon and trout Farmed strains of coho salmon can grow faster than non-selected native strains (Devlin et al. 2001; McClelland et al. 2005; Tymchuk et al. in press). Differentiation in growth rate in domestic compared with wild strains can occur in a relatively short period of time as coho salmon have demonstrated a 60% increase in mass after only four generations under a selective breeding program (Hallerman and Kapuscinski 1992). Domestic rainbow trout have been shown to grow much faster than wild rainbow trout strains (Devlin et al. 2001; Tymchuk and Devlin 2005) and are able to maintain their growth advantage when reared in natural lakes (Ayles and Baker 1983; Flick and Webster 1976; Lachance and Magnan 1990). General trends in growth Environmental factors such as availability of food resources and temperature will , of course, have an influence on growth of fish. However, there can also be large differences in growth between cultured and wild strains that are due to genetic differences between the strains. The magnitude of the growth differences caused by genotype will be dependent on the purpose and history of the cultured strain. Aquacultured strains that have been intensely selected for enhanced growth will show a larger shift in growth phenotype from the founding line relative to a strain that has not experienced directed selection. There is still limited knowledge on how environment will act on the inherent genetic differences (i.e. will environmental conditions affect different genotypes in distinct ways through genotype x environment interactions?). Metabolism and physiology There is limited information on the physiological differences between farmed and wild strains (Table 5). It is often thought that altered feed conversion efficiency underlies the ability of farmed strains of fish to growth faster. This has been supported by studies on Atlantic salmon (Thodesen et al. 1999) although no differences in conversion efficiency were found between wild and selected strains of brown trout (Sanchez et al. 2001). Altered endocrine regulation, particularly that involving changes of the growth hormone / insulin-like growth factor axis, may also explain the enhanced growth of selected strains of fish (Fleming et al. 2002). Selection in a culture environment may also alter smolting physiology. Atlantic salmon reared in the hatchery had altered timing of metabolic changes for smolting relative to wild fish (Leonard and McCormick 2001), and smolts from wild strains were better able to tolerate transfer to seawater (Handeland.et al. 2003). Wild fish were found to have an increased tolerance to saltwater relative to hatchery conspecifics (Shrimpton et al. 1994), although these differences 209 may be due to rearing environment. Ugedal et al. (1998) found that both wild and hatchery fish migrating downstream had well -developed seawater tolerance. Upon release, hatchery honmasu parr had elevated levels of growth hormone while those of wild parr did not increase. Additional differences detected between farmed and wild fish that may lead to physiological differences include altered swimbladder (Poppe et al. 1997) in farmed Atlantic salmon and altered heart morphology in both Atlantic salmon and rainbow trout (Poppe et al. 2003). It is suggested that the observed differences may lead to a decrease tolerance of stress in the farmed fish. Domestic strains of rainbow and brown trout were found to have altered critical thermal maxima relative to wild strains (Carline and Machung 2001). These differences were suggested to be genetically determined. Physiology may not always differ between farmed and wild strains. For example, no differences in swimming and cardiac performance were observed between farmed and wild Atlantic salmon (Dunmall and Schreer 2003). Table 5. Differences in metabolism and physiology between cultured and wild salmonids SPECIES RESULT SOURCE Rainbow trout and brook trout Altered critical thermal maxima Carline et al. 2001 Atlantic salmon No differences in swimming and cardiac performance Dunmall and Schreer 2003 Atlantic salmon F had higher levels of pituitary growth hormone Fleming et al. 2002 Atlantic salmon W smolts had better tolerance of seawater Handeland et al. 2003 Atlantic salmon and rainbow Altered heart morphology trout Atlantic salmon Brown trout Atlantic salmon Altered swimbladder No differences in feed conversion efficiency F had higher feed conversion efficiency Poppe et al. 2003 Poppe etal. 1997 Sanchez et al. 2001 Thodesen et al. 1999 F = farmed, D = domestic, Hy = hybrid, W = wild In general, cultured fish will tend to show physiological differences relative to wild fish. However, it is often not clear whether these differences are a cause or a consequence of other phenotypic differences such as growth or behavior between the strains. It is therefore difficult to clarify whether physiological differences have a genetic basis per se, or if they are a product of the environment. Any observed differences are likely due to a combination of these factors, unless cultured strains have been specifically selected for altered physiological characteristics such as higher feed conversion efficiencies. 210 Aggressive behavior Aggressive behavior is often affected by artificial rearing (Table 6). A study on the agonistic behavior of domestic and wild steelhead trout suggested that four to seven generations of domestication resulted in behavioral divergence of the two populations (Berejikian et al. 1996), although Johnsson et al. (1996) found no difference in dominance between fish selected for enhanced growth and wild individuals. Studies on domestic farmed Atlantic salmon found genetically based increases in aggression level relative to wild fish (Einum and Fleming 1997; Fleming et al. 2002; McGinnity et al. 1997, 2003). In one study examining brown trout, hatchery-reared fish of hatchery genotype were more aggressive than hatchery-reared fish of wild genotype introduced into a stream at the same time, but less aggressive than resident wild fish. Concurrently, the hatchery-reared fish had a lower rate of growth relative to both the resident and introduced wild fish (Deverill et al. 1999). In contrast, hatchery coho salmon parr dominated wild parr even though the wild parr had a prior resident advantage (Rhodes and Quinn 1998). Similar results were obtained by (Swain and Riddell 1990) where hatchery juvenile coho salmon were more aggressive than wild fish and in this case, experimental design revealed that differences were caused by genetic rather than environmental effects. Interestingly, other research has found that selection for enhanced growth results in indirect selection for tameness, as opposed to aggression (Doyle and Talbot 1986). For example, wild Atlantic salmon were found to make more aggressive attacks toward farmed individuals than farmed made on wild (Mork et al. 1999). In an attempt to explain the discrepancy relating to the genetic link between aggressive behavior and growth, (Ruzzante and Doyle 1991) concluded that agonistic behavior will be inversely proportional to growth when selection occurs in an environment with forced social interaction and unlimited food resources. Table 6. Differences in aggression between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon F more aggressive Einum & Fleming 1997 Atlantic salmon F dominated W in tank, W dominated F in semi-natural habitat Fleming & Einum 1997 Atlantic salmon F competitively displaces W downsteam McGinnity etal. 1997 Atlantic salmon W displayed higher levels of agonistic behaviour relative to F Mork etal. 1999 F = farmed, D = domestic, Hy = hybrid, W = wild Cultured and wild fish do show differences in their levels of aggression towards conspecifics, although there has not been a consistent trend as to whether aggression increases or decreases under culture. The common assertion is that aggression will decrease under culture when fish are reared in crowded conditions and do not have to fight for limited food resources. Knowledge is lacking on the expression of genetically determined differences in aggression under varied environmental conditions and the extent of behavioral plasticity that can be expressed by the different strains. 211 Predation risk Several studies indicate that anti-predator behavior has been altered through domestication so that domesticated fish are more willing to risk predation to feed and consequently may have higher mortality rates (Table 7). Farmed Atlantic salmon show increased risk-taking behavior relative to wild fish (Einum and Fleming 1997; Fleming et al. 2002; Fleming and Einum 1997) and a reduced response to predators, as measured by flight and heart rate response (Johnsson et al. 2001). Juvenile farmed coho salmon begin feeding sooner after a simulated predator attack relative to wild coho (Tymchuk et al. 2006) and domestic/wild hybrid rainbow trout with enhanced growth rate have been observed to be more willing to risk predation to feed relative to wild individuals (Johnsson and Abrahams 1991). Similarly, brown trout selected for enhanced growth showed reduced anti-predator behavior relative to wild fish (Johnsson et al. 1996), and growth-enhanced transgenic salmonids have been observed in some, but not all, cases to expose themselves to greater predation risk and suffer higher levels of predation mortality (Abrahams and Sutterlin 1999; Sundstrom et al. 2003, 2004, and 2005; Vandersteen Tymchuk et al. 2005). Table 7. Differences in predation risk between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon F and Hy less responsive to predation risk Einum & Fleming 1997 Atlantic salmon F less responsive to predation risk Fleming & Einum 1997 Atlantic salmon D less responsive to predation risk Johnsson et al. 2001 Brown trout F less responsive to predation risk Johnsson et al. 1996 Steelhead trout Hy less responsive to predation risk Johnsson & Abrahams 1991 Coho salmon F less responsive to predation risk Tymchuk et al. in press F = farmed, D = domestic, Hy = hybrid, W = wild A genetically determined reduced response to predators seems to be a consistent trend in domestic strains across several species. However, these studies have tested the predator response in an artificial environment with no real risk of being consumed by a predator. Few studies directly test whether the domestic strains would respond to a real predator in the wild or, more importantly, to overall fitness if their realized mortality would be any different from that experienced by the wild fish. Foraging behavior Currently, there are few studies on foraging behavior directed specifically at farmed fish (Table 8), so the following examples include studies on hatchery fish. In addition to being more 212 willing to risk predation to feed, domestic fish may have altered foraging strategies relative to wild individuals. These foraging strategies may not be as suitable for the natural environment as they have evolved within the structure of the culture environment. Hatchery reared brown trout have been observed to feed less and move more, generating more foraging costs overall relative to wild brown trout in the same stream (Bachman 1984). Bachman (1984) postulated that high-energy costs are a prominent cause for high mortality of released fish into the wild. Hatchery-reared brown trout demonstrated less effective foraging ability initially after release from culture, but showed a rapid learning curve when subsequently foraging for wild prey. After approximately one week, they were feeding nearly as well as the wild fish. This effect was most apparent in May when food sources were plentiful; there was greater discrepancy between hatchery and wild foraging consumption during periods when food was low (Johnsen and Ugedal 1986). Hatchery honmasu salmon released into the wild began feeding immediately on natural prey, however their stomach fullness was lower than that of wild fish for one week after release (Munakata et al. 2000). Stomach analysis of escaped farmed Atlantic salmon in the Pacific indicates that these fish are capable of successful feeding (McKinnell et al. 1997). Escaped farm Atlantic salmon off the Faroe Islands show feeding patterns typical of wild fish (Jacobsen and Hansen 2001). Table 8. Differences in foraging behaviour between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon No differences in diet Einum & Fleming 1997 Atlantic salmon No difference in feeding pattern of escaped Jacobsen & Hansen 2001 and wild fish F = farmed, D = domestic, Hy = hybrid, W = wild There is no evidence for a strong genetic basis for any observed differences in foraging strategy. Environment would likely have a strong effect on this phenotypic characteristic, and in nature would be expected to quickly overcome any genetic differences that had evolved within a culture environment (since i f an individual did not eat, it would die). A genetic basis for this trait would probably be due to phenotypic expression of other phenotypic traits such as growth or morphology that would drive foraging behavior characteristics. Habitat selection and dispersal Differences in habitat selection (Table 9) may reduce the potential for interaction between wild and cultured fish, or limit the ability of escaped fish to successfully locate and acquire food resources to meet their metabolic requirements. Hatchery cutthroat trout spent a greater proportion of time in riffles relative to wild fish and were more aggressive (Mesa 1991). This elevated expenditure of energy (unnecessary aggression and use of fast-flowing water) may be expected to limit the ability of domestic fish to survive in the wild. Fifth-generation farmed Atlantic salmon have been found to spend more time feeding pelagically and used more of the water column than did wild fish, who spent more time hiding (Mork et al. 1999). Similarly, 213 hatchery-reared Atlantic salmon were found to maintain a higher position within the water column relative to wild individuals (Dickson and MacCrimmon 1982). Dispersal clearly plays an important role in determining the fitness of salmon in nature, and differences in this trait (Table 9) have been found between hatchery and wild salmonids. Domestic masu salmon released as unfed alevins and eyed eggs tended to disperse upstream, while wild fish tended to disperse downstream and also tended to disperse more extensively than domestic fish (Nagata et al. 1994). Hatchery-reared trout had lower dispersal and upstream movement relative to concurrently stocked wild fish ( J 0 r g e n s e n and Berg 1991). Differences have also been observed in the use of a river estuary by hatchery and wild chinook salmon (Levings et al. 1986), with hatchery fish spending approximately half the amount of time in the estuary as did wild. Table 9. Habitat selection and dispersal differences between cultured and wild salmonids. TRAIT RESULT SPECIES SOURCE Habitat selection No differences in current or depth Atlantic salmon Einum & Fleming 1997 occupied by F and W F spent more time feeding Atlantic salmon Morketal. 1999 pelagically and used more of the water column Dispersal D fish tended to disperse upstream Masu salmon Nagata etal. 1994 and less extensively than W fish, that tended to disperse downstream F = farmed, D = domestic, Hy = hybrid, W = wild As with foraging behavior, there are limited data on which to make conclusions regarding any genetically based differences in habitat selection and dispersal between cultured and wild fish. Although the method of feeding within the culture environment may alter use of habit, differences observed for this trait could also be due to underlying physiological differences and not direct selection on the trait itself. Survival Escaped domesticated salmon can survive in the wild, although recapture rates of farmed fish vary considerably (Table 10). The ability of farmed fish to survive upon escape and successfully mature will depend on several factors including the timing of escape and whether the escape location is adjacent to spawning streams (Hansen and Jonsson 1994). The recapture rate of wild Atlantic salmon (2.3%) in the Faroese fishery was significantly higher than the recapture rate of escaped farmed fish at 1.2% (Hansen et al. 1997). Survival of farmed, wild, and hybrid Atlantic salmon parr was similar after release to a stream (Einum and Fleming 1997), but progeny of naturally spawned farmed Atlantic salmon had reduced survival (McGinnity et al. 2003; McGinnity et al. 1997). Recapture rates of domestic, hybrid, and wild brook trout planted in seven lakes were similar in three lakes, while in another four lakes, domestic fish had lower recapture rates than the hybrid and wild fish (Fraser 1981). Hybrid fingerling brook trout had 214 better survival than wild and domestic fish released into a pond (Keller and Plosila 1981). Domestic brook trout had better over-winter survival, but experienced higher mortality during the summer fishing season relative to wild fish (Flick and Webster 1964; Mason et al. 1967). Similar trends in survival have been indicated by studies on hatchery fish. Wild steelhead trout had the highest rate of survival in a stream environment (Reisenbichler and Mclntyre 1977). This was in contrast to a hatchery environment, where hatchery fish had the highest survival. Survival of wild brown trout populations was three times that of hybrid or hatchery fish (Skaala et al. 1996) and similar results were found when comparing hatchery-reared and native trout (Weiss and Schmutz 1999). Wild brown trout smolts had higher survival rates relative to domesticated smolts (Aarestrup et al. 2000), even though they were smaller when released into a river from the hatchery. Returns of stocked Scottish salmon were significantly lower than the returns of the wild fish (Verspoor and De Leaniz 1997). Importantly, not all studies have indicated that domesticated fish will incur higher mortality rates in nature. Hatchery-reared Atlantic salmon smolts did not suffer increased mortality relative to wild smolts in an estuary of the River Orkla in Norway (Hvidsten and Lund 1988). Such effects may depend highly on specific environmental conditions. For example, upstream from a hatchery, cormorants were found to consume only wild salmonids, while downstream predation was restricted to hatchery-reared fish (Kennedy and Greer 1988). Table 10. Differences in survival. SPECIES RESULT SOURCE Brown trout Survival better for W than D Aerestrup et al. 2000 Atlantic salmon Survival similar for F, Hy and W Einum & Fleming 1997 Brook trout D fish had lower over-summer survival Flick & Webster 1964 Atlantic salmon Brook trout Brook trout Brook trout Brook trout relative to W, no difference in over-winter survival D, Hy, and W had similar survival in three lakes; D had lower survival than Hy and W in four lakes Hy had better survival than D or W fingerlings stocked into a pond D had lower survival than W when released into lakes, with Hy having intermediate survival Survival of D higher over winter, W better over summer when fishing was open Survival of F less than W McGinnity et al. 1997, 2003 Lachance & Magnan 1990 Fraser 1981 Keller & Plosila 1981 Mason et al. 1967 F = fanned, D = domestic, Hy = hybrid, W = wild 215 Survival is influenced by all other phenotypic traits and the environment in which they are expressed. Cultured fish, either through a plastic response to their environment or through an adaptive response to altered selection pressures, tend to express phenotypic characteristics best suited for the culture environment. Consequently, they tend to not have as high survival as wild fish in a natural environment. However, it is still not clear whether cultured fish that experience a natural environment throughout their life history will still show decreased survival relative to the wild fish. It is not known how strong the genetic basis of survival is, or whether the cultured fish still have the ability to show a phenotypically plastic response to the environment that will maximize their ability to survive. Sexual maturation and spawning Genetic effects of farmed fish on wild populations will depend in part on the reproductive behavior of farmed fish in the wild. Evidence suggests that farmed fish have the ability to breed successfully in the wild, although contradicting results do occur. There are generally significant differences in breeding potential (Table 11) of cultured and wild fish (Fleming and Gross 1992, 1993; Fleming et al. 1996), although other studies have found similar reproductive success for hatchery and native fish in the wild (Dannewitz et al. 2004; Palm et al. 2003). Morphology and life history traits related to reproductive behavior respond evolutionarily to altered selection regime in the hatchery environment (Fleming 1994; Fleming and Gross 1989). The genetic effects of aquacultured salmon on wild populations will depend not only on the size of the wild population, but also on variation in breeding success (Fleming and Petersson 2001). The following sections describe aspects of reproductive success in cultured and wild salmonids. Table 11. Differences in reproductive behaviour and success between cultured and wild salmonids. TRAIT RESULT SPECIES SOURCE Maturation and migration F adults migrated to nearby streams Atlantic salmon in a random manner F adults did not migrate as far up theAtlantic salmon river Reproductive physiology D had higher gonadosomatic index Brook trout relative to W and Hy No difference in fecundity between Chinook salmon F and W Spawning behaviour F spawned earlier Atlantic salmon Heggberget et al. 1993 Webb etal. 1991 Lachance & Magnan 1990 Bryden et al. 2004 Fleming et al. 2000 F males were less aggressive, Atlantic salmon courted less, and participated in fewer spawnings relative to W F redds had a greater number of egg Atlantic salmon pockets with fewer eggs per pocked relative to W F females moved more and F fish Atlantic salmon spent less time in spawning area Fleming et al. 1996 Lura etal. 1993 Oklandetal. 1995 216 TRAIT RESULT SPECIES SOURCE F spawned earlier Atlantic salmon Sa?grov etal. 1997 Reproduction F and W parr may or may not show Atlantic salmon differences in aggression At high density, F had reduced Atlantic salmon success relative to W; at low density, there was no difference in success 25-35% of eggs in river of F origin Atlantic salmon Lifetime reproductive success of F Atlantic salmon less than W Lifetime reproductive success of F Atlantic salmon less than W W fish had higher reproductive Atlantic salmon success than F at high densities, no difference at low densities F males had reduced spawning Atlantic salmon success relative to W males escaped F females hybrized more Atlantic salmon frequently with trout than W Weir et al. 2004 Clifford et al. 1998a,b Carr etal. 1997 Fleming et al. 2000 Fleming et al. 1996 Sasgrov etal. 1997 Weir et al. 2004 Youngson et al. 1993 F = farmed, D = domestic, Hy = hybrid, W = wild Age of maturation and spawning migration Migration and maturation timing have demonstrated high heritabilities and genetic correlation for populations of chinook salmon, indicating that these traits would respond rapidly to selection and may accelerate selection for other traits by producing reproductively isolated populations (Quinn et al. 2000), The breeding system of Atlantic salmon evolves through selection on viability and sexual selection for breeding opportunities. Introduced salmon alter the frequency of different breeding phenotypes (due to differences in morphology), which will alter selection and thereby influence the breeding system (Fleming 1998). Although this is not a direct genetic affect, over time the altered selection parameters may cause a concurrent change in the genetic profile that defines the breeding strategy. Hatchery-reared Atlantic salmon can have altered age at maturity relative to wild fish, with the hatchery fish producing a higher proportion of mature grilse (Kallio-Nyberg and Koljonen 1997) due in part to the elevated growth rates induced by food availability in the hatchery. Comparisons of hatchery and wild summer and winter steelhead trout indicated that hatchery fish returned earlier than wild in both the summer and winter returns (Leider et al. 1986), and further, the hatchery fish had a lower number of saltwater age categories relative to the wild fish. Hatchery and wild summer steelhead were found to be more similar than hatchery and wild winter steelhead, indicating that the life history strategies of the local populations must be considered when assessing differences, and potential interaction, between domesticated and wild fish. 217 To reproduce successfully, farmed salmon must be capable of finding suitable spawning areas in streams. The farmed salmon may have more difficulty in homing to appropriate spawning streams, as they may not have been imprinted on the local freshwater streams (Hansen et al. 1987). One study found that farmed Atlantic salmon did migrate to nearby streams, albeit in a random manner (Heggberget et al. 1993), while the wild salmon released concurrently homed to the appropriate river more precisely and in a shorter period of time. Consquently, the largest salmon streams may receive the highest numbers of escaped farmed salmon, which are also the most important salmon streams in terms of potential for reproduction. A study examining the spawning behavior of escaped sub-adult farmed Atlantic salmon in Scotland using radio tagging and direct observation revealed that farmed salmon tend to spawn later in the year and do not migrate as far upstream as do wild fish. It is possible that the lower migration of the farmed salmon occurred because some were raised in the hatchery, which is located in the lower reaches of the river to which they would have imprinted on (Webb et al. 1991). Delayed spawning may arise due to differences in genetic origin of the farmed and wild fish, since different stocks of Atlantic salmon are known to have divergent average spawning dates. Thus, genetic introgression of farmed genotypes into wild population may affect a specific demographic subset of a population preferentially. Maturation morphology Secondary sexual characteristics are reduced in cultured males relative to wild males (Hard et al. 2000) but the extent of genetic or environmental influence on these observations is unclear. The morphology of fish at maturity can be strongly influenced by environment (Bessey et al. 2004) making tests of genetic divergence of this trait difficult, particularly when the strains cannot be reared in a natural environment. Reproductive physiology Variation in egg mass and fecundity in brown trout is, in part, genetically determined (Jonsson and Jonsson 1999) and investment by hatchery fish was found to be greater than wild strains. Similarly, domestic fish from a lake had a higher gonadosomatic index relative to hybrid and wild brook trout, and concurrently higher fecundity (Lachance and Magnan 1990). Again, this is not universally observed since the fecundity of lake trout was not found to differ between hatchery and wild strains (Peck 1988), nor were differences observed between farmed and wild strains of chinook salmon (Bryden et al. 2004). Spawning behavior One study found that escaped farmed Atlantic salmon have reduced spawning success because farmed females moved more during spawning and the farmed fish spent less time in the spawning area relative to the wild fish (Okland et al. 1995). In a different study, escaped farmed female Atlantic salmon demonstrated differences in spawning behavior (redds of farmed fish contained more egg pockets and fewer eggs per pocket) although all components of normal spawning behavior were present (Lura and Sa^grov 1993). It seems that the most significant differences in spawning behavior may be exhibited by the males. Fleming et al. (1996) found farmed males to be less aggressive than wild males, with a. consequent reduction in the number of females courted and in breeding success. Further studies indicated conflicting results when 218 testing for differences in aggression between farmed and wild parr (Weir et al. 2004). The presence of differences among the strains of fish may be dependent on life-history stage in addition to genotype and environment; Captively reared coho salmon of both sexes showed a full range of reproductive behavior demonstrated by wild fish (albeit reduced from normal levels). They also had the ability to spawn naturally in the wild (Berejikian et al. 1997; Bessey et al. 2004), although there may be a competitive advantage for wild females as they constructed more nests per individuals relative to the cultured females. Reproductive success Atlantic salmon: Some escaped farmed Atlantic salmon in N W Ireland were successful in completing their life cycle and breeding or interbreeding with wild fish (Clifford et al. 1997), although at an average success rate of only 7%, with a maximum breeding frequency of 70% in one river, (Clifford et al. 1998). At high densities, wild Atlantic salmon had better reproductive success than did escaped farmed fish, whereas there was no observed difference in reproductive success between farmed and wild fish when densities were low (presumably since mate choice was limited). In the River Vosso in Norway, escaped farmed fish were more prevalent (81%) and spawned earlier than wild fish and most fry in the river were of farmed origin. Using minisatellite D N A profiling, the reproductive success of secondary males (subdominant adults or parr, which mature in freshwater) was measured and compared for wild and sea-ranched Atlantic salmon (Thompson et al. 1998). Sea-ranched fish secondary males had higher reproductive success (48.2%o) than wild secondary males (28.9%), The Magaguadavic River located in the center of the New Brunswick salmon aquaculture industry has been monitored to study the effect of escaped farmed salmon entering the river. Over a period of four years, the numbers of wild fish decreased (to very low numbers of <10) while the numbers of escaped farmed fish increased. Analysis of the eggs indicated that 20%) to 35%> of the eggs were of farmed origin (Carr et al. 1997a). The farmed fish were found to enter the river later than the wild fish, and most of the farmed fish were sexually immature (Carr et al. 1997b). Examination of the gonadal tissue demonstrated that the farmed female salmon showed no sign of maturing upon the year of entry to freshwater, although some farmed males would be capable of interbreeding with wild females (Lacroix et al. 1997). Some fish may over-winter in the river and mature, thereby subsequently becoming capable of spawning. After tracking the reproductive success of farmed and wild salmon released into a river, it was found that the farmed salmon, particularly the males, were reproductively inferior to the wild fish (Fleming et al. 1996, 2000). The overall reproductive success of the farmed fish was 16%) of that of the wild fish in one study. However, the productivity of the native population was depressed by 30%>, indicating resource competition and competitive displacement. Other results indicated that farmed females had less than one-third the reproductive success of wild females, with farmed males indicating only 1-3% of the reproductive success of wild males (Fleming et al. 1996). These differences are usually due to inferior competitive ability on the spawning ground for the farmed males, and lower fecundity and inept egg care for the wild females. Hatchery experiments on farmed and wild fish support the observation that farmed males have reduced spawning success compared with wild males (Weir et al. 2004) However, not all adults in all experiments were derived from the same environment (wild or culture) so that effects likely are not due to genetic differences alone. A different study demonstrated that farmed and hybrid Atlantic salmon parr reared in a near-natural environment had higher breeding and fertilization success than wild individuals 219 (Garant et al. 2003). The authors suggest that introgression of domestic genes past the initial generation following sea-pen escape could be mediated by early maturing farm and hybrid males. Although initial gene flow is typically attributed to matings between wild males and domestic females, this trend may be reversed in consequent generations due to precocious parr. Discrepancies between the two previous studies are quite likely due to an environmental effect on the genetic differences between domestic and wild fish. Brown trout: Studies with brown trout have shown that reproductive success may differ between male and female fish. Asymmetry between the male and female reproductive success was found by comparing measurements of inbreeding (as indicated by protein-coding loci and mitochondrial haplotypes) with the domestic females showing higher rates of introgression with the native fish (Poteaux et al. 1998b). Hatchery brown trout were found to interbreed with wild fish, with male hatchery fish contributing more to interbreeding than hatchery females (Hansen et al. 2000). Pacific salmon: There is a lack of knowledge on the ability of farmed strains of Pacific salmon to spawn successfully in the wild. Some studies on hatchery strains of Pacific salmon may provide some indication of the expected results for farmed strains. There would likely be an even stronger effect of culture on their ability to reproduce successfully in the wild. Comparisons of spawning success of hatchery and wild coho salmon found trends similar to those observed for Atlantic salmon where the hatchery fish had reduced success (from both environmental and genetic causes), which was more severe in males (Berejikian et al. 1997, 2001; Fleming and Gross 1993; Fleming and Petersson 2001). Similarly, wild strains reared in culture conditions have reduced reproductive success (Berejikian et al. 1997; Bessey et al. 2004). General conclusions Consistent trends have been observed in that cultured fish often have the physiological ability to spawn, but altered spawning behavior can limit their success. Even though the reproductive success of farmed fish may be low, the potential for significant gene flow still exists since the population of farmed fish often outnumbers the population of resident wild fish (at least in the case for Atlantic salmon) at times by as much as 3:1 (Lund et al. 1994; Lura and 0kland 1994). We do not have any data on the ability of farmed and wild Pacific salmon to spawn in nature, but comparisons between hatchery and wild coho salmon indicate that trends observed for Atlantic salmon may be typical of the phenotypic changes expected during domestication. W H Y A R E C U L T U R E D A N D W I L D FISH D I F F E R E N T ? Genotype, environment, and genotype x environmental (G x E) interactions Phenotypic differences between farmed and wild salmonids may arise from a combination of genetic and environmental effects but the origin of the difference is not well defined in most cases. To assess genetic effects, experiments must be performed by rearing fish of different origins in a common environment. Environmental effects (i.e. phenotypic plasticity) can be tested by rearing fish of a common genetic background in different environments. As will 220 be discussed below, cultured fish may have an altered genotype in response to selection pressures from an artificial environment, leading to genetically based phenotypic differences between the cultured and wild strains. Alternatively, since salmonids are so phenotypically plastic (Hutchings 2004), they may have an altered phenotype in response to the altered environmental conditions in which they are reared. These environmentally based phenotypic differences would not be passed along to offspring as they do not have a genetic basis. Thus, environmental effects are anticipated to have single generation effects arising directly from escaped fish, whereas genetic differences are those with the potential to affect a species on a longer time frame. The sensitivity of a phenotypic trait to environmental conditions (i.e. phenotypic plasticity) may itself have been altered through domestication. It is therefore critical to separate the influence of genotype and environment, and to understand genotype by environment interactions to fully predict the genetic effects of interaction between wild and cultured fish. Genetic mechanisms leading to differentiation of cultured and wild fish The mechanisms causing genetic changes in cultured fish include inbreeding, genetic drift and selection. Selection can be further broken down into three categories: artificial selection, domestication, and relaxation of natural selection. It is generally accepted that farming of salmon generates genetic change due to intentional and unintentional selection in culture (Fleming and Einum 1997; Reisenbichler and Mclntyre 1977). In many cases, the goal of selection is to produce a homogeneous line that demonstrates constancy in the desirable traits such as enhanced growth (Gjedrem 2000). In cases here diversity is desired, domestication effects would be lower. Genetic changes may also occur for traits other than those that are the focus of the selective breeding program. For example, fifth-generation farmed Atlantic salmon differed significantly from wild populations in loci other than those chosen for the selective breeding program (MJ0lner0d et al. 1997). The characteristic of the genetic changes caused by indirect selection during domestication can vary according to the method of fish culture and the extent of time that fish spend in the artificial environment (Utter and Epifanio 2002). For example, characteristics of the culture environment such as rearing density and the source, of food into the tank may affect the evolution of aggressive behavior or increased motivation for surface feeding. Documented changes in genotypic variability of cultured fish Differences are often found when comparing the genotype (of selectively neutral alleles) of cultured and wild fish (Table 12). A variety of methods have been used including, indirect methods (e.g. use of fluctuating asymmetry as a measure of genetic robustness and thus developmental stability), direct assessments of genetic variation (at the D N A level using single nucleotide polymorphisms detected by sequencing or restriction fragment polymorphism as well as micro- and minisatellite variation), or methods involving phenotypic assessments of progeny derived from genetic crosses. 221 Table 12. Differences in genetic variability between cultured and wild salmonids. SPECIES RESULT SOURCE Atlantic salmon F had lower heterozygosity over three minisatellite loci Clifford etal. 1998 F had lower heterozygosity and fewer alleles Cross & King 1983 No difference in heterozygosity but F had lower gene diversityDanielsdoftir et al. 1997 F had lower genetic variability relative to W Mjolnerod etal. 1997 F had lower genetic variability relative to W Wislonetal. 1995 F had lower heterozygosity and fewer alleles Verspoor 1988 Pacific salmon and trout F had fewer polymorphic loci, fewer number of alleles per loci Allendorf & Phelps 1980 and lower heterozygosity Fl Hy (between genetically isolated strains) had increased genetic variability Gharrett & Smoker 1991 F2 Hy demonstrated increased bilateral asymmetry Gharrett & Smoker 1991 F maintained high heritability for weight, indicating no reduction in genetic variability relative to W Hershberger et al. 1990b F = farmed, D = domestic, Hy = hybrid, W = wild Atlantic salmon Artificially reared Atlantic salmon, selected for growth and disease resistance, were found to possess lower genetic variability as measured by mean heterozygosity and mean number of alleles at six enzyme loci (Cross and King 1983). It was argued that these genetic changes were caused by founder effects and genetic drift rather then selection. First-generation cultured Atlantic salmon had 26% less heterozygosity and 12% fewer alleles relative to wild fish, consistent with a loss of genetic variability expected from random drift, probably caused by using a small number of adults for the broodstock (Verspoor 1988). Other studies have further supported the trend that farmed Atlantic salmon typically have lower measures of genetic variability (14-45%) reduction) relative to wild populations (MJ0lner0d et al. 1997; Wilson et al. 1995). One study on the genetic variation between farmed and wild populations of Atlantic salmon found no significant difference in heterozygosity, but the farmed fish possessed lower total gene diversity (Danielsdottir et al. 1997). The genetic distance between the farmed population and the wild populations was as great as the genetic differences between separate wild populations. 222 Pacific salmon Loss of genetic variation was observed for isozyme loci in a cultured hatchery population of cutthroat trout revealing a 57% decrease in polymorphic loci, 29% reduction in average number of alleles per loci, 21% reduction in heterozygosity, and changes in allelic frequencies between age-class relative to the wild population from which they were derived 14 years earlier (Allendorf and Phelps 1980). Genetic differences between the reared and wild populations were as large as differences between natural populations. Loss of genetic variability in domesticated fish has not been consistently observed. Coho salmon in a 10-year selection and breeding program demonstrated an increase of more than 60% increase in weight while maintaining high measures of heritability for weight (Hershberger et al. 1990). This study indicates that long-term selection programs can produce improvements in traits such as growth without dramatically reducing genetic variation. While direct molecular assessment of variation has not been reported for these populations, the retention of high heritability presumably arises from the overall breeding strategy undertaken in the domestication program for coho salmon in this case. MECHANISMS OF GENETIC INTERACTION BETWEEN AQUACULTURED AND WILD FISH Genetic effects of domestic fish may be direct or indirect. Direct genetic effects include the alteration of the wild genome (introgression) due to interbreeding between wild and domesticated fish or the production of sterile hybrids. Indirect effects include the effect of reduced effective population size or altered selection pressure arising from competition or the introduction of pathogens (Krueger and May 1991; Skaala et al. 1990; Waples 1991). Interbreeding between cultured and native populations has generally been found to be disadvantageous when the genetic effects alter fitness-related traits (Hindar et al. 1991). However, genetic effects of hybridization between farmed and wild salmon are somewhat unpredictable and may differ between populations. Most studies have focused on the fitness of the Fi generation when exploring the effects of interbreeding between domestic and wild strains. Such hybrids may in fact have enhanced fitness due to hybrid vigor. A clearer picture would be provided by following the fitness of the F 2 and later generations, as this in when outbreeding depression will begin to be expressed (see below). There is also limited information on the ability of introgressed genotypes to show an evolutionary response to environmental pressures. It is not known whether hybrid populations that phenotypically appear to be "wild" would evolve in a manner similar to true wild fish populations. The genetic effect of escaped cultured fish on wild populations will depend on the demographics of the wild population, the magnitude and frequency of the escape, and the extent of introgression of aquacultured genotypes into the wild population (Hutchings 1991). The phenotype of wild and farmed hybrids may vary depending on the source of the wild population (for example see Einum and Fleming 1997). Additive genetic effects There is a growing body of literature indicating that phenotypic differences between domestic and wild fish are largely a result of additive genetic differences. Therefore, the phenotypic effects of domestication tend to be diluted with repeated backcrossing into wild 223 populations. Interbreeding between farmed and native Atlantic salmon generally resulted in hybrids with intermediate expression of traits such as aggression, growth, and anti-predator behavior (Einum and Fleming 1997). McGinnity et al. (1997, 2003) found that first-generation and second-generation backcross hybrids were intermediate between wild and farmed Atlantic salmon in growth, survival, and parr maturity rates. For coho salmon, there is a strong correlation between the proportion of domestic genes within the genotype and measures of growth (McClelland et al. 2005; Tymchuk et al. 2006) and anti-predator behavior (Tymchuk et al. 2006). Introgression of domesticated and wild rainbow trout (Ayles and Baker 1983; Tymchuk and Devlin 2005) and chinook salmon (Bryden et al. 2004) show similar trends for growth and other measures of physiology such as disease resistance. Heterosis Crossing different strains with low genetic variability may re-establish lost alleles and allelic combinations, which could lead to hybrid vigor, or heterosis, wherein the Fi generation would have increased fitness relative to the parental stocks. Heterosis is most likely to occur i f the parental stocks are inbred and not highly genetically divergent. However, i f the parent stocks came from different habitats, the resulting progeny may not be well adapted for either habitat. An allele that is advantageous in one genome or environment may be disadvantageous to overall fitness in another genome or environment. Cross-breeding experiments on five Norweigian strains of Atlantic salmon did not find significant heterosis for survival or body weight (Gjerde and Refstie 1984). Further experiments support a lack of heterosis for growth rate in Atlantic salmon fry (Friars et al. 1979). Heterosis for competitive ability may have been detected in hybrids of Atlantic and farm salmon in one instance, but the results are not clear (Einum and Fleming 1997). However, heterosis for growth and survival was detected when two different strains of brook trout were crossed (Webster and Flick 1981). Other studies on rainbow trout (including crosses among inbred lines) have detected evidence of heterosis for body weight (Ayles and Baker 1983; Gjerde 1988; Gjerde et al. 1983; Wangila and Dick 1996). The introduction of low numbers of genetically novel fish into large populations may. in some cases, be beneficial by providing otherwise unavailable natural variation to the population which can be acted on by natural selection. Outbreeding depression Outbreeding depression could cause a negative effect by disrupting co-adapted allele complexes, which would be indicated by the F 2 generation. The route of genetic introgression is likely to be through hybridization rather than by pure farm stock displacing pure wild populations, and thus outbreeding depression may not be apparent until recombination has separated co-adapted genotypes. Studies indicate that outbreeding depression can occur in fish populations. After following hybrids of genetically isolated pink salmon, the Fi generation had increased genetic variability relative to the control fish, followed in the F 2 generation by very low survival and increased bilateral asymmetry indicating that outbreeding depression had occurred (Gharrett and Smoker 1994). Recent studies have confirmed the previous observation that outbreeding depression is possible in populations of pink salmon (Gharrett et al. 1999; Gilk et al. 2004). However, a study by McGinnity et al. (1997) undertaken to examine the survival and growth of farmed, wild, and hybrid Atlantic salmon in a natural environment, found no 224 evidence for outbreeding depression as the fitness of the hybrid genotypes was intermediate between the farmed and wild genotypes. No evidence of outbreeding depression was found in crosses of wild coho salmon from different populations (Smoker et al. 2004) or during introgression of domestic and wild rainbow trout (Tymchuk and Devlin 2005) and coho salmon (Tymchuk et al. 2006). The difficulty in obtaining high statistical power in these experiments for estimates of outbreeding depression suggests that these negative results should be interpreted with caution until further data have been developed. Effects on population size and inbreeding depression Escaped cultured fish (possessing low genetic variability) may reduce the effective population size (Ne), leading to a population bottleneck and a concurrent founders effect. However, escapees could also increase effective population size i f novel alleles are being added to a breeding group from distant populations. Analysis of microsatellite loci in wild and hatchery-reared brown trout used for enhancement purposes indicated reduced Ne and a loss of genetic variability (Hansen et al. 2000). A potential effect of reduced population size could be that native fish need to mate with genetically similar fish, leading to inbreeding depression. Inbreeding depression can occur in fish species as demonstrated in a study by (Gjerde et al. 1983) in which rainbow trout showed inbreeding depression for survival. Empirical studies have shown that inbreeding has a negative effect on fitness by causing a shift in fitness-related phenotypic traits. Another effect of increased difficulty in finding mates is the positive correlation between hybridization rates (for example between Atlantic salmon and brown trout) and the proportion of escaped farmed fish (Hindar and Balstad 1994). EXAMPLES OF GENETIC INTROGRESSION OF CULTURED GENOTYPES INTO WILD POPULATIONS Significant evidence now exists demonstrating the presence of aquacultured salmonids in nature. For example, initial rates of escaped farmed Atlantic salmon found in Norwegian rivers were as high as 70%, accounting for 1-15% of fish sampled in north and south Norway, respectively (Gausen and Moen 1991). Similarly, as salmon culture developed in SW Ireland, the number of escaped farmed fish entering salmon rivers was found to be substantial, making up to 70%) of the catch in one river (Gudjonsson 1991), and in Northern Ireland, annual average proportions of farmed fish in the fishery ranged from 0.26^1.04% (mean 2.4%) in the marine environment and from 0.13- 2.62% (mean 0.89%) in freshwater (Crozier 1998). Since that time, continued evidence for the introgression of aquacultured salmonid genes into wild populations has been reported, including examples of native and non-native strains and species (Table 13). 225 Table 13. Examples of introgression of cultured genes into wild populations. SPECIES RESULT SOURCE Atlantic salmon No genetic effect of F escape except in two small rivers Clifford et al. 1998 near escape Shift in allele frequencies of the W population towards that Crozier 1993 of the D; changes still present after two generations Introgression rates between 0.5% and 17.8% Webb et al. 1993a Introgression rates between 1.9% and 12.3% Webb etal. 1993b Brown trout W population genetically altered by interbreeding with D Fritzner et al. 2001 Evidence of genetic contribution from D in W population Hansen et al. 2001 F = farmed, D = domestic, Hy = hybrid, W = wild The effects of genetic introgression of farmed genotypes into wild populations will depend on how the natural genetic variation is distributed. Effects of introgression will be worse if much of the variation is distributed among populations and is associated with adaptive traits or reproductive barriers between stocks (Skaala et al. 1990). Anadromous populations of salmonids tend to be more resistant to introgression due to their more complex life histories (e.g. overlapping maturation age classes) as well as the sharing of genetic information that occurs through straying (Utter and Epifanio 2002). In addition, genetic distance between the two populations does not seem to be a reliable indicator of the potential effects of introgression (Utter and Epifanio 2002). Atlantic salmon Following extensive escapes of farmed salmon in Scotland, a study was done to examine the extent to which these fish entered rivers and successfully spawned (Webb et al. 1993b). Fry were sampled from 16 rivers near known escapes (River Polla in the previous year) to over 200 km away (throughout which range many other farms are also found). Using the presence of canthaxanthin (a pigment added to salmon feed, which is deposited in eggs) in the fry as an indicator of being derived from farmed mothers, 14 of 16 rivers showed evidence of farmed parentage. The average frequency of farmed progeny was 5.1%, ranging from 0.5-17.8%) of the progeny sampled in different rivers. The values obtained in this study may underestimate the true introgression of farmed genotypes into populations, since contributions from paternal sources (which do not provide canthaxanthin to the embryo) would not be detected (Webb et al. 1993b). A follow-up study was undertaken to determine the extent of introgression in the second year after an escape of farmed Atlantic salmon in Scotland (Webb et al. 1993a). Less than 0.5% of the escaped fish returned to the river to spawn, but the data indicated that fish may return for multiple years following an escape. Of 54 redds examined, five were found to contain progeny with canthaxanthin indicating farmed origin. An examination of rivers adjacent to the river Polla revealed the presence of canthaxanthin-bearing fry at frequencies of 1.9-12.3 %> in the year of 226 the escape as well as the year after. The authors feel that these levels of farmed progeny may be the background level of farmed spawning that is occurring in the rivers due to escapes other than the documented escape under study. Farmed salmon escaped in the Glenarm River in Northern Ireland in 1990. A n allozyme study indicated a shift in allele frequencies towards those in the farmed salmon, indicating that interbreeding had occurred (Crozier 1993). A follow-up study in 1997 indicated that even after two generations, the wild population was still significantly different from the pre-escape population. In addition, new alleles not previously detected were found, leading to the conclusion that further influxes of farmed salmon had occurred. Alternately, another study (Clifford et al. 1998) found little long-term (three years) genetic effect of a large-scale escape other than in two small rivers near the escape. Brown trout Brown trout populations in Denmark have been found to be altered genetically by breeding with domesticated trout, as indicated by genetic differentiation analysis using microsatellite markers (Fritzner et al. 2001). In a similar study of brown trout, results indicated that while there was an absence of domesticated trout in the river, there was evidence of genetic contribution by the domesticated fish in the wild individuals (Hansen et al. 2001). In Spain, threatened populations of native brown trout are stocked with hatchery-reared fish and it has been found that the hatchery and wild fish do interbreed. Consequently, the genetic distance between the two strains (domestic and wild) has been seen to be decreasing at a rate of approximately 5% each year (Garcia-Marin et al. 1999). Similar effects have been noted in Norway, where the release of hatchery fish altered the genetic make-up of wild brown trout populations (Skaala 1992). Brown trout populations in the Mediterranean were examined for evidence of introgression from stocked domestic brown trout. Introgression rates varied from 0% -77% among test sites. Little natural variability was found to exist between native populations from the different sites, although introgression is increasing genetic variability and thereby increasing the variability between populations. The concern is that native populations will eventually be replaced by uniform domestic stocks (Berrebi et al. 2000). No reproductive barriers to genetic exchange were indicated after electrophoretic analysis of hybridization between introduced domestic and native populations of brown trout in the Mediterranean (Barbat-Leterrier et al. 1989). Rates of introgression ranged from 0^ 10%>. Other studies support the observance of introgression between hatchery and wild populations of brown trout in the Mediterranean (Poteaux et al. 1999) although they argue that natural selection limits introgression by acting against hatchery and/or hybrid genes. However, several years post-stocking have still not eliminated the genetic differences due to introgression during stocking. Introgression between domestic and wild fish did occur in a stocked population of brown trout, and when compared with a population that had been previously stocked, there was a decrease in the presence of domestic alleles in the non-stocked population (Poteaux et al. 1998a). Allozyme analysis of four stocked and four unstocked populations of brown trout indicated varying levels of hybridization and introgression in both the stocked and protected areas, with evidence that. hybrids can successfully reproduce past the first generation (Cagigas et al. 1999). 227 Atlantic salmon X brown trout hybrids In the 16 rivers examined for presence of farmed salmon described in Webb et al. 1993a, evidence of brown trout x Atlantic salmon hybridization was also observed. A l l hybrids were the progeny of Atlantic salmon females by brown trout males based on analysis of mtDNA, and the incidence of hybrid vs. pure progeny from Atlantic salmon mothers was 4.3% hybrids for wild Atlantic mothers and 35% for farmed mothers (based on presence of canthaxanthin in progeny). This higher incidence of hybrid progeny from farmed mothers suggests a lower fidelity of spawning than for wild mothers (Youngson et al. 1993). Another study also found that hybridization between Atlantic salmon and brown trout varied with the status of the salmon population (Hindar and Balstad 1994). Pacific salmon Over 70%) of the fish farmed on the west coast are Atlantic salmon. There is little evidence that escaped Atlantic fish are capable of establishing naturally spawning populations, although there is some evidence of escaped farmed Atlantic salmon occasionally breeding in nature (Volpe et al. 2000). Laboratory-based studies indicate that the Atlantic escapees have such limited ability to interbreed with wild Pacific salmon that this is not considered a risk (Devlin unpublished, cited in Noakes et al. 2000). Interbreeding between cultured and wild Pacific salmon could potentially be a concern, but currently there is a lack of knowledge on the extent of introgression of farmed Pacific salmon into wild populations due to the difficulties of identifying cultured, wild and hybrid individuals. An examination of allele frequencies of rainbow trout in the Yakima River provided evidence of introgression from nonanadromous domesticated hatchery strains of rainbow trout (Campton and Johnston 1985). Similarly, evidence for introgression between native and hatchery rainbow trout was found in Metolius River, a tributary of the Deschutes River, Oregon (Currens et al. 1997). Introgression may not always occur, as one study found no evidence of introgression between hatchery and wild chum salmon populations (LeClair et al. 1999). For coho salmon, wild spawners in the Samish and lower Nooksack rivers have been introgressed or replaced by hatchery-produced fish, whereas wild spawners in the upper Nooksack River show little influenced by gene flow from the hatchery strains (Small et al. 2004). Recent laboratory studies on introgression between cultured and wild Pacific salmonids have revealed that a large proportion of phenotypic divergence between the strains arises from additive genetic differences (Tymchuk et al. 2005, 2006; McClelland et al. 2005). THE IMPORTANCE OF GENETICS TO CONSERVATION OF SALMONIDS Genotype, in addition to environment, determines the adaptive phenotypic characteristics of salmonids, and as such it is likely that disruption of this genetic structure may have both short-term and long-term effects on individual fitness as well as the future resilience of populations to natural and anthropogenic pressures. 228 The genetic character of wild salmonid populations defines their current and future adaptive phenotypes Significant evidence now exists demonstrating the unique genetic character of populations of several salmonid species. Native salmonid populations can be considered as genetically distinct stocks that have evolved adaptations to maximize fitness under selection regime of their local environments. Genetic differentiation among salmonid populations is well documented (King et al. 2001; Stahl 1987) and in some cases can be correlated with adaptive phenotypes (Beacham and Murray 1987; Beacham et al. 1988; Clarke et al. 1995; Groot and Margolis 1991; Quinn 2005; Taylor 1991). Some genetically determined life-history traits of adaptive significance are known. For example, the age of smoltification in stream-type chinook salmon (Clarke et al. 1994), the populations of which tend to migrate to sea after spending at least one year in streams inland compared with coastal populations of ocean-type salmon, which migrate to the ocean in their first year. Another good example of a life-history adaptation is the nonanadromous form of sockeye salmon, kokanee. The two forms of sockeye can be found together in the same population, but seem to maintain their genetic identity through assortative mating (Foote et al. 1989) and reduced fitness of hybrids (Wood and Foote 1996) Genetic differentiation of populations at the molecular level has been detected using allozymes and D N A markers in the mitochondrial and nuclear genomes, and although such genetic differentiation among populations does not necessarily arise from selection, in many cases it does. For Atlantic salmon in Europe, genetic diversity both within and among populations has been correlated with performance characteristics relevant to survival and recruitment (Bourke et al. 1997; Verspoor 1997). Similarly, an examination of the genetic structure of wild and hatchery brown trout indicated that more than 60% of the genetic variability in the wild populations was due to differences between the populations (Garcia-Marin et al. 1991). In contrast, only 3% of the genetic variability for the hatchery fish was explained by differences between populations, demonstrating that the hatchery fish represent a much more homogenous population with a narrower genetic base. Several species of Pacific salmon in British Columbia show clear evidence of genetic differentiation down to the subpopulation level (Beacham and Murray 1987; Beacham et al. 2002; Beacham and Withler 1985) and in some cases this differentiation is correlated with evolutionary significant units (ESUs) for the species (Wood and Foote 1996). The homing behavior of salmon may facilitate genetic differentiation (which may occur by genetic forces in addition to selection), and indeed populations forced into genetic isolation (e.g. chinook salmon introduced into New Zealand from California approximately 90 years ago) already display evidence of phenotypic divergence between populations (Kinnison et al. 1998). Thus, natural populations carry unique reservoirs of genes and gene combinations, which are specially suited for providing adaptive phenotypes in the environment in which they evolved. Consequently, interactions between strains of fish with different genetic backgrounds are anticipated in some cases to genetically alter local native populations and reduce their viability by altering local phenotypic adaptations (Bams 1976; Verspoor 1998) and threatening intraspecific genetic diversity (Ryman 1997). 229 Genetic variation cushions effect of deleterious genes In addition to providing a cushion against extinction from environmental change, genetic variability in fish populations can hide deleterious recessive genes. Under non-equilibrium conditions, as the frequency of heterozygotes in the population decreases there is a loss of alleles and an increase in the expression of recessive alleles that may reduce survival when expressed. Indeed, heterozygosity, a measure of genetic variability, has been observed to be positively correlated with measures of fitness in several instances (Wang et al. 2002). For example, a study on six strains of hatchery-reared rainbow trout showed that higher heterozygosity related to faster development rate (as measured by hatching time) and egg size (Danzmann et al. 1985; Danzmann et al. 1986; Ferguson et al. 1985). In chinook salmon (Arkush et al. 2002) and rainbow trout (Ferguson and Drahushchak 1990) higher heterozygosity was associated with greater disease resistance. However, heterozygosity is not always correlated with fitness (Hutchings and Ferguson 1992). GENERAL CONCLUSION Escaped farmed salmon have made their presence known in the wild, and although their numbers may have a large variation, in general there is a positive correlation between the number of escaped farmed fish in the population and the size of the local aquaculture industry. Farmed fish demonstrate fitness-related differences relative to wild fish including less responsiveness to risk of predation and altered foraging strategies that tend to incur higher metabolic costs. For these reasons, and other considerations as discussed above, farmed salmon tend to have lower survival than wild fish. However, some farmed fish do survive and evidence indicates that these fish may be capable of successfully reproducing in the wild. As a consequence, the farmed fish may successfully pass along their genetic material with a general trend of lower genetic variability in addition to altered fitness-related traits. The genetic effects on interaction between cultured and wild fish populations will depend on the frequency and magnitude of escapes, their ability to interbreed with conspecifics, the size of the receiving population, and, most critically, the fitness of the domesticated genotypes in nature. Based on the present studies, the effect of a small introgression of farm alleles into a wild population would be anticipated to be diluted with repeated backcrosses in the absence of selection, and little phenotypic effect would be detectable after just two or more generations. In contrast, escapes of large numbers of farmed individuals into small populations, or repeated escapes of moderate numbers over several generations, would be anticipated to have an effect on the phenotype of the receiving population. The critical question - as yet unanswered - is whether natural selection can restore into such introgressed populations, genotypes yielding phenotypes and fitness well adapted for current and future environmental conditions. 230 GLOSSARY Aquacultured - fish reared in culture throughout their entire life, usually for commercial purposes Cultured - broad term referring to all fish not entirely reared in the wild; often applied to strains that have been reared throughout their entire life history in aquaria, such as in captive brood programs Domestic - A formal definition of domestication was proposed by (Price 1984) as "that process by which a population of animals becomes adapted to man and to the captive environment by some combination of genetic changes occurring over generations and environmentally induced developmental events reoccurring during each generation"; domestication effects can be acquired very rapidly (within a single generation), and thus essentially all cultured fish may be at least partly domestic Farmed - synonymous with aquacultured Founder effect - chance change in the frequency of some genetic variants in populations as a result of a relatively small number of initial founders of the population Genetic drift - random change in allelic frequencies that results from the sampling of gametes from generation to generation Genotype x environment (G x E) interactions - Non-linear responses of phenotypes to environmental conditions observed among genotypes. Hatchery - fish reared for part of their life history in hatchery facilities; usually produced from artificially spawned adult fish captured after returning from the wild, and are released back into the wild at various life history stages (depending on the species) Heterosis - fitness of hybrids that exceeds the mean performance of the parental lines Hybrid vigor - synonymous with heterosis Inbreeding - non-random mating (with respect to genotype) where the mating individuals are more closely related than those drawn from the population by chance Inbreeding depression - decline in mean fitness with increasing homozygosity within populations Ne - effective population size; effective number of breeding individuals; number of individuals that would give rise to the calculated sampling variance if they bred as an idealized population 2 3 1 Norm of reaction - the phenotypic expression of a genotype under different environmental conditions Outbreeding - non-random mating (with respect to genotype) where the mating individuals are less closely related than those drawn from the population by chance Outbreeding depression - fitness of hybrids is below that of the parental lines Plasticity - the environmentally sensitive production of alternative phenotypes by given genotypes. 232 R E F E R E N C E S Aarestrup, K., C. Nielsen, and S.S. Madsen. 2000. Relationship between gill Na+, K+-ATpase activity and downstream movement in domesticated and first-generation offspring of wild anadromous brown trout (Salmo trutta). Can. J. Fish. Aquat. Sci. 57: 2086-2095. Abrahams, M.V. , and A. Sutterlin. 1999. The foraging and anti-predator behaviour of growth-enhanced transgenic Atlantic salmon. Anim. Behav. 58: 933-942. Allendorf, F.W. 1991. Ecological and genetic effects of fish introductions: Synthesis and recommendations. Can. J. Fish. Aquat. Sci. 48: 178-181. Allendorf, F.W., and S.R. Phelps. 1980. Loss of genetic variation in a hatchery stock of cutthroat trout. Trans. Am. Fish. Soc. 109: 537-543. Altukhov, Y.P. , and E.A. Salmenkhova. 1990. Introductions of distinct stocks of chum salmon, Oncorhynchus keta (Walbaum) into natural populations of the species. J. Fish Biol. 37: • 25-33. Arkush, K.D. , A.R. Giese, H.L. Mendonca, A . M . McBride, G.D. Marty, and P.W. Hedrick. 2002. Resistance to three pathogens in the endangered winter-run chinook salmon (oncorhynchus tshawytscha): Effects of inbreeding and major histocompatibility complex genotypes. Can. J. Fish. Aquat. Sci. 59: 966-975. Ayles, G.B., and R.F. Baker. 1983. Genetic differences in growth and survival between strains and hybrids of rainbow trout (salmo gairdneri) stocked in aquaculture lakes in the Canadian Prairies. Aquaculture 33: 269-280. Bachman, R.A. 1984. Foraging behavior of free-ranging wild and hatchery brown trout in a stream. Trans. Am. Fish. Soc. 113: 1-32. Bams, R.A. 1976. Survival and propensity for homing as affected by presence or absence of locally adapted paternal genes in two transplanted populations of pink salmon (Oncorhynchus gorbuscha). J. Fish. Res. Board Can. 33: 2716-2725. Barbat-Leterrier, A. , R. Guyomard, and F. Krieg. 1989. Introgression between introduced domesticated strains and Mediterranean native populations of brown trout (Salmo trutta L.). Aquat. Living. Resour. 2: 215-223. Beacham, T.D., and R.E. Withler. 1985. Heterozygosity and morphological variability of chum salmon (Oncorhynchus keta) in southern British Columbia. Heredity 54: 313-322. Beacham, T.D., and C B . Murray. 1987. Adaptive variation in body size age, morphology, egg size and developmental biology of chum salmon (Oncorhynchus keta) in British Columbia. Can. J. Fish. Aquat. Sci. 44: 244-261. Beacham, T.D., R.E. Withler, C B . Murray, and L.W. Barner. 1988. Variation in body size, morphology, egg size, and biochemical genetics of pink salmon in British Columbia. Trans. Am. Fish. Soc. 117: 109-126. Beacham, T.D., K.J . Supernault, M . Wetklo, B. Deagle, K. Labaree, J.R. Irvine, J.R. Candy, K . M . Miller, R.J. Nelson, and R.E. Withler. 2002. The geographic basis for population structure in fraser river chinook salmon (Oncorhynchus tshawytscha). Fish. Bull. 101: 229-242. Berejikian, B.A. , S.B. Mathews, and T.P. Quinn. 1996. Effects of hatchery and wild ancestry and rearing environments on the development of agonistic behaviour in steelhead trout (Oncorhynchus mykiss) fry. Can. J. Fish. Aquat. Sci. 53: 2004-2014. 233 Berejikian, B.A. , E.P. Tezak, S.L. Schroder, C M . Knudsen, and J.J. Hard. 1997. Reproductive behavioral interactions between wild and captively reared coho salmon (Oncorhynchus kisutch). Academic Press, London, U K . Berejikian, B.A. , E.P. Tezak, L. Park, E. LaHood, S.L. Schroder, and E. Beall. 2001. Male competition and breeding success in captively reared and wild coho salmon (Oncorhynchus kisutch). Can. J, Fish. Aquat. Sci. 58: 804-810. Berg, S. and J0rgensen, J. 1991. Stocking experiments with 0+ and 1+ trout parr, Salmo trutta L . , of wild and hatchery origin: 2. Post-stocking movements. J. Fish Biol. 39: 171-180. Berrebi, P., C. Poteaux, M . Fissier, and G. Cattaneo-Berrebi. 2000. Stocking impact and allozyme diversity in brown trout from Mediterranean southern France. J. Fish Biol. 56: 949-960. Bessey, C , R.H. Devlin, N.R. Liley, and C A . Biagi. 2004. Reproductive performance of growth-enhanced transgenic coho salmon. Trans. Am. Fish. Soc. 133: 1205-1220. Bourke, E.A., J. Coughlan, H. Jansson, P. Galvin, and T.F. Cross. 1997. Allozyme variation in populations of Atlantic salmon located throughout europe: Diversity that could be compromised by introductions of reared fish? ICES J. Mar. Sci. 54: 974-985. Bryden, C.A., J.W. Heath, and D.D. Heath. 2004. Performance and heterosis in farmed and wild chinook salmon (Oncorhynchus tshawytscha) hybrid and purebred crosses. Aquaculture 235: 249-261. Campton, D.E. and J.M. Johnston. 1985. Electrophoretic evidence for a genetic admixture of native and non-native rainbow trout in the Yakima River, Washington. Trans. Am. Fish. Soc. 114: 782-793. Cagigas, M.E. , E. Vazquez, G. Blanco, and J.A. Sanchez. 1999. Genetic effects of introduced hatchery stocks on indigenous brown trout (Salmo trutta L.) populations in Spain. Ecol. Freshw. Fish 8: 141-150. Carline, R.F., and J.F. Machung. 2001. Critical thermal maxima of wild and domestic strains of trout. Trans. Am. Fish. Soc. 130: 1211-1216. Carr, J.W., J .M. Anderson, F.G. Whoriskey, and T. Dilworth. 1997a. The occurrence and spawning of cultured Atlantic salmon (Salmo salar) in a Canadian river. Academic Press, London, U K . Carr, J.W., G.L. Lacroix, J.M. Anderson, and T. Dilworth. 1997b. Short communication: Movements of non-maturing cultured Atlantic salmon (Salmo salar) in a Canadian river. Academic Press, London ,UK. Clarke, W.C., R.E. Withler, and J.E. Shelbourn. 1994. Inheritance of smolting phenotypes in backcrosses of hybrid stream-type x ocean-type chinook salmon (Oncorhynchus tshawytscha). Estuaries 17: 13-25. Clarke, W.C., C. Groot, and L. Margolis. 1995. Physiological ecology of pacific salmon. U B C Press, Vancouver. Clifford, S.L., P. McGinnity, and A. Ferguson. 1997. Genetic changes in an Atlantic salmon population resulting from escaped juvenile farm salmon. J. Fish Biol. 52: 118-127. Clifford, S.L., P. McGinnity, and A. Ferguson. 1998. Genetic changes in Atlantic salmon (Salmo salar) populations of northwest Irish rivers resulting from escapes of adult farm salmon. Can. J. Fish. Aquat. Sci.55: 358-363. Cotter, D., V . O'Donovan, N . O'Maoileidigh, G. Rogan, N . Roche, and N.P. Wilkins. 2000. A n evaluation of the use of triploid Atlantic salmon (Salmo salar L.) in minimising the impact of escaped farmed salmon on wild populations. Aquaculture 186: 61-75. 234 Cross, T.F., and J. King. 1983. Genetic effects of hatchery rearing in Atlantic salmon. Aquaculture 33: 33-40. Crossman, E.J. 1991. Introduced freshwater fishes: A review of the North American perspective with emphasis on Canada. Can. Tech. Rep. Fish. Aquat. Sci. 48: Crozier, W.W. 1993. Evidence of genetic interaction between escaped farmed salmon and wild Atlantic salmon (Salmo salar L.) in a northern Irish river. Aquaculture 113: 19-29. Crozier, W.W. 1998. Incidence of escaped farmed salmon, Salmo salar L . , in commercial salmon catches and fresh water in Northern Ireland. Fish. Manage. Ecol. 5: 23-29. Currens, K.P., A.R. Hemmingsen, R.A. French, D.V. Buchanan, C B . Schreck, and H.W. L i . 1997. Introgression and susceptibility to disease in a wild population of rainbow trout. N . Am. J. Fish. Manage. 17(4): 1065-1078. Danielsdottir, A . K . , G. Marteinsdottir, F. Arnason, and S. Gudjonsson. 1997. Genetic structure of wild and reared Atlantic salmon (Salmo salar L.) populations in Iceland. ICES J. Mar. Sci. 54: 986-997. Dannewitz, J., E. Petersson, J. Dahl, T. Prestegaard, A. -C. Lof, and T. Jarvi. 2004. Reproductive success of hatchery-produced and wild-born brown trout in an experimental stream. J. Appl. Ecol. 41: 355-364. Danzmann, R.G., M . M . Ferguson, and F.W. Allendorf. 1985. Does enzyme heterozygosity influence developmental rate in rainbow trout? Heredity 56: 417-425. Danzmann, R.G., M . M . Ferguson, F.W. Allendorf, and K . L . Knudsen. 1986. Heterozygosity and developmental rate in a strain of rainbow trout (Salmo gairdneri). Evolution 40: 86-93. Deverill, J.L, C E . Adams, and C W . Bean. 1999. Prior residence, aggression and territory acquisition in hatchery-reared and wild brown trout. J. Fish Biol. 55: 868-875. Devlin, R.H., and E . M . Donaldson. 1992. Containment of genetically altered fish with emphasis on salmonids. World Scientific, Singapore (Singapore). Devlin, R.H., C A . Biagi, T.Y. Yesaki, D.E. Smailus, and J.C. Byatt. 2001. Growth of domesticated transgenic fish. Nature 409: 781-782. Devlin, R.H., C A . Biagi, and T.Y. Yesaki. 2004. Growth, viability and genetic characteristics of gh transgenic coho salmon strains. Aquaculture 236: 607-632. Dickson, T.A., and H.R. MacCrimmon. 1982. Influence of hatchery experience on growth and behavior of juvenile Atlantic salmon (Salmo salar) within allopatric and sympatric stream populations. Can. J. Fish. Aquat. Sci. 39: 1453-1458. Doyle, R.W., and A.J . Talbot. 1986. Artificial selection on growth and correlated selection on competitive behaviour in fish. Can. J. Fish. Aquat. Sci. 43: 1059-1064. Dunmall, K . M . , and J.F. Schreer. 2003. A comparison of the swimming and cardiac performance of farmed and wild Atlantic salmon, Salmo salar, before and after gamete stripping. Aquaculture 220: 869-882. Einum, S., and L A . Fleming. 1997. Genetic divergence and interactions in the wild among native, farmed and hybrid Atlantic salmon. J. Fish Biol. 50: 634-651. Einum, S., and L A . Fleming. 2001. Implications of stocking: Ecological interactions between wild and released salmonids. Nord. J. Freshw. Res. 75: 56-70. Ferguson, M . M . , and L.R. Drahushchak. 1990. Disease resistance and enzyme heterozygosity in rainbow trout. Heredity 64: 413-417. Ferguson, M . M . , R.G. Danzmann, and F.W. Allendorf. 1985. Developmental divergence among hatchery strains of rainbow trout (Salmo gairdneri), part 2. Hybrids. Canadian Journal of Genetics and Cytology 27: 298-307. 235 Fleming, L A . 1994. Captive breeding and the conservation of wild salmon populations. Conserv. Biol. 8: 886-888. Fleming, I. A . 1998. Pattern and variability in the breeding system of Atlantic salmon (salmo salar), with comparisons to other salmonids. Can. J. Fish. Aquat. Sci. 55 (Supplement 1): 59-76. Fleming, I.A., and M.R. Gross. 1989. Evolution of adult female life history and morphology in a pacific salmon (coho: Oncorhynchus kisutch). Evolution 43: 141-157. Fleming, I. A. , and M.R. Gross. 1992. Reproductive behavior of hatchery and wild coho salmon (Oncorhynchus kisutch): Does it differ? Aquaculture 103: 101-121. Fleming, I.A., and M.R. Gross. 1993. Breeding success of hatchery and wild coho salmon (Oncorhynchus kisutch) in competition. Ecol. Appl. 3: 230-245. Fleming, I.A., and S. Einum. 1997. Experimental tests of genetic divergence of farmed from wild Atlantic salmon due to domestication. ICES J. Mar. Sci. 54: 1051-1063. Fleming, I. A. , and E. Petersson. 2001. The ability of released, hatchery salmonids to breed and contribute to the natural productivity of wild populations. Nord. J. Freshw. Res. 75: 71-98. Fleming, I.A., B. Jonsson, and M.R. Gross. 1994. Phenotypic divergence of sea-ranched, farmed, and wild salmon. Can. J. Fish. Aquat. Sci.51: 2808-2824. Fleming, I. A. , B. Jonsson, M.R. Gross, and A. Lamberg. 1996. An experimental study of the reproductive behaviour and success of farmed and wild Atlantic salmon (Salmo salar). J. Appl. Ecol. 33: 893-905. Fleming, I.A., K. Hindar, I.B. Mjoelneroed, B. Jonsson, T. Balstad, and A . Lamberg. 2000. Lifetime success and interactions of farm salmon invading a native population. Proc. R. Soc. Lond., B. 267: 1517-1523. Fleming, I.A., T. Agustsson, B. Finstad, J.L Johnsson, and B.T. Bjornsson. 2002. Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 59: 1323-1330. Flick, W.A., and D.A. Webster. 1964. Comparative first year survival and production in wild and domestic strains of brook trout, Salvelinus fontinalis. Trans. Am. Fish. Soc. 93: 58-69. Flick, W.A., and D.A. Webster. 1976. Production of wild, domestic, and interstrain hybrids of brook trout (Salvelinus fontinalis) in natural ponds. J. Fish. Res. Board Can. 33: 1525— 1539. Foote, C.J., C C . Wood, and R.E. Withler. 1989. Biochemical genetic comparison of sockeye salmon and kokanee, the anadromous and nonanadromous forms of Oncorhynchus nerka. Can. J. Fish. Aquat. Sci.6: 149-158. Fraser, J.M. 1981. Comparative survival and growth of planted wild, hybrid, and domestic strains of brook trout (salvelinus fontinalis ) in Ontario lakes. Can. J. Fish. Aquat. Sci. 38: 1672-1684. Friars, G.W., J.K. Bailey, and R.L. Saunders. 1979. Considerations of a method of analyzing diallel crosses of Atlantic salmon. Canadian Journal of Genetics and Cytology 21: 121-128. Fritzner, N.G. , M . M . Hansen, S.S. Madsen, and K. Kristiansen. 2001. Use of microsatellite markers for identification of indigenous brown trout in a geographical region heavily influenced by stocked domesticated trout. J. Fish Biol. 58: 1197-1210. 236 Garant, D., J.J. Dodson, and L. Bematchez. 2003. Differential reproductive success and heritability of alternative reproductive tactics in wild Atlantic salmon (Salmo salar 1.). Evolution 57: 1133-1141. Garcia-Marin, J.L., P.E. Jorde, N . Ryman, F . M . Utter, and C. Pla. 1991. Management implications of genetic differentiation between native and hatchery populations of brown trout (Salmo trutta) in Spain. Aquaculture 95: 235-249. Garcia-Marin, J.L., N . Sanz, and C. Pla. 1999. Erosion of the native genetic resources of brown trout in spain. Ecol. Freshw. Fish 8: 151-158. Gausen, D., and V. Moen. 1991. Large-scale escapes of farmed Atlantic salmon (salmo salar) into Norwegian rivers threaten natural populations. Can. J. Fish. Aquat. Sci. 48: 426-428. Gharrett, A.J. , and W.W. Smoker. 1991. Two generations of hybrids between even- and odd-year pink salmon (Oncorhynchus gorbuscha). Can. J. Fish. Aquat. Sci. 48: 1744-1749. Gharrett, A.J. , W.W. Smoker, R.R. Reisenbichler, and S.G. Taylor. 1999. Outbreeding depression in hybrids between odd- and even-broodyear pink salmon. Aquaculture 173: 117-130. Gilk, S.E., LA. Wang, C.L. Hoover, W.W. Smoker, S.G. Taylor, A . K . Gray, and A.J . Gharrett. 2004. Outbreeding depression in hybrids between spatially separated pink salmon, Oncorhynchus gorbuscha, populations: Marine survival, homing ability, and variability in family size. Environ. Biol. Fish. 69: 287-297. Gjedrem, T. 1979. Selection for growth-rate and domestication in Atlantic salmon. J. Anim. Breed. Genet. 96: 56-59. Gjedrem, T. 1983. Genetic variation in quantitative traits and selective breeding in fish and shellfish. Aquaculture 33: 51-72. Gjedrem, T. 2000. Genetic improvement of cold-water fish species. Aquae. Res. 31: 25-33. Gjerde, B. 1986. Growth and reproduction in fish and shellfish. Aquaculture 57: 37-56. Gjerde, B. 1988. Complete diallele cross between six inbred groups of rainbow trout, Salmo gairdneri. Aquaculture 75: 71-87. Gjerde, B., and T. Refstie. 1984. Complete diallel cross between five strains of Atlantic salmon. Livestock Production Science 11: 207-226. Gjerde, B., K. Gunnes, and T. Gjedrem. 1983. Effect of inbreeding on survival and growth in rainbow trout. Aquaculture 34: 327-332. Groot, C , and L. Margolis. 1991. Pacific salmon life histories. 564 pp. Gudjonsson, S. 1991. Occurrence of reared salmon in natural salmon rivers in Iceland. Aquaculture 98: 133-142. Hallerman, E .M. , and A.R. Kapuscinski. 1992. Ecological implications of using transgenic fishes in aquaculture. ICES J. Mar. Sci. Symp. Handeland, S.O., B.T. Bjornsson, A . M . Arnesen, and S.O. Stefansson. 2003. Seawater adaptation and growth of post-smolt Atlantic salmon (Salmo salar) of wild and farmed strains. Aquaculture 220: 367-384. Hansen, L.P., and B. Jonsson. 1994. Development of sea ranching of Atlantic salmon, Salmo salar 1., towards a sustainable aquaculture strategy. Aquae. Fish. Manage. 25: 199-214. Hansen, L.P., K . B . Doving, and B. Jonsson. 1987. Migration of farmed adult Atlantic salmon with and without olfactory sense, released on the Norwegian coast. J. Fish Biol. 30: 713-721. 237 Hansen, L.P., J.A. Jacobsen, and R.A. Lund. 1997. The incidence of escaped farmed Atlantic salmon, salmo salar /., in the faroese fishery and estimates of catches of wild salmon. ICES, Copenhagen, Denmark. Hansen, M . M . , E.D. Ruzzante, E.E. Nielsen, and D.K. Mensberg. 2000. Microsatellite and mitochondrial D N A polymorphism reveals life-history dependent interbreeding between hatchery and wild brown trout {Salmo trutta L.). Mol . Ecol. 9: 583-594. Hansen, M . M . , E.E. Nielsen, D. Bekkevold, and K . L . Mensberg. 2001. Admixture analysis and stocking impact assessment in brown trout (Salmo trutta), estimated with incomplete baseline data. Can. J. Fish. Aquat. Sci. 58: 1853-1860. Hard, J.J., B.A. Berejikian, E.P. Tezak, S.L. Schroder, C M . Knudsen, and L.T. Parker. 2000. Evidence for morphometric differentiation of wild and captively reared adult coho salmon: A geometric analysis. Environ. Biol. Fish. 58: 61-73. Heggberget, T.G., F. 0kland, and O. Ugedal. 1993. Distribution and migratory behaviour of adult wild and farmed Atlantic salmon (Salmo salar) during return migration. Aquaculture 118: 73-83. Hershberger, W.K., J .M. Myers, R.N. Iwamoto, W . C Macauley, and A . M . Saxton. 1990. Genetic changes in the growth of coho salmon (Oncorhynchus kisutch) in marine net-pens, produced by ten years of selection. Aquaculture 85: 187-197. Hindar, K. , and T. Balstad. 1994. Salmonid culture and interspecific hybridization. Conserv. Biol. 8: 881-882. Hindar, K. , N . Ryman, and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Can. J. Fish. Aquat. Sci.48: 945-957. Hutchings, J.A. 1991. The threat of extinction to native populations experiencing spawning intrusions by cultured Atlantic salmon. Aquaculture 98: 119-132. Hutchings, J.A. 2004. Norms of reaction and phenotypic plasticity in salmonid life histories, p. 510. In A. P. Hendry and S. C. Stearns, eds. Evolution illuminated: Salmon and their relatives. Oxford University Press, Inc., New York. Hutchings, J.A., and M . Ferguson. 1992. The independence of enzyme heterozygosity and life-history traits in natural populations of Salvelinus fontinalis (brook trout). Heredity 69: 496-502. Hvidsten, N.A. , and R.A. Lund. 1988. Predation on hatchery-reared and wild smolts of Atlantic salmon, Salmo salar L. , in the estuary of river Orkla, Norway. J. Fish Biol. 33: 121-126. Hynes, J.D., E.H. Brown, J.H. Helle, N . Ryman, and D.A. Webster. 1981. Guidelines for the culture of fish stocks for resource management. Can. J. Fish. Aquat. Sci. 38: 1867-1876. Jacobsen, J.A., and L.P. Hansen. 2001. Feeding habits of wild and escaped farmed Atlantic salmon, salmo salar 1., in the northeast Atlantic. ICES J. Mar. Sci. 58: 916-933. Jacobsen, J.A., L.P. Hansen, and R.A. Lund. 1992. Occurrence of farmed salmon in the norwegian sea. Ices, Copenhagen (Denmark). Johnsen, B.O., and O. Ugedal. 1986. Feeding by hatchery-reared and wild brown trout, salmo trutta 1., in a Norwegian stream. Aquacult. Fish. Manage. 17: 281-287. Johnsson, J.L, and M . V . Abrahams. 1991. Interbreeding with domestic strain increases foraging under threat of predation in juvenile steelhead trout (Oncorhynchus mykiss): An experimental study. Can. J. Fish. Aquat. Sci. 48: 243-247. Johnsson, J.L, E. Petersson, E. Jonsson, B.T. BjSrnsson, and T. Jarvi. 1996. Domestication and growth hormone alter antipredator behaviour and growth patterns in juvenile brown trout, Salmo trutta. Can. J. Fish. Aquat. Sci.53: 1546-1554. 238 Johnsson, J.L, J. Hojesjo, and L A . Fleming. 2001. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can. J. Fish. Aquat. Sci. 58: 788-794. Jonsson, N . , and B. Jonsson. 1999. Trade-off between egg mass and egg number in brown trout. J. Fish Biol. 55: 767-783. Kallio-Nyberg, I., and M . L . Koljonen. 1997. The genetic consequence of hatchery-rearing on life-history traits of the Atlantic salmon (Salmo salar L.): A comparative analysis of sea-ranched salmon with wild and reared parents. Aquaculture 153: 207-224. Kazakov, R.V., and O.V. Semenova. 1986. Morphological characteristics of hatchery-reared and wild young Atlantic salmon Salmo salar L. Proc. Zoological Institute, Russian Academy of Science. Keller, W.T., and D.S. Plosila. 1981. Comparison of domestic, hybrid and wild strains of brook trout in a pond fishery. N . Y . Fish Game J. 28: 123-137. Kennedy, G.J.A., and J.E. Greer. 1988. Predation by cormorants, Phalacrocorax carbo (L.), on the salmonid populations of an.Irish river. Aquacult. Fish. Manage. 19: 159-170. King, T.L., S.T. Kalinowski, W.B. Schill, A.P. Spidle, and B.A. Lubinski. 2001. Population structure of Atlantic salmon (Salmo salar 1.): A range-wide perspective from microsatellite D N A variation. Mol. Ecol. 10: 807-821. Kinnison, M . , M . Unwin, N . Boustead, and T. Quinn. 1998. Population-specific variation in body dimensions of adult chinook salmon (Oncorhynchus tshawytscha) from new Zealand and their source population, 90 years after introduction. Can. J. Fish. Aquat. Sci. 55:554-563. Krueger, C C , and B. May. 1991. Ecological and genetic effects of salmonid introductions in North America. Can. J. Fish. Aquat. Sci. 48: 66-77. Lachance, S., and P. Magnan. 1990. Performance of domestic, hybrid, and wild strains of brook trout, Salvelinus fontinalis, after stocking: The impact of intra- and interspecific competition. Can. J. Fish. Aquat. Sci. 47: 2278-2284. Lacroix, G.L., B.J. Galloway, D. Knox, and D. MacLatchy. 1997. Absence of seasonal changes in reproductive function of cultured Atlantic salmon migrating into a Canadian river. ICES J. Mar. Sci. 54: 1086-1091. LeClair, L .L . , S.R. Phelps, and T.J. Tynan. 1999. Little gene flow from a hatchery strain of chum salmon to local wild populations. N . Am. J. Fish. Manage. 19(2): 530-535. Leider, S.A., M.W. Chilcote, and J.J. Loch. 1986. Comparative life history characteristics of hatchery and wild steelhead trout (Salmo gairdneri) of summer and winter races in the kalama river, Washington. Can. J. Fish. Aquat. Sci. 43: 1398-1409. Leonard, J.B.K., and S.D. McCormick. 2001. Metabolic enzyme activity during smolting in stream- and hatchery-reared Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 58: 1585-1593. Levings, C D . , C D . McAllister, and B.D. Change. 1986. Differential use of the campbell river estuary, British Columbia, by wild and hatchery-reared juvenile chinook salmon (Oncorhynchus tshawytscha). Can. J. Fish. Aquat. Sci.43: 1386-1397. Lund, R.A., L.P. Hansen, and F. Okland. 1994. Escaped farmed salmon and geographical zones established for wild fish protection. NINA Oppdragsmelding 303: 15. Lura, H. , and H. Saegrov. 1991. A method of separating offspring from farmed and wild Atlantic salmon (Salmo salar) based on different ratios of optical isomers of astaxanthin. Can. J. Fish. Aquat. Sci. 48: 429^133. 239 Lura, H. , and H. Sasgrov. 1993. Timing of spawning in cultured and wild Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) in the river Vosso, Norway. Ecol. Freshw. Fish 2: 167-172. Lura, H. , and F. 0kland. 1994. Content of synthetic astaxanthin in escaped farmed Atlantic salmon, Salmo salar, 1. Ascending Norwegian rivers. Fish. Manage. Ecol. 1: 205-216. Mason, J.W., O.M. Brynilds, and P.E. Degurse. 1967. Comparative survival of wild and domestic strains of brook trout in streams. Trans. Am. Fish. Soc. 96: 313-&. McClelland, E.K., J .M. Myers, J.J. Hard, L .K. Park, and K . A . Naish. 2005. Two generations of outbreeding in coho salmon (Oncorhynchus kisutch): Effects on size and growth. Can. J. Fish. Aquat. Sci. 62: 2538-2547. McGinnity, P., C. Stone, J.B. Taggart, D. Cooke, D. Cotter, R. Hynes, C. McCamley, T. Cross, and A. Ferguson. 1997. Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: Use of D N A profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J. Mar. Sci. 54: 998-1008. McGinnity, P., P. Prodohl, A . Ferguson, R. Hynes, N . 6. Maoileidigh, N . Baker, D. Cotter, B. O'Hea, D. Cooke, G. Rogan, J. Taggart, and T. Cross. 2003. Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proc. R. Soc. Lond., B. 270: 2443-2450. McKinnell, S., A.J . Thomson, E.A. Black, B.L. Wing, C M . Guthrie, III, J.F. Koerner, and J.H. Helle. 1997. Atlantic salmon in the north pacific. Aquacult. Res. 28: 145-157. Mesa, M.G. 1991. Variation in feeding, aggression, and position choice between hatchery and wild cutthroat trout in an artificial stream. Trans. Am. Fish. Soc. 120: 723-727. Mjelnerod, I.B., U.H. Refseth, E. Karlsen, T. Balstad, K.S. Jakobsen, and K . Hindar. 1997. Genetic differences between two wild and one farmed population of Atlantic salmon (Salmo salar) revealed by three classes of genetic markers. Hereditas 127: 239-248. Mork, O.I., B. Bjerkeng, and M . Rye. 1999. Aggressive interactions in pure and mixed groups of juvenile farmed and hatchery-reared wild Atlantic salmon Salmo salar L . in relation to tank substrate. Aquacult. Res. 30: 571-578. Morton, A. , and J.P. Volpe. 2002. A description of escaped farmed Atlantic salmon Salmo salar captures and their characteristics in one pacific salmon fishery area in British Columbia, Canada, in 2000. Alaska Fishery Research Bulletin 9: 102-110. Munakata, A. , B.T. Bjornsson, E. Jonsson, M . Amano, K. Ikuta, S. Kitamura, T. Kurokawa, and K. Aida. 2000. Post-release adaptation processes of hatchery-reared honmasu salmon parr. J. Fish Biol. 56: 163-172. Nagata, M . , M . Nakajima, and M . Fujiwara. 1994. Dispersal of wild and domestic masu salmon fry (Oncorhynchus masou) in an artificial channel. Journal of Fish Biology 45: 99-110. Noakes, D.J., R.J. Beamish, and M . L . Kent. 2000. On the decline of pacific salmon and speculative links to salmon farming in British Columbia. Aquaculture 183: 363-386. Okland, F., T.G. Heggberget, and B. Jonsson. 1995. Migratory behaviour of wild and farmed Atlantic salmon (Salmo salar) during spawning. J. Fish Biol . 46: 1-7. Palm, S., J. Dannewitz, T. Jarvi, E. Petersson, T. Prestegaard, and N . Ryman. 2003. Lack of molecular genetic divergence between sea-ranched and wild sea trout (Salmo trutta). Mol. Ecol. 12: 2057-2071. Peck, J.W. 1988. Fecundity of hatchery and wild lake trout in Lake Superior. Journal of Great Lakes Research 14: 9-13. 240 Peterson, R.G. 1999. Potential genetic interaction between wild and farm salmon of the same species. Pp. 23. The Office of the Commissioner for Aquaculture Development, Fisheries and Oceans Canada. Petersson, E., and T. Jarvi. 1995. Evolution of morphological traits in sea trout (Salmo trutta) parr (0+) through sea-ranching. Nord. J. Freshw. Res. Drottningholm 70: 62-67. Petersson, E., T. Jarvi, N.G. Steffher, and B. Ragnarsson. 1996. The effect of domestication on some life history traits of sea trout and Atlantic salmon. J. Fish Biol. 48: 776-791. Poppe, T.T., H. Hellberg, D. Griffiths, and H. Meldal. 1997. Swimbladder abnormality in farmed Atlantic salmon Salmo salar. Dis. Aquat. Org. 30: 73-76. Poppe, T.T., R. Johansen, G. Gunnes, and B. T 0 r u d . 2003. Heart morphology in wild and farmed Atlantic salmon Salmo salar and rainbow trout Oncorhynchus mykiss. Dis. Aquat. Org. 57: 103-108. Poteaux, C , D. Beaudou, and P. Berrebi. 1998a. Temporal variations of genetic introgression in stocked brown trout populations. J. Fish Biol. 53: 701-713. Poteaux, C , F. Bonhomme, and P. Berrebi. 1998b. Differences between nuclear and mitochondrial introgressions of brown trout populations from a restocked main river and its unrestocked tributary. Biological Journal of the Linnean Society 63: 379-392. Poteaux, C , F. Bonhomme, and P. Berrebi. 1999. Microsatellite polymorphism and genetic impact of restocking in Mediterranean brown trout (Salmo trutta L.). Heredity 82: 645-653. Price, E.O. 1984. Behavioral aspects of animal domestication. Quarterly Review of Biology 59: 1-32. Quinn, T. 2005. The behavior and ecology of Pacific salmon and trout. American Fisheries Society, Bethesda, Maryland, USA / U B C Press, Vancouver, B C , Canada. Quinn, T.P., M.J. Unwin, and M.T. Kinnison. 2000. Evolution of temporal isolation in the wild: Genetic divergence in timing of migration and breeding by introduced chinook salmon populations. Evolution 54: 1372-1385. Reisenbichler, R.R., and J.D. Mclntyre. 1977. Genetic differences in growth and survival of juvenile hatchery and wild steelhead trout, Salmo gairdneri. J. Fish. Res. Board Can. 34: 123-128. Rhodes, J.S., and T.P. Quinn. 1998. Factors affecting the outcome of territorial contests between hatchery and naturally reared coho salmon parr in the laboratory. J. Fish Biol. 53: 1220-1230. Robison, B.D., P.A. Wheeler, and G.H. Thorgaard. 1999. Variation in development rate among clonal lines of rainbow trout (Oncorhynchus mykiss). Aquaculture 173: 131-141. Robison, B.D., P.A. Wheeler, K. Sundin, P. Sikka, and G.H. Thorgaard. 2001. Composite interval mapping reveals a major locus influencing embryonic development rate in rainbow trout (Oncorhynchus mykiss). J. Hered. 92: 16-22. Ruzzante, D.E., and R.W. Doyle. 1991. Rapid behavioural changes in medaka (Oryzias latipes) caused by selection for competitive and noncompetitive growth. Evolution 45: 1936-1946. Ryman, N . 1997. Minimizing adverse effects of fish culture: Understanding the genetics of populations with overlapping generations. ICES J. Mar. Sci. 54: 1149-1159. Sanchez, M.P., B. Chevassus, L . Labbe, E. Quillet, and M . Mambrini. 2001. Selection for growth of brown trout (Salmo trutta) affects feed intake but not feed efficiency. Aquat. Liv. Res. 14:41-48. 241 Shrimpton, J.M., N.J. Bernier, and D.J. Randall. 1994. Cortisol dynamics and smolting in hatchery and wild coho salmon. Fish Physiology Association, Vancouver, B C (Canada). Skaala, 0. 1992. Genetic population structure of Norwegian brown trout. J. Fish Biol. 41: 631-646. Skaala, 0., G. Dahle, K .E . J 0 r s t a d , and G. Naevdal. 1990. Interactions between natural and farmed fish populations: Information from genetic markers. J. Fish Biol. 36: 449—460. Skaala, 0., K . E . J 0 r s t a d , and R. Borgstram. 1996. Genetic impact on two wild brown trout (Salmo trutta) populations after release of non-indigenous hatchery spawners. Can. J. Fish. Aquat. Sci.53: 2027-2035. Small, M.P., A .E . Pichahchy, J.F. Von Bargen and S.F. Young. 2004. Have native coho salmon (Oncorhynchus kisutch) persisted in the Nooksack and Samish rivers despite continuous hatchery production throughout the past century? Conservation Genetics 5: 367-379. Smoker, W.W., L A . Wang, A.J . Gharrett, and J.J. Hard. 2004. Embryo survival and smolt to adult survival in second-generation outbred coho salmon. J. Fish Biol. 65: 254-262. Stahl, G. 1987. Genetic population structure of Atlantic salmon. Pp. 121-140 in N . Ryman and F. Utter, eds. Population genetics and fishery management. University of Washington Press, Seattle. Sundstrom, F.L., R.H. Devlin, J. I. Johnsson, and C A . Biagi. 2003. Vertical position reflects increased feeding motivation in growth hormone transgenic coho salmon (Oncorhynchus kisutch). Ethology 109: 701-712. Sundstrom, L.F., M . Lohmus, and R.H. Devlin. 2005. Selection on increased intrinsic growth rates in coho salmon, Oncorhynchus kisutch. Evolution 59: 1560-1569. Sundstrom, L.F., M . Lohmus, R.H. Devlin, J.I. Johnsson, C A . Biagi, and T. Bohlin. 2004. Feeding on profitable and unprofitable prey: Comparing behaviour of growth-enhanced transgenic and normal coho salmon (Oncorhynchus kisutch). Ethology 110: 381-396. Swain, D.P., and B.E. Riddell. 1990. Variation in agonistic behavior between newly emerged juveniles from hatchery and wild populations of coho salmon, Oncorhynchus kisutch. Can. J. Fish. Aquat. Sci. 47: 566-571. Swain, D.P., B.E. Riddell, and C B . Murray. 1991. Morphological differences between hatchery and wild populations of coho salmon (oncorhynchus kisutch ): Environmental versus genetic origin. Can. J. Fish. Aquat. Sci. 48: 1783-1791. Taylor, E.B. 1991. A review of local adaptation in Salmonidae, with particular reference to Pacific and Atlantic salmon. Aquaculture 98: 185-207. Thodesen, J., B. Grisdale-Helland, S.J. Helland, and B. Gjerde. 1999. Feed intake, growth and feed utilization of offspring from wild and selected Atlantic salmon (Salmo salar). Aquaculture 180: 237-246. Thompson, C.E., W.R. Poole, M . A . Matthews, and A. Ferguson. 1998. Comparison, using minisatellite D N A profiling, of secondary male contribution in the fertilisation of wild and ranched Atlantic salmon (Salmo salar) ova. Can. J. Fish. Aquat. Sci. 55: 2011-2018. Thorpe, J.E. 2004. Life history response of fishes to culture. J. Fish Biol. 65: 263-285. Tymchuk, W.E., and R.H. Devlin. 2005. Growth differences among first and second generation hybrids of domesticated and wild rainbow trout (Oncorhynchus mykiss). Aquaculture 245: 295-300. Tymchuk, W.E., C A . Biagi, R.E. Withler, and R.H. Devlin. 2006. Growth and behavioural consequences of introgression of a domesticate aquaculture genotype into a native strain of coho salmon (Oncorhynchus kisutch). Trans. Am. Fish. Soc. 135: 442-445. 242 Ugedal, O., B. Finstad, B. Damsgard, and A. Mortensen. 1998. Seawater tolerance and downstream migration in hatchery-reared and wild brown trout. Aquaculture 168: 395-405. Utter, F., and J. Epifanio. 2002. Marine aquaculture: Genetic potentialities and pitfalls. Reviews in Fish Biology and Fisheries 12: 59-77. Vandersteen Tymchuk, W.E., M . V . Abrahams, and R.H. Devlin. 2005. Competitive ability and mortality of growth-enhanced transgenic coho salmon fry and parr when foraging for food. Trans. Am. Fish. Soc. 134: 381-389. Verspoor, E. 1988. Reduced genetic variability in first-generation hatchery populations of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci.45: 1686-1690. Verspoor, E. 1997. Genetic diversity among Atlantic salmon (Salmo salar L.) populations. Proceedings of an ICES/NASCO Symposium held in Bath, England, 18-22 April 1997, Academic Press, London (UK), Dec 1997, vol. 54, no. 6 Verspoor, E. 1998. Genetic impacts on wild Atlantic salmon (Salmo salar 1.) stocks from escaped farm conspecifics: An assessment of risk. 20 pp. Verspoor, E., and C.G. De Leaniz. 1997. Stocking success of Scottish Atlantic salmon in two Spanish rivers. J. Fish Biol. 51: 1265-1269. Volpe, J.P., E.B. Taylor, D.W. Rimmer, and B.W. Glickman. 2000. Evidence of natural reproduction of aquaculture-escaped Atlantic salmon in a coastal British Columbia river. Conservation Biology 14: 899-903. Wang, S., J. J. Hard, and F. Utter. 2002. Genetic variation and fitness in salmonids. Conservation Genetics 3: 321-333. Wangila, B.C.C., and T.A. Dick. 1996. Genetic effects and growth performance in pure and hybrid strains of rainbow trout, Oncorhynchus mykiss (Walbaum) (order: Salmoniformes, family: Salmonidae). Aquae. Res. 27: 35-41. Waples, R.S. 1.991. Genetic interactions between hatchery and wild salmonids: Lessons from the Pacific Northwest Webb, J.H., D.W. Hay, P.D. Cunningham, and A.F. Youngson. 1991. The spawning behaviour of escaped farmed and wild adult Atlantic salmon (Salmo salar 1.) in a northern Scottish river. Aquaculture. 98: 97-110. Webb, J.H., I.S. McLaren, M.J. Donaghy, and A.F . Youngson. 1993a. Spawning of farmed Atlantic salmon, Salmo salar L. , in the second year after their escape. Aquacult. Fish. Manage. 24: 557-561. Webb, J.H., A.F. Youngson, C E . Thompson, D.W. Hay, M.J. Donaghy, and I.S. McLaren. 1993b. Spawning of escaped farmed Atlantic salmon, Salmo salar L. , in western and northern Scottish rivers: Egg deposition by females. Aquacult. Fish. Manage. 24: 663-670. Webster, D.A., and W.A. Flick. 1981. Performance of indigenous, exotic, and hybrid strains of brook trout (Salvelinus fontinalis) in waters of the Adirondack mountains, New York. Can. J. Fish. Aquat. Sci. 38: 1701-1707. Weir, L .K . , J.A. Hutchings, L A . Fleming, and S. Einum. 2004. Dominance relationships and behavioural correlates of spawning success in farmed and wild male Atlantic salmon, Salmo salar. Journal of Animal Ecology 73: 1069-1079. Weiss, S., and S. Schmutz. 1999. Performance of hatchery-reared brown trout and their effects on wild fish in two small Austrian streams. Trans. Am. Fish. Soc. 128: 302-316. 243 Wilson, I.F., E.A. Bourke, and T.F. Cross. 1995. Genetic variation at traditional and novel allozyme loci, applied to interactions between wild and reared Salmo salar L. (Atlantic salmon). Heredity 75: 578-588. Wood, C.C.,.and C.J. Foote. 1996. Evidence for sympatric genetic divergence of anadromous and nonanadromous morphs of sockeye salmon (Oncorhynchus nerka). Evolution 50: 1265-1279. Youngson, A.F. , J.H. Webb, C E . Thompson, and D. Knox. 1993. Spawning of escaped farmed Atlantic salmon (Salmo salar): Hybridization of females with brown trout (Salmo trutta). Can. J. Fish. Aquat. Sci. 50: 1986-1990. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share