Labrus bergylta (ballan wrasse) have recently emerged as a key resource to aquaculture through proven efficacy in controlling infestations of sea lice (Leclercq et al., 2014a). However, due to complex ecology, and a complete lack of sexual dimorphism gender identification endures as a key restriction to optimising broodstock management therefore male selection and establishing optimal sex ratios is difficult (Talbot et al., 2012). L. bergylta, are noted to demonstrate unusually coloured plasma ranging in hue from green to blue with the haem catabolite biliverdin established as the causal pigment in the majority of cases (Abolins, 1961). As most vertebrates excrete biliverdin, or rapidly metabolise it to prevent toxicity, accumulation to such excess is a phenomenon which merits attention. Notably, correlation between plasma biliverdin and gender has been reported in some Labridae. Although patterns vary between species, the abundance or characteristics were such that sexual identity could be established (Gagnon, 2006). Pigment analysis was therefore proposed as a potential sex-marker in L. bergylta. In the initial experimental phase (Chapter 3), the ultimate aim was to isolate and identify the blue pigment from L. bergylta plasma, and to develop a method of quantification. The initial phase confirmed the target pigment was biliverdin IXα by visible spectroscopy, TLC, HPLC, MSMS, and a series of reactions. Following this, a protocol was developed (Chapter 2) to quantify the pigment. This method was applied accross plasma sampled from four geographically distinct wild populations with established biometrics including age, mass, length, gender and external phenotype. Subsequent analysis revealed that although pigment abundance did not vary relative to ontogeny, and there was no difference in concentration between the binary genders, plasma biliverdin was depleted in individuals undergoing sex change. Although this conclusion was complicated by significant biliverdin variation relative to origin and phenotype, which were interrelated based on relative distributions across populations, further analysis of plasma pigment in related species identified that biliverdin accumulation was associated with protogynous species. Considering the anti-oxidant capacity of biliverdin and other potentially relevant functions, this was indicative of association with the tissue remodelling processes which accompany inversion. During Chapter 3 it was noted that the biliverdin appeared tightly bound to a protein moiety. Based on the hypothesis that the pigment was actively managed and accumulated in L. bergylta plasma by this association, the next phase of experiments (Chapter 4) was an exploration of biliverdin and its binding protein in L. bergylta. The experiments revealed plasma biliverdin comigrated with the protein such that it was depleted from solution at the same rate indicating that all of the pigment was associated. Subsequent electrophoretic experiments using the fractionation products supported this, and UV fluourescence identified fragments of interest in the 25-28 kDa region. To confirm observations from the previous cross species comparison, the study was similarly expanded to include other Labrini. This revealed that although the 25 kDa band was common to all species, and genders, the 28 kDa band was collocated with the protogynous, and as such hyperbiliverdinaemic species. The 28 kDa band was sequenced using MSMS, and was identified as similar to the lipocalin Apolipoprotein A1. In combination with the properties of biliverdin, and considering that ApoA1 is analogous to serum albumin in many telesots, this supported the chromoprotein association as the main mechanism of biliverdin accumulation in such species. Further to the proposed function of biliverdin with inversion processes, and considering relevant literature, the active properties of ApoA1 suggested additional associations with prolonged altered states of metabolism which considering the ecology of L. bergylta would include gender transition, overwintering torpor and prolonged micronutrient limitation, all of which occur simultaneously. Other potential roles include modulating inflammatory responses, inhibiting pathogenic incursions and acting as an external point of contact innate immune response. From this, it was concluded that the data fully supported the previous assertions of biliverdins relevance in protogynous species, and identified a number of properties which could be of great interest to the industry in terms of welfare. The final experimental phase (Chapter 5) had two main aims. The first was to establish whether protogynous inversion could be artificially induced in L. bergylta as a means of generating male fish, and whether size had any effect on the process. The second was then to utilise controlled induction for tracking biliverdin mobilisation across the process to test the previous hypothesis. The preliminary trial demonstrated that both androgen inhibition and non-aromatisable testosterone could stimulate inversion in female L. bergylta. From this, the second trial then determined that although there was a dose dependant effect in that high androgen dosages appeared to compress the inversion process, relative size was not a factor. Gonad histology was used to create a unified scale of protogynous transition which could be expressed as a gradient to structure the biliverdin analysis. Although the biliverdin data demonstrated cryptic trends at the higher resolution gender scales, when the endpoint was condensed back to the binary gender scale employed previously (Chapter 3), the prior assertion of depletion during transition, and therefore the association with sex change associated tissue remodelling was supported. Ultimately this thesis revealed links between the biliverdin macromolecule and the highly unusual metabolic and physiological demands of gender transition in sequentially protogynous hermaphroditic temperate wrasse species.

The ectoparasitic monogenean fluke, Gyrodactylus salaris, is a parasite known to be highly pathogenic to Atlantic salmon (Salmo salar). Although present in the environment of several neighbouring European countries, the UK is thought to be G. salaris-free, but, if national contingency plans to control this parasite are to be effective, it is vital that we understand the factors underlying its transmission from host to host. This study demonstrates that the majority of parasites transferring to new hosts are mature parasites that have reproduced at least once. Since, exploration and host transfer strategies pose a risk to survival; the parasite will endeavour to pass on its genes before attempting to transfer from one host to another. This study has also shown that when pregnant parasites are forced to leave their hosts, their offspring are aborted prematurely to ensure the survival of the mature parasite. Gyrodactylids do not possess a free-swimming stage in their life cycle, which allows for their migration between hosts. In spite of this, they are able to rapidly colonise naïve hosts, even in non-shoaling populations of fish. This study investigates the transmission strategies employed by detached parasites in the colonisation of new hosts. Observations of gyrodactylids collected from 3-spine sticklebacks, Gasterosteus acuelatus, suggest that their activity increases as a stickleback approaches, alerting the host to its presence. The parasite is then ingested directly by the prospective host. A time series of experimental exposures and specimens prepared for Scanning Electron Microscopy (SEM) suggest that once ingested, the parasites attach to the lining of the buccal cavity and then migrate out to their preferred colonisation site on the outer surface of the fish. It is proposed that this may be an alternative route for host infection. Similarly, direct ingestion by the scavenging on infected hosts by 3-spine sticklebacks suggests another route of infection of new hosts. Although these routes of transmission may be of lesser significance, infections in the buccal cavity may be an important indicator for detection of infection and those personnel involved in screening fish for gyrodactylids should be aware that this is an area in which infections can occur. This study also demonstrated that the use of the anaesthetic 2-phenoxyethanol does not affect the number of gyrodactylids which leave the host to colonise a new host. Additionally, observations of the transmission process suggest that turbulence produced by the movement of the fish’s fins may facilitate the transfer of detached parasites from the substrate. While this hypothesis appears to be supported by video evidence and photographic stills gathered throughout the duration of this study, further work should be conducted using particle tracking techniques to determine the efficacy of using a vortex effect as a means of colonising new hosts. Field sampling processes may have an effect on this type of research, giving rise to problems with the accurate diagnosis, management and control of gyrodactylids in a variety of fish. Gyrodactylus infected specimens of 3-spine stickleback (Gasterosteus aculeatus L.), minnows (Phoxinus phoxinus L.) and stone loach (Barbatula barbatula L.) from one Scottish river were cohabited. The study found that small numbers of Gyrodactylus do transfer to atypical hosts. This study highlights that personnel involved in fish disease surveillance programmes should be aware of the consequences of transporting multiple species in the same transport vessel as gyrodactylids may infect species previously thought to be resistant. Equally, diagnosticians should be aware of the fact that atypical species may act as temporary hosts and that their gyrodactylid fauna should not be assumed. Non-feeding life-cycle stages, such as the dispersal stages of parasites, are dependant for survival upon finite energy reserves gathered during feeding phases. Thus, those individuals with more limited reserves will die sooner and consequently have less time available to find a new host once detached. At this stage, the principal energy reserves in gyrodactylids are stored as large lipids droplets. Confocal laser scanning microscopy (CLSM) has been used to investigate the distribution of lipid droplets in Gyrodactylus, which have migrated off their fish host, testing the hypothesis that these droplets function as a proxy for the nutritional state. This study, demonstrated that the lipid droplets were particularly associated with the gut and that there is a significant variability in the volume of stored lipid carried out by each individual. Transmission Electron Microscopy (TEM) showed that gyrodactylids carry lipid droplets at all stages of their life cycle, including at release from the birth pore. It is likely that transferring worms require stored energy reserves to survive in the event of failure to establish contact with a new host. These reserves could allow the parasite to survive without a host for several days. As gyrodactylids appear to respond to a range of stimuli including vibration and chemicals released from the host, the presence or absence of such cues may have consequences on the rates of Gyrodactylus transmission. If these chemical stimuli can be identified and then mimicked or blocked, then this may offer potential opportunities for the control of gyrodactylid behaviour and for disrupting their transmission to new hosts. Baseline gyrodactylid behaviour, in the absence of a host, was determined under white light and infrared. This was achieved using a specially constructed arena and purpose written image analysis software to analyse parasite movement under different lighting conditions. The study found that gyrodactylids were more active in the dark than in light conditions, typically displaying longer, more sinuous tracks under red light than under white light. To begin investigating the effect of chemical presence on gyrodactylid behaviour, the activity of octopaminergic agonists and antagonist which bind to muscle receptors and alter muscle activity, were assessed. The impact of octopamine hydrochloride, clonidine hydrochloride, amitraz and, a toxic reference, chlordimeform, over a range of concentrations (0.2 to 3.2µM/L) were assessed on gyrodactylid behaviour. All of the four chemicals affected Gyrodactylus and produced muscle tetanus, causing muscle spasms when extension was attempted. Prolonged exposure resulted in death. Only the highest concentration of chlordimeform, the toxic reference, affected 100% of Gyrodactylus after 24 hours. After 48 hours, all of the Gyrodactylus treated with chlordimeform were either affected, moribund or dead. Amitraz was more toxic than chlordimeform with 80% of Gyrodactylus being dead after 24 hours at the highest concentration. After 48 hours 100% of Gyrodactylus exposed to 3.2 µm/L amitraz were dead, and up to 80% were dead in those exposed to lower concentrations; with no parasites being left unaffected. Although these particular compounds are toxic to fish, the effect of these agonistic chemicals on Gyrodactylus behaviour and survival is interesting and suggests that a closely related compound that is safe for use against fish may offer a potential treatment for the control of G. salaris infections in rivers. An ultrastructure study was undertaken to contribute to the current understanding of gyrodactylid ultrastructure. The findings of this research require broad understanding of gyrodactylid behaviour for their interpretation. Photographic evidence was gathered using transmission and electron microscopy. From these results, it is clear that Gyrodactylus gasterostei on a three-spine stickleback host will respond to a range of stimuli (i.e. vibration or chemical cues released from the host) in their assessment of host suitability.

The wide use of nanoscale materials in food and health care products raises the concern of their possible uptake across the gastrointestinal tract, but very limited data are available on their uptake kinetics, and the potential hazards for humans. In this study, the uptake mechanism of titanium dioxide (TiO2) across the isolated perfused fish intestine and human intestinal Caco-2 cells were evaluated. The in vitro preparation of the whole gut sac and the isolated perfused intestine of rainbow trout were performed using both bulk and nano TiO2 in a concentration of 1 mg l-1 for up to 4 h. The results showed that the Ti from both bulk and TiO2 NPs were mainly accumulated in the mid and hind intestine, with 80% or more of the accumulation in the mucosa rather than the underlying muscularis. Perfused intestines showed a saturable, time-dependent accumulation of the Ti from TiO2 and the uptake of Ti from exposure to NPs was faster than that of the bulk form. The uptake of Ti from exposure to TiO2 NPs increases 10 fold when the CO2 in the gas mixture was lowered to 0.5%. Subsequently, further investigation on the mechanisms of uptake of TiO2 was applied using different kinds of inhibitors. Adding 10 mmol l-1 cyanide did not stop Ti uptake from TiO2 exposures, and 100 µmol l-1 vanadate (ATPase inhibitor) caused a 2.8 fold reduction in the net uptake rate of Ti for the TiO2 NP exposure. Luminal additions of 120 IU ml-1 nystatin (endocytosis inhibitor) blocked the uptake of Ti from both bulk and TiO2 NPs treatments. The results indicate that Ti accumulation from TiO2 exposures was sensitive to both nystatin and vanadate; the former suggesting that there is an endocytosis involvement in the uptake of TiO2 across the intestinal epithelium. Human intestinal Caco-2 cell showed a steady, saturable and time-dependent accumulation of Ti over 24 h exposures to 1 mg l-1 TiO2 (for all forms of TiO2). A scanning electron microscope study indicated the appearance of the particles underneath the cells, increasing the evidence of the Ti uptake from different forms of TiO2 by Caco-2 cells. Both nystatin and vanadate increase the accumulation of TiO2 which suggests interference of these drugs with endocytic pathways. All the data in the thesis demonstrates Ti uptake across the intestinal epithelium from TiO2 exposures involving CO2-dependent and nystatin-sensitive mechanisms. The results in this thesis have contributed to some understanding on the behaviour, uptake and effects of the TiO2 NPs across the intestine; and highlight the possible dietary hazard of the NPs to human health.

Typescript (Photocopy) / Includes 3 published papers by the author in back of volume 2 / Bibliography: leaves 510-541 of vol. 1 / 2 v. : ill ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--Dept. of Zoology, University of Adelaide, 1990

Marked shifts in the life history traits of fish have been reported in many exploited fish stocks, with a particular trend towards decreasing size and age at maturity. Though other environmental and behavioural factors have been implicated, the key driver of these changes links to fishing pressure, through both the direct selective effects of fishing itself, and indirectly through the manipulation of important biological and environmental factors. Although reproduction itself has been well described in teleosts, the mechanisms of environmental and endogenous entrainment of maturation remain unclear and it was the principal aim of this thesis to improve current understanding of these systems in gadoids. Photoperiod has been identified as the strongest environmental cue for entraining seasonal behaviours, including seasonal reproduction. Over the last decade, several of the key drivers involved in the photoneuroendocrine cascade have been elucidated in mammals and birds, with the Eya3 pathway merging as an important mechanism for entraining maturation. However, little is yet known of their influence on maturation in fish. In the first study, the photoperiodic regulation of the Eya3-Tshβ-Dio2 cascade was analysed in Atlantic cod exposed to either continuous light (reproductive inhibition) or simulated natural photoperiod (reproductive stimulation) from July to December. Monthly expression was measured through QPCR, demonstrating a strong activation of pituitary Eya3 under declining photoperiod. As this coincided with the onset of secondary gametogenesis, these results suggest that Eya3 may play a stimulatory role in the photoneuroendocrine cascade of Atlantic cod. Although photoperiod represents the most reliable and noise free proximate signal to entrain the reproductive process, it is clear that a minimum growth and energetic state must be reached for maturation to progress. This directed the second line of study – a series of diet restriction trials on haddock and cod designed to investigate the influence of naturally occurring lipid levels in the diet on growth and reproduction, define the “critical window” in which fish assess their energetic state and how this is analysed before commencing secondary gametogenesis, and to assess whether Eya3 is regulated by the growth axis in cod. The results of these experiments indicate that overall size around the autumn equinox is the most accurate indicator of maturation commitment in cod. Additionally, Eya3 expression was elevated in maturing fish indicating a role for this marker in linking the energetic signal with entrainment of the reproductive axis. Finally, an analysis of the physiological and genetic stock structure of cod from the North Sea IVa stock region and westward into VIa was performed. The results of this analysis support previous genetic studies, indicating further fine-scale structuring of these stocks, reflective of the structure indicate by the differences in maturation strategies of the component populations. The results identified both environmental and harvest related pressures which may be driving the current stock structure. The results of these studies greatly improves our understanding of the key drivers and mechanisms regulating maturation in cod, highlighting new avenues for future research.

Rainbow trout represents the most prominent species in freshwater farming in UK aquaculture. One of the common diseases constraining rainbow trout production and increasingly causing problems in Atlantic salmon (Salmo salar L.) hatcheries worldwide is rainbow trout fry syndrome (RTFS) or bacterial cold water disease (BCWD). During the last 20 years, the development of a commercial vaccine against RTFS has been hindered by the prevalence of a wide range of the fish pathogen F. psychrophilum, thus the current treatment of choice is the use of antibiotics. Studies involved in understanding the innate and adaptive immune response of vaccinated rainbow trout fry using inactivated whole cell are still lacking. Therefore, the aim of this thesis is to characterise the strain diversity and antibiotic susceptibility of UK F. psychrophilum isolates, evaluate the efficacy of a whole-cell formalin-killed polyvalent vaccine, which was developed based on the characterisation results of this study, and investigate the immune response in trout fry following the immersion vaccination via the changes in expression of relevant immune genes. A total of 315 F. psychrophilum isolates, 293 of which were collected within the UK, were characterised using four genotyping methods and a serotyping scheme. A high strain diversity was identified among the isolates with 54 pulsotypes, ten (GTG)5-PCR types, two 16S rRNA allele lineages, seven plasmid profiles and three serotypes. The predominant profile observed within the F. psychrophilum isolates examined was PFGE cluster II – (GTG)5-PCR type r1 – 16S rRNA lineage II – serotype Th (n= 70/156, 45%). The characterisation results not only revealed the wide distribution within the UK and the persistence within a site of predominant pulsotypes, but also the presence of unique genotypes in certain sites or countries. Co-existence of genetically and serologically heterogeneous isolates within each farm was detected, highlighting the reasons this disease is so difficult to control, especially by vaccination. The occurrence over time of F. psychrophilum pulsotypes within a site could provide important epidemiological data for farm management and the development of site-specific vaccines. The antimicrobial susceptibilities of 140 F. psychrophilum strains, 125 of which were from the UK, were evaluated by the broth microdilution (MIC) and disc diffusion methods. There was evidence of reduced susceptibilities to three of the main antimicrobials used in UK aquaculture. Broth microdilution testing showed that only 12% of 118 UK isolates tested were WT to oxolinic acid (MIC COWT 0.25 mg L-1), 42% were WT for oxytetracycline (MIC COWT 0.25 mg L-1), and 66% were WT for amoxicillin. In contrast, all the isolates tested were WT (MIC COWT 2 mg L-1) for florfenicol, the antimicrobial of choice for RTFS control in the UK. Despite the imprecision of disc diffusion-based COWT values due to high standard deviations, there was a high categorical agreement between the classification of the strains (into WT or NWT) by MIC and disc diffusion methods for florfenicol (100%), oxolinic acid (99%), amoxicillin (97%) and oxytetracycline (94%). In general, this study showed that the UK F. psychrophilum isolates examined remain susceptible to florfenicol and also stresses the importance of performing susceptibility testing using standardised methods and COWT values. Several statistically significant associations between genotypes and the reduced susceptibilities of F. psychrophilum strains were revealed. A whole-cell formalin killed polyvalent vaccine against RTFS/BCWD was developed by combining three genetically and serologically divergent strains, recently collected from UK farms. The efficacy of this polyvalent vaccine was evaluated after immersion vaccination in 5 g trout and bath challenge using hydrogen peroxide as a pre-stressor with a virulent heterologous isolate of F. psychrophilum strain. Significant protection was achieved with an RPS of 84%. The combination of exposure to hydrogen peroxide prior to bath challenge may be an alternative to an injection challenge with 12 g trout, although further standardisation and optimisation of the challenge model is required. Changes in the innate immune response of trout fry following the initial vaccination included the up-regulation of the interleukin 1 β (IL-1β) gene in head kidney at 4 h and the up-regulation of toll-like receptor-2 (TLR-2) in skin at day 2. While the expression levels of C3 was unchanged, the down regulation of CD8-α in head kidney and spleen and CD4-1 in spleen were documented. IgM and IgT transcripts were found to be up-regulated in hind-gut two days post-vaccination. Understanding the strain diversity and the antibiotic susceptibility of UK F. psychrophilum isolates could help improve the control strategies, such as preventing the spreading of pathogenic F. psychrophilum clones between fish farms, reducing the use of antibiotics in RTFS/BCWD treatment and monitoring the development of acquired antibiotic resistance mechanisms. Moreover, strain characterisation data of UK F. psychrophilum species has assisted in selecting suitable candidates for developing an effective RTFS vaccine.

Freshwater sport fisheries contribute substantially to the economies of England and Wales. However, many trout fisheries rely partly or entirely on stocking farmed trout to maintain catches within freshwater fisheries. Farmed trout often differ genetically from their wild counterparts and wild trout could be at risk of reduced fitness due to interbreeding or competition with farmed fish. Therefore, to protect remaining wild brown trout (Salmo trutta L) populations and as a conservation measure, stocking policy has changed. Legislation introduced by the Environment Agency (EA, 2009) will now only give consent to stocking of rivers and some stillwaters with sterile, all-female triploid brown trout. There are reliable triploidy induction protocols for some other commercially important salmonid species however; there is limited knowledge on triploid induction in brown trout. Previously, triploid brown trout have been produced by heat shocks although reduced survivals were obtained suggesting that an optimised heat shock had not been identified, or that heat shock gives less consistent success than hydrostatic pressure shock (HP), which is now recognised as a more reliable technique to produce triploid fish. Thus the overall aim of this thesis was to conduct novel research to support the aquaculture and freshwater fisheries sector within the United Kingdom by optimising the production and furthering the knowledge of triploid brown trout. Firstly, this PhD project investigated an optimised triploidy induction protocol using hydrostatic pressure (Chapter 2). In order to produce an optimised hydrostatic pressure induction protocol three experiments were conducted to (1) determine the optimal timing of HP shock application post-fertilisation, (2) define optimal pressure intensity and duration of the HP shock and (3) study the effect of temperature (6-12 °C) on triploid yields. Results indicated high survival to yolk sac absorption stage (69.2 - 93.6 %) and high triploid yields (82.5 - 100 %) from the range of treatments applied. Furthermore, no significant differences in triploid rates were shown when shock timings and durations were adjusted according to the temperature used. In all treatments deformity prevalence remained low during incubation (<1.8 %) up to yolk sac absorption (~550 degree days post hatch). Overall, this study indicated that the optimised pressure shock for the induction of triploidy in brown trout delivering high survival and 100 % triploid rate (a prerequisite to brown trout restocking) is a shock with a magnitude of 689 Bar applied at 300 Centigrade Temperature Minutes (CTM) for 50 CTM duration. Regarding the assessment of triploid status, the second experimental chapter tested the accuracy and efficacy of three ploidy verification techniques (Chapter 3). Techniques studied were erythrocyte nuclei measurements (Image analysis), flow cytometry (Becton Dickinson Facscalibur flow cytometer) and DNA profiling (22 polymorphic microsatellite loci) to assess the effectiveness of triploidy induction in brown trout. Results indicated the validity of using erythrocyte indices major nuclear axis measurements, flow cytometric DNA distributions expressed as relative fluorescence (FL2-Area), and polymorphic microsatellite loci (Ssa410UOS, SSa197, Str2 and SsaD48) for assessing ploidy status in brown trout. Accuracy of each technique was assessed and indicated that all techniques correctly identified ploidy level indicating 100 % triploid rate for that commercial batch of brown trout. These techniques may be utilised within aquaculture and freshwater fisheries to ensure compliance with the legislation introduced by the EA. As a result of the legislation introduced by the Environment Agency triploid brown trout will freely interact with diploid trout therefore there is a need to assess feeding response and behavioural differences between diploid and triploid trout prior to release. Therefore, in the third experimental chapter (Chapter 4) diploid and triploid brown trout were acclimated for six weeks on two feeding regimes (floating/sinking pellet). Thereafter, aggression and surface feeding response was compared between pairs of all diploid, diploid and triploid and all triploid brown trout in a semi natural stream (flume). In each pairwise matching, fish of similar size were placed in allopatry and rank determined by the total number of aggressive interactions initiated. Dominant individuals initiated more aggression than subordinates, spent more time defending a territory and positioned themselves closer to the food source (Gammarus pulex) whereas subordinates occupied the peripheries. When ploidy was considered, diploid trout were more aggressive than triploid, and dominated their siblings when placed in pairwise matchings. However, surface feeding did not differ statistically between ploidy irrespective of feeding regime. Triploids adopted a sneak feeding strategy while diploids expended more time defending a territory. In addition, an assessment of whether triploids exhibited a similar social dominance to diploids when placed in allopatry was conducted. Although aggression was lower in triploid pairs than in the diploid/triploid pairs, a dominance hierarchy was observed between individuals of the same ploidy. Dominant triploid fish were more aggressive and consumed more feed items than subordinate individuals. Subordinate fish displayed a darker colour index than dominant fish suggesting increased stress levels. However, dominant triploid fish seemed more tolerant of subordinate individuals and did not display the same degree of invasive aggression as observed in the diploid/diploid or diploid/triploid matchings. These novel findings suggest that sterile triploid brown trout feed similarly but are less aggressive than diploid trout and therefore may provide freshwater fishery managers an alternative to stocking diploid brown trout. In addition to research at the applied level in triploid brown trout, this thesis also examined the fundamental physiological effects of ploidy in response to temperature regime. Triploid salmonids have been shown to differ in their tolerance to environmental temperature. Therefore the fourth experimental chapter (Chapter 5) investigated whether temperature tolerance affected feed intake and exercise recovery. Diploid and triploid brown trout were exposed to an incremental temperature challenge (10 and 19 °C) and subsequent survival and feed intake rates were monitored. Triploids took longer to acclimate to the increase in temperature however feed intake were significantly greater in triploids at high temperature. In a follow on study, we investigated post-exercise recovery processes under each temperature regime (10 and 19 °C). Exhaustion was induced by 10 minutes of forced swimming, with subsequent haematological responses measured to determine the magnitude of recovery from exercise. Plasma parameters (alkaline phosphatase, aspartate aminotransferase, calcium, cholesterol, triglycerides, phosphorous, total protein, lactate, glucose, pH, magnesium, osmolality, potassium, sodium, chloride, lactate dehydrogenase) were measured for each ploidy. Basal samples were taken prior to exercise and then at: 1; 4, and 24 hours post-exercise. Contrary to previous studies, there was no triploid mortality during or after the exercise at either temperature. Although diploid and triploid brown trout responded metabolically to the exercise, the magnitude of the response was affected by ploidy and temperature. In particular, triploids had higher levels of plasma lactate, osmolality, and lower pH than diploids at 1 hour post exhaustive exercise. By 4 hours post-exercise plasma parameters analysed had returned to near basal levels. It was evident that the magnitude of the physiological disturbance post-exercise was greater in triploids than diploids at 19 °C. This may have implications where catch and release is practiced on freshwater fisheries. Overall, this work aimed to develop and/or refine current industry induction and assessment protocols while better understand the behaviour and physiology of diploid and triploid brown trout. The knowledge gained from this work provides aquaculture and freshwater fisheries with an optimised protocol, which delivers 100 % triploid rates and profitability without compromising farmed trout welfare, thus ultimately leading towards a more sustainable brown trout industry within the United Kingdom.

Commercial capture-based aquaculture of the Atlantic bluefin tuna (ABT), Thunnus thynnus (L.), has been prominent in the Mediterranean for over a decade. Owing to several limitations encountered in working with the species, including its high commercial value, there has been little research carried out relating to this species. The objective of this study was to examine several health parameters of captive ABT. These included an examination of coronary artery lesions, haematology, plasma biochemistry, assessment of immune function and changes in fatty acid (FA) flesh content through the on-growing period. Arteriosclerosis in fish is a pathologic condition of uncertain etiology and involves the main coronary artery in teleosts. Apart from reports of their widespread occurrence in salmonids, they have been described from a restricted number of wild ABT specimens but have not received further attention. This investigation analysed the effect of size and period of net-pen rearing on the prevalence and severity of arteriosclerotic lesions in ABT. Coronary arteries from wild and captive fish were investigated and prevalence was 100 %, but increasing structural degradation was observed with increasing fish size, suggesting that lesions progress throughout the life of the fish. Due to the limited availability of wild specimens, the effect of captivity on arteriosclerosis in ABT could not be adequately quantified, although observations suggest that the farming process has no major effect on arteriosclerotic lesions in ABT. Studies on the haematology, plasma biochemistry and immunology of ABT are limited. Haematological and plasma biochemical indices are useful in animal health assessment but use of these requires the establishment of species-specific ranges. Blood was collected from captive ABT specimens of varying weight (61-361 kg) and the major haematological (n = 45), plasma biochemical (n = 30) and immunological parameters (n = 45) were quantified. Size-based differences were found in haematological indices between experimental sub-groups including increased erythrocyte number and haemoglobin level in smaller ABT. No differences were found in immunological parameters except for total IgM levels, which were higher in the smaller individuals. Preliminary investigations indicated that disease prevalence in captive ABT is very low. Epidermal mucus is an important interface between fish and their environment and comprises immunological components which act as a first barrier against pathogen entry or colonisation. Mucus was collected from captive ABT and analysed for innate immune components. The presence of IgM was detected in the mucus of ABT by an enzyme-linked immunosorbent assay and several different enzymes were detected with an API-ZYM kit assay. Zymography experiments confirmed the presence of protease-like enzymes in the mucus, while enzyme assays quantified alkaline phosphatase, protease, esterase and cathepsin B activities. Lysozyme levels were high. The mucus agglutinated sheep erythrocytes but did not demonstrate complement or bacteriolytic activity. There is restricted information on the fatty acid composition of farmed ABT or how this is influenced when the fish are held under commercial aquaculture conditions. This study investigated the FA composition of farmed ABT, its variation by dorsal muscle region and the correlation between dietary FA composition with that of the fish. Analysis of flesh samples retrieved from farmed ABT did not reveal significant differences in the FA composition of experimental sub-groups irrespective of size, time held in captivity or diet. These results indicate that FA metabolism in ABT is substrate-selective. Gene expression measurements from several organs of ABT showed that expression of Δfad5 and elovl5, genes involved in FA metabolism, were highest in the brain followed by the liver but no expression of these genes was detected in the spleen. The findings of this research address aspects of health evaluation and nutritional status in farmed ABT and are discussed in terms of farming practice. Conclusions from some of these studies suggest that the practice of holding wild-captured stock in cages for periods of up to 18 months does not result in significant impact on ABT.

Flesh colour and fillet fat percentage are the two most important attributes to salmon fillet quality. A medium genetic component to body lipid percentage within commercial lines has previously been shown (h2 = 0.17-0.24). A low level of heritability (h2 = 0.16) has also been reported for flesh colour in Atlantic salmon. To investigate whether this genetic component includes loci of major effect, a genome-wide QTL scan was performed with commercially bred Atlantic salmon (Landcatch Natural Selection). Five large full-sib families (10 parents with 153 offspring) were genotyped using microsatellite markers. To utilize the large difference between sire and dam recombination rate, a two-stage genotyping was employed. Initially, the parents and offspring were genotyped for two microsatellite markers per linkage group, and sire based QTL analysis was used to detect linkage groups with significant effects on those flesh quality traits. A linear-regression based interval as analytical method was applied for QTL detection. The results revealed evidence of QTLs affecting percentage fat percentage and flesh colour on linkage groups LNS16 and LNS1 respectively. To confirm the QTL and to provide an improved estimate of position, a dam-based analysis was then employed. One major QTL was located on the genome-wide significance level for percentage fat percentage. Microsatellite marker Ssa0016NVH (at position of 1.3 cM) was found to be tightly linked to QTL affecting percentage fat percentage. In addition, a QTL affecting flesh colour was found to be flanked by microsatellite markers Ssa9.44NUIG at position of 68.7 cM and Ssa0021NVH at position of 50.6 on linkage group LNS16. The evidence for suggestive QTL affecting flesh colour on linkage group LNS1 was also revealed. In order to increase marker density within these and other linkage groups, AFLP markers were employed, 24 primer combinations resulted in a total of 489 polymorphic fragments. Among 11 fragments that were found to be linked to the microsatellite markers on linkage group LNS16, four fragments (AAG-CAC328, AGG-CAG447, AGG-CTA237 and AGG-CTC237) were tightly linked to microsatellite marker Ssa9.44NUIG, but none were found to be linked to microsatellite Ssa0021NVH. Moreover, none of the AFLP markers were found to be linked to microsatellites residing on linkage group LNS1. Using a constructed map of microsatellite and AFLP markers for linkage group LNS16, the dam based analysis revealed a significant QTL for flesh colour at the location of 189 cM, while the sire based analysis detected a significant QTL for fat percentage at the location of 80 cM. Considering the dominant nature and clustering character of AFLP markers, it was concluded that a certain primer combination in AFLP markers could be of limited use for fine mapping and QTL detection in Atlantic salmon.

The aim of this study was to investigate the basis to male mating decisions in two related species of bitterling: Rhodeus ocellatus and R. amarus. Bitterling have a resource-based mating system; females lay eggs in the gills of live freshwater mussels and males fertilize the eggs by releasing sperm into the inhalant syphon of the mussel. Male bitterling perform courtship behaviour and aggressively defend mussels in a territory from which they exclude other males. Using laboratory and field experiments it was shown that male aggressive behaviour is inherited through additive maternal genes. Male aggression is also influenced by the number of conspecific males encountered in competition for a mussel, and by the degree of clustering of mussels. Limited availability of mussels results in stronger selection on traits males use in mating context: hence they are more aggressive, larger and more colourful. The differences in mating behaviours in different environments may indicate a conflict between male dominance and female choice, but have not led to reproductive isolation. Resource availability during ontogenesis and male density during embryogenesis, however, do not exert an effect on male aggressive behaviour. Red carotenoid-based nuptial coloration functions as an inter- and intra-sexual signal and undergoes rapid variation in response to changes in mating context. Male bitterling do not modulate their courtship and aggressive behaviour in response to variation in female size, and their choice of mussel species is influenced by, and consistent with, female oviposition choice.