The objective of the work described in the current thesis was to provide a better understanding of some of the key factors associated with sea louse, Lepeophtheirus salmonis, infection of farmed Atlantic salmon. In Chapter 2, initial work focused on establishing the vertical and horizontal distribution of sea lice copepodids and spatial patterns of on-farm infections. The louse distribution was investigated along the main current gradient across adjacent salmon production pens at three commercial sites. A depth profile for the distribution of larval lice was also established for the top 12 m of the water column at three different locations in close proximity to commercial salmon farms. Within all multi-pen sites there were clear patterns of distribution and infection along the main water current gradient with the abundance of lice in end pens at each site appearing to be different from the central pens. The vertical distribution pattern of free swimming L. salmonis larvae (nauplii and copepodids) showed that the surface 6 m harboured 85.5 ± 1.6 % of the lice present in the water body analysed (0 – 12 m depth), irrespective of sampling date and location. In Chapter 3, further environmental effects / influences on attachment success of the copepodids were analysed using controlled infection challenges. A flume with adjustable flow rates, and controlled light conditions was designed for this study. Flume current velocity was observed to be a significant factor in infection success, with higher infection levels observed at lower current velocities, while higher current velocities were demonstrated to reduce settlement success. At fixed velocity, higher copepodid exposure levels gave rise to higher infection levels, this having a linear relationship suggestive of a lack of competitive effects for space on the fish. Light was also shown to play an important role in host settlement. A positive correlation between increasing light intensity and higher louse attachment success was found for all tested light spectra / wavelengths (white - Halogen, blue 455 nm, green 530 nm and red 640 nm). Observation of an infecting cohort of copepodids showed maximal infection at four days post-moulting with a tail-off of infection by six days post-moulting. However, even under the optimal conditions represented by a flume challenge, including linear water flow, the constraint of copepodids to pass close to the salmon host and the very high exposure levels of copepodids per fish, louse attachment success was still relatively low. Chapter 4 examined implementation of a possible management approach based upon some of the environmental influences observed. This chapter described a study in which environmental manipulation of salmon swimming depth was employed on-site in an attempt to reduce farm infection of Atlantic salmon. The effects of submerged artificial lighting in combination with submerged feeding were tested with respect to salmon swimming depth and sea lice infection, following the hypothesis that L. salmonis infection in a commercial salmon population could be reduced through exposure to deep lighting and feeding. The results of the study suggest that swimming depth manipulation can indeed be used at a commercial scale to reduce salmon lice burdens on Atlantic salmon by physically minimising spatial interactions between the two animals. In the final research chapter (Chapter 5), this thesis examines the question of whether ploidy of the host impacts on sea louse infection levels and whether susceptibility of individual fish is consistent between replicate infections. Results showed that triploid salmon are not subject to higher sea louse infection levels under experimental challenge and farm infection conditions compared to diploid hosts. In addition, triploid fish subject to initial infection, did not become more or less resistant to infection compared to diploids when comparing repeated sea louse infections. In summary, this thesis describes work conducted to analyse key infection pathways and factors influencing infection of Atlantic salmon by sea lice and suggestions made as to how findings may be exploited to reduce louse burdens in Atlantic salmon farming. The practical solutions presented to exploit the results found in this work are currently under consideration by the Scottish salmon industry.

Water samples were collected at four sites on the White River, Muncie, Indiana, on a biweekly basis from April 1980 to April 1981, to determine seasonal variation in A. hydrophila density. In conjunction with water sampling, temperature, dissolved oxygen, conductivity and pH were determined at each site. During this period a yearly mean of 589 colony forming units (CFU) per ml was obtained with a range of 0-6350. Peaks in mean A. hydrophila numbers occurred during the summer (1808 CFU ml-1) and early spring (3946 CFU ml-1) with lows occurring in winter (42 CFU ml-1). No significant correlations could be found between CFU values and the physico-chemical parameters studied over the calendar year. However, the compiling of data from May 1979 to April 1981 did yield a significant correlation between A. hydrophila density values and the following water quality parameters: temperature, conductivity, and dissolved oxygen.Fish were collected during June through October 1980 at various sites along the White River, Delaware County, Indiana, for the determination of past or present exposure to A. hydrophila infection by serological and direct isolation methods. Of 102 different fish sera tested, 36% (37/102) had a detectable titer to A. hydrophila (isolate #113). However, the direct isolation method only yielded seventeen percent (8/47) of the fishes sampled to be positive forA. hydrophila systemically. Agglutinins to A. hydrophila were demonstrated in eleven different fish species, whereas five different species revealed no detectable antibody to the specific particulate antigen employed. Estimated CFU ml-1 of A. hydrophila was the only parameter which correlated with the presence of agglutinating antibody to this organism in river fish populations.Ball State UniversityMuncie, IN 47306

Proliferative Kidney Disease (PKD) caused by the Malacosporean parasite Tetracapsuloides bryosalmonae, is presently the most economically damaging disease of British rainbow trout farming, costing the industry in excess of £2.5 million per annum in the UK alone. With no vaccine or prophylactic treatment available, and only management techniques currently adopted to minimise the stress and mortality associated with the disease, alternative approaches must now be considered. This document investigates if selective breeding for PKD resistance is possible by assessing the level of additive genetic variation, and calculating the subsequent estimates of heritability, for commercial strains of rainbow trout. During a PKD outbreak on a commercial farm, 1500 communally reared juvenile rainbow trout from two strains (Houghton Spring and Isle of Man) were sampled on a single day, their body weight and fork length measured, and severity of kidney swelling scored according to the scale of Clifton-Hadley et al. (1987). Fish were assigned to individual families using microsatellite parentage assignment. Significant additive genetic variation was observed in the population, and families were ranked according to estimated breeding values. A combined estimate of heritability (h2 = 0.19 ± 0.08) for kidney score suggests the population will respond well to selective breeding for kidney score, which may be deemed a measure of resistance, whilst the favourable genetic correlations between kidney score and the production traits measured suggest simultaneous selection for kidney score and growth traits should also be effective. In order to support the findings of the initial research, controlled challenge experiments were conducted. Using the family EBV information on kidney score from the IoM strain (due to its certification as a disease-free site), four females, two with high and two with low response to PKD, were each crossed with a randomly selected neomale to produce twenty two families for PKD challenge experiments. The PKD experimental challenges showed evidence of additive genetic variation to kidney score over an eleven week period, supporting initial findings. A low score was deemed as evidence of greater resistance to the parasite in this study. Although female EBV was taken into consideration in the statistical model, there was found to be no significant difference in resistance according to family. Immunohistochemistry stained kidney sections from each individual involved in the challenges proved kidney score correlated significantly to the number of parasites in the kidney, suggesting that the scale of Clifton-Hadley et al. (1987) is a sufficient and accurate basis on which to describe the severity of PKD, and infection level in rainbow trout. Having discovered evidence that furunculosis, causative agent Aeromonas salmonicida, plays a major role in the mortality of fish suffering from PKD in the field, the bacterial disease was investigated to assess the resistance of the same families used in the PKD challenges. Twenty one of the families were used to discover that additive genetic variation for resistance to furunculosis is apparent when assessed as both a binary and longitudinal trait, suggesting significant genetic improvement can be made to increase resistance to furunculosis in the IoM stock. No significant correlation was observed between kidney score, EBV, and resistance to this bacterium, but there was a positive phenotypic correlation found between furunculosis resistance and size, suggesting simultaneous selection for performance and resistance is possible within this population.

Of the 430+ extant species of Gyrodactylus, ectoparasitic monogenetic flukes of aquatic vertebrates, Gyrodactylus salaris Malmberg, 1957 is arguably the most well-known. Following the introduction of this species into Norway in the 1970s with consignments of infected Atlantic salmon smolts, Salmo salar L., this species has had a devastating impact on the Norwegian Atlantic salmon population, decimating wild stocks in over 40 rivers. Gyrodactylus salaris is the only OIE (Office International des Epizooties) listed parasitic pathogen of fish and has been reported from 19 countries across Europe, though many of these records require confirmation. The UK, Ireland and some selected watersheds in Finland are currently recognised as G. salaris-free states; however, the threat that this notifiable parasite poses to the salmon industry in the UK and Ireland is of national concern. Current British contingency plans are based on the assumption that if G. salaris were to be introduced, the parasite would follow similar dynamics to those on salmonid stocks from across Scandinavia, i.e. that Atlantic strains of Atlantic salmon would be highly susceptible to infection, with mortalities resulting; that brown trout, Salmo trutta fario L., would be resistant and would lose their infection in a relatively short period of time; and that grayling, Thymallus thymallus (L.), would also be resistant to infection, but would carry parasites, at a low level, for up to 143 days. Two of the objectives of this study were to confirm the current distribution of G. salaris across Europe, and then, to investigate the relative susceptibility of British salmonids to G. salaris, to determine whether they would follow a similar pattern of infection to their Scandinavian counterparts or whether, given their isolation since the last glaciation and potential genetic differences, they would exhibit different responses. It has been almost six years since the distribution of G. salaris across Europe was last evaluated. Some of the European states identified as being G. salaris-positive, however, are ascribed this status based on misidentifications, on partial data resulting from either morphological or molecular tests, or according to records that have not been revisited. Additional Gyrodactylus material from selected salmonids was obtained from several countries to contribute to current understanding regarding the distribution of G. salaris across Europe. From the work conducted in the study, G. salaris is reported from Italy for the first time, alongside three other species, and appears to occur extensively throughout the central region without causing significant mortalities to their rainbow trout, Oncorhynchus mykiss (Walbaum), hosts. The analysis of archive material from G. salaris-positive farms would suggest that G. salaris has been in the country for at least 12 years. Material obtained from rainbow trout from Finland and Germany was confirmed as G. salaris supporting existing data for these countries. No specimens of G. salaris, however, were found in the additional Gyrodactylus material obtained from Portuguese and Spanish rainbow trout, only Gyrodactylus teuchis Lautraite, Blanc, Thiery, Daniel et Vigneulle, 1999, a morphologically similar species was found. Gyrodactylus salaris is now reported from 23 out of ~50 recognised states throughout Europe, only 17 of these however, have been confirmed by either morphology or by an appropriate molecular test, and only ten of these records have been confirmed by a combination of both methods. To assess the susceptibility of English and Welsh salmonids to G. salaris, a number of salmonid stocks of wild origin, were flown to the Norwegian Veterinary Institute (NVI) in Oslo, where they were experimentally challenged with G. salaris. Atlantic salmon from the Welsh River Dee, S. trutta fario from the English River Tyne and T. thymallus from the English River Nidd, raised from wild stock in government hatcheries, were flown out and subsequently challenged with G. salaris haplotype A. After acclimation, each fish was infected with ~50–70 G. salaris and marked, so that parasite numbers on individual fish could be followed. The dynamics on individual fish were followed against a control (Lierelva Atlantic salmon). The experiment found that the number of G. salaris on S. salar from the River Dee continued to rise exponentially to a mean intensity (m.i.) of ~3851 G. salaris fish-1 (day 40 post-infection). These salmon were highly susceptible, more so than the Norwegian salmon control (m.i. ~1989 G. salaris fish-1 d40 post-infection) and were unable to regulate parasite numbers. The S. trutta fario and T. thymallus populations, although initially susceptible, were able to control and reduce parasite burdens after 12 (m.i. ~146 G. salaris fish-1) and 19 (m.i. ~253 G. salaris fish-1) days, respectively when peak infections were seen. Although the latter two hosts were able to limit their G. salaris numbers, both hosts carried infections for up to 110 days (i.e. when the experiment was terminated). The ability of S. trutta fario and T. thymallus to carry an infection for long periods increases the window of exposure and the potential transfer of G. salaris to other susceptible hosts. The potential role that brown trout may play in the transmission and spread of G. salaris in the event of an outbreak, needs to be considered carefully, as well as the interpretation of the term “resistant” which is commonly used when referring to brown trout’s susceptibility to G. salaris. The current British surveillance programmes for G. salaris are focused on the screening of Atlantic salmon and on the monitoring of the rainbow trout movements. The findings from this study demonstrate that G. salaris can persist on brown trout for long periods, and suggest that brown trout sites which overlap with Atlantic salmon or rainbow trout sites are also included within surveillance programmes and that the role that brown trout could play in disseminating infections needs to be factored into contingency/management plans. Throughout the course of the study, a number of parasite samples were sent to the Aquatic Parasitology Laboratory at Stirling for evaluation. Some of these samples represented Gyrodactylus material that were associated with fish mortalities, but the species of Gyrodactylus responsible appeared to be new to science. A further aspect of this study was, therefore, to investigate these Gyrodactylus related mortalities in aquaculture stock and to describe the species found in each case, which may represent emerging pathogens. The two new species, Gyrodactylus orecchiae Paladini, Cable, Fioravanti, Faria, Di Cave et Shinn, 2009 and Gyrodactylus longipes Paladini, Hansen, Fioravanti et Shinn, 2011 on farmed gilthead seabream, Sparus aurata L., were collected from several Mediterranean farms. The finding of G. orecchiae in Albania and Croatia was associated with 2–10% mortality of juvenile stock and represents the first species of Gyrodactylus to be formally described from S. aurata. Subsequently, G. longipes was found in Bosnia-Herzegovina and Italy, and at the Italian farm site, it occurred as a mixed infection with G. orecchiae, but these infections did not appear to result in any loss of stock. Unconfirmed farm reports from this latter site, however, suggest that a 5–10% mortality of juvenile S. aurata was also caused by an infection of Gyrodactylus, which is suspected to be G. longipes. Additional samples of Gyrodactylus from a gilthead seabream farm located in the north of France have been morphologically identified as G. longipes, extending the geographical distribution of this potentially pathogenic species to three countries and three different coasts. In addition to these samples, some specimens of Gyrodactylus from a Mexican population of rainbow trout were sent for evaluation.

Acute hepatopancreatic necrosis disease (AHPND) is an emerging bacterial infection in shrimp that has been widespread across the major world shrimp producing countries since 2009. AHPND epizootics have resulted in a huge loss of global shrimp production, similar to that caused by white spot disease in the 1990’s. The epizootiological understanding of the spread of AHPND is still in its early stages, however, and most of the currently published research findings are based on experimental studies that may struggle to capture the potential for disease transmission at the country scale. The main aim of this research, therefore, is to develop epizootiological tools to study AHPND transmission between shrimp farming sites. Some tools used in this research have already been applied to shrimp epizoology, but others are used here for the first time to evaluate the spread of shrimp diseases. According to an epizootiological survey of AHPND in Thailand (Chapter 3), the first case of AHPND in the country was in eastern shrimp farms in January 2012. The disease was then transmitted to the south in December 2012. The results obtained from interviews, undertaken with 143 sample farms were stratified by three farm-scales (large, medium and small) and two locations (east and south). Both the southern location and large-scale farming were associated with a delay in AHPND onset compared with the eastern location and small- and medium-scale farming. The 24 risk factors (mostly related to farming management practices) for AHPND were investigated in a cross-sectional study (Chapter 3). This allowed the development of an AHPND decision tree for defining cases (diseased farms) and controls (non-diseased farms) because at the time of the study AHPND was a disease of unknown etiology. Results of univariate and unconditional logistic regression models indicated that two farming management practices related to the onset of AHPND. First, the absence of pond harrowing before shrimp stocking increased the risk of AHPND occurrence with an odds ratio () of 3.9 (95 % CI 1.3–12.6; P‑value = 0.01), whereas earthen ponds decreased the risk of AHPND with an of 0.25 (95 % CI 0.06–0.8; P‑value = 0.02). These findings imply that good farming management practices, such as pond-bottom harrowing, which are a common practice of shrimp farming in earthen ponds, may contribute to overcoming AHPND infection at farm level. For the purposes of disease surveillance and control, the structure of the live shrimp movement network within Thailand (LSMN) was modelled, which demonstrated the high potential for site-to-site disease spread (Chapter 4). Real network data was recorded over a 13-month period from March 2013 to March 2014 by the Thailand Department of Fisheries. After data validation, c. 74 400 repeated connections between 13 801 shrimp farming sites were retained. 77 % of the total connections were inter-province movements; the remaining connections were intra-province movements (23 %). The results demonstrated that the LSMN had properties that both aided and hindered disease spread (Chapter 4). For hindering transmission, the correlation between and degrees was weakly positive, i.e. it suggests that sites with a high risk of catching disease posed a low risk for transmitting the disease (assuming solely network spread), and the LSMN showed disassortative mixing, i.e. a low preference for connections joining sites with high degree linked to connections with high degree. However, there were low values for mean shortest path length and clustering. The latter characteristics tend to be associated with the potential for disease epidemics. Moreover, the LSMN displayed the power-law in both and degree distributions with the exponents 2.87 and 2.17, respectively. The presence of power-law distributions indicates that most sites in the LSMN have a small number of connections, while a few sites have large numbers of connections. These findings not only contribute to a better understanding of disease spread between sites, therefore, but also reveal the importance of targeted disease surveillance and control, due to the detection of scale-free properties in the LSMN. Chapter 5, therefore, examined the effectiveness of targeted disease surveillance and control in respect to reducing the potential size of epizootics in the LSMN. The study untilised network approaches to identify high-risk connections, whose removal from the network could reduce epizootics. Five disease-control algorithms were developed for the comparison: four of these algorithms were based on centrality measures to represent targeted approaches, with a non-targeted approach as a control. With the targeted approaches, technically admissible centrality measures were considered: the betweenness (the number of shortest paths that go through connections in a network), connection weight (the frequency of repeated connections between a site pair), eigenvector (considering the degree centralities of all neighbouring sites connected to a specified site), and subnet-crossing (prioritising connections that links two different subnetworks). The results showed that the estimated epizootic sizes were smaller when an optimal targeted approach was applied, compared with the random targeting of high-risk connections. This optimal targeted approach can be used to prioritise targets in the context of establishing disease surveillance and control programmes. With complex modes of disease transmission (i.e. long-distance transmission like via live shrimp movement, and local transmission), an compartmental, individual-based epizootic model was constructed for AHPND (Chapter 6). The modelling uncovered the seasonality of AHPND epizootics in Thailand, which were found likely to occur between April and August (during the hot and rainy seasons of Thailand). Based on two movement types, intra-province movements were a small proportion of connections, and they alone could cause a small AHPND epizootic. The main pathway for AHPND spread is therefore long-distance transmission and regulators need to increase the efficacy of testing for diseases in farmed shrimp before movements and improve the conduct of routine monitoring for diseases. The implementation of these biosecurity practices was modelled by changing the values of the long-distance transmission rate. The model demonstrated that high levels of biosecurity on live shrimp movements (1) led to a decrease in the potential size of epizootics in Thai shrimp farming. Moreover, the potential size of epizootics was also decreased when AHPND spread was modelled with a decreased value for the local transmission rate. Hence, not only did the model predict AHPND epizootic dynamics stochastically, but it also assessed biosecurity enhancement, allowing the design of effective prevention programmes. In brief, this thesis develops tools for the systematic epizootiological study of AHPND transmission in Thai shrimp farming and demonstrates that: (1) at farm level, current Thai shrimp farming should enhance biosecurity systems even in larger businesses, (2) at country level, targeted disease control strategies are required to establish disease surveillance and control measures. Although the epizootiological tools used here mainly evaluate the spread of AHPND in shrimp farming sites, they could be adapted to other infectious diseases or other farming sectors, such as the current spread of tilapia lake virus in Nile tilapia farms.

The increasing incidence of infections due to antibiotic resistant bacteria has led to renewed interest in bacteriophages (= phages) and phage therapy. Although phage therapy has been applied to control bacterial diseases in plants, poultry, livestock and humans, its application in aquaculture is still relatively limited. The emergence of phage-resistant bacterial mutants has been considered to be one of the major limitations of phage therapy. This study aimed to (i) isolate and characterise phages; (ii) select phages and their bacterial hosts to set up in vivo phage therapy models with aquaculture animals, and estimate the efficiency of phage therapy; (iii) investigate the generation and characteristics of phage-resistant mutants, and thus estimate the consequence of applying phage therapy when phage-resistant mutants emerge; and (iv) discuss the prospects for application of phages in aquaculture. Two Vibrio isolates and their phages were isolated from a Scottish marine fish farm. Based on the results of conventional phenotype testing and 16S rRNA gene sequencing analysis, the two vibrios, V9 and V13, were identified as Vibrio splendidus and Vibrio cyclitrophicus, respectively. The bacterial characteristics including morphology, temperature and salinity range of growth, production of extracellular enzymes, and the possession of virulence genes were examined. According to the morphological characteristics observed using transmission electron microscopy by negative staining, phage PVS9 of V. splendidus V9 was identified as a myophage, while phage PVC13 of V. cyclitrophicus V13 was identified as a siphophage. The phages could only lyse one bacterial host strain and their genomic DNA was double stranded with a size of ~46 kb. The two Vibrio isolates were found to be non- or of low virulence to rainbow trout, goldsinny wrasse and Artemia in pathogenicity experiments. Thus an in vivo phage therapy model could not be set up using these Vibrio isolates and their phages. Two phages pAS-3 and pAS-6 were isolated using the Aeromonas salmonicida subsp. salmonicida Hooke strain as the host. Phages pAS-3 and pAS-6 had a similar genome size of ~50 kb, and the same relatively narrow host range within A. salmonicida subsp. salmonicida strains. The siphophage pAS-3 formed clear plaques and inhibited A. salmonicida Hooke growth in vitro completely for at least 18 hours when using MOI = 1,000, whereas the podophage pAS-6 formed turbid plaques and weakly inhibited Hooke growth. Rainbow trout exposed by intraperitoneal injection with 0.1 mL of the raw phage preparations at a concentration of 108 PUF mL-1 showed no adverse effects over 14 days. In the phage therapy trial, fish were firstly injected with 1 x 102 CFU fish-1 of A. salmonicida Hooke, then immediately injected with phage preparations of pAS-3 and pAS-6, respectively, using MOI = 10,000. Compared with the control group (which did not receive phage treatment), phage treated groups showed a delay in the time to death, and lower mortalities. However, the mortalities and time to death between phage treated and non-treated groups were not significantly different. Phage-resistant mutants of pathogenic A. salmonicida strain Hooke were induced by repeatedly challenging with phage pAS-3. One of the mutants, termed HM, was chosen to compare the characteristics with the parental wild-type strain Hooke. Test results including the formation of ‘smooth’ colonies on TSA, autoagglutination negative, the formation of creamy colonies on Coomassie Brilliant Blue agar, and the degradation of a thick/furry layered structure on the cell surface indicated a deficiency of the A-layer in the phage-resistant mutant HM. Therefore, it was deduced that the A-layer either directly acted as the receptor of A. salmonicida phage pAS-3, or was affected indirectly by the change of an unknown phage receptor. The greater wax moth larvae model was used to compare the virulence of the phage-resistant mutant HM and the parental wild-type strain Hooke, as it is an ethically acceptable animal model, which has the advantages of being low cost and convenient for injection, and is also a recognised alternative model for bacterial pathogens of fish. The results showed that virulence of the phage-resistant mutant HM did not decline in the greater wax moth larvae model compared with that of the parental wild-type strain Hooke. In conclusion, different approaches were used to isolate and characterise phages from different aquaculture environments for potential use in phage therapy. A rainbow trout model was set up using pathogenic A. salmonicida strain Hooke and two A. salmonicida phages pAS-3 and pAS-6. The use of phage treatment led to lower cumulative mortalities and delay to the time of death, although the differences between the groups were not significant, futher work is required to determine if these phages have potential in phage therapy. The consequence of applying phage therapy when phage-resistant mutants emerge was estimated based on their characteristics and virulence, and no decline in virulence of the phage-resistant mutant from this study indicates the importance of fully testing the virulence of phage-resistant mutants before carrying out large scale field trials of phage therapy. It appears feasible to use phage therapy as an alternative approach to control bacterial infections in aquaculture, but further studies are required to focus on improving effectiveness, and also to overcome the concrete limitations and hurdles in application and commercialisation. Moreover, a broader range of applications of phages in aquaculture should be explored.

M.Sc. (Zoology) / Please refer to full text to view abstract

No description available.

Gyrodactylus salaris Malmberg, 1957, is a notifiable freshwater ecto-parasite that infects both wild and farmed populations of Atlantic salmon (Salmo salar, L.). It has caused catastrophic damage to wild salmon stocks in Norway since its accidental introduction in 1975, reducing salmon density in some rivers by 98% over a period of five years. It is estimated that G. salaris has cost the Norwegian salmon industry more than 500 million EUR. Currently the UK has G. salaris free status under EU law, however, it is believed that if G. salaris emerged in the UK the impact would be similar to that witnessed in Norway. The aim of this thesis is to develop mathematical models that describe the salmon-G. salaris system in order to gain a greater understanding of the possible long-term impact the parasite may have on wild populations of Atlantic salmon in G. salaris-free territories such as the UK. Mathematical models, including deterministic, Leslie matrix and individual based models, were used to investigate the impact of G. salaris on Atlantic salmon at the individual and population level. It is known that the Atlantic strain of Atlantic salmon, examples of which occur naturally in Norway and the UK, does not have any resistance to G. salaris infections and the parasite population is able to quickly grow to epidemic levels. In contrast, the Baltic strain of Atlantic salmon, examples of which occur naturally in Sweden and Russia, exhibits some form of resistance and the parasite is unable to persist. Thus, baseline models were extended to include immunity to infection, a trade-off on salmon reproductive rate, and finally, to consider interactions between populations of G. salaris and multiple strains of salmon exhibiting varying levels of immunity from fully susceptible to resistant. The models proposed predict that in the absence of host resistance or an immune response infections by G. salaris will result in an epidemic followed by the extinction of the salmon host population. Models also predict that if salmon are able to increase their resistance to G. salaris infections through mutations, salmon population recovery after the epidemic is indeed possible within 10-15 years post introduction with low level parasite coexistence. Finally, models also highlight areas where additional information is needed in order to improve predictions and enable the estimation of important parameter values. Model predictions will ultimately be used to assist in future contingency planning against G. salaris outbreaks in the UK and possibly as a basis for future models describing other fish/ecto-parasite systems.

M.Sc. (Zoology) / Please refer to full text to view abstract