DIPLOID AND TRIPLOID CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA (WALBAUM): AN EXPLORATION OF INDUCTION EFFICACY, PERFORMANCE AND GENOMIC ARCHITECTURE by Rachael M. Johnson B.Sc., The University of Northern British Columbia, 2000 THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in NATURAL RESOURCES AND ENVIRONMENTAL STUDIES © Rachael M. Johnson, 2003 THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA February 2003 All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author. 1 ^ 1 National Library of Canada Bibliothèque nationale du Canada Acquisitions and Bibliographic Services Acquisitions et services bibliographiques 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue Wellington Ottawa ON K1A0N4 Canada Y o u r ^ Votn référence Naîreréiérence The author has granted a non­ exclusive hcence allowing the National Library o f Canada to reproduce, loan, distribute or sell copies o f this thesis in microform, paper or electronic formats. L’auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique. The author retains ownership of the copyright in this thesis. Neitiier the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author’s permission. L’auteur conserve la propriété du droit d’auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation. 0 612 80661-8 - CanadS - APPROVAL Name: Rachael M. Johnson Degree: Master of Science Thesis Title: DIPLOID AND TRIPLOID CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA (WALBAUM): AN EXPLORATION OF INDUCTION EFFICACY, PERFORMANCE AlW GENOMIC ARCHITECTURE Examining Committee: Chair; Dr. Robert Tait Dean of Graduate Studies UNBC Co-Supervisor: Dr. Mark Shrimpton Assistant Professor, Biology Program ÜNBC / Co-j^pervisor: Dr. Daniel Heath Adjunct Professor, Biology Program UNBC '"Committee Member: Dr. Assistant Professor, Biol UNBC lurray Program Committee Member: Dr. Robert Devlin Head, Molecular Biology Program Fisheries and Oceans Canada (West Vancouver, BC) -T V L E xa te r ^ Examiner: Dr. Tillmann Benfey e r^ Ex£ Professor - Department of Biology University of New Brunswick Date Approved: Abstract Triploid salmon are sterile and thus may comprise part of an overall plan to minimize potential genetic disturbances to wild populations caused by the rearing of farmed fish in open seawater netpens. Despite the potential benefits of sterility, triploids are not widely reared for aquacultural purposes in North America mainly due to variable and inconsistent performance. While triploidization is being explored in an increasing number of species, the effect of triploid ization in chinook salmon {Oncorhynchus tshawytscha) has rarely been investigated. In this study, chinook salmon were triploidized in order to assess 1) the efficacy of two triploid induction techniques 2) the utility of triploid chinook salmon for commercial aquaculture and 3) to develop a description of the genomic architecture of triploids. In Chapter One, a comparative examination of triploid ization success and whole organism performance (survival, growth and the antibody response to vaccination) in diploid, heat-shock induced triploid and pressure-shock induced triploid full-sib family groups was carried out in terms of the effect of treatment, genotype (family) and treatment by genotype (family) interactions. In Chapter Two, a comparative examination of performance (survival, growth, and the lysozyme activity response to vaccination) in terms of the distribution and magnitude of phenotypic variance was carried out using a quantitative genetic framework and a paternal half-sib experimental design. Variance was partitioned into additive genetic and a combined epistasis, dominance and maternal effects component and narrow sense heritability values were calculated. Minimal differences in triploidization success, growth and immune functioning were found between heat- and pressure-shock treated family groups. Although pressure-shock treated fish survived better than heat-shock induced triploids, triploids did not survive or grow as well as diploids. Survival of treatment groups was significantly 11 influenced by treatment and family effects while growth traits and antibody response to vaccination were more strongly influenced by the effect of family. Interaction effects were most prevalent for immune function. Triploidization increased total phenotypic as well as additive genetic variance but this was associated with an unexpected and counter-intuitive decrease in the influence of the non-additive component (combined epistatic, dominance and maternal effects) indicating that triploidy may not have increased the genetic complexity of relationships among alleles or loci and that the primary effect of triploidization was additive and dominant. This was also highly suggestive of an overall ploidy dependent regulation of gene expression. The obvious dichotomy between high and low performing families regardless of treatment/ploidy status, the existence of significant family components for many of the performance variables combined with increased heritability values for the measured traits indicated that selective breeding of diploids for increased triploid performance might be successful. However, the presence of family by treatment interactions (although explaining a relatively low amount of variance) observed in the Chapter One study, the increased range of phenotypic variance and profoundly different pattern of variance partitioning found in the Chapter Two study suggest that the production of a uniform fish product, at least during the freshwater period of growth might be compromised. Ill TABLE OF CONTENTS Abstract..........................................................................................................................ii Table of Contents.......................................................................................................... iv List of Tables..................................................................................................................vi List of Figures................................................................................................................ vii Acknowledgements.................................................................................. viii General Introduction.....................................................................................................1 CHAPTER ONE COMPARISONS AMONG DIPLOID, HEAT-SHOCK INDUCED TRIPLOID AND PRESSURE-SHOCK INDUCED TRIPLOID CHINOOK SALMON {ONCORHYNCHUS TSHAWYTSCHA) FOR SURVIVAL, GROWTH, IMMUNE FUNCTION AND TRIPLOIDIZATION EFFICACY 1.0 Abstract.................................................................................................................... 7 1.1 Introduction.............................................................................................................. 9 1.2 Methods....................................................................................................................12 1.2.1 Fish & Rearing Conditions..............................................................................12 1.2.2 Ploidy Determination.......................................................................................14 1.2.3 Survival............................................................................................................ 16 1.2.4 Growth..............................................................................................................16 1.2.5 Vaccination & EnzymeLinked Immunosorbent Assay (ELISA)....................17 1.2.6 Statistical Analysis.......................................................................................... 19 1.3 Results..................................................................................................................... 20 1.3.1 Triploidization Success...................................................................................20 1.3.2 Incubation Survival..........................................................................................23 1.3.3 Survival After Exogenous Feeding................................................................ 29 1.3.4 Size-at-age...................................................................................................... 30 1.3.5 Relative Growth R a te.................................................................................... 34 1.3.6 Response to Vaccination; Enzyme-Linked Immunosorbent Assay............ 37 1.4 Discussion............................. 40 IV CHAPTER TWO QUANTITATIVE GENETIC ANALYSIS OF DIPLOID AND TRIPLOID CHINOOK SALMON {ONCORHYNCHUS TSHAWYTSCHA) PERFORMANCE CHARACTERISTICS 2.0 Abstract.................................................................................................................... 46 2.1 Introduction.............................................................................................................. 47 2.2 Methods....................................................................................................................51 2.2.1 Fish Breeding Design & Treatment Details...................................................51 2.2.2 Fish Rearing & Husbandry............................................................................. 52 2.2.3 Ploidy Determination.......................................................................................53 2.2.4 Survival............................................................................................................ 53 2.2.5 Growth............................................................................................................. 54 2.2.6 Vaccination & Serum Lysozyme Activity Assay........................................... 56 2.2.7 Statistical Analysis.......................................................................................... 57 2.3. Results................ 60 2.3.1 Patterns of Variance Distribution Among & Within Families....................... 60 2.3.2 Sib-Analysis, Narrow Sense Heritability & Maternal Effects........................60 2.3.2.1 Size-at-age..............................................................................................65 2.3.2.2 Relative Growth Rate............................................................................. 66 2.3.2.3 Response to Vaccination: SerumLysozyme Activity............................ 68 2.3.2.4 Triploidization Success.......................................................................... 69 2.3.2.5 Sign Test of Heritability Differences......................................................69 2.3.3 Performance & TriploidizationSuccess..........................................................70 2.4 Discussion.............................................................................................................. 71 General Conclusions..................................................................................................... 80 Literature Cited.............................................................................................................. 82 List of Tables Chapter One Table 1.1, Analysis of variance components and omega squared values (w^) for red blood cell nuclear analysis..................................................................................22 Table 1.2. Analysis of variance components for survival to the eyed stage of development.....................................................................................................................24 Table 1.3. Analysis of variance components for freshwater survival after the eyed stage of development to the end of incubation...................................................24 Table 1.4. Analysis of variance components and omega squared values (w^) for survival after the start of exogenous feeding.......................................................... 29 Table 1.5. Analysis of variance components and omega squared values (w^) for size-at-age results......................................................................................................31 Table 1.6. Analysis of variance components and omega squared values (oo^) for relative growth rate....................................................................................................35 Table 1.7. Analysis of variance components and omega squared (co^) values for antibody titer.............................................................................................................. 37 Chapter Two Table 2.1. Summary of measured performance variables and in-text Abbreviations...................................................................................................................55 Table 2.2. Among and within family phenotypic variance of measured traits in diploid control and triploid pressure-shock treated family groups...........................61 Table 2.3. Variance components for the heritability analysis of triploidization, size-at-age and serum lysozyme activity measured in diploid control and triploid pressure-shock family treatment groups....................................................................... 62 Table 2.4. Variance components for the heritability analysis of relative growth rate and triploidization success measured in diploid control and triploid pressure-shock family groups........................................................................................ 63 Table 2.5. Sire component heritability (h^± SE) and maternal effect estimates (Vm, %) for measured performance traits..................................................................... 65 VI List of Figures Chapter One Figure 1.1. Frequency distribution of mean red blood cell nuclear length across treatment groups............................................................................................... 22 Figure 1.2 Reaction norms for family effects on mean treatment group triploidization success levels........................................................................................ 23 Figure 1.3. Mean cumulative survival (with 95% confidence intervals) of treatment groups through incubation........................................................................... 26 Figure 1.4. Reaction norms for family effects on mean treatment group survival during incubation............................................................................................. 27 Figure 1.5. Mean survival (95% Cl) of treatment groups from the onset of exogenous feeding to just prior to saltwater transfer................................................. 30 Figure 1.6. Mean size-at-age (weight in grams ± 95% 01) of family treatment groups at three freshwater time points after the onset of exogenous feeding.......... 32 Figure 1.7. Reaction norms for family effects on mean treatment group size-at age after the onset of exogenous feeding (95% 01)........................................34 Figure 1.8. Reaction norms (95% 01) for family effects on mean treatment group relative growth rate..............................................................................................36 Figure 1.9. ELISA titration curves (95% confidence intervals)...................................38 Figure 1.10. Reaction norms (95% 01) for family effects on mean treatment group antibody titre as determined by indirect ELISA................................................ 39 Chapter Two Figure 2.1. Difference between mean diploid heritability (all traits combined) and individual performance trait heritability estimates of triploids............................. 74 Figure 2.2. Difference between mean diploid maternal effects (all traits combined) and individual performance trait maternal effects estimates of triploids...................................................................................................... 77 VII Acknowledgements My absolute gratitude is extended primarily to Dr Dan Heath for his expertise, guidance, support, patience and blunt honesty. I am completely indebted to him for providing me with multiple opportunities to pursue research-based projects in his lab. Thanks, Dan. Much appreciation and many thanks are also extended to Dr Mark Shrimpton for his expertise, guidance and support as co-supervisor but equally for his kindness and humor. I also thank my committee. Dr Bob Devlin (DFO), Dr Brent Murray (UNBC) and my external examiner, Dr Tillman Benfey (UNB) for sharing their knowledge and insight. My gratitude also extends to Drs John and Ann Heath of Yellow Island Aquaculture Ltd. (YIAL), not only for supplying critical financial and infrastructural support but also for their expertise, fundamental interest and absolute glee for all things research-related. Financial support was gratefully received from the Natural Sciences and Engineering Research Council of Canada (NSERC Industrial Postgraduate Scholarship) with operational funding from a grant to Yellow Island Aquaculture Ltd. from the Science Council of B.C. (Technology BC) and an NSERC Collaborative Research and Development Grant to Dr Heath. Special thanks for field and lab assistance is extended to the following people: G. Cho, A. Clark, M. Heath, G. Heath, E. Heath, J. Heath, A. Heath, G. Osbourne, C. Busch, C. Bryden, C. Richardson, A. Newhook, D. McCabe, L. Rankin, M. Shrimpton, D. Heath, M. Stuyt, H. Robert, P. Thibault, A. Hubbesty (U of Windsor), R. Beecroft (ImmunoPrecise Antibodies), G. Prosperi-Porta (PBS-DFO), Bill narrower (BC Ministry of Agriculture, Food & Fisheries), S. St-Hilaire (PBS-DFO), T. Yesaki (BC Ministry of Water, Land & Air Protection), D. Stanton. Finally, my profound appreciation is extended to my amazingly smart and beautiful children, Sabrina, Cassandra and Caitlin for their incredible patience and support. vm General Introduction Polyploidy, the state of having three or more complete sets of nuclear chromosomes, has recently received considerable attention in the fields of genetics, evolution and aquaculture. This interest is in part attributable to the usefulness of a polyploid platform for dissecting patterns of genetic expression and regulation and also because of the insights that might be gained by understanding the processes affecting the evolutionary fate of duplicated genes (Spring 1997; Galitski etal. 1999; Suzuki et a!. 1999; Force 2000; Otto and Whitton 2000; Otto and Yong 2002). While polyploidy has long been known to have been a major evolutionary force involved in the diversification and adaptation of many plant species (Soltis and Soltis 1995; Ramsey and Schemske 1998) there is strong and accumulating evidence suggesting that two rounds of genome duplication occurred in the ancestral vertebrate lineage (Ohno 1970; Lundin 1993; Force 2000; Gibson and Spring 2000; Furlong and Holland 2002). Thus, polyploidy may have played an important role in defining both the rate and form of species diversification by providing the genetic structure and variation necessary for rapid adaptive evolutionary change (Ohno et al. 1968; Ramsey and Schemske 1998; Force et al. 1999; Otto and Whitton 2000). Polyploidy is specifically known to have played an important role in the evolution of a number of fish families including the catfish Corydoras-Aspidoras-Crochis species group (within the family Callichthyidae), suckers (Catostomidae; Ferris and Whitt 1978; Ferris 1984) and the salmon and trout (Salmonidae; Schultz 1979; Allendorf and Thorgaard 1984). While tetraploidy in catostomids is thought to have arisen via hybridization (allopolyploidy) followed by a relatively rapid re-diploidization, salmonids exhibit evidence of intraspecific genomic doubling (autopolyploidy) followed by an incomplete transition back to the diploid state (e.g., residual tetrasomie inheritance and 1 the presence of multivalent pairing of chromosomes during meiosis) (Allendorf and Thorgaard 1984; Devlin and Nagahama 2002). Salmonids provide intriguing models for the study of genetic change after polyploidization; however, because they are especially tolerant of many different kinds of chromosome set manipulations, including artificially induced polyploid states they also provide an interesting opportunity to study potential phenotypic or genomic effects of perturbation. Ploidy manipulation in fish has been studied since the 1950s, originally because of its usefulness in elucidating cytological function but also because of a perceived possibility for increased growth potential in salmonids reared for aquacultural purposes (reviewed by Ihssen etal. 1990). More recently, ploidy manipulation has received increased attention because of its applications to genetic mapping, studies of sex determination mechanisms, production of isogenic lines, inbreeding as well as whole organism performance (Ihssen et al 1990; Thorgaard 1992; Bongers et ai. 1998; Aral 2001). Ploidy manipulation to induce triploidy (the state of having three sets of chromosomes) has been of particular interest because it is relatively easy to induce in salmon and because of the potential for increased growth offish reared for aquaculture. The production of monosex all-female triploid salmon is particularly useful in an aquacultural context because they are sterile. Sterility in female triploids is thought to be due to disrupted endocrinological functioning as well as problems with homologous chromosome pairing and segregation during gametogenesis which result in predominantly aneuploid germ cells (Benfey etal. 1989; Carrasco et al. 1998; Benfey 1999). Triploid females do not experience the typical sex steroid stimulation provided by the hypothalamus-pituitary-gonadal axis that diploids and triploid males respond to during maturation and so do not produce the vitellogenin and other factors necessary for proper oocyte development or exhibit the gonadal growth, conditional degradation, sexual precocity or breeding behaviour normally experienced or displayed by diploids and triploid males (Benfey et al. 1986; Piferrer et al. 1994; Amano et al. 1998). This lack of maturation means that commercially-farmed fish may be grown on an extended schedule, potentially to larger sizes than diploids and harvested at any time. Perhaps more importantly, the lack of successful gametogenesis and spawning behavior means that all-female triploid fish farm escapees are unable to breed with wild stocks thus minimizing the potential threat that genetic introgression may pose for intraspecific diversity. Along with sterility, all-female triploid salmonids exhibit morphological characteristics typical of an induced triploid state in vertebrates. In relation to diploids, these include: 1) nuclear enlargement with triploids containing 50% more DNA, 2) larger cell size with approximate maintenance of the nuclear:cytoplasmic ratio along with a reduction in the surface areaivolume ratio, 3) fewer cells, but maintenance of diploid tissue, organ and body size and 4) increased allelic diversity with potentially three different alleles at each locus (Swarup 1959a; Leary et al. 1985; Benfey 1999). These conditions suggest that triploids may experience potentially positive fitness or production related effects due to increased allelic diversity, higher genetic expression of growth related loci and somatic reallocation of energy normally invested in the maturational process. While investigation of the physiological consequences of induced triploidy in fish is incomplete, especially at the cytological and genetic levels of study, exploration of physiological parameters such as hematology, oxygen consumption and aerobic capacity, a limited number of immunocompetence determinants, osmoregulation and stress response, energetics and development indicate either relatively subtle differences (e.g., intolerance of hypoxic conditions, Ojolick etal. 1995; reduced sensory perception, Allah etal. 1990) or an overall similarity of function rather than overt differences (reviewed by Benfey 1999). There is a growing body of published work comparing the performance of diploids and triploids in terms of survival and growth in an increasingly diverse array of fish species (e.g. Benfey 1989; Pandian and Koteeswaran 1998; Felip etal. 2001; Aral 2001). Despite an overall trend indicating that triploids usually experience lower survival than diploids especially during the juvenile stages of development (Solar et al. 1984; Happe et al. 1988), effects of triploidization on growth tend to be more variable between studies with results including growth less than, greater than and equal to that of diploid fish. Despite similarities between diploids and triploids in physiological parameters and the variability of triploid performance, direct comparison of studies is often confounded by differences in study design and conditions (including strain, sex, age, size and genetic origin offish (Benfey 1999) and possibly induction method) as well as potential species-specific differences in the tolerance of or response to triploidization (Thorgaard 1986; Ihssen et al. 1990). Additionally, very few studies have focused on Pacific salmon species such as chinook salmon (Oncorhynchus tshawytscha). Triploid salmon can be produced by breeding tetraploids with diploids (e.g.,Chourrout et al. 1986) but are more commonly induced by applying a short thermal or pressure-based shock to fertilized ova (Ihssen et al 1990; Felip, Zanuy, Carrillo, and Piferrer 2001; Hu lata 2001). Because salmon eggs are ovulated after completion of meiosis I and in a state of arrested metaphase II, only the second meiotic division is available for manipulation. Fertilization triggers a meiotic reactivation cascade that facilitates the transition to anaphase (Holloway et al. 1993; Ciosk et al. 1998). If a shock is applied to the egg shortly after fertilization but before completion of anaphase II, the haploid chromosome set normally extruded as the second polar body can be retained within the egg cytoplasm. If fusion of this chromosome set with the maternal and paternally contributed chromosome sets occurs prior to the first mitotic cleavage, the embryo will develop as a triploid. Interestingly, the specific cytological mechanisms by which heat and pressure treatments cause second polar body retention have not been examined in detail, despite evidence to suggest that the meiotic response to heat and pressure may differ (e.g., Fankhauser and Godwin 1948; Swarup 1959b; Chourrout 1986). It is also not known if heat and pressure shock-treatments applied to fertilized fish eggs affect subsequent embryonic development or cellular functioning in a differential manner or whether there are lasting effects of treatment on cellular or organismal performance (however, see Malison etal. 1993). This may be partly due to a strong research focus on production-based goals (primarily triploidization success and growth) and difficulty in separating treatment effects from the effects of triploidization perse. Additionally, there has been no published work investigating potential phenotypic effects on quantitative genetic parameters after triploidization in salmon. The current research includes two quite different, but interrelated, perspectives. The first study, as outlined in Chapter One, has a very applied aquacultural focus and explores the efficacy of triploidization methods as well as the comparative freshwater performance of diploid and triploid chinook salmon full-sib families generated using heat and pressure-shock methodologies. The impact of triploidization on performance was examined in terms of the effects of genotype (family), treatment and genotype by treatment interactions on growth and immune characteristics. The study outlined in Chapter Two has a less applied focus but extends the range of Chapter One by exploring the effect of triploidization on both the phenotypic variance and the additive genetic control of specific performance variables using a quantitative genetic analysis framework. CHAPTER 1 COMPARISONS AMONG DIPLOID, HEAT-SHOCK INDUCED TRIPLOID AND PRESSURE-SHOCK INDUCED TRIPLOID CHINOOK SALMON (ONCORHYNCHUS TSHAWYTSCHA) FOR SURVIVAL, GROWTH, IMMUNE FUNCTION AND TRIPLOIDIZATION EFFICACY 1.0. Abstract The production of sterile salmon for aquaculture would be an effective strategy with which to minimize potential genetic risks associated with fish farm escapes. All­ female triploid salmon are sterile and can be relatively easily generated by applying a heat or pressure-based shock to eggs. However, while triploidization in fish has been examined for some time, this technology has not been widely implemented in North American fish farms. This is mainly due to inconsistent or inferior performance of triploids as compared to diploids but there is also a lack of species-specific research especially for chinook salmon (Oncorhynchus tshawytscha). Additionally, the comparative efficacy of heat and pressure-based induction protocols and the effect of treatment and genotype on fitness related traits have rarely been examined or controlled for in the same study, despite strong evidence indicating that these are important variables. In this study, the effect of treatment and genotype on freshwater survival, growth, immune function and triploidization success was examined in full-sib groups of heatand pressure-shock induced chinook salmon triploids and diploids originating from each of five families. It was found that triploidization rates were highest for pressure-based induction, but few differences between the two treatment groups for survival, growth and antibody response to intraperitoneal vibrio (Vibrio anguillarum) vaccination were observed. While triploids had lower survival and poorer performance than diploids. 7 significant differences occurred mainly during the embryonic or larval stages of development. Survival was mainly influenced by treatment (heat-shock, pressure-shock) and family (genotype) while growth characteristics and the immune response to vaccination were mainly influenced by family (genotype). Significant interactions between family (genotype) and treatment were found. This interaction explained a relatively low proportion of the total variation in survival and growth but a large proportion of the variation in immune response. Despite the occurrence of interactions, separation of high and low performing families was evident regardless of ploidy and family effects were substantial indicating that selection of high performance diploid broodstock for triploidization purposes would be part of an effective strategy for improvement of triploid offspring performance. 1.1. Introduction Teleosts are tolerant of a wide range of chromosomal manipulations and artificially induced polyploid states (e.g., see Arai 2001 1997; Pandian and Koteeswaran 1998). Triploidy (the genomic state of having three complete sets of chromosomes) is especially easy to induce and salmonids appear to be quite tolerant of this condition (Thorgaard 1992; Benfey 1991; Ihssen et al. 1990). All-female triploid salmon are of particular interest because they are sterile and so may have considerable economic and environmental benefits for aquaculture (Cotter et al. 2000; Wilkins at al. 2001). Some potential advantages of using sterile salmon for aquaculture include: 1) the circumvention of sexual maturation and associated conditional degradation, 2) increased somatic growth, 3) extended g row-out, 4) wider harvest windows and 5) minimization of possible genetic and ecological threats to wild populations. The technology, however, has not been widely implemented in North American salmon farms. This is predominantly due to inconsistent or highly variable performance-based results, and a general trend indicating that triploids usually experience lower survival than diploids and may experience reduced growth (Benfey 1999). The precise causes of impaired performance are not known [but may include treatment induced stress effects (e.g., Swarup 1959; Malison etal. 1993) and ploidy related differences in cellular level physiology (Benfey 1999), or genetic disruptions]. These factors may be most important to triploid survival during the critical stages of emybryonic development. Despite growing evidence indicating the importance of species and family-specific genetic differences in the tolerance of triploidization in salmon (see Withler etal. 1995; W ithlerefa/. 1998; Bonnet etal. 1999; Blanc etal. 2001), direct comparisons across published performance-based studies are often confounded by differences in study conditions (i.e., strain, sex, age, size; Benfey 1999) or a lack of species or populationspecific induction protocol optimization. Triploidization of all-female salmon is most commonly and successfully induced by applying either a heat or pressure-based shock to ova fertilized with sperm from hormonally sex-reversed females (i.e., neomales). It is possible that the trauma associated with these treatments may be contributing to the impaired performance of triploids; however, neither the specific treatment effects associated with the application of heat and pressure, nor the potential for long-term effects of treatment on cytological, genetic and molecular aspects have been investigated, and the effects of triploidy per se on these parameters as well as whole organism functioning has only been investigated very superficially in fish (i.e., primarily in terms of traits that are of economic importance to aquaculture). Suggestions that heat and pressure may have different effects on both amphibian and fish oocyte meiotic and mitotic structure and overall functioning exist in the literature (e.g., Fankhauser and Godwin 1948; Swarup 1959; Chourrout 1986; Diter etal. 1993). It is known that heat and hydrostatic pressure can have severe structural effects on microtubule and cytoskeletal structural integrity; however, these effects may be reversible (Wilson etal. 2001b; Begg etal. 1983). Disruption of microtubule structure during meiosis is thought to be the mechanism by which the second polar body is forced to remain in the egg after triploidization treatment (however, see Fankhauser and Godwin 1948); heat and pressure may also have other effects on eggs subjected to triploidization treatments. Pressure is known to disrupt protein structure and important cellular regulatory processes as well as change protein and mRNA distribution patterns (Begg, Salmon, and Hyatt 1983; Crenshaw et al. 1996; Wilson etal. 2001a; Wilson, Zimmerman, and Zimmerman 2001b). Heat tends to have acute effects on cell 10 structures such as membranes (fluidity) and proteins (dénaturation) but may also result in critical impairments of cell and genetic functioning (Hildebrandt et al. 2002). While applications of heat and pressure to oocytes have the potential to seriously interfere with subsequent embryonic development and cellular functioning, application at the intensity and duration used to triploidize fish is probably not severe enough to cause acute structural damage to the activated oocyte but might interfere with subsequent cellular expression patterns and protein function. While identifying the least traumatic method for triploidization is important, the adequate performance of triploids under conditions of culture is a necessary requirement for the adoption of triploid technology by the aquaculture industry. There is also a need to determine if increases in performance gained by using traditional selective breeding programs and molecular-based selection techniques can be transferred to fish after triploidization. As part of this evaluative process, I generated fullsib chinook salmon {Oncorhynchus tshawytscha) family groups consisting of heatinduced triploids, pressure-induced triploids, and diploid controls. In this way I was able to determine which of the two induction methods was most effective for triploidizing chinook salmon while comparing the effects of treatment, family (genotype) and the interaction between treatment and family (treatment x family) on specific fitness-related performance characteristics during freshwater rearing. I chose to follow the family treatment groups only through the freshwater stages of development because the early life history of salmon consists of a series of critical stages when selective pressures are intense and at which family, treatment and family by treatment interaction effects on performance variables might be most evident. Inferior performance during the critical stages of juvenile development can potentially have long-term negative effects on overall fish performance (Rossiter 1998). The goal of this study was therefore to 11 supplement existing literature evaluating the utility of triploids for aquacultural purposes by providing clear evidence on the impact of treatment and genotype on triploid chinook salmon performance throughout the freshwater phase of development. 1.2 Methods 1.2.1 Fish & Rearing Conditions Mature monosex female chinook salmon from Yellow Island Aquaculture, Ltd. (YIAL; Quadra Island, B.C.) were used to generate the offspring families for this study. Five phenotypic males (hormonally masculinized genotypic females) were bred 1:1 with 5 females to form five full-sib families. Spawning of all broodfish took place at YIAL hatchery facilities on Quadra Island, between October 31 and November 5, 2000. Broodstock were maintained in tanks supplied with hatchery water (mean water temperature ± SE = 8.05 °C ± 0.01 °C) prior to collection of gametes. Eggs from each full-sib cross were collected and dry fertilized separately. Time was tracked immediately upon addition of milt to the egg masses. The egg-milt mixture was left for 2 minutes to allow for fertilization before hatchery water at 8.05 °C ± 0.01 °C was added to induce swelling and micropyle closure and to wash excess milt and debris from the eggs. Three ~250 ml sub-samples of eggs from each fertilized egg mass were randomly sampled and subjected to one of the following three treatments: i) Hydrostatic pressure-shock, five minutes at 6.89 x lO'^kPa (10 000 psi) of pressure, 30 minutes after fertilization (mean water temperature ± SE = 8.05 °C ± 0.01 °C). Pressure was applied using a modified 30-ton H-frame hydraulic press and a custom built 1.6 L experimental egg pressure cylinder ii) Heat-shock, ten minutes submerged in a 29.0 ± 1.0 ° C uniformly heated, aerated waterbath, 25 minutes after fertilization followed by 30 minutes of aircooling. Eggs were placed in plastic mesh incubation boxes (Viberg) for the heat 12 treatment. Boxes were then hung on hooks in the hatchery for 30 minutes of air-cooling (mean air temperature ± SE = 8.4 °C ± 0.3 °C) iii) Control. Fertilized egg sub-samples from each female were left untreated and transferred to incubation trays immediately after the water hardening process. The particular treatment protocols used were based on extensive experimentation carried out by YIAL to optimize triploidization protocols for Chinook salmon under the specific hatchery conditions at their facilities (YIAL has been experimenting with triploidization since 1993), and on a survey of treatment protocols designed for salmonid induction outlined in the literature (e.g., Hill, Nickerson et al. 1982; Benfey and Sutterlin 1984; Guoxiong, Solar etal. 1989). Family treatment groups were incubated in separate compartments of vertical stack incubation trays (Heath Tecna Corp.). Trays were divided into twelve compartments (10 cm x 10 cm x 5cm) with each compartment holding an average of 700 eggs. Offspring families were assigned to incubation stacks and trays randomly. Water temperature within the stacks was monitored using a digital data logger (Onset Computer Corp.) and development stage of the fish was tracked using Accumulated Temperature Units (ATUs; calculated as the cumulative total of daily mean temperatures). Mean water temperature (± SB) during the incubation period was 7.72 °C ± 0.02 °C. Mean flow within the stacks was 13 L/minute. When eggs reached the eyed stage of development (the point at which the eye spots of the developing embryo are visible through the egg shell and at which % yolk vascularization has occurred; November 28-December 11, 2000, 280-296 ATUs) they were mechanically shocked and sorted using a Jensorter machine (Model JM4C, Jensorter, Inc.) in order to get total egg counts and to separate live from dead eggs within each family treatment group. As alevins completed yolk-sac absorbtion (March 7-11, 2001 ; 973-982 ATUs), 13 two sets of fifty alevins each were randomly selected from the treatment groups within each of the five families (two replicate groups from each family treatment groups) and transferred from the vertical incubation stack tray compartments to 140 L aerated rearing tanks for the onset of exogenous feeding. Two sets of fish were randomly assigned to each tank regardless of ploidy at a starting density of 100 fish/tank (-0.71 fish/L; mean density = 0.27g/L). One set of fish in each tank was fin clipped for identification purposes (either the upper or lower caudal fin lobe was removed). Flow rate of water to the tanks was approximately 3 L/minute and the mean temperature (± SE) was 8.79 °C ± 0.02 °C. Fish were handfed to satiation multiple times per day with commercial feed (Ewos, Canada, Ltd.). 1.2.2. Ploidy Determination Two techniques, flow cytometry and red blood cell (rbc) nuclear analysis, were used to determine the level of triploidization success within heat and pressure treated groups of fish. Both methods are commonly used to determine the ploidy level of fish subjected to triploidzation treatments. While flow cytometry is a highly sensitive method and allows accurate measurement of the double-stranded DNA content of individual cells or nuclei, the measurement of rbc nuclear dimensions is strongly and positively correlated with genome size (C-value) in fish (Gregory and Hebert 1999), is simple and does not require expensive equipment. Both methods are able to reliably distinguish between triploid and diploid salmonids (e.g., Allen 1983; Benfey et al. 1984; Small and Benfey 1987; Johnson etal. 1984; Teplitz etal. 1994). Flow cytometry was used in this study to verify the results of the rbc nuclear analysis. Erythrocyte nuclear length was used to determine family-specific triploidization success (Welters et al. 1982; Beck and Biggers 1983; Benfey, Sutterlin, and Thompson 14 1984). Whole blood smears were made from all terminally sampled fish. Slides were fixed in methanol and stained with Wright-Giemsa (Sigma). Visualization and measurement of erythrocyte nuclei was accomplished under oil immersion (lOOOx magnification) using an Olympus BX-50 compound microscope (Olympus Optical Co.) equipped with a Qlmaging Retiga 1300 Monochromatic digital camera (Quantitative Imaging Corp.) and the Northern Eclipse, version 6.0 imaging program (Empix Imaging Inc.). The length of ten randomly chosen erythrocyte nuclei was measured to the nearest 0.01pm in each of 778 whole blood smears sampled from heat (n = 230), pressure (n = 245) and control (n = 303) treatment groups within each of the five study families. Two hundred fish (twenty fish per family treatment group) from the five study familes were also terminally sacrificed at 186-200 days post-fertilization for flow cytometric analysis. Blood samples were prepared using a modified version of Allen’s (Allen 1983) protocol (G. Osbourne, UBC Biomedical Research Centre). Briefly, 1.5-5.0 pi of blood was collected from the caudal vasculature of each fish and ejected into 0.5 m l of a 50 mg/L solution of propidium idodide (pH 7.2; 50 mg propidium iodide, 25 ml citrate acid dextrose, 8.5 mg RNAse A, 1.0 ml IGEPAL-630, phosphate buffered saline to 1 L) kept on ice. Each sample was vortexed and refrigerated at 4 °C overnight to allow membrane disruption and intercalcation of the propidium iodide dye with dsDNA. In the morning, 4% paraformaldehyde solution was added to each sample. Samples were kept on ice and transported to the Multiuser Flow Cytometry facility in the Biomedical Research Centre at the University of British Columbia for analysis on the FACSCAN (Becton Dickinson) flow cytometer. Blood sampled from 10 diploid control fish was used as a series of external standards and DNA indexes [modal DNA content (channel #) of test sample /modal DNA content (channel #) diploid standard] were 15 calculated to determine pioidy status. 1.2.3. Survival Incubation survival of family treatment groups was monitored from fertilization to the eyed stage of development (280-296 ATUs) and then followed through to the alevin stage, just prior to transfer of the fish to freshwater rearing tanks (mean ATUs = 970). Embryo mortalities from fertilization to the eyed stage (3/4 yolk vascularization) were assessed after the eggs were mechanically shocked and sorted using a Jensorter machine. Total egg number and the number of live eggs for each family treatment group were evaluated at this time. After the initial eyed egg count, mortality was monitored at least every two days. Survival (as a proportion of the initial live eyed egg count) was determined for all groups at 510 ATUs (post-hatch), 616 ATUs, 746 ATUs, 789 ATUs, 853 ATUs, and 970 ATUs (just prior to ponding). Survival of the five experimental families (i.e., heat, pressure and control treatment groups from each family plus replicates totaling 30 treatment groups) was monitored at least every second day after transfer to rearing tanks. Survival was determined as the number of live fish remaining immediately prior to experimental vaccination treatment divided by the total number of fish originally fin-clipped and released into the rearing tank. Experimentally sacrificed fish (i.e. for flow cytometry) within each group were excluded from the final survival assessment. 1.2.4 Growth Weight (in grams) was determined by non-terminal sampling at five developmental time points during the freshwater growth of the ponded family groups, at ponding (mean ATUs = 985; -127 days post-fertilization), at 1 200 ATUs (mean ATUs = 16 1224; ~155 days post-fert.), at 1 700 ATUs (mean ATUs = 1729; ~214 days post-fert.), at 1 800 ATUs (mean ATUs = 1843; -227 days post-fert), and at 1 900 ATUs (mean ATUs = 1941; -236 days post-fert.). At ponding, 100 alevins from each family treatment group were weighed in water as they were transferred to rearing tanks. At all other sample times, weights of individual fry were recorded. Forty fish per family treatment group (heat, pressure, control + replicates) were weighed at the 1 200 and 1 700 ATU sample points. At thel 800 and 1 900 ATU sample points, approximately 6 6 fish were sampled from each treatment group (treatment group and replicate) within all families. Blood smears for pioidy determination were taken from all fish sampled at the 1 800 ATU and 1 900 ATU sample points but not at the 1 200 and 1 700 ATU sample points. Relative growth rate of family treatment groups was assessed from ponding to 1 900 ATUs (mean ATUs = 985 -1941), an average time interval of 112 days. Because individual fish were not tracked, mean family treatment group weight at ponding was used for the first sampling point and individual fish weights were used for the second sampling point. Relative growth rate was calculated using: {(Y2-Y^)^\s(t2-ti))x100 Where Ya = individual fish weight at the second sampling point; Yi = mean family treatment group weight at ponding, ti - ta = the mean time interval in days between the first and second sampling points. 1.2.5. Vaccination & Enzyme Linked Immunosorbent Assay (ELISA) To determine if treatment groups differed in their serum antibody response to vaccination, a vaccination trial using four of the original five families was conducted. Within a family, each fish within one replicate treatment group received a 0.1 mL dose 17 of a commercial water-based vibrio vaccine (Alpha Dip 2100, Alpharma NW Inc.; Vibrio anguilla rum, serotype 01 and V. ordalii bacterin) by intraperitoneal injection. Each fish in the second family treatment replicate group received 0.1 mL of a phosphate buffered saline solution. All fish within each treatment group were anaesthetized, weighed, injected and returned to the appropriate rearing tanks for ten days to allow for establishment of an antibody response. Mean treatment group weights at the time of injection were: heat = 3.91 g, pressure = 4.19 g, control = 4.58 g; the mean temperature at injection was 9.53 °C and ranged from 9.19 - 9.53 °C. On the eleventh day after vaccination, all fish were euthanized and weighed. Blood was taken from the caudal vasculature with uncoated microhematocrit tubes and a blood smear was made for pioidy determination via red blood cell analysis. Serum was transferred to microcentrufuge tubes after overnight refrigeration (4 °C) and centrifugation (5 minutes at 7100 rpm, 5125 x g). All serum samples were held at -2 0 °C and then at -8 0 °C until determination of antibody titre by enzyme linked immunoassay (ELISA). Two hundred diploid YIAL production fish (mean weight at injection = 5.80 g) were used to generate a positive control serum for use as a control standard in the immunoassay. Fish were injected with the Alpha Dip 2100 vaccine, as noted above and were terminally sampled on the 11^ day after vaccination (temperature at injection = 8.93 °C; mean temperature over serum generation period = 9.29 °C, range = 8.93 - 9.76 °C). Specific antibodies against Vibrio anguillarum were detected in fish sera after vaccination using an indirect ELISA protocol (R. Beecroft, Immuno-Precise Antibodies, Ltd.). Microtitre plates (96 wells; Nunc Maxi Sorb) were coated with a Vibrio anguillarum cell suspension (culture provided by G. Prosperi-Porta, Pacific Biological Station, Nanaimo, BC). Plates were incubated overnight at 37 °C and blocked with a 3% normal 18 goat serum phosphate buffered saline solution. Salmon anti-vibrio immune serum was diluted 1/20-1/2560 across the plates. Four salmon anti-vibrio serum samples as well as positive (pooled control serum), negative (family treatment group specific negative control serum-sham injected) and blank controls were run on each plate. Rabbit anti­ salmon immunoglobulin (H+L chain; ImmunoPrecise Antibodies, Inc.; 1/4000) 2 ° antibody and goat anti-rabbit iummunoglobulin G (H + L), 3 ° antibody labeled with alkaline phosphatase (Caltag, Inc.; 1/2000) were used for antibody detection. Alkaline phosphatase substrate solution was added and the optical density ( O D 4 0 5 n m ) determined after 30 minutes using a VERSAmax tunable plate reader with Softmax Pro 4.0 software (Molecular Devices, Corp.). Four salmon anti-vibrio immune serum samples from each of the three family treatment groups within four of the five study families as well as negative control serum samples (sham-injected) from each family treatment group were analyzed by ELISA. 1.2.6 Statistical Analysis Data normality was assessed visually and with normal probability plots. Hartley’s Fmax test (Hartley, 1940, 1950) was used to assess homogeneity of variance among treatment types for each measurement of performance. Data transformation was used if data non-normality or variance differences were extreme. Generally, two-way mixed model analysis of variance (ANOVA) was used to detect treatment (fixed effect), family (random) and the treatment by family interaction effects on performance measurements. Multiple comparisons using Tu key’s Honest Significant Difference (HSD) test were performed when significant effects (P < 0.05) were detected. Reaction norms were plotted to visually clarify and interpret significant interactions and effect sizes. Effect size of the main factors and the interaction was also determined using omega squared (w^, 19 proportion of total variance). Omega squared was calculated using the formulas of Dodd and Schultz (1973) for fixed, random and interaction effects (reported in Olejnik and Algina 2000) and Cohen’s (1992) scale of effect size (i.e., 0.0099 2 small < 0.0588; 0.0588 ^ medium < 0.1379; large > 0.1379). One-way AN OVA was used to detect treatment effects at the eyed stage of development, tank or incubation position effects or the effect of tank pioidy composition (diploid, triploid or mixed) on measured variables even though placement of family groups was completely randomized. Analyses were performed using Systat Version 10.0 (SPSS Inc. 2000) and SPSS Version 10.0.1 (SPSS Inc. 1999). 1.3 Results 1.3.1 Triploidization Success Overall triploidization success levels measured using red blood cell nuclear analysis and flow cytometry were high for both heat- and pressure-shock treatments. Flow cytometry data confirmed the results obtained by the red blood cell nuclear analysis. Ninety-seven percent of both heat and pressure-shock treated fish were found to be triploid by flow cytometry, while rbc nuclear analysis indicated that overall triploid levels were 94% for heat-shock treatment groups and 96% for pressure-shock treatment groups; all sampled control fish were identified as diploid. There was some unexpected overlap between the diploid and triploid nuclear length distributions (Figure 1.1). Therefore samples with mean rbc nuclear lengths between 8.25 pm and 8.75 pm (between dotted lines in Figure 1.1 ; n = 8 or 1.0% of sampled fish) were eliminated from the analysis because of uncertainty in pioidy designation. It is possible that these fish were aneuploid but this could not be established as blood was not taken for flow cytometric analysis from these specific fish. Aneuploid salmon do not usually survive 20 very long past hatching (Chourrout 1986) and no evidence of aneuploidy was detected in the samples analyzed with flow cytometry. The overlap between measurements may reflect natural variability in rbc nuclear size. Red blood cell nuclear length was significantly different between treatment groups (P < 0.001). Diploid control groups had significantly smaller rbc nuclear lengths than pressure and heat-shock groups (P < 0 .0 0 1 ) however, pressure and heat-shock groups did not differ significantly from each other (P = 0.89; mean rbc nuclear length ± SO, 2N-Control = 7.33 pm ± 0.38 pm; 3NHeat = 9.32 pm ± 0.54 pm; 3N-Pressure = 9.30 pm ± 0.51 pm). While pressure-shock treatment was more successful than heat-shock treatment at inducing triploidization according to the rbc analysis (mean ± SD, 3N-Heat = 0.94 ± 0.03, 3N-Pressure = 0.96 ± 0.04), the difference between treatment success rates was small (2%) and not significant. Results from the mixed model ANOVA using family triploidization proportions (arcsinV transformed) indicated that the main effects of treatment and family were not significant (Table 1.1) but there was a large and statistically significant interaction (P < 0.001), indicating that triploidization levels varied across families dependent on the nature of the applied treatment (Figure 1.2). The magnitude of the effect size of the interaction was large (Cohen 1977) (i.e., the proportion of the total variation in triploidization success that was explained by the interaction was high; refer to omega squared values w^. Table 1.1) however, the overall family-based level of variation in triploidization success was low for both treatments (89%-100%) suggesting that differences in the efficacy of treatment were not large. 21 Table 1.1. Analysis of variance components and omega squared values (w^) for red blood cell nuclear analysis. Sum-of-Squares Source of Variation Treatment 0.0558 Family 0.1140 Treatment x Family 0.0095 Residual 0.0002 indicates significance at the P < 0.001 level 100 df 1 4 4 10 Mean-Square 0.05576 0.02854 0.02366 0.00002 F-ratio 2.357 1.206 1217.465“ * w? 0.02 0.31 0.53 " 75- S' c 3 O2 50" 25" 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.0010.5011.00 Mean rbc nuclear length (urn) Figure 1.1. Frequency distribution of mean red blood cell nuclear length across treatment groups. Nuclei > 8.5 |xm = 3N, nuclei < 8.5 p.m = 2N. Samples with mean rbc nuclear length between 8.25 |im and 8.75 ^m (between dotted lines; n = 8 , 1.0% of sampled fish) were eliminated from the analysis because of uncertainty in pioidy designation. 22 1.00 0.98c 0 1 0.95- Tî O a c 0.93- 0.90- Heat Pressure T reatment Figure 1.2. Reaction norms for family effects on mean treatment group triploidization success levels. Triploidization success levels were determined using the length of 10 red blood cell nuclei for each sampled fish (heat = 230 fish; pressure = 245 fish). Symbols identify family treatment groups. 1.3.2 Incubation Survival A one-way ANOVA was used to test for the effect of treatment on survival (arcsin V transformed) to the eyed stage of development. Treatment groups differed significantly in survival to this stage of development (-288 ATUs; P = 0.031) and a large proportion of the variation in survival was associated with the treatment effect (Table 1.2). Mean survival of the two triploid groups did not differ significantly (P = 0.997). 23 Diploids experienced the highest mean survival, but diploid mean survival was only significantly higher than that of the heat-shock treated triploids (P = 0.047) (mean back transformed values ± SD; 2N-Control = 0.95 ± 0.09; 3N-Pressure = 0.68 ± 0.12, 3NHeat = 0.67 ± 0.02). Table 1.2. Analysis of variance components and omega squared value (co^) for survival to the eyed stage of development. Source of Variation Sum-of-Squares 1573.312 T reatment 1994.816 Residual Indicates significance at the P < 0.05 level df 2 12 Mean-Square 786.656 166.235 F-ratio 4.732* uF 0.34 The overall magnitude of the effect of triploidization on survival (arcsin V transformed) through the rest of incubation (after the eyed stage of development) was large and clearly negative with heat and pressure-shock treatment groups experiencing overall mean survival to the end of incubation that was 42% and 32%, respectively, below that of diploid control groups (mean back transformed values ± SD: 2N-Control = 0.83 ± 0.01 ; 3N-Pressure = 0.51 ± 0.03; 3N-Heat = 0.41 ± 0.01). The mean survival of the two triploidized groups (heat and pressure) did not differ significantly from each Table 1.3. Analysis of variance components for survival after the eyed stage of development to the end of incubation. Source of Variation Sum-of-Squares df Mean-Square F-ratio T reatment 7785.680 2 3892.840 11.067** 0.32 5497.017 Family 1374.254 4 3.907* 0.30 Treatment X Family 2814.084 8 351.761 0.19 8.231"* Residual 1.058 75 0.014 indicates significance at the P < 0.05 level; indicates significance at the P < 0.005; indicates significance at the P < 0.001 level; other at any developmental stage during incubation at which survival was determined (i.e., at 510 ATUs , P = 0.971; 616 ATUs , P = 0.769; 746 ATUs, P = 0.812; 789 ATUs , P = 0.794; 853 ATUs , P = 0.866; 970 ATUs, P = 0.478; Figure 1.3). The mean survival 24 of heat-shock treatment groups was always significantly lower than that of control groups (except just after hatching, at 510 ATUs when the mean survival of heat, pressure and control groups did not differ significantly from each other, P = 0.064) (616 ATUs , P = 0.041: 746 ATUs, P = 0.030; 789 ATUs , P = 0.032; 853 ATUs , P = 0.030; 970 ATUs, P < 0.001) the mean survival of pressure-shock treatment groups was only significantly lower than the mean diploid control group survival at the end of incubation, just prior to ponding (970 ATUs; P = 0.020; Figure 1.3). A significant interaction between family and treatment on incubation survival (Table 1.3) was found to exist but this interaction only described 19% of the total variance in survival (w^. Table 1.3). To describe the interaction, reaction norms showing the effect of family of origin on treatment group survival at each developmental stage were plotted (Figure 1.4). The graphs clearly show that the response to treatment varies depending on family of origin and that treatment and family effects cannot be interpreted without considering this interaction. Interestingly, an obvious distinction can be seen between families that survived relatively well regardless of applied treatment (i.e., high performance families) and those that survived poorly as triploids (heat-shock or pressure-shock); however, this distinction appears to become less important by the end of incubation (970 ATUs) when treatment effects tend to dominate (Figure 1.4 f). The effect size of all factors (main and interaction) was large but the treatment and family effects explain more of the variation in overall survival than does their interaction (i.e., just prior to ponding at 970 ATUs, Figure 1.4f; Table 1.3, values). Survival was arcsine transformed to attain normality for analysis; reported means and SDs are backtransformed values. Interestingly, there appears to be more variation in survival in response to pressure- than to heat shock treatments. There were no significant effects of the section position within treatment groups within the incubation trays on survival at 25 any developmental stage during incubation (P > 0.05). 1 .2 5 - 1.00 5 0.75 0 .5 0 ! 0.25 - 500 ...I... 600 700 800 900 Developmental stage (ATUs) 1000 Figure 1.3. Mean cumulative survival (with 95% confidence intervals) of treatment groups through incubation. Survival was calculated as the proportion of surviving embryos/live embryos after mechanical shocking and sorting at the 510, 616, 746, 789, 853 and 970 ATU developmental stages. Circles = 2N-Control, squares = 3N-Pressure, triangles = 3N-Heat. 26 Figure 1.4. Reaction norms for family effects on mean treatment group survival to the 510, 616, 746, 789, 853 and 970 ATU developmental stages during larval incubation (af). Survival is calculated as a percentage of live eggs at the eyed stage of development, (actual figure occurs on next page). 27 1.00 1.001 0.90- 0.80- 0.80£ 0.70- (0 0.600.600.50- 0.40Heat Pressure Heat Control Pressure Treatment Treatment a) 510 ATUs b) 616 ATUs 0.90- 0.90- 0.80- 0.80- 5 0.70- « 0.70- 3 0.60- 0.60- 0.50- 0.50- 0.40- 0 .40 Heat Pressure Heat Control Pressure Treatment Treatment c) 749 ATUs d) 789 ATUs Control Control 0.90- 0.70- « 0.70J > I W 0.50- CO 0.50- 0.300.30 Heat Pressure Control Heat Pressure Treatment Treatment e) 853 ATUs f) 970 ATUs 28 Control 1.3.3 Survival After Exogenous Feeding Treatment had a significant and large effect on survival after ponding, as did family (P < 0.001 and P < 0.01; Table 1.4). While diploid control groups maintained significantly higher mean survival than heat-shock treated groups during this period (P < 0 .0 0 1 ), control and pressure-shock group mean survival did not differ significantly (mean ± SD; 2N-Control = 0.91 ± 0.07, 3N-Pressure = 0.88 ± 0.08, 3N-Heat = 0.75 ± 0.10; Figure 1.5). Pressure-shock treated fish had significantly higher mean survival than did heat-shock treated fish during this period (P = 0.003).The effect of heat-shock treatment on survival through freshwater was clearly negative; heat-shock treatment groups had a mean survival that was 16% below that of control and 13% below that Table 1.4. Analysis of variance components and omega squared values (w^) for freshwater Source of Variation Treatment Family Treatment x Family Residual Sum-ofSquares 0.294 0.208 0.02967 0.119 df Mean-Square F-ratio 2 0.147 0.05197 0.003708 0.0079 39.580™ 0.42 14.015** 0.32 0.469 0 4 8 15 of pressure-shock treatment groups (P = 0.002). While there was a significant and large effect of family on survival (P = 0.001), it did not appear to be associated with as much of the variation in survival (w^ = 0.32) as the effect of treatment (w^ = 0.42); there was no significant family by treatment interaction on freshwater survival. There was also no significant effect of rearing tank or the pioidy composition of tank companion groups on survival. Survival was arcsine transformed to attain normality for analysis (Zar 1996); reported means and SDs are non-transformed values. 29 1.00 0 .9 0 - 0.60 Heat Pressure Treatment Control Figure 1.5. Mean survival (95% Cl) of treatment groups from the onset of exogenous feeding (ponding) to just prior to saltwater transfer. 1.3.4 Size-at-age. At ponding, the mean weight of heat- and pressure-shock treated groups were not significantly different from each other (P = 0.113) but they were significantly lower than that of the mean weight of diploid control groups (both heat and pressure-shock, P < 0.001). Significant effects of family (P < 0.001) and the interaction between family and treatment (P < 0.001) on size-at-age existed at each sampled developmental stage after ponding (1224, 1729, 1843 and 1941 ATUs; Table 1.5). Unlike the effect of treatment on 30 survival, the effect of treatment on size-at-age was not a significant factor at any but the 1729 ATU developmental stage (P = 0.005). The mean weight of diploid control groups Table 1.5. Analysis of variance components and omega squared values (oj^) for size-at-age results. Size-at-age was determined at four developmental stages (1224,1729,1843 and 1941 Source 1224 ATUs Sum-of-Squares df Mean-Square F-ratio Treatment Family Treatment x Family Residual 0.119 5.858 0.528 3.626 2 0.05972 T465 0.06596 0.006 0.919 22.388*“ 10.260*“ 0 21.747 98.131 7.947 211.549 2 10.873 24.533 0.993 0.378 11.059“ 24.851*“ 2.625“ 0.03 0.33 16.295 74.627 4.602 0.704 3.569 16.353*** 6.541*“ 16.194 85.846 5.619 0.985 2.963 15.525*** 5.705“ * 4 8 564 0.61 0.05 1729 ATUs Treatment Family Treatment x Family Residual 4 8 559 0 .0 2 1843 ATUs T reatment Family Treatment x Family Residual 32.591 298.509 36.818 685.979 2 4 8 975 0 .0 1 0.33 0 .0 2 1941 ATUs Treatment Family Treatment x Family Residual 32.388 343.384 44.951 884.370 2 4 8 898 0 .0 1 0.30 0.03 if* indicates significance at the P < 0.05 level; Indicates significance at the P < 0.01 level, "indicates significance at the P ^ 0.001 level. was always significantly higher than the mean weight of triploid groups but the difference in weight was never more than 0.50 grams (1224 ATUS, P = 0.01, heat; P < 0.001, pressure; 1729 ATUs, P < 0.001, heat and pressure; 1843 ATUs, P < 0.001, heat and pressure; 1953 ATUs, P < 0.001, heat and pressure). Heat and pressure shock 31 treatment group mean weights were never significantly different from each other except at the second freshwater developmental stage (1729 ATUs) when the mean weight of heat-shock treatment groups was 0.15 g heavier than that of pressure-shock groups (Figure 1.6). 5.50 5 .0 0 4.50 4.00 E 3.50 D) 3.00 O) © 2.50 2.00 1.50 1.00 0.50 —I - T ~ I I f ' ”1— <— I I— 1— 1— T - p r T — , 1— I— r - 1 — I— 1 r “f - r 1200 1300 1400 1500 1600 1700 1800 Developmental stage (ATUs) - |- ~ i“ t " t 1900 j 2000 Figure 1.6. Mean size-at-age (weight in grams ± 95% Cl) of family treatment groups at three freshwater time points after the onset of exogenous feeding, 1729, 1843 and 1941 ATUs. Triangles = 3N-heat-shock, squares = 3N-pressure-shock, circles = 2N-control. Reaction norms describing the effect of family on treatment group size-at-age are shown in Figure 1.7. Response to treatment differed significantly across families with an obvious dichotomy establishing early in development between low and high performing 32 families. Those families that performed poorly as diploids also performed poorly as triploids regardless of treatment and this pattern persisted across the developmental stages at which size-at-age was determined. Surprisingly, the negative response to pressure-shock treatment in size-at-age was most pronounced in families that grew well as diploids and as heat-induced triploids, suggesting that treatment effects may be manifested differently across performance traits; this relationship probably influenced the size of the treatment effect which remained small throughout the freshwater period (as did the interaction effect size) compared to the effect size of family at each stage of development. The effect of family explained 61%, 33%, 33% and 30% of the total variance at each of the size-at-age sample points (Figure 1.7; Table 1.5, refer to values). Significant tank effects were found at each developmental stage (P < 0.05). To further clarify the nature of tank effects, the effect of the pioidy composition of fish within the tanks on size-at-age was investigated. There was a significant effect of pioidy tank composition on size-at-age at 1224 ATUS (P = 0.001), 1789 ATUs (P < 0.001), 1853 ATUs ((P < 0.001) and 1941 ATUs (P < 0.001); however tanks composed of two triploidized treatment groups (heat, pressure or mixed heat and pressure) were never significantly different than tanks composed of a mixture of diploid control and triploidized treatment groups (1224 ATUs, P = 0.507; 1789 ATUs, P = 0.944; 1853 ATUs, P = 0.554; 1941 ATUs, P = 0.175). Considering that triploidization success rates for heat and pressure treated groups were 97% by flow cytometry (94% and 95% by rbc nuclear analysis), this suggests that rearing fish in low-density mixed pioidy groups did not affect growth. The existence of tank effects even after complete randomization of group placement at ponding may indicate sensitivity of treatment fish to subtle environmental differences inherent in tank position or an interaction between genotype and environment. 33 0.804.00- 0.502 . 00 - 0.40Heat Pressure Control Heat Treatment Pressure Control Treatment b) 1729 ATUs a) 1224 ATUs 6 .0 0 - 5. 00-' # 4 .0 0 3.003.00Heat Pressure Control Heat Treatment Pressure Control Treatment c) 1843 ATUs d) 1941 ATUs Figure 1.7. Reaction norms for family effects on mean treatment group size-at age through freshwater rearing (95% Cl). Graphs a-d depict data for five families at 1224, 1729, 1843 and 1941 ATUs (~ 155, 214, 227 and 236 days post-fertilization) after the onset of exogenous feeding. 1.3.5 Relative Growth Rate. There were significant effects of family (P ^ 0.05) and the interaction between treatment and family (P s 0.001) on relative growth rate but the effect of treatment was not significant (P = 0.891) over the ~100-day period for which freshwater growth was determined (Table 1.6). However, the effect size of the interaction between treatment and family {u / = 0.04) was smaller than the effect size of family (Figure 1.8 and u / 34 values, Table 1.6), which explained 29% of the variation in relative growth rate. Treatment group mean relative growth rates did not differ significantly from each other (P > 0.05; 2N-Diploid, mean ± SD = 14.17% ± 3.48%; 3N-Pressure = 14.28% ± 3.41%; 3N-Heat = 14.73% ± 4.06%). Table 1.6. Analysis of variance components and omega squared values (w ) for relative growth rate. Sum-of-Squares df MeanSource of Variation F-ratio Square 2 Treatment 0.005719 0.00286 0.117 0 4 Family 0.123 0.29 0.490 4.935* 8 Treatment x Family 0 .2 0 2 0.02528 10.380*** 0.04 898 0 . 0 0 1 Residual 2.187 Indicates significance at the P ^ 0.001 level, P < 0.05 Reaction norms showing the effect of family on the relative growth rate of treatment groups reflect the pattern exhibited by size-at-age (i.e., low and high performing families) and are plotted in Figure 1.8. Treatment group performance clearly varied with family-of-origin. Families that had low relative growth rates performed poorly regardless of treatment, as did those with relatively high growth rates however, the high performing families appeared to exhibit poor growth when pressure-shock was used for triploidization. 35 0.18 -J 0.16 - 5 0.14 - O) ■4= 0 .1 2 - 0.10 - Heat Pressure Treatment Control Figure 1.8. Reaction norms (95% 01) for family effects on mean treatment group relative growth rate. Relative growth rate was arcsine transformed (arcsinV) to attain normality for analysis (Zar 1996) but reported means and SDs are non-transformed. There were significant effects of tank (P < 0.001) and the pioidy composition of tanks (P < 0.001) on relative growth rate. Mixed treatment tanks (heat-shock/diploid-control or pressureshock/diploid-control) had a significantly lower mean relative growth rate (P < 0.001) than did tanks with either 2 diploid groups, 2 heat-shock treated groups or 2 pressure- 36 shock treated groups (P < 0.001) (mean ± SD; all diploid tanks = 15.45% ± 3.29%; triploid treatment tanks = 14.99% ± 3.41%; mixed treatment tanks = 13.67% ± 3.77%) 1.3.6 Response to Vaccination: Enzyme-Linked Immunosorbent Assay Antibody titers were generally low. Mean serum titration curves for the treatment groups are shown in Figure 1.9. While the mean antibody values for the heat-shocked groups were consistently lower than pressure and control treatment groups over the complete dilution range, the serum dilution of 1 / 2 0 was found to be the dilution that best distinguished between mean antibody values of vaccinated (vibrio vaccine) and unvaccinated (sham; PBS) groups. Only a significant effect of the interaction between treatment and family (P < 0.001) on antibody titer at the 1/20 dilution was found (Table 1.7) but the interaction explained 26% of the total variation in response versus 46% explained by the effect of family (w^ values; Table 1.7). The mean OD405 values of Table 1.7. Analysis of variance components and omega squared (w^) values for antibody titer. Source of Variation Sum-of-Squares 0.045 T reatment 0.293 Family 0.197 Treatment x Family 0.235 Residual Indicates significance at the P < 0.001 level. MeanSquare F-ratio 2 0 .0 2 2 3 0.098 0.033 0.007 0.680 2.979 5.030*** df 6 36 0 .0 0 0.46 0.26 sham-injected treatment groups were significantly different from vaccinated treatment group means but the mean O D 4 0 5 value of vaccinated pressure-shock, heat-shock and diploid control groups were not significantly different from each other (mean O D 4 0 5 ± SD, 3N-Heat = 0.30 ± 0.18, 3N-Pressure = 0.39 ± 0.29, 2N-Control = 0.37 ± 0.29; Figure 1.9). Heat-shock treatment groups appear to vary less between families than do control 37 1.000 0.750 I0.500 o '4O o 0.250 0.000 — — -j.......-........ I.......... —.. Y ............ j.................I.......... ..... j — 1/20 1/40 1/80 1/160 1/320 1/640 1/128(5 Dilution Figure 1.9. ELISA titration curves (95% confidence intervals). Anti-vibrio immune serum titration curves for heat (solid triangles), pressure (solid squares) and control (solid circles) treatment groups as tested by ELISA using Vibrio anguillarum as antigen. Lower curves are mean values for titrations of serum from sham-injected treatment groups (PBS); upper curves are mean values for titrations of serum from vaccinated groups (vibrio vaccine). Error bars indicate 95% confidence intervals. and pressure-shock treatment groups in their response to vaccination but the response is highly variable for all groups (Figure 1.10). Mean weight did not differ significantly between treatment groups (P = 0.124; mean ± SD; heat = 3.90 g ± 1.20; pressure = 3.88 g ± 0.94 g; control = 4.48 g ± 0.73 g) and the correlation between OD value andweight was not significant (Pearson correlation = 0.218, 38 = 2.222, P = 0.14). OD405 values were log transformed to attain normality for analysis; reported means and SDs are non-transformed values. Red blood cell nuclear analysis confirmed that all heat and pressure shock treated fish tested for an immune response were triploids and all control fish were diploid. There was a significant tank effect on OD405 nm (P < 0.001) but no significant effect of the pioidy composition of tank (diploid, triploidized, mixed diploid/triploidized) on OD405 nm- 1. 1 0 - 1.00 E 0 .9 0 J 0 .8 0 ° 0 .7 0 0.60 0.50 0.40 0.30 0.20 Heat Pressure Control Treatment Figure 1.10. Reaction norms (95% Cl) for family effects on mean treatment group antibody titre as determined by indirect ELISA. Fish were vaccinated using a commercial vibrio vaccine; Vibrio anguiliarum was used as antigen in the ELISA. 39 1.4 Discussion The first objective of this study was to examine the relative effectiveness of heat and pressure-based induction treatments. Triploidization was most successful when pressure-shock was used for induction, although both methods resulted in high levels of triploidized fish (96% vs 94% triploidization success). However, the slightly lower rate obtained with heat may have occurred due to differences in the consistency of exposure. This may be especially relevant if intra-female egg size is not constant since the surface to volume ratio of eggs will be variable and eggs will differ in the intensity of shock that is received. Interestingly, there was a significant and large interaction between treatment and family that was associated with 53% of the variation in family triploidization success. To my knowledge, this is the first time such an interaction has been reported. Because inter-treatment variability was essentially nil, an interaction of this magnitude is most likely due to genotypic variability in response to treatment potentially caused by inter-female differences in meiotic timing or susceptibility for retention of the polar body (Diaz et al. 1993). Differences in triploidization success between females has often been noted (e.g., in newt, coho salmon, and yellow perch Fankhauser and Watson 1942; Habicht at ai. 1994; Withler, Beacham, Solar, and Donaldson 1995; Withler, Clarke, Blackburn, and Baker 1998; Malison, Procarione, Held, Kayes, and Amundson 1993). It is unlikely that this interaction was due to environmentally determined differences in egg "quality" due to over-ripeness. Maturing females were checked regularly for ovulation and eggs were harvested, fertilized and subjected to triploidization treatments immediately after they were stripped from the females. In this study 31% of the total variation in triploidization success was associated with differences between families (Table 1.1) suggesting the existence of a significant genotypic component. 40 The second objective of this study was to examine the impact of treatment and family on survival and performance. Triploids are fundamentally different from diploids at a number of different biological levels of organization (i.e., cell size, genetic content). The existence of such radical differences in primary biological structure (Benfey 1999) suggests that integrated physiological functioning might be impaired but that these negative effects might be offset by positive fitness related effects of increased genetic diversity (Leary et al. 1985; Allendorf and Leary 1984), reallocation of energy to somatic growth, gene dosage effects (genetic expression directly proportional to ploidy) and/or protection from mutagenic events (Thorgaard et ai. 1999). Alternatively, disruptions to regulatory and epigenetic pathways (i.e., inherited changes in the patterns of genetic functioning that are not explained by DNA mutation (Russo et al. 1996; Spencer 2000; Bird 2002), or inverse dosage effects (i.e., a gene expression activity level that is not directly proportional to gene dosage in a positive manner but rather is reduced as gene copy number increases (Devlin, Holm, Grigliatti 1982; Devlin, Holm, Grigliatti 1988; Birchler, Bhadra, et al 2001) might be compounded with impaired physiological functioning. Additionally, treatment effects caused by the stress of induction might also affect performance of triploids by inflicting structural damage or interfering with the availability or function of embryonic or maternally supplied substances (e.g., mRNA, IgM, proteins) during early development. Because significant interactions were found to exist, main effects were interpreted only when the omega squared effect sizes (w^) associated with them were larger than the effect size of the interaction (Sokal and Rohlf 1995). Treatment was the main factor affecting survival of experimental fish during all stages of development and growth (i.e., to the eyed stage, through the rest of incubation until the start of exogenous feeding, and from the onset of exogenous feeding until the 41 time of saltwater transfer) and was significantly associated with 34%, 32% and 42% of total variation in survival during the study. However, the effect of family was also large and significant during incubation (after the eyed stage) and after the start of exogenous feeding, explaining 30 % and 32% of the total variation. The interaction effects on survival were significant only during incubation after the eyed stage of development, and accounted for 19% of total variation (versus the 32% and 30% accounted for by treatment and family). The reaction norms showed meaningful interactions throughout incubation, however the magnitude of the interaction decreased by the end of incubation indicating that a shift in the relative importance of factors had occurred. This shift might have been caused by differences in the ability of families to cope with treatment and/or a delayed or threshold embryo response to treatment. Importantly, there was a relatively consistent ranking of families in terms of survival through incubation and a strong dichotomy between those families performing well as diploids and pressure-induced triploids and those families doing relatively poorly regardless of treatment. High and low performance families were still distinguishable after ponding. This dichotomy suggests that it might be possible to improve survival of triploids using family based selection. However, maternal effects (i.e., the influence of the maternal phenotype and environment on offspring phenotype that is independent of the maternal genetic contribution), cannot be separated from the genetic effects of family using this study design and they are expected to have a large effect on juvenile survival (Heath and Blouw 1998). In contrast with survival, the effects of treatment and the interaction between treatment and family on growth parameters were minimal while family effects were large (30-61% of the variation in size-at-age and 29% of the variation in relative growth rate). Reaction norms show that the interaction was predominantly due to a pronounced 42 depression in growth that occurred in three pressure-shock treatment family groups that grew well as heat-shock induced triploids and diploids. This indicates that pressureshock treatment may affect growth differently from heat-shock. Pressure-shock treatments have been shown to reduce protein synthesis in oocytes and cultured cells (Wilson, Trogadis, Zimmerman, and Zimmerman 2001a; Begg, Salmon, and Hyatt 1983; Symington et al. 1991) but extended effects on subsequent juvenile growth seem unlikely unless growth is delayed immediately upon emergence and fish are unable to compensate (i.e., growth rates were depressed in these groups). The large effect of family on weight and growth rate is probably explained by the presence of genotypic effects and partly by the presence of maternal effects that influence offspring size during early development (Heath and Blouw 1998; Berg at a i 2001). However, study fish were sampled after the onset of exogenous feeding, approximately 154-258 days post­ fertilization when maternal effects are known to be either negative (i.e., offspring tend to resemble paternal phenotypes more than maternal phenotypes) or not significantly different from zero (Heath et a i 1999). Ranking of families was relatively consistent throughout freshwater growth and a clear dichotomy between low and high performing families was evident so that those families that grew well as diploids were the ones that also grew well as triploids. Individual fish serum antibody titre response against Vibrio anguillarum antigen varied widely within family treatment groups (but less within heat-shocked family groups) and no significant effect of treatment was detected. These results agree with those of Kusada (1991) who also found no difference in the ability of fish (ayu; Plecoglossus altivelis) to respond to vaccination; however agglutination techniques were used in that study. Comparative measurement of diploid-triploid immune system parameters has been limited in salmonids, and most have focused on measurements of 43 non-specific immunity (Yamamoto and lida 1995; Kusada, Salati, et al. 1991; Benfey 1999). However, the majority of these studies have concluded that diploids and triploids do not differ in their ability to mount effective non-specific immune responses (e.g., haemolytic, bactericidal, neutrophil activity and phagocytosis in rainbow trout and leukocyte profiles in tench, Yamamoto and lida 1995; Svobodova at al. 2001) and are equally responsive to vaccination as measured by mortality after challenge or natural outbreak (e.g., in ayu and African catfish, Inada at al. 1990; Na-Nakorn and Lakhaanantakun 1993). However, there is a substantial amount of anecdotal evidence suggesting that triploids are more susceptible to disease than diploids (Ojolick, Cusack, at al 1995; Langston, Johnstone, and Ellis 2001; J.W. Heath, personal communication) and recently, differences in the timing and recovery of triploid complement system activity and the hypofaerraemic response to lipopolysaccharide injection were found in Atlantic salmon (Langston at al. 2001) suggesting that subtle differences in immune functioning may be discovered by using temporal sampling techniques. In summary, the full-sib Chinook salmon families in this study grew well as triploids and did not exhibit overt differences from diploids in their ability to respond to vaccination. While triploid family groups performed less well than diploids and experienced considerably higher mortality, these differences were either within acceptable bounds for hatchery-reared fish or occurred during embryonic or larval development prior to the onset of exogenous feeding, when selective pressures are intense and the financial investment is low. Overall differences in growth and immune function between heat and pressure-shock treated family groups were found to be minimal indicating that the choice of induction treatment may not be as significant a factor as the quality of diploid broodstock used to generate them. Analysis of overall yield (calculated as the mean replicate group weight x number of surviving fish) prior to 44 saltwater transfer indicated that the lower survival experienced by heat-shock treated family groups after ponding resulted in a significantly lower yield than was obtained from diploid control groups (P=0.003). Mean yield obtained from pressure-shock treated family groups was not significantly different from that of heat-shock treated groups (P=0.308) or diploid control groups (P = 0.095) indicating that a slight advantage in yield might be gained by using pressure-based induction techniques. While the industrial application of triploid technology depends on the performance of triploid fish within the commercial environment, it also depends on how well improvements made using the selective breeding of diploids are maintained after triploidization. The obvious dichotomy between high and low performing families evident after triploidization and the large effect of family on survival, growth and the response to vaccination in this study indicates that family selection would be effective. This may be offset somewhat by significant treatment by family interactions and a trend suggesting more variable performance in triploid treatment groups. While this variation might be explained by the increased genetic potential in triploids for either beneficial or detrimental allelic combinations and higher levels of interlocus epistatic interactions, it complicates selection strategies and the ability to predict performance. Identification of the potential sources of this variation (e.g., does triploidization affect the size of the additive genetic component or the magnitude of maternal effects?) is necessary to determine the feasibility of selective breeding programs. While this study design prevented the quantification of additive genetic variation and maternal effects, these issues are addressed in the study included in Chapter Two of this thesis. 45 Chapter 2 QUANTITATIVE GENETIC ANALYSIS OF DIPLOID AND TRIPLOID CHINOOK SALMON PERFORMANCE CHARACTERISTICS 2.0 Abstract Monosex all-female Chinook salmon families bred using a paternal half-sib breeding design (62 females and 31 males) were used to test whether triploidization resulted in changes in: 1) the distribution and magnitude of phenotypic variation, 2) narrow-sense heritability and 3) maternal effects, of specific fitness-related parameters (i.e., of survival, size-at-age, relative growth rate and serum lysozyme activity) measured during the freshwater phase of the lifecycle. Analysis was performed separately for diploid and triploid family groups. It was found that triploidization resulted in significantly higher levels of phenotypic variance and profoundly different patterns of variance distribution, although this relationship was reversed for lysozyme activity. Additive genetic variance accounted for much more of the total phenotypic variance in triploids and this resulted in significantly higher narrow sense heritability values for triploid groups. However, maternal effects estimates were substantially lower in triploids than in diploids. These results indicate that the main effects of adding an extra set of chromosomes to the Chinook salmon genome are primarily additive and dominant and that, somewhat counter-intuitively, the relative magnitude of the combined effect of dominance, epistasis and maternal effects is not increased. This is highly suggestive of an overall ploidy dependent mode of gene expression. 46 2.1 Introduction Successful triploidization results in a balanced or euploid chromosomal state because an entire set of chromosomes is retained by the zygote. In triploid Chinook salmon this means that the chromosome number increases from 68 (the diploid number) to 102 chromosomes (Simon 1963; Phillips and Rab 2001). Unlike aneuploidy, in which a single chromosome or gene construct is added to the genome, triploidy does not always result in potentially catastrophic genomic imbalance. However, triploidy does increase bulk DNA content, the number of alleles at each locus and, potentially, the interactions among loci. These fundamental changes may modify relationships within (dominance) and between (epistasis) loci, with resultant alterations in gene expression and ultimately phenotype. Gene expression may be altered in a number of different ways including regulatory factor effects, RNA mediated interference and homology dependent recognition and silencing (Wassenegger 2002a; Wassenegger 2002b). Modulation of gene expression may result in gene dosage effects or dosage compensation. When a dosage effect occurs, gene expression is correlated with the number of copies of the structural gene (ploidy in the case of euploids). For example, haploids, diploids and triploids would have gene expression levels of 50%, 100% and 150% (positive gene dosage effect) or 200%, 100% and 67% (inverse gene dosage effect) (Birchler et al. 2001). Dosage compensation may also occur whereby a positive gene dosage effect is compensated by an inverse effect of another regulatory product on the structural gene leading to gene expression at diploid levels regardless of genomic ploidy state (Birchler et a /2001). The addition of a complete set of chromosomes probably does not disrupt gene expression patterns to the same degree as genomic manipulations that generate 47 aneuploidies because overa// stoichiometric relationships are not disrupted and the cytoplasmic:nuclear ratio is generally preserved, so that the concentration of regulatory factors is likely maintained (Birchler et al. 2001). However, this has not been investigated in a vertebrate ploidy series (i.e, groups of organisms in which the number of complete chromosome sets is varied sequentially). Evidence from plant and non­ vertebrate ploidy series generated using corn (Zea mays), fruitfly {Drosophila melanogaster) and yeast {Saccharomyces cerevisiae) have found that gene expression in polyploids tends to be positively associated with ploidy so that expression of specific genes increases in a linear manner as ploidy is increased (i.e., a positive gene dosage effect is exhibited) (Guo, Davis, and Birchler 1996; Lucchesi and Rawls 1973; Birchler at al. 1990; Galitski at al. 1999). Although some genes were found to have unusually high or low expression patterns outside of the range of simple gene dosage or dosage compensation effects (e.g., Guo at al. 1996), positive gene dosage effects appear to be the most prevalent form of modified expression in ploidy series experiments. However, a recent study using a silkworm {Bombyx mon) ploidy series suggests that a more complex relationship exists between ploidy state, parental origin of chromosome sets and parental specific regulatory factor influences on expression (Suzuki at al. 1999). Triploidization can be induced when a shock (typically heat, pressure or chemical) is applied to a fertilized egg just prior to second polar body extrusion. This shock, if applied successfully, causes the set of chromosomes within the polar body to be retained within the egg. Shock-induced triploidization is a stressful and highly perturbational event. Salmon zygotes subjected to induction must cope with potential treatment-related trauma as well as possible developmental, cellular, regulatory and phenotypic perturbations related to the forced transformation of genetic background caused by retention of an extra set of chromosomes. Despite this, typical cellular-level 48 compensatory responses to triploidization are displayed by salmonids (e.g., increased nuclear and cell size with apparent maintenance of the diploid nuclear to cytoplasmic ratio) and physiological parameters are remarkably similar to those of diploids (reviewed by Benfey 1999). Genetic changes associated with shock-induced triploidy in salmonids have not been thoroughly investigated. However, it is known that in addition to the increase in DNA quantity, shock-induced triploidy increases allelic diversity and number (i.e. three versus two alleles at each locus and potentially an additional different allele per locus (depending on recombination rate) (Thorgaard et al. 1983; Allendorf and Leary 1984; Leary et al. 1985). This increased genetic diversity might be expected to have positive fitness-related effects (e.g., deleterious alleles may have a higher probability of being masked or a synergistically favorable combination of alleles may occur (Garnier-Gere et al. 2002; Wang et al. 2002). However, the increased structural complexity implicit in triploid genomic architecture may alter allele and gene interactions (i.e., dominance and epistasis), alter specific regulatory factor stoichiometry, epigenetic, or developmental gene regulation so that gene expression patterns might be affected in a detrimental or stochastic manner. Furthermore, since genotype and phenotype are fundamentally linked through the patterns of gene expression during development, changes in genetic architecture (e.g., modifications of epistatic relationships among sets of developmentally important genes) caused by triploidization may modify expression by changing regulatory control of transcription patterns during development, and thus modify phenotypic potential. It is unknown if gene expression patterns change after triploidization or if epigenetic regulation occurs in salmon species. Gene expression studies of triploid salmon should be able to quantify the detailed effect of triploidization on expression 49 patterns of a specific sub-set of genes and clarify the role of parental or strain specific regulatory factor influences on gene expression. Quantitative genetic analysis on the other hand, would allow a direct estimation of the average phenotypic change (over all alleles at contributing loci) in additive genetic variance attributable to ploidy modification. This analysis would specifically entail the decomposition of phenotypic variance and comparative diploid/triploid estimates of narrow sense heritability (h^, the additive genetic component of the phenotypic variance of a trait). If complete or partial dosage compensation is occurring in triploids, then heritability values might be expected to be similar to those of diploids; however, if there is incomplete dosage compensation then heritability values might be expected to be significantly larger in triploids. Results obtained in Chapter One and recent published work (Bonnet et al. 1999; Blanc at al. 2001 and Friars at al. 2001) suggest that phenotypic variance is increased in triploid salmon. If all or most gene action is additive there should be a linear relationship between phenotypic variance, allelic or genetic diversity and additive genetic variation (Falconer and Mackay 1996; Reed and Frankham 2001). Such a relationship would predict higher triploid phenotypic variance and heritability values relative to diploids. Triploidization may also modify dominance related interactions between alleles (the phenotypic effect of the interaction of alleles at single loci. Falconer and Mackay 1996) as well as epistatic interactions among loci (the phenotypic effect of gene interactions, Cheverud and Routman 1995). Epistasis and dominance may contribute to additive genetic variance and inflate heritability estimates under certain conditions and allele frequencies (e.g., perturbation of genetic background, population bottlenecks) (Willis and Orr 1993; Whitlock at al. 1993; Cheverud and Routman 1995; Lynch and Walsh 1998). In quantitative genetic analyses, epistatic variance components are usually considered negligible and so are generally ignored, and if main effects are 50 primarily additive or dominant then this may be valid (Falconer and Mackay 1996; Roff 1997; Wade 2002). In the present study, a suite of monosex all-female Chinook salmon families {Oncorhynchus tshawytscha) bred using a paternal half-sib mating design was used to test whether triploidization resulted in changes in: 1) the distribution or magnitude of phenotypic variation, 2) narrow-sense heritability and 3) maternal effects. Maternal effects occur when the phenotype or genotype of the mother, or the environment she experiences has a phenotypic effect on her offspring (Rossiter 1996; Mousseau and Fox 1998; McAdam et al. 2002). Although this analysis is primarily designed to test for changes in the nature of quantitative trait expression in diploid and triploid salmon, the results will have relevance for aquaculture as well. The potential for significant changes in the inheritance patterns of performance traits in triploid offspring from a highperformance broodstock has serious implications for the application of triploid sterilization in commercial salmon aquaculture. 2.2 Methods 2.2.1 Fish, Breeding Design & Treatment Details Monosex all-female chinook salmon broodstock from Yellow Island Aquaculture Ltd. (YIAL; Quadra Island, B.C.) were mated using a paternal half-sib design. In this mating scheme, each of 31 hormonally masculinized phenotypic males (neomales) were mated to two independent, non-related and randomly chosen females (females were bred once and only to one male). The breeding design resulted in a total of 62 fullsib families nested within 31 paternal half-sib groups. The fertilized eggs from each fullsib family were divided into two 250 ml sub-samples, each of which was subjected to one of the following two treatments: i) Hydrostatic pressure-shock. 6.89 x 1 51 kPa (10 000 psi) of pressure applied for 5 minutes, 30 minutes after fertilization or, ii) Control. Eggs were left untreated and transferred to incubation trays immediately after the water hardening process. 2.2.2. Fish Rearing & Husbandry Eggs from each family treatment group were incubated in separate compartments of vertical stack incubation trays (Heath Techna Corp.). When eggs reached the eyed stage of development (the point at which the eye spots of the developing embryo are visible through the egg shell and at which % yolk vascularization has occurred; November 28-December 11, 2000, -280-296 ATUs) they were mechanically shocked and sorted using a Jensorter machine (Model JM4C, Jensorter, Inc.) and returned to the incubation stacks. Trays were divided into twelve compartments (10 cm x 10 cm x 5cm) with each compartment holding approximately 700 eggs. Water temperature within the stacks was monitored using a digital data logger (Onset Computer Corp.) and development stage of the fish was tracked using Accumulated Temperature Units (ATUs; calculated as the cumulative total of daily mean temperatures). Mean water temperature during the incubation period was 7.72 °C ± 0.02 °C. Mean flow within the stacks was 13 L/minute. As alevins completed yolk-sac absorbtion (February 24-March 11, 2001 ; -9271006 ATUs), 100 alevins were randomly selected from the treatment groups (pressure and control) within each of the 62 full-sib families and transferred from the vertical incubation stack tray compartments to 140 L aerated rearing tanks for the onset of exogenous feeding. Two sets of fish were randomly assigned to each tank at a starting density of 200 fish per tank (-1.43 fish/L). One set of fish in each tank was fin clipped for identification purposes (either the upper or lower caudal fin lobe was removed). 52 Flow rate of water to the tanks was approximately 3 L/minute and the mean temperature (February 24‘^-July 1) was 8.79 °C (7.23 -10.18 °C). Fish were handfed to satiation multiple times per day with commercial feed (Ewos, Canada, Ltd.). 2.2.3. Ploidy Determination Erythrocyte nuclear length was used to determine family-specific and overall triploidization success (Wolters et al. 1982; Beck and Biggers 1983; Benfey et al. 1984). The validity of this measurement had been specifically tested against flow cytometric data during preliminary analysis of juvenile chinook at YIAL and was confirmed again for the analysis of chapter one data. A nuclear length of 8.5 pm was found to be the threshold measurement that most reliably distinguished diploid from triploid individuals. Approximately 11-20 fish were terminally sampled from each family treatment group (pressure = 1084 fish; control = 997) and whole blood smears were made for each fish. The length of ten randomly chosen erythrocyte nuclei per smear was measured to the nearest 0.01pm and the mean nuclear length used to determine the ploidy status of each fish. Slides were fixed in methanol and stained with Wright-Giemsa (Sigma). Visualization and measurement of erythrocyte nuclei was accomplished under oil immersion (lOOOx magnification) using an Olympus BX-50 compound microscope (Olympus Optical Co.) equipped with a Qlmaging Retiga 1300 Monochromatic digital camera (Quantitative Imaging Corp.) and the Northern Eclipse, version 6.0 imaging program (Empix Imaging Inc.). 2.2.4 Survival Incubation survival of family treatment groups was monitored from fertilization to the eyed stage of development (S-1) and then followed through to the alevin stage, just 53 prior to transfer of the fish to freshwater rearing tanks. Embryo mortalities from fertilization to the eyed stage (3/4 yolk vascularization) were assessed after the eggs were mechanically shocked and sorted using a Jensorter machine. Total egg number and the number of live eggs for each family treatment group were evaluated at this time. After the initial eyed egg count, mortality was monitored at least every two days. Incubation survival was determined for all groups at the following developmental stages, S-2, S-3, S-4, S-5, S-6, S-7. Refer to Table 2.1 below, for a summary of the performance variables measured in this study and the corresponding in text abbreviations. Survival after transfer to rearing tanks (8-8) was also monitored every second day and was determined as the number of live fish remaining immediately prior to the experimental vaccination treatment divided by the total number offish originally fin-clipped and released into the rearing tank. Experimentally sacrificed fish within each group (i.e., those fish terminally sampled to determine triploidization success) were excluded from the final survival assessment. 2.2.5 Growth Weight (in grams) was determined by non-terminal sampling at five time points during the freshwater growth of the ponded family groups: W-0, W -1, W-2, W-3 and W4 (Refer to Table 2.1 for specific definitions). At ponding (W-0), 100 alevins from each family treatment group were weighed in water as they were transferred to rearing tanks so that a mean family weight was obtained at this sample point. At all other sample times, weights of individual fish were recorded. Approximately twenty fish per family treatment group (pressure, control) were weighed at the W-1 (n = 2460) and W-2 (n = 2568) sample points. At the W-3 (n = 1290) and W-4 (n = 1157) sample points, 40 fish 54 per family treatment group were weighed. Table 2.1. Summary of measured performance variables and in-text abbreviations Abbreviation Performance Variable Incubation survival (staae-SDecific) S-1 S-2 S-3 S-4 S-5 S-6 S-7 Fert-288 ATUs (eyed stage) -288- 510 ATUs -510- 616 ATUs -616- 746 ATUs -746- 789 ATUs -789- 853 ATUs -853-970 ATUs Survival after oondina to rearina tanks S-8 -970-2008 ATUs (-127-244 days post-fert) Size-at-age (weight, grams) W-0 W-1 w-2 W-3 W-4 -970 ATUs (-127 days post-fert)* -1228 ATUs (-159 days post-fert) -1729 ATUs (-215 days post-fert) -1835 ATUs (-235 days post-fert) -1943 ATUs (-244 days post-fert) Relative growth rate rgr-c -970-1943 ATUs (-127-244 days post-fert) rgr-1 rgr-2 rgr-3 - 970-1228 ATUs (-127-159 days post-fert) -1228-1729 ATUs (-159-215 days post-fert) -1729-1943 ATUs (-215-244 days post-fert) Serum lysozyme activity (EU/5 gl) SLR Pre-post-vaocination difference * NOTE: W-0 = the mean weight of 100 alevins per family treatment group (weighed in water at the time of ponding) 55 Relative growth rate of replicate family treatment groups was assessed for the complete period of growth after ponding, a time interval o f -100 days (rgr-c in Table 2.1) but was also calculated for specific sub-periods; rgr-1 (32 days), rgr-2 (56 days) and rgr3 (29 days) in Table 2.1. Because individual fish were not tracked, mean replicate treatment group weight data were used for the first sampling point and individual fish weights were used for the second sampling point. Relative growth rate was calculated using the following formula: (( Y2 -Ÿ 7)+ Ÿ^ (t2 -ti )) x1 00 where, Yg = individual fish weight at the second sampling point, Yi = mean treatment group weight at the first sampling point and tg - ti = the mean time interval in days between the first and second sampling points. 2.2.6. Vaccination & Serum Lysozyme Activity Assay To determine if pre- and post-vaccination serum lysozyme activity levels differed between ploidy types, approximately 15 fish per family treatment group were terminally sampled just prior to (June 6-14*^, 2001) and 10 days after vaccination (June 24-25^) with a commercial vibrio vaccine (Alpha-Dip 2100, Vibrio anguillarum, serotype 01 and V. ordalii bacterin; Alpharma NW Inc.). Vaccine was diluted 1:9 with hatchery water and fish were immersed for 30 seconds in the aerated solution before being returned to rearing tanks (vaccination occurred June 14*'^-15“^). Pre-immunization serum lysozyme activity levels were determined for 769 fish (2N-Control = 379, 3N-Pressure = 390) and post-immunization serum lysozyme activity levels were determined for 842 fish (2NControl = 390, 3N-Pressure = 452) using the modified microplate assay protocol of Rungruangsak -Torrissen (Rungruangsak -Torrissen et al. 1999) based on Ellis (1993). 56 Briefly, 5 |jl of undiluted serum were placed into wells of a 96-well microplate; 95 pi of a 0.21 mg/mL Miccrococcus lysodeikticus -0.05 M phosphate buffered saline solution was then added quickly to all wells using a multi-channel pipettor and the absorbance at 450nm was measured after 1 and 5 minutes at 25 °C using a VERSAmax tunable plate reader with Softmax Pro 4.0 software (Molecular Devices, Corp.). Thirty-two serum samples were run in duplicate on each plate with two columns of wells used to run a series of hen egg white controls (4-1000 pg/mL HEWL diluted in 0.05 M PBS), one column of wells was left as a series of blank controls, and one column of wells was run as internal PBS controls. Serum samples with absorbance values outside the range of the HEWL controls were run again in dilution. The enzymatic activity of the HEWL standards was determined using a quality control assay. One unit of lysozyme activity (EU) was defined as the amount of enzyme causing a decrease in A 45 onm of 0.001/minute. The response to vaccination was determined as the individual fish post­ vaccination enzyme activity level minus the mean family pre-vaccination enzyme activity level. Sample activity levels were expressed as enzyme un its/m L (EU/mL) and were logio transformed to attain normality for statistical analysis. 2.2.7 Statistical Analyses One-way analysis of variance (ANOVA) was used to determine the effect of treatment on measured performance parameters. One-way ANOVA was again used to determine how total phenotypic variance of each performance trait was partitioned within and among families. Pitman's procedure for correlated populations was used to test for significant differences between variances (Zar 1996). Analysis was run separately for each treatment group/performance trait combination. To partition phenotypic variance for the quantitative genetic analysis, nested 57 ANOVA was used with sire and dam factors treated as random effects (Model II). Variance was partitioned between sires, between dams nested within sire, [dam(sire)], and between offspring nested within dam, [offspring(dam)]) using the following linear model; Yijk = M+ Ai + Bij + Eijk where, Yyk is the phenotype of the kth offspring from the the jth dam, p is the parametric mean of the population. ith sire. By is the random effect of the jth dam mated family of the ith sire mated to A; is to the ith the randomeffect of the sire and Sykis the residual deviation (Sokal and Rohlf 1995). Sire and dam additive genetic components (heritabilities, h^sire, h^dam) were estimated using the appropriate mean squares from the nested ANOVAs and variance components calculated using standard formulas as outlined in Roff (1997) and Falconer and Mackay (1996). Standard errors of sire and dam heritabilities were estimated using the appropriate intraclass correlation coefficients using the techniques of Robertson (1959). The sire heritability was used as the best estimate of the additive genetic component because it is not inflated by variance due to dominance or maternal effects and probably only minimally inflated by epistatic effects (Roff 1997), at least for diploids. Sire heritabilities were considered significantly different from zero when the F-value derived from the analysis of variance indicated a significant sire effect regardless of whether the 95% confidence interval of the h^ estimate (derived as 1.96 x SE of the h^ value) encompassed zero (Roff 1997). It is important to recognize that an assumption of negligible epistatic effects has been made in the analysis. This may not be a valid assumption for the calculation of heritability for triploids where the probability for epistatic interactions is increased. But, as Roff (1997) states, epistatic interactions will 58 tend to inflate the additive and dominance components of variance and so would probably appear as inflated sire or dam heritability components. The paternal half-sib model employed in this study does not allow for the specific decomposition of epistatic, maternal effects or dominance variances. Despite this, a general estimate of the magnitude of maternal effects was calculated using the difference between the dam and sire causal components of variance divided by the total phenotypic variance for each performance trait. The dam component contains all variance due to maternal effects (genetic and non-genetic), one quarter of the variance attributable to dominance and small proportions of epistatic variance (i.e., 3/16 additive X additive, 1/8 additive x dominance and 1/16 dominance x dominance). If dominance and/or epistasis is present (or inflated by triploidization) it will be confounded with the maternal effects estimate (Roff 1997). The study design thus allows maternal effects to be detected and generally estimated, but the maternal effects estimate cannot be decomposed and the presence of dominance and/or epistasis may inflate the estimate (Roff 1997). Analysis was performed separately for diploid control and triploid pressureshock treatment groups for each trait. Differences between diploid and triploid heritabiity estimates were examined and a paired sample nonparametric sign test was used to identify the probability of obtaining the observed distribution of differences (Zar 1996). Sire and dam heritabilities were estimated for triploidization success (a threshold trait) in pressure-shock treated fish first on a 0, 1 scale (where, 1 = triploid and 0 = diploid) according to Roff (1997) and then on the underlying or liability (Falconer and Mackay 1996) scale according to Dempster and Lerner (1950) using Hamaker’s (1978) exact approximation of the z value. Standard errors and maternal effects estimates were calculated as above. 59 2.3 Results 2.3.1 Patterns of Variance Distribution Among and Within Families Triploidization resulted in increased overall phenotypic variance (Table 2.2). Triploids exhibited significantly higher (P < 0.05) phenotypic variance among families than diploids for most traits (i.e., for survival, both during incubation and after ponding, size-at-age and relative growth rate) with the exception of serum lysozyme activity. However, diploids did have significantly higher among family variance for incubation survival over the fourth and last sample periods, S-4 and S-7, and for relative growth rate during the last sample period, rgr-3. Within family variance of treatment group sizeat-age was not significantly different between ploidy types at any of the post-fertilization sample points (W-1 to W-4) but triploids had significantly higher within family variance for the first and third sample periods (rgr-1, rgr-3 and rgr-c). Diploid within family variance was significantly higher for relative growth rate over the complete study period (rgr-c). Interestingly, the trend indicating increased variance among and within triploid families was reversed for the difference between pre- and post vaccination serum lysozyme activity. Diploids exhibited significantly higher among (P < 0.05) and within family variance (P < 0.001) than did triploids. 2.3.2 Sib analysis, Narrow Sense Heritability and Maternal Effects A number of interesting differences between diploids and triploids were noted in the partitioning of phenotypic variance among sires, among dams nested within sires,and among progeny within dams during the sib analysis (Table 2.3 and 2.4) and estimation of heritability (Table 2.5). Results are reported in terms of the distribution of variance, the relative contribution of causal components along with the heritability and maternal effects estimates (Table 2.5) separately for each performance trait below. 60 Table 2.2. Among and within family phenotypic variance of measured traits in diploid control and triploid pressure-shock treated family groups. *, P < 0.05; ** P < 0.01, ***, P < 0.001. Trait abbreviations are as defined in Table 2.1. N/A = not applicable. Trait Among Family Phenotypic Variance 2N - Control 3N - Pressure Within Family Phenotypic Variance 2N - Control 3N - Pressure Incubation survival fstaae-soecific) 3.337 X 10'^ 1.343 X 10^ 3.855 X 10'^ 7.092 X 10'^ 3.691 X 10^ 3.795 X 10^ 7.889 X 10'^* 5.521 X 10'^* 4.624 X 10'^*** 5.781 X 10'^** 5.253 X 10-2* 4.827 X 10'^** 5.218 X 10'^** 7.542x10=' N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 3.434 X 10'^ 6.293 X 10^** N/A N/A 6.794 X 10'^ 1.206 X 10'^ 1.073 X 10'^ 1.847 X 10'^ 1.011 X 10'^* 2.010x10""* 1.597 X 10""* 2.834 X 10""* 9.761 X 10® 0.436 0.605 1.201 1.101 X 10'^ 0.457 0.526 1.152 Rgr-c 7.989 X 10'® 1.031 X 10-^* 6.501 xIO'^* 6.388x10"* Rgr-1 Rgr-2 Rgr-3 1.830 X 10 ® 5,323 X 10 ® 5.483 X 10 ®* 2.135 X 10 ®* 9.443 X 10'®* 4.925 X 10 ® 4.596 X 10 ® 2.354x10^' 9.767 X 10 ® 5.092 X 10 ®* 2.228 X 10"* 9.872 X 10 ®* 0.9868* 0.8646 13.954*** 3.205 S-1 S-2 S-3 S-4 S-5 S-6 S-7 Survival after oondina to rearina tanks S-8 Size-at-age (weight, grams) W-1 W-2 W-3 W-4 Relative growth rate Serum lysozyme activity (EU/5 fjl) SLR 61 Table 2.3. Results of nested Model II ANOVA and calculated variance components for the heritability analysis of size-at-age and serum lysozyme activity measured in diploid control and triploid pressure-shock family treatment groups. *, P < 0.05; **, P< 0.01;*** P< 0.001. Trait Size-at-age W-1 W-2 W-4 Lysozyme activity SLR Source Triploid Pressure-Shock Diploid Controi SS df MS Van SS df MS Van Sire Dam(sire) Offspring 15.046 10.841 11.111 29 30 1142 0.519 0.361 0.010 0.004 0.018*** 0.010 28.455 15.063 12.137 30 31 1172 0.948 0.486 0.010 0.0116* 0.0238*** 0.0100 Sire Dam(sire) Offspring 198.968 284.379 494.618 29 30 1139 6.861 9.479 0.434 -0.065 0.452*** 0.434 377.653 439.169 535.803 30 31 1175 12.588 14.167 0.456 -0.039 0.686*** 0.456 Sire Dam(sire) Offspring 192.417 201.986 324.637 28 29 523 6.872 6.965 0.621 -0.005 0.634*** 0.621 338.050 204.546 269.725 28 29 521 12.073 7.053 0.518 0.251 0.654*** 0.518 Sire Dam(sire) Offspring 371.881 325.343 650.349 28 29 522 13.281 11.219 1.246 0.103* 0.997*** 1.246 696.174 393.470 522.665 28 29 515 24.863 13.568 1.015 0.565* 1.255*** 1.015 Sire Dam(sire) Offspring 1162.682 2178.403 6022.107 17 18 199 68.393 121.022 30.262 -4.048 12.966*** 30.262 1738.701 1670.575 4222.691 21 22 253 82.795 75.935 16.690 0.490 8.464*** 16.690 62 Table 2.4. Results of nested Model II ANOVA and calculated variance components for the heritability analysis of relative growth rate and triploidization success measured in diploid control and triploid pressure-shock family groups. *, P < 0.05; ** P < 0.01, ***, P < 0 .001 . Trait Source SS Diploid Control MS df Var. SS Triploid Pressure-Shock Var. df MS Relative growth rate Rgr-c Sire Dam(sire) Offspring 0.334 0.309 0.769 28 29 522 0.012 0.011 0.001 5.0 X 10'® 0.001*** 0.001 0.549 0.241 0.755 28 29 518 0.020 0.008 0.001 0.0006** 0.007*** 0.001 Rgr-1 Sire Dam(sire) Offspring 0.392 0.372 0.633 29 30 1142 0.014 0.012 0.001 5.0 X 10'® 5.0 X 10^*** 0.001 0.435 0.332 0.703 30 31 1168 0.014 0.011 0.001 0.001 5.0 X 10"**** 7.5x10" Sire Dam(sire) Offspring 0.293 0.331 1.028 29 30 1140 0.010 0.011 0.001 -2.5 X 10'® 5.0 X 10'“*** 0.001 0.857 0.231 1.117 30 31 1176 0.029 0.007 0.001 5.5 X 10"**** 3.0 X 10"**** 0.001 Sire Dam(sire) Offspring 0.185 0.164 0.654 28 29 468 0.007 0.006 0.001 0.001 5.0 X 10"^*** 5.0 X 10'® 0.152 0.099 0.558 28 29 483 0.005 0.003 0.001 1.0x10"* 2.0x10"**** 0.001 Sire Dam(sire) Offspring N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 26.091 15.016 67.832 28 29 998 0.932 0.518 0.068 0.0112 0.025*** 0.068 Rgr-3 Triploidization 63 Table 2.5. Sire component heritability (h^± SE) and maternai effect estimates (expressed relative to total variance, Vm, %, and as an absolute value) for measured performance traits. Heritabilities considered significantly different from zero are marked as: *, P < 0.05, **P < 0.01, *** P :< 0.001. Triploid ± SE values highlighted in bold type are significantly larger than comparable diploid values. NA Trait h x ire ±SE Design Vm relative, absolute Dams, sires, n Triploidization 2N-Control NA NA 3N-Pressure 0.43 ± 0.20 (0,1 scale) 0.32 ±0.14 (underlying) 2N-Control N/A 3N-Pressure 13% UB: 58, 29, 1056 Size at age W-1 0.50 ± 0.21 1.02 ±0.35* 44%, 0.0140*** 27%, 0.0122 B: 60, 30, 1200 W-2 -0.32 ±0.10 -0.14 ±0.29 51%, 0.452 60%, 0.686*** B: 60, 30, 1200 W-3 -0.01 + 0.07 0.71 ±0.29 51%, 0.634*** 28%, 0.403 B: 58, 29, 580 W-4 0.18 ±0.14* 0.80 ± 0.33* 38%, 0.894*** 24%, 0.690 B: 58, 29, 580 Relative growth rate rgr-c 0.10±0.11 1.04 ±0.31** 46%, 0.00095 74%, 0.0064*** B: 58, 29, 580 rgr-1 0.13 ±0.08 0.19±0.11 29%, 0.00045 0%, -0.0005 B: 60, 30, 1200 rgr-2 -0.07± 0.02 1.19 ±0.39*** 33%, 0.0005 0%, -0.00025 B: 60, 30, 1200 rgr-3 0.13 ±0.12 0.31 ±0.18 0%, -0.0005 8%, 0.0001 B: 58, 29, 580 Serum lysozyme activity SLR-response to vaccination -0.41 ±0.10 0.08 ±0.15 30%, 12.966*** 31%, 7.974 UB: 36,18, 235 64 2.3.2.1 Size-at-age While the component of variance in size-at-age associated with differences between the progeny of different sires (among sires component) was considerably smaller than that associated with differences between the progeny of different dams (nested within sires) at all sample points in both diploids and triploids, the among sires component of variance in triploids was on average four times larger than the comparable diploid value. A mean of 16% (15.78%) of the variance in size-at-age of triploid progeny and 4% (4.22%)of the variance in diploid progeny was attributable to differences among sires (Table 2.3). This result suggested that a larger proportion of the phenotypic variance in size-at-age of triploid progeny was due to additive genetic effects. As can be seen in Table 2.5, the estimates of additive genetic variance (as a proportion of total phenotypic variance) or heritability, were consistently higher in triploids at each sample point (except for W-2 when both estimates were negative); this translated to an average (over all size-at-age sample points) of 63% of total phenotypic variance explained by additive genetic variance in triploids versus 17% in diploids (calculated as the mean of 4 x among sire variance). Heritability estimates of size-atage at W-1 and W-4 in triploids and W-4 in diploids were found to be significantly different from zero. However, the large standard errors associated with the heritability estimates for size-at-age result in 95% confidence intervals that rendered differences between diploid and triploid estimates predominantly non-significant. The proportion of the phenotypic variance in size-at-age attributable to differences among dams was surprisingly similar between ploidy types. Of total phenotypic variance at each of sample point (W-1, W-2, W-3 and W-4), 56%, 51%, 51% and 42% in diploids and 55%, 60%, 46% and 44% in triploids was attributable to differences among progeny of females mated to the same male suggesting that considerably less variance was associated with 65 maternal, dominance and epistatic effects In triplolds than in diploids. Maternal effects estimates (relative, Vm, %) for size-at-age were substantially lower for triploids than for diploids at each sample point (except at W-2;Table 2.5) and were on average 11% lower than diploid estimates. The absolute value of the variance attributable to maternal effects was found to be significantly higher in diploid groups than in triploid groups as tested with Pitman's procedure for correlated populations, although this trend was reversed at W-2. 2.3.2.2 Relative growth rate The distribution of variance in relative growth rate was somewhat more complex when considered as three distinct periods of growth (rgr-1, rgr-2, rgr-3) than when overall relative growth rate was considered (rgr-c). Similar to the trend in size-at-age, substantially more variance in the overall relative growth rate was attributable to differences among progeny of different sires (among sire variance) in triploids than in diploids. Among sire variance accounted for 7% of total phenotypic variance in triploids and only 2% in diploids; this resulted in a much larger additive genetic variance component in triploids than in diploids as is evident by the heritability estimates for rgr-c in Table 2.5 (h^ ± SE, 2N = 0.10 ± 0.11; 3N = 1.04 ± 0.31). Only the triploid complete relative growth rate heritability estimate was found to be significantly different from zero. While large standard errors again made diploid and triploid heritability estimates not significantly different from each other, there is a distinct indication of greater genetic determination of relative growth rate in triploids than in diploids. The distribution of variance among dams (within sires) and among progeny (within dams) was substantially different between ploidy types in terms of magnitude (2N; 49% among dams, 49% among progeny; 3N; 81%% among dams, 12% among progeny) suggesting 66 that there was more intra-family variation in diploids but potentially more influence of dominance, epistasis and maternal effects in triploids than in diploids. The relative maternal effects estimate ( V m , %; Table 2 . 5) confirmed the substantially higher influence of the maternal effects component in triploids for this trait. The relative value of maternal effects for triploids was 28% higher than that of diploids ( 2 N - V m = 46%, 3 N - V m = 74%). Absolute variance attributable to maternal effects in triploids was significantly higher in triploids than in diploids (P < 0.001). When relative growth rate was considered as three distinct periods of growth (rgr-1, rgr-2, rgr-3), diploids and triploids exhibited opposite temporal trends in the distribution and magnitude of variance components over the study period (Table 2.4). In diploids, the among sire component of variance accounted for 3% and 0% of the total phenotypic variance in growth periods rgr-1 and rgr-2 and rose to 65% of the total phenotypic variance during rgr-3, just prior to saltwater transfer. The opposite trend in among sire variance was exhibited by triploids. Triploid among sire variance accounted for 63%, 30% and 8% of the variance among progeny of different sires in rgr-1, rgr-2 and rgr-3. The among dam component of variance remained stable in diploids accounting for between 32-33% of the total phenotypic variance during each growth period while the triploid among dam component declined (accounting for 32%, 16% and 15% of the total phenotypic variance in each growth period). Opposite trends in the among progeny (within dam) component of variance were also apparent between ploidy types, with diploid variance decreasing and triploid variance increasing over time. Relative maternal effects estimates ( V m , %; Table 2.5) were substantially lower in triploids than in diploids for each growth period when considered separately (averages of 3%-3N and 21%-2N). Maternal effects estimates for diploids initially increased and then declined over the experimental period (rgr-1 = 29%, rgr-2 = 33%, rgr-3 = 0%; Table 67 2.5) while triploid estimates were essentially zero (rgr-1 and rgr-2) and then experienced a slight increase (rgr-3) (rgr-1 = 0%, rgr-2 = 0%, rgr-3 = 8%; Table 2.5). Absolute maternal effects variances were either negative or very low (Table 2.5). Pitman's test for significant differences in variance could not be used due to negative or zero variance values. The general trends in variance distribution suggest that triploid relative growth rate was becoming more variable over time while diploid variability declined. Additive genetic variance appeared to be quite variable in triploid groups over the relative growth rate periods and remained fairly constant for diploid groups, but all heritability estimates for these growth periods, at least for diploids were not significantly different from zero. Heritability in triploids was significantly different from zero for rgr-2; this estimate was also significantly different from the diploid heritability estimate for this period (h^± SE, 2N, 0.07 ±0.02; 3N, 1.1910.39). 2.3.2.3 Response to Vaccination: Serum Lysozyme Activity In contrast to the growth parameters detailed above, the phenotypic variance of serum lysozyme activity response to vaccination (SLR; Table 2.3) was distributed similarly in diploids and triploids. There was no variance associated with differences among progeny of different sires in diploids and only 2% of total phenotypic variance was associated with the among sire component in triploids. The remaining phenotypic variance was predominantly distributed at the inter-individual level so that 70% (diploid) and 65% (triploid) of the total phenotypic variance was associated with differences among individual progeny within families. Variance attributable to among dam differences accounted for the remaining variance, with 30% and 33% associated with the among dam component in diploids and triploids respectively. Heritability estimates of the difference between pre- and post-vaccination lysozyme activity (the response to 68 vaccination) in diploids and triploids were not significantly different from zero and were not significantly different from each other since 95% confidence intervals overlapped (h^ ± SE, 2N, -0.41 ± 0.10; 3N, 0.08 ± 0.15). While the relative maternal effects estimates of the response to vaccination were not different (2N = 30%, 3N = 31%: Table 2.5), absolute variance attributable to maternal effects was significantly higher in diploids than in diploids (Table 2.5). 2.3.2 4 Triploidization Success As with the serum lysozyme acitivity response to vaccination, variation in triploidization success was partitioned mainly among progeny (within dams) (65%) and among dams (within sires) (24%). Although more variance was associated with differences among progeny of different dams bred to the same sire than with differences among progeny of the same sire, 11% of the total phenotypic variance was associated with the among sire component suggesting that a substantial proportion of the variability in triploidization success was under additive genetic control. The underlying heritability estimate, while not associated with a significant sire effect (in the AN OVA) was moderate (h^ ± SE, 0.32 ± 0.14) and the 95% confidence interval did not encompass zero. Maternal effects were present but were relatively low (13%-relative estimate; absolute variance = 0.0138) suggesting that the combined environmental and genetic components of the effect of the maternal phenotype, dominance and epistatic effects on triploidization of offspring was not large. 2.3.2.5 Sign Test of Heritability Differences Triploid heritabilities were consistently higher than those of diploids and the sign test confirmed this by rejecting the null hypothesis that the median difference between 69 triploid and diploid heritability estimates was zero [P(X < 0 or X > 9) = 0.0019; Co.o5(2),9 = 1 , n-Co.o5(2),9 - 8; 0.02 > P > 0.001]. 2.3.3 Performance and Triploidization Success As performance-based differences between treatment groups were not the main focus of this paper, I review these results only briefly below. As was found in Chapter One, diploid families outperformed triploids but the differences were minimal. While triploids experienced survival that was 13-12% lower than that of diploids during incubation and after ponding, relative growth rate over the freshwater rearing period did not differ significantly between treatment groups. By the time of saltwater transfer, mean weight of diploids and triploids was also not significantly different. Additionally, no significant difference in the non-specific immune parameter, serum lysozyme activity was found either within or between ploidy types, before or after immersion vaccination. Pre-vaccination lysozyme activity however was lower than post-vaccination activity for both diploid and triploid family groups although levels after vaccination were not significantly higher than before exposure to the vaccine. Interestingly, while the overall mean lysozyme activity level of pressure-shock triploid family groups was 3% higher than that of diploid control groups prior to vaccination, it was 10% lower than that of the control groups after vaccination, but not significantly so. Overall triploidization success was 88%. A small percentage of samples (n = 21 fish or 1.0% of those analyzed) were discarded because of uncertainty in the determination of ploidy due to overlap between the diploid and triploid mean red blood cell nuclear length distributions. Family-specific analysis of induction success indicated that triploid levels ranged from 22% -100%, with 71% of the 62 families (44/62) having triploidization rates between 95% (13 families) and 100% (31 families). Of the remaining 70 families, 7 had triploidization rates between 80-89%, 5 between 71-78%, 2 between 6568%, 3 between 30-40% and 1 family had a triploidization rate of 22%. The presence of diploids in samples (up to 78% in one family) is an acknowledged limitation of this study. However, all known diploids were dropped from analysis of size-at-age at the W-3 and W-4 sample points (-235 and 244 days post-fertilization) as well as the respective relative growth rate sample periods and only known triploid and diploid samples were included in the serum lysozyme activity level assays, so that the inclusion of pressureshocked but diploid individuals in analysis was only confounded with survival and the W-1 and W-2 (-159 and 215 day post-fertilization) size at age measurements. 2.4. Discussion Triploidization is expected to increase total phenotypic variance. Using a quantitative genetic framework, the total phenotypic variance can also be decomposed into genetic effects (i.e., additive, dominance and epistatic components) and environmental effects (e.g., a maternal effects component). However, the specific experimental design used in this study, a paternal half-sib design, can only decompose total variance into an additive genetic component versus a combined epistatic, dominance and maternal effects component. Despite this limitation, the differences between the patterns of distribution of the phenotypic variation are useful and can provide insight into the relative importance and distribution of the casual components. The observed increased phenotypic variability among, and to a lesser extent, within, triploid families shows that the addition of an extra set of chromosomes directly affects the resulting phenotypic variance. Among family variance was expected to be larger than within family variation because of the 2/3rds maternal genetic contribution. This phenotypic effect is probably most evident because environmental variance has 71 been minimized and genetic composition of families has been kept constant. The expansion of phenotypic variance might be explained within a developmental context, especially if the developmental process is thought of as a set of modules consisting of networks of interacting transcriptional genes. If these networks have evolved to be increasingly complex as a result of selection for developmental stability (Frank 1999; Siegal and Bergman 2002), with the suppression of genetic variation (cannalization) as a by-product (Wagner et al. 1997), then perturbing the regulatory control of development by forcing triploidization could release phenotypic variation (Stearns 2002). While increased phenotypic variation was evident in growth related traits in triploids, significantly less variation occurred in the triploid lysozyme response to vaccination than was evident in diploids. It is unknown why cannalization appears to have been disrupted in the growth-related parameters but not in the immune parameter. It is possible that this relates to differences in the regulatory control complexity and/or parental specific genetic expression patterns of growth-related and non-specific immune genes. Significantly higher among and within family phenotypic variance in growth parameters (of length and weight) was also recently noted for triploid full-sib Atlantic salmon families (Friars et al. 2001). The authors attributed the higher levels of variance to a triploidization-induced disruption of uniformity but did not hypothesize a mechanism. Phenotypic variation of immune parameters has not been looked at in detail before. Total phenotypic variance was also partitioned into among sire, among dam and among progeny components for the heritability analysis. Among sire variance is used to estimate additive genetic variance and heritability while the difference between the among dam and among sire variance is used to estimate the combined contribution of dominance, epistasis and maternal effects. Additive genetic variance and heritability are 72 important because they describe the genetic resemblance between relatives and can be used to estimate the short term response to selection (Lynch and Walsh 1998; Wang et al 2002); however such considerations are irrelevant in triploids because of sterility. However, because additive genetic variance also reflects the variance of the average effect of alleles of the parents as expressed in the offspring (Falconer and Mackay 1996), these estimates can provide information about the genetic architecture and gene expression patterns of triploids. Heritability estimates of performance parameters in this study were significantly higher in triploids than in diploids (Table 2.5; sign test results); this is clearly illustrated in Figure 2.1 as the difference between triploid individual trait heritability values and the mean diploid heritability value (over all traits). A simple increase in genetic material should cause significant changes in the amount of genetically controlled phenotypic variance if there is an overall additive relationship between the transcription of alleles and their affect on the phenotype. If dosage compensation does not occur, higher heritabilities in triploids for growth traits that typically show moderate heritability values in diploids (e.g., Gjedrem 1983; Kinghorn 1983) would be consistent with the occurrence of an overall ploidy dependent regulation (i.e., positive gene dosage effects) of gene expression averaged over all growth related loci. While an increase in genetic content may cause higher heritability values in triploids, it must be kept in mind that an inflation of heritability might also occur because of an increase in epistatic interactions (i.e., additive x additive or the interaction between homozygous loci). This is possible because the covariance between half-sibs, which is the most statistically valid component with which to estimate heritabilty (as it contains 1/4 of the phenotypic variance attributable to the inherited action of genes; i.e. the additive genetic variance) also contains 1/16th of the variance attributable to additive by 73 1.25 Ë 0.00 I W-1 I I I I I I I I W-2 W-3 W-4 rgr-c rgr-1 rgr-2 rgr-3 SLR Trait Figure 2.1. Difference between mean diploid heritability (for all traits combined) and individual performance trait heritability estimates of triploids. Differences are marked by a plus sign if the value of h^ripioid > the h^dipioio value and by a minus sign if h^tnpioid < h^dipioid • W-1 to W-4 represent weights, rgr-c, and rgr-1-3 represent relative growth rates and SLR represents serum lysozyme activity response to vaccination. See Table 2.1 for details. additive epistasis (Falconer and Mackay 1996; Roff 1997). This interaction is generally ignored because it is assumed to be very small (Falconer and Mackay 1996; Roff 1997). However, if in triploids there is a higher level of additive-by-additive interaction between loci this might result in inflated heritability estimates especially since the additive effect 74 of a gene will change dependent on the frequencies of its epistatic partners (Cheverud et al. 1999; Wade 2002). Maternal effects are defined as the non-genetic influences of the maternal phenotype, genotype, and environment on the phenotype of the offspring (Mousseau and Fox 1998; Falconer and Mackay 1996). Maternal effects in Chinook salmon are mainly transmitted through prezygotic allocation, since there is negligible maternal care. For example, the mother's nutritional status, hormonal ovarian environment or susceptibility to disease may affect egg quality (e.g.,yolk nutritional value, mRNA or maternal protein content, organelle metabolism), egg size, meiotic status at fertilization and juvenile mortality (Rossiter 1996; Heath and Blouw 1998; Wade 1998). Maternal effects are estimated in a paternal half-sib experimental design by subtracting the among sire variance from the among dam (nested within sire) variance and expressing it as a proportion of the total variance (Falconer and Mackay 1996). However, because the among dam component of variance includes dominance and specific proportions of the variance due to epistatic interactions (i.e., interactions between alleles, as well as interactions between homozygous loci, heterozygous loci and interactions between homozygous and heterozygous loci) and all of the variance due to maternal effects (in addition to 1/4 of the additive genetic variance) (Falconer and Mackay 1996), this estimate will be inflated. Although the among dam variance is therefore not considered useful for estimating the narrow sense heritability, it may reflect the relative importance of specific epistatic interactions and can be used to detect the presence of maternal effects, albeit confounded with dominance and epistatic effects. However, because the epistatic effects included in the among dam component of variance are generally considered to be quite small in diploids, they are usually 75 ignored (Falconer and Mackay 1996; Roff 1997). Such an approach may not be entirely appropriate for triploid offspring, since the relative contribution from dominance and epistatic effects is simply unknown; however, maternal effects are known to be very large for diploid chinook salmon fry and are expected to swamp the non-additive genetic variance components during early development (Heath and Blouw 1998; Heath et al. 1999). Both relative and absolute maternal effects estimates were found to be generally lower for triploids than for diploids for most performance traits (Table 2.5; note the w-2 and rgr-c exceptions). In Figure 2.2, the mean maternal effects value for diploid traits (all traits combined) is subtracted from individual triploid maternal effects values. This Is especially interesting for a number of reasons. First, while the addition of 2 maternal sets of chromosomes would tend to increase the maternal additive component, it might also inflate or magnify the influence of maternal effects and thus potentially have a greater influence on offspring phenotype. This does not seem to have occurred. Second, because triploids have higher among sire values it indicates that dominance, epistasis and maternal effects have less effect on the triploid phenotype. However, if total phenotypic variance is increased in triploids but the combined contribution of maternal effects, dominance and epistasis remains relatively stable then the maternal effects estimate would be expected to be lower. Absolute maternal effects variance values (independent of total variance) in triploids however, exhibited a trend towards lower values (Table 2.5). 76 30.00re 20.00- o E M 10.00- z m 0. 0 0 - 1 - 10. 00- 2 I -20.00Q -30.00-40.00W-1 W-2 W-3 W-4 rgr-c Trait rgr-1 rgr-2 rgr-3 Figure 2.2. Difference between individual performance trait maternal effects estimates of triploids and the mean value of diploid maternal effects (for all traits combined). Differences are marked by a plus sign if the value of h^tripioid > the h^dip lo id value and by a minus sign if h‘ trip lo id h^dipioid - W-1 to W-4 represent weights, rgr-c, and rgr-1-3 represent relative growth rates and SLR represents serum lysozyme activity response to vaccination. See Table 2.1 for details. Growth traits and survival of young diploid salmon are strongly influenced by maternal effects during the egg, larval and early juvenile stages (Kinghorn 1983; Heath and Blouw 1998; Heath, Fox, and Heath 1999; Nagler et al. 2000). However, the magnitude of maternal effects on offspring growth in farm-reared chinook salmon decrease as juveniles develop, becoming negative in a compensatory manner, and then 77 become not significantly different from zero by approximately 150 days post-fertilization (Heath and Blouw 1998; Heath, Fox, and Heath 1999). If the difference between sire and dam variances is mainly a result of dominance and epistatic genetic effects rather than predominantly maternal effects, the results suggest that these influences are less in triploid offspring. This is unexpected since increased structural complexity in triploid genomes is expected to increase the phenotypic effects of dominance and epistasis. However, if the main effects of triploidy on phenotype are additive and magnified due to the increased copy number of nuclear genetic material, the relative effects of dominance and epistatic interactions might not be detected if they are not as drastically influenced by triploidization. Another interesting result of this study was that triploidization success was shown to have a moderate genetic basis (h^± SE = 0.32 ± 0.14). This is the first estimate of heritability reported for this trait and it suggests that there is a genetic component to the cellular response to pressure-shock. This means that a breeding program to improve triploidization success might be successful; however the logistics would be difficult as a proportion of each spawned family would have to be retained as diploids to serve as broodstock. The distribution of variance among causal components in the sib-analysis of triploidization success was unexpected as the majority of variance was attributable to differences among progeny (65%) rather than differences among dams (24%). This was surprising as the eggs from both females bred to the same sire were subjected to the pressure-based triploidization treatment simultaneously. If inter-individual variation within dams was predominantly responsible for whether or not triploidization by pressure was successful then it may mean that differences in the susceptibility of individual eggs to polar body retention within females exist. Whether this is due to 78 differences in intra-female timing of meiosis is unknown. Family differences in triploidization success of salmonids has often been noted and is most likely due to the combination of genetic differences between dams and environmental effects. For example genetic differences in meiotic timing, microtubule structure or egg provisioning might exist or interact with environmental factors such as differences in pre-spawning rearing environments of females, the timing between ovulation and spawning or the consistency of triploidization treatment (Levanduski et al. 1990; Diaz et al. 1993; Teskeredzic et al. 1993; Galbreath and Samples 2000). In summary, triploidization increased phenotypic variation of growth traits both among and within families of chinook salmon. Additionally, the data clearly indicated that the proportion of phenotypic variance attributable to the average additive effects of alleles (additive genetic variance and heritability estimates) increased after triploidization while the relative size of dominance, epistatic and maternal effects probably did not. This pattern of variance distribution indicates that the primary effects of adding an extra set of chromosomes to the salmonid genome are additive and this, in turn is highly suggestive of a predominantly ploidy-dependent mode of gene expression. Dosage effects appear to be present in triploids; and specific gene expression patterns should show a general up-regulation in triploids. In terms of the utility of triploidy for aquacultural purposes, triploidization appears to disrupt the normal inheritance of performance gains made through the selective breeding of diploids. This decreases the utility of triploids because the trend of increased phenotypic variation makes the prediction of performance difficult. It is recommended that individual fish farms weigh the potential ecological and marketing benefits of producing sterile stock with the disadvantages of producing a genetically less reliable animal for market. 79 General Conclusions Triploidization is an abrupt and traumatic genetic perturbation that causes profound changes in genome and cell size. These changes appear to directly affect the phenotype of chinook salmon by increasing the range of performance responses. While this expansion of phenotypic variance does not result in better performance than diploids, the results of this study suggest that performance is not substantially compromised either, at least in terms of growth and immune response. Increased phenotypic variance however, translates into an increase in additive genetic variance indicative of an overall ploidy dependent pattern of gene expression. Triploidization was also accompanied by an apparent decrease in the relative influence of dominance, epistasis and maternal effects, which was unexpected because it was assumed that triploidization would result in a more complex genomic architecture. Despite profound differences between diploids and triploids in the partitioning of phenotypic variance and potentially genetic expression, chinook salmon responded quite well to triploidization especially when pressure rather than heat was used for polar body retention. Growth as well as the specific and non-specific immune responses to vaccination was not substantially different between diploids and triploids. As expected, survival after triploidization was compromised but this occurred mainly during the embryonic and larval stages of development. The obvious dichotomy between high and low performing families regardless of treatment/ploidy status and the existence of significant family components for many of the performance variables (Chapter One) indicate that selective breeding of diploids for increased triploid performance might be possible. However, the presence of family by treatment interactions (although explaining a relatively low amount of variance) observed in the Chapter One study and the increased range of phenotypic variance and 80 profoundly different pattern of variance partitioning found in the Chapter Two study suggest that the effectiveness of a selective breeding of diploid stock to increase triploid performance might be limited by prediction difficulties. These studies have contributed to salmonid research by providing: 1) the first comparative quantitative genetic analysis of a vertebrate triploid model 2) the first heritability estimates for a triploid vertebrate 3) the first estimate of the heritability of triploidization success and, 4) the first comprehensive comparative assessment of triploid chinook salmon performance The results described in this thesis suggest a number of areas for future study. Primarily, this would involve detailed marker based analysis followed by a survey of gene expression patterns using microarray technology. An ongoing study of this nature is currently underway involving a team of researchers from the University of Windsor, UNBC, UBC and DFO, funded by NSERC-CRD. A detailed comparison of diploid and triploid non-specific and specific immune parameters and expression patterns would be especially interesting. Additional research utilizing a more extensive breeding design for quantitative genetic analysis (e.g., factorial, partial diallel or Eisen design), while logistically difficult, would also be useful for separating and estimating the genetic and environmental influence of maternal and epistatic components. 81 Literature Cited Aliah,R.s., Yamaoka.K., Inada.Y., and Taniguchi.N. 1990. Effects of triploidy on tissue structure of some organs of ayu. Nippon Suisan Gakkaishi 56; 569-575. Allen,S.K.Jr. 1983. Flow Cytometry: Assaying Experimental Polyploid Fish and Shellfish. Aquaculture 33; 317-328. Allendorf,F.W. and Leary,R.F. 1984. Heterozygosity in gynogenetic diploids and triploids estimated by gene-centromere recombination rates. Aquaculture 43; 413-420. Allendorf.F.W. and Thorgaard,G.H. 1984. Tetraploidy and the evolution of salmonid Fishes. In Evolutionary Genetics of Fishes. Edited by B.J.Turner. Plenum Press, New York pp. 1-53. Amano,M., Kobayashi,M., Okumoto,N., and Aida,K. 1998. Low GnRH levels in the brain and the pituitary in triploid female sockeye salmon. Fisheries Science 64; 340-341. Arai,K. 2001. Genetic improvement of aquaculture finfish species by chromosome manipulation techniques in Japan. Aquaculture 197; 205-228. Beck,M.L. and Biggers,C.J. 1983. Erythrocyte measurements of diploid and triploid Ctenopharyngodon idella x Hypophthalmichthys nobilis hybrids. Journal of Fish Biology 22; 497-502. Begg,D.A., Salmon,E.D., and Hyatt,H.A. 1983. The changes in organization of actin in the sea urchin egg cortex in response to hydrostatic pressure. The Journal of Cell Biology 97; 1795-1805. Benfey,T.J. 1989. A bibliography of triploid fish, 1943 to 1988. Canadian Technical Report of Fisheries and Aquatic Sciences 1682; 1-33. Benfey,T.J. 1991. The physiology of triploid salmonids in relation to aquaculture. Canadian Technical Report of Fisheries and Aquatic Sciences 1789; 73-80. Benfey,T.J. 1999. The physiology and behavior of Triploid fishes. Review of Fisheries Science 7; 39-67. 82 Benfey,T.J., Dye,H.M., Solar,1.1., and Donaldson,E.M. 1989. The growth and reproductive endocrinology of adult triploid Pacific salmonids. Fish Physiology and Biochemistry 6: 113-120. Benfey,T.J., Solar,l.l., De Jong,G., and Donaldson,E.M. 1986. Flow-cytometric confirmation of aneuploidy in sperm from triploid rainbow trout. Transactions of the American Fisheries Society 115: 838-840. Benfey,T.J., Sutterlin,A.M., and Thompson,R.J. 1984. Use of erythrocyte measurements to identify triploid salmonids. Canadian Journal of Fisheries and Aquatic Science 41: 980-984. Benfey,T.J., Sutterlin,A.M. 1984. Triploidy induced by heat shock and hydrostatic pressure in landlocked Atlantic salmon {Salmo salar L.). Aquaculture 36: 359-367 Berg,O.K., Hendry,A.P., Svendsen,B., Bech,C., Arnekleiv,J.V., and Lohrmann,A. 2001. Maternal provisioning of offspring and the use of those resources during ontogeny: variation within and between Atlantic Salmon families. Functional Ecology 15: 13-23. Birchler,J.A., Hiebert,J.C., and Paigen,K. 1990. Analysis of autosomal dosage compensation involving the alcohol dehydrogenase locus in drosophilamelanogaster. Genetics 124: 677-686. Birchler.J.A., Bhadra,U., Bhadra.M.P., and Auger,D.L. 2001. Dosage-dependent gene regulation in multicellular eukaryotes: Implications for dosage compensation, aneuploid syndromes, and quantitative traits. Developmental Biology 234: 275288. Bird,A. 2002. DNA méthylation patterns and epigenetic memory. Genes & Development 16: 6-21. Blanc,J.M., Poisson,H., and Vallee,F. 2001. Covariation between diploid and triploid progenies from common breeders in rainbow trout, Oncorhynchus mykiss (Walbaum). Aquaculture Research 322: 507-516. Bongers,A.B.J., Sukkel,M., Gort,G., Komen,J., and Richter,C.J.J. 1998. Development and use of genetically uniform strains of common carp in experimental animal research. Laboratory Animals 32: 349-363. 83 Bonnet,s., Haffray.F., Blanc,J.M., Vailee,F., Vauchez,C., Faure,A., and Fauconneau,B. 1999. Genetic variation in growth parameters until commercial size in diploid and triploid freshwater rainbow trout {Oncorhynchus mykiss) and seawater brown trout {Salmo trutta). Aquaculture 173: 359-375. Carrasco,L.A.P., Doroshov,S., Penman,D.J., and Bromage,N. 1998. Long-term, quantitative analysis of gameteogenesis in autotriploid rainbow trout, Oncorhynchus mykiss. Journal of Reproduction and Fertility 113: 197-210. Cheverud,J.M. and Routman,E.J. 1995. Epistasis and its contribution to genetic variance components. Genetics 139: 1455-1461. Cheverud,J.M., Vaughn,T.T., Pletscher,L.S., King-Ellison,K., Bailiff,J., Adams,E., Erickson,0., and Bonislawski,A. 1999. Epistasis and the evolution of additive genetic variance in populations that pass through a bottleneck. Evolution 53: 1009-1018. Chourrout,D. 1986. Techniques of chromosome manipulation in rainbow trout: a new evaluation with karyology. Theoretical and Applied Genetics 72: 627-632. Chourrout,D., Chevassus,B., Krieg,F., Happe,A., Burger,G., and Renard,P. 1986. Production of second generation triploid and tetraploid rainbow trout by mating tetraploid males and diploid females-potential of tetraploid fish. Theoretical and Applied Genetics 72: 193-206. Ciosk,R., Zachariae,W., Michaelis,C., Shevchenko,A., Mann,M., and Nasmyth,K. 1998. An ESP1/PDS1 complex regulates loss of sister chromatid cohesion at the metaphase to anaphase transition in yeast. Cell 93: 1067-1076. Cohen, J. 1992. A Power Primer. Psychological Bulletin 112: 155-159. Cotter,D., 0'Donovan,V., 0'Maoileidigh,N., Rogan,G., Roche,N., and Wilkins,N.P. 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: 6175. Crenshaw,H.C., Allen,J.A., Skeen,V., Harris,A., and Salmon,E.D. 1996. Hydrostatic pressure has different effects on the assembly of tubulin, actin, myosin II, vinculin, talin, vimentin and cytokeratin in mammalian tissue cells. Experimental Cell Research 227: 285-297. 84 Dempster,E.R. and Lerner.I.M. 1950. Heritability of threshold characters. Genetics 35: 212-236. Devlin,R.H. and Nagahama,Y. 2002. Sex determination and sex differentiation in fish: an overview of genetic, physiological, and environmental influences. Aquaculture 208: 191-364. Devlin,R.H., Holm, D.G., Grigliatti, T.A. 1988. The influence of whole-arm trisomy on gene expression in Drosophila. Genetics 118: 87-101. Devlin,R.H., Holm, D.G., Grigliatti, T.A. 1982. Autosomal dosage compensation in Drosophila melanogaster strains trisomie for the left arm of chromosome 2. Proceedings of the National Academy of Sciences of the United States of America. 79:1200-1204. Diaz,N.F., lturra,P., Veloso,A., Estay,F., and Colihueque,N. 1993. Physiological factors affecting triploid production in rainbow trout, Oncorhynchus mykiss. Aquaculture 114. 33-40. Diter,A., Quillet,E., and Chourrout,D. 1993. Suppression of 1st egg mitosis induced by heat shocks in the rainbow-trout. Journal of Fish Biology 42: 777-786. Ellis,A.E. 1993. Lysozyme Assays. In Techniques in Fish Immunology. Edited by J.S.Stolen, T.C.Fletcher, D.P.Anderson, B.S.Robertson, and W.B.van Muiswinkel. SOS Publications, Fair Haven pp. 101-103. Falconer,D.S. and Mackay,T.F.C. 1996. Introduction to Quantitative Genetics. Addison Wesley Longman Limited, Harlow. Fankhauser,G. and Godwin,D. 1948. The cytological mechanism of the triploidy inducing effect of heat on eggs of the newt, Triturus viridescens. Proceedings of the National Academy of Sciences of the United States of America 34: 544-551. Fankhauser,G. and Watson,R.C. 1942. Heat-induced triploidy in the newt, Triturus viridescens. Proceedings of the National Academy of Sciences of the United States of America 28: 436-440. 85 F e lip A , Zanuy.S-, Carrillo,M., and Piferrer.F. 2001. Induction of triploidy and gynogenesis in teleost fish with emphasis on marine species. Genetica 111: 175195. Ferris,S.D. 1984. Tetraploidy and the evolution of the Catostomid fishes. In Evolutionary Genetics of Fishes. Edited by B.J.Turner. Plenum Press, New York pp. 55-93. Ferris,S.D. and Whitt.G.S. 1978. Phylogeny of tetraploid Catostomid fishes based on the loss of duplicate Gene expresssion. Systematic Zoology 27: 189-206. Force,A., Lynch,M., Pickett,F.B., Amores,A., Yan,Y., and Postlethwait,J. 1999. Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151: 1531-1545. Force,A.G. The preservation of duplicate genes by complementary, degenerative mutations and the origin of organismal complexity. Doctor of Philosophy thesis. University of Oregon. Frank,S.A. 1999. Population and quantitative genetics of regulatory networks. Journal of Theoretical Biology 197: 281-294. Friars,G.W., McMillan,I., Quinton,V.M., 0'Flynn,F.M., McGeachy,S.A., and Benfey,T.J. 2001. Family differences in relative growth of diploid and triploid Atlantic salmon {Salmo salar L.). Aquaculture 192: 23-29. Furlong,R.F. and Holland,P.W.H. 2002. Were vertebrates octoploid? Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 357: 531-544. Galbreath,P.P. and Samples,B.L. 2000. Comparison of thermal shock protocols for induction of triploidy in brook trout. North American Journal of Aquaculture 62: 249-259. Galitski,T., Saldanha,A.J., Sytyles,C.A., Lander,E.S., and Fink,G.R. 1999. Ploidy regulation of gene expression. Science 285: 251-254. Garnier-Gere,P.H., Naciri-Graven,Y., Bougrier,S., Magoulas,A., Heral,M., Kotoulas,G., Hawkins,A., and Gerard,A. 2002. Influences of triploidy, parentage and genetic diversity on growth of the Pacific oyster Crassostrea gigas reared in contrasting natural environments. Molecular Ecology 111: 1499-1514. 86 Gibson,T.J. and Spring,J. 2000. Evidence in favour of ancient octaploidy in the vertebrate genome. Biochemical Society Transactions 28; 259-264. Gjedrem,T. 1983. Genetic variation in quantitative traits and selective breeding in fish and shellfish. Aquaculture 33: 51-72. Gregory,T.R. and Hebert,P.D.N. 1999. The modulation of DNA content: proximate causes and ultimate consequences. Genome Research 9: 317-324. Guo,M., Davis,P., and Birchler,J.A. 1996. Dosage effects on gene expression in a maize ploidy series. Genetics 142: 1349-1355. Guoxiong,C., Solar,1.1., Donaldson,E.M. 1989. Comparison of heat and hydrostatic pressure shocks to induce triploidy in steelhead trout {Oncorhynchus mykiss). Canadian Technical Report of Fisheries and Aquatic Sciences 1718. Habicht, C., Seeb, J. E., Gates, R. B., Brock, I. R., and Olito, C. A. Triploid coho salmon outperform diploid and triploid hybrids between coho salmon and chinook salmon during their first year. 1994. Juneau, AK (USA). International Symposium on Genetics of Subarctic Fish and Shellfish. 1993. Hartley,H.O. 1950. The maximum F-ratio as a short-cut test for heterogeneity of variance. Biometrika 37: 308-312. Hartley,H.O. 1940. Testing the homogeneity of a set of variances. Biometrika 31: 249255. Hamaker,H.C. 1978. Approximating the cumulative normal distribution and its inverse. Applied Statistics 27: 76-77. Happe,A., Quillet,E., and Chevassus,B. 1988. Early life history of triploid rainbow trout {Salmo gairdneri Richardson). Aquaculture 71: 107-118. Heath,D.D. and Blouw,D.M. 1998. Are maternal effects in fish adaptive or merely physiological side effects? In Maternal Effects as Adaptations. Edited by T.A.Mousseau and C.W.Fox. Oxford University Press, New York pp. 178-201. Heath,D.D., Fox,C.W., and Heath,J.W. 1999. Maternal effects on offspring size: variation through early development of chinook salmon. Evolution 53: 1605-1611, 87 Hildebrandt.B., Wust.P., Ahlers,0., DieingA-, Sreenivasa,G., Kerner.T-, Felix,R., and Riess.H. 2002. The cellular and molecular basis of hyperthermia. Critical Reviews in Oncology Hematology 43: 33-56. Hill.J.M., Nickerson,A., Sheldon,D.L., Warren,K.A. 1982. Production of triploid chinook salmon Oncorhynchus tshawytscha using flow throough heat shock and subsequent gradual cooling. 33rd Annual Meeting of the EAAP Commission of Animal Genetics, Leningrad, USSR: 1-9 Holloway,S., Glotzer,M., King,R.W., and Muray,A.W. 1993. Anaphase is initiated by proteolysis rather than by the inactivation of maturation-promoting factor. Cell 73: 1393-1402. Hulata,G. 2001. Genetic manipulations in aquaculture: a review of stock improvement by classical and modern technologies. Genetica 111: 155-173. lhssen,P.E., McKay,L.R., McMillan,I., and Phillips,R.B. 1990. Ploidy manipulation and gynogenesis in fishes: cytogenetic and fisheries applications. Transactions of the American Fisheries Society 119: 698-717. lnada,Y., Yoshimura,K., and Taniguchi,N. 1990. Study on the resistance to vibriosis in induced triploid ayu Plecoglossus altivelis. Nippon Suisan Gakkaishi/Bull.Jap.Soc.Sci.Fish. 56: 1587-1591. Johnson,O.W., Rabinovich,P.R., and Utter,F.M. 1984. Comparison of the reliability of a coulter counter with a flow cytometer in determining ploidy levels in Pacific salmon. Aquaculture 43: 99-103. Kinghorn,B.P. 1983. A review of quantitative genetics in fish breeding. Aquaculture 31: 283-304. Kusada,R., Salati.F., Hamaguchi,M., Kawai,K. 1991. The effect of triploidy on phagocytosis, leucocyte migration, antibody and complement levels of ayu, Plecoglossus altivelis. Fish & Shellfish Immunology 1: 243-249. Langston,A,L., Johnstone,R., and Ellis,A.E. 2001. The kinetics of the hypoferraemic response and changes in levels of alternative complement activity in diploid and triploid Atlantic salmon, following injection of lipopolysaccharide. Fish & Shellfish Immunology 11: 333-345. 88 Leary,R.F., Allendorf.F.W., Knudsen,K.L., and Thorgaard.G.H. 1985. Heterozygosity and developmental stability in gynogenetic diploid and triploid rainbow trout. Heredity 54; 219-225. Levanduski,M.J., Beck.J.C., and Seeb,J.E. 1990. Optimal thermal shocks for induced diploid gynogenesis in chinook salmon {Oncorhynchus tshawytscha). Aquaculture 90: 239-250. Lucchesi,J.C. and Rawls,J.M., Jr. 1973. Regulation of gene function a comparison of enzyme activity levels in relation to gene dosage in diploids and triploids of Drosophila melanogaster. Biochemical Genetics 9: 41-51. Lundin.L.G. 1993. Evolution of the vertebrate genome as reflected in paralogous chromosomal regions in man and the house mouse. Genomics 16: 1-19. Lynch,M. and Walsh,B. 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland. Malison,J.A., Procarione,L.S., Held.J.A., Kayes,I.B., and Amundson,C.H. 1993. The influence of triploidy and heat and hydrostatic pressure shocks on the growth and reproductive development of juvenile yellow perch {Perea flavescens). Aquaculture 116: 121-133. McAdam,A.G., Boutin,S., ale,D., and Berteaux.D. 2002. Maternal effects and the potential for evolution in a natural population of animals. Evolution 56: 846-851. Mousseau,T.A. and Fox,C.W. 1998. Maternal Effects as Adaptations. Oxford University Press, New York. Na-Nakorn,U. and Lakhaanantakun,A. 1993. Comparison between the performance of diploid and triploid Clarias macrocephalus. Fish Genetics 52: 79-86. Nagler,J.J., Parsons,J.E., and Cloud,J.G. 2000. Single pair mating indicates maternal effects on embryo survival in rainbow trout, Oncorhynchus mykiss. Aquaculture 184: 177-183. Ohno,S. 1970. Evolution by Gene Duplication. Springer-Verlag, Berlin. Ohno,S., Wolf.U., and Atkin,N.B. 1968. Evolution from fish to mammals by gene duplication. Hereditas 59: 169-187. 89 Ojolick.E.J., Cusack,R., Benfey.T.J., and Kerr,S.R. 1995. Survival and growth of all­ female diploid and triploid rainbow trout {Oncorhynchus mykiss) reared at chronic high temperature. Aquaculture 131: 177-187. Olejnik,S., Algina,J. 2000. Measures of effect size for comparative studies: applications, interpretations, and limitations. Contemporary Educational Psychology 25: 241286. Otto,8.P. and Whitton.J. 2000. Polyploid incidence and evolution. Annual Review of Genetics 34: 401-437. Otto,S.P. and Yong,P. 2002. The evolution of gene duplicates. Advances in Genetics 46: 451-483. Pandian,T.J. and Koteeswaran,R. 1998. Ploidy induction and sex control in fish. Hydrobiologia 384: 167-243. Phillips,R. and Rab,P. 2001. Chromosome evolution in the Salmonidae (Pisces): an update. Biological Reviews 76: 1-25. Piferrer,F., Benfey,T.J., and Donaldson,E.M. 1994. Gonadal morphology of normal and sex-reversed triploid and gynogenetic diploid coho salmon (Oncorhynchuskisutch). Journal of Fish Biology 45: 541-553. Ramsey,J. and Schemske,D.W. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 467-501. Reed,D.H. and Frankham,R. 2001. How closely correlated are molecular and quantitative measures of genetic variation? A meta-analysis. Evolution 55: 10951103. Robertson,A. 1959. Experimental design in the evaluation of genetic parameters. Biometrics 15: 219-226. Roff,D.A. 1997. Evolutionary Quantitative Genetics. Chapman and Hall, New York. Rossiter,M.C. 1996. Incidence and Consequences of Inherited Environmental Effects. Annual Review of Ecology and .Systematics 27: 451-476. 90 Rungruangsak -Torrissen,K., Wergeland,H.l., Glette,J., and Waagbo.R. 1999. Disease resistance and immune parameters in Atlantic salmon {Salmo salar L.) with genetically different trypsin isozymes. Fish & Shellfish Immunology 9: 557-568. Russo,V.E.A., Martienssen,R.A., and Riggs,A.D. 1996. Epigenetic mechanisms of gene regulation. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Schultz. 1979. Role of polyploidy in the evolution of fishes. In Polyploidy: Biological Relevance. Edited by W.H.Lewis. Plenum Press, New York pp. 313-340. Siegal,M.L. and Bergman,A. 2002. Waddington's canalization revisited: Developmental stability and evolution. Proceedings of the National Academy of Sciences of the United States of America 99: 10528-10532. Simon,R.C. 1963. Chromosome morphology and species evolution in the five North American species of Pacific salmon (Oncorhynchus ). Journal of Morphology 112: 77-97. Small,S.A. and Benfey,T.J. 1987. Cell size in triploid salmon. The Journal of Experimental Zoology 241: 339-342. Sokal,R.R. and Rohlf,F.J. 1995. Biometry. W.H. Freeman and Company, New York. Solar,l.l., Donaldson,E.M., and Hunter,G.A. 1984. Induction of triploidy in rainow trout (Salmo gairdneri Richardson) by heat shock , and investigation of early growth. Aquaculture 42 : 57-67. Soltis,D.E. and Soltis,P.S. 1995. The dynamic nature of polyploid genomes. Proceedings of the National Academy of Sciences of the United States of America 92: 8089-8091. Spencer,H.G. 2000. Population genetics and evolution of genomic imprinting. Annual Review of Genetics 34: 457-477. Spring,J. 1997. Vertebrate evolution by interspecific hybridisation-are we polyploid? Federation of European Biochemical Societies Letters 400: 2-8. Stearns,S.C. 2002. Progress on canalization. Proceedings of the National Academy of Sciences of the United States of America 99 : 10229-10230. 91 Suzuki,M.G., Shimada.T., Yokoyama.T., and Kobayashi,M. 1999. The influence of triploidy on gene expression in the silkworm, Bombyx mori. Heredity 82: 661-667. Svobodova,Z., Flajshans.M., Kolarova.J., Modra,H., Svoboda,M., and Vajcova,V. 2001. Leukocyte profiles of diploid and triploid tench, Tinea tinea (L). Aquaculture 198: 159-168. Swarup,H. 1959a. Effect of triploidy on the body size, general organization and cellular structure in Gasterosteus aeuleatus (L). Journal of Genetics 56: 129-142. Swarup,H. 1959b. Production of triploidy in Gasterosteus aeuleatus (L.). Journal of Genetics 56: 129-142. Symington,A., Zimmerman,S., Stein,J., Stein,G., and Zimmerman,A.M. 1991. Hydrostatic pressure influences histone mRNA. Journal of Cell Science 98: 123129. Teplitz,R.L., Joyce,J.E., Doroshov,S.I., and Min,B.H. 1994. A prelimnary ploidy analysis of diploid and triploid salmonids. Canadian Journal of Fisheries and Aquatic Science 51: 38-41. Teskeredzic,E., Donaldson,E.M., Teskeredzic,Z., Solar,l.l., and McLean,E. 1993. Comparison of hydrostatic pressure and thermal shocks to induce triploidy in coho salmon {Oneorhynehus kisuteh). Aquaculture 117: 47-55. Thorgaard,G.H. 1986. Ploidy manipulation and performance. Aquaculture 57: 57-64. Thorgaard,G.H. 1992. Application of genetic technologies to rainbow trout. Aquaculture 100: 85-97. Thorgaard,G.H., Allendorf,F.W., and Knudsen,K.L. 1983. Gene-centromere mapping in rainbow trout: High interference over long map distances. Genetics 103: 771783. Thorgaard,G.H., Arbogast,D.N., Hendricks.J.D., Pereira,C.B., and Bailey,G.S. 1999. Tumor Suppression in Triploid Trout. Aquatic Toxicology 46: 121-126. Wade,M.J. 1998. The evolutionary genetics of maternal effects. In Maternal Effects as Adaptations. Edited by T.A.Mousseau and C.W.Fox. Oxford University Press, New York pp. 5-21. 92 Wade.M.J. 2002. A gene's eye view of epistasis, selection and spéciation. Journal of Evolutionary Biology 15: 337-346. Wagner,G.P., Booth,G., and Bagheri-Chaichian,H. 1997. A population genetic theory of canalization. Evolution 51: 329-347. Wang,S., Hard,J.J., and UTTER,F. 2002. Genetic variation and fitness in salmonids. Conservation Genetics 3: 321-333. Wassenegger,M. 2002a. Gene silencing-based disease resistance. Transgenic Research 11: 639-653. Wassenegger,M. 2002b. Gene silencing. International Review of Cytology - A Survey of Cell Biology, Vol 219 219: 61-113. Whitlock,M.C., Phillips,P.C., and Wade,M.J. 1993. Gene interaction affects the additive genetic variation in subdivided populations with migration and extinction. Evolution 47: 1758-1769. Wilkins,N.P., Cotter,D., and 0'Maoileidigh,N. 2001. Ocean migration and recaptures of tagged, triploid, mixed-sex and all-female Atlantic salmon {Salmo salar L.) released from rivers in Ireland. Genetica 111 : 197-212. Willis,J.H. and Orr,H.A. 1993. Increased heritable variation following population bottlenecks: the role of dominance. Evolution 47: 949-957. Wilson,R.G., Trogadis,J.E., Zimmerman,A.M., and Zimmerman,8. 2001a. Hydrostatic pressure induced changes in the cytoarchitecture of pheochromocytoma (PC-12) cells. Cell Biology International 25: 649-665. Wilson,R.G., Zimmerman,S., and Zimmerman,A.M. 2001b. The effects of hydrostatic pressure-induced changes on the cytoskeleton and on the regulation of gene expression in pheochromocytoma (pc-12) cells. Cell Biology International 25: 667-677. With 1er,R.E., Beacham,T.D., Solar,l.l., and Donaldson,E.M. 1995. Freshwater growth, smolting, and marine survival and growth of diploid and triploid coho salmon {Oncorhynchus kisuteh). Aquaculture 136: 91-107. 93 With 1er,R.E., Clarke,W.C., Blackburn,J., and Baker,I. 1998. Effect of triploidy on growth and survival of pre-smolt and post-smolt coho salmon {Oncorhynchus kisuteh). Aquaculture 168 : 413-422. Wolters,W.R., Chrisman,C.L., and Libey,G.S. 1982. Erythrocyte nuclear measurements of diploid and triploid channel catfish, Ictalurus punctatus (Rafinesque). Journal of Fish Biology 20; 253-258. Yamamoto,A. and lida,T. 1995. Non-specific defense activities of triploid rainbow trout. Fish Pathology 30: 107-110. Zar,J.H. 1996. Biostatistical Analysis. Prentice-Hall, London. 94