The Scottish salmon industry is facing challenges in the control of aquatic infectious disease, as is the case in other countries such as Chile and Norway. Disease outbreaks can have an enormous economic impact and possibly affect wild fish populations. Disease transmission in an aquatic environment is complex and there are several transmission routes (vertical transmission, natural reservoirs, hydrodynamic transmission and long-distance movements). Effective control methods such as vaccines are not available in all cases and therefore disease prevention remains a priority. In livestock, epidemiological network models have been proven to be a highly useful tool to investigate the role of different transmission routes on the course of epidemics and have the potential to provide the same utility for aquatic networks. Understanding the complex contact network will result in more effective disease prevention, surveillance systems and control strategies. The aim of this thesis was to investigate the Scottish live fish movement network and its consequences for pathogen transmission between farms in order to develop and optimize control strategies for epidemics. The main objective of chapter 3 was to investigate the effect of different fallowing strategies on the spread of diseases with different transmission properties. A network model was constructed that included both local transmission and long-distance transmission. The basic structure of this network was a ring model where neighbours within a management area could infect each other and non-local transmission occurred at random. The results showed that when long-distance transmission was under reasonable control in comparison with local transmission risk, synchronized fallowing at the management area level was potentially a highly effective tool in the control of infectious diseases. Chapter 4 presents a detailed description of the number of live fish movements and their timing for Atlantic salmon (Salmo salar) in Scottish aquaculture. For this, movement records from 2002 to 2004 were provided by Marine Scotland, Aberdeen. Salmon are anadromous and have a freshwater (FW) and seawater phase (SW). Scottish live fish movements can be divided in FW-FW, FW-SW, SW-SW, SW-FW and “other” movements. The latter are mainly movements from and to research sites. This study showed that the contact structure and timing of live fish movements are seasonal and differ largely between production phases. Disease control measures should take these differences into account to optimize their strategies. In chapter 4, live fish movements were shown to be seasonal; therefore in chapter 5 the main aim was to quantify the effects of seasonality of live fish movements on the course of epidemics. The results showed that the sequence of salmon movements is important for the course of an epidemic. Seasonality is important when local transmission is higher than 0.05 per contact per week and when the movements are not clustered and when movements do not occur in a specific order based on the specific assumptions made in this model. In conclusion, this thesis described the complex live fish movement structure of salmon in Scotland and showed that biosecurity in SW farms is good but could be further improved if all management areas apply synchronized fallowing. The results of this study suggest that biosecurity between freshwater sites could be improved by the application of a system similar to management areas in SW farms.

Bottlenose dolphins (Tursiops truncatus) kept in zoos and dolphinarias rarely get an outlet for their echolocation abilities as their pool environment is often quite barren. Not much research has been carried out on enrichments promoting echolocation for dolphins in human care. In the present study a setup with live fish was compared to a setup with air-filled floats (providing strong sonar targets, similar to the swim bladders of large fish) and a control setup. A PCL (porpoise click logger) was used to record the echolocation click trains produced by the dolphins and aimed at the three setups. Behavioural data was also collected from video footage. Both the PCL data and all the behavioural observations indicated that the fish setup was more interesting than the float and the control setup, for the dolphins to echolocate towards. However, there were some contradictions with some parameters, where the floats and control seemed to be more interesting. This was probably due to the location of the PCL hydrophone in relation to the floats and fish, and not because the dolphins had a real bigger interest in these setups. To increase the possibility for dolphins to perform more echolocation in human care and increase their welfare, live fish can be recommended as echolocation enrichment.

The diploma thesis is focused on the creation of an economically feasible business plan for the establishment of a company dealing with the breeding and sale of live fish in Slovakia. The theoretical part contains a definition of the issues related to the main idea of the plan and its validation in the form of selected analytical and evaluation tools, the structure and content of individual parts of the business plan. The analytical part is devoted to the validation of the business model and analysis of the market and general environment, including the analysis of own resources and capabilities to identify key factors and evaluate them with subsequent decision-making on the appropriate market strategy. The design part describes the selected strategy of the company and the business model, which is elaborated into partial parts of the business plan.

This thesis investigates the computer vision application of live fish recognition, which is needed in application scenarios where manual annotation is too expensive, when there are too many underwater videos. This system can assist ecological surveillance research, e.g. computing fish population statistics in the open sea. Some pre-processing procedures are employed to improve the recognition accuracy, and then 69 types of features are extracted. These features are a combination of colour, shape and texture properties in different parts of the fish such as tail/head/top/bottom, as well as the whole fish. Then, we present a novel Balance-Guaranteed Optimized Tree with Reject option (BGOTR) for live fish recognition. It improves the normal hierarchical method by arranging more accurate classifications at a higher level and keeping the hierarchical tree balanced. BGOTR is automatically constructed based on inter-class similarities. We apply a Gaussian Mixture Model (GMM) and Bayes rule as a reject option after the hierarchical classification to evaluate the posterior probability of being a certain species to filter less confident decisions. This novel classification-rejection method cleans up decisions and rejects unknown classes. After constructing the tree architecture, a novel trajectory voting method is used to eliminate accumulated errors during hierarchical classification and, therefore, achieves better performance. The proposed BGOTR-based hierarchical classification method is applied to recognize the 15 major species of 24150 manually labelled fish images and to detect new species in an unrestricted natural environment recorded by underwater cameras in south Taiwan sea. It achieves significant improvements compared to the state-of-the-art techniques. Furthermore, the sequence of feature selection and constructing a multi-class SVM is investigated. We propose that an Individual Feature Selection (IFS) procedure can be directly exploited to the binary One-versus-One SVMs before assembling the full multiclass SVM. The IFS method selects different subsets of features for each Oneversus- One SVM inside the multiclass classifier so that each vote is optimized to discriminate the two specific classes. The proposed IFS method is tested on four different datasets comparing the performance and time cost. Experimental results demonstrate significant improvements compared to the normal Multiclass Feature Selection (MFS) method on all datasets.