Abstract
Τhe decline of wild fish stock and the increasing demand for food and fish protein have accelerated the development of fish farming. The intense fish production in limited space (e.g. sea cages) has potentially deteriorated the aquatic environments (nutrient accumulation from uneaten feed and excreta cause eutrophication, pollution, chemical compounds, disease outbreaks, escapes etc.) gaining the scientific and societal attention. The environmental impact of aquaculture in combination with the competition for land use (leisure activities, aesthetic impact, tourism, residential development etc.) and the desire for a profitable business have contributed to a shift to land-based aquaculture systems. Recirculating aquaculture systems (RAS) are closed-containment systems where fish are farmed in reused water and provide a stable annual fish production. Due to the controlled farming environment, where water quality control systems and waste treatment technologies are installed, the fish quality and growth are improved while the risk of fish escapees and miscellaneous incidents are minimized. However, because of the recirculation (up to 99%) and the increased stocking density, waste derived from fish feed and fish excreta is accumulated. The control of organic matter in a RAS is of high importance for the good management of the facility. High organic matter loading in the water deteriorates the water quality, favouring the microbial blooming which might lead to disease outbreak directly affecting the fish. Additionally, increased levels of organic matter reduce the efficiency of the various water treatment processes. Several technologies have been installed to remove waste, and consequently organic matter, from the water stream of a closed system. However, waste removal processes are not fully optimized and the interactions between the waste treatment units are not well understood. Ozone has been implemented in several water treatment applications as a secondary treatment step to improve the water quality by oxidizing the organic matter and miscellaneous dissolved compounds in the water, having also bactericidal properties ensuring simultaneously disinfection. However, ozone is also toxic for aquatic organisms in extremely low concentrations (0.01 mg O3/L). Therefore, the risk of losing fish because of overdosing inhibits the full implementation of ozone in aquaculture industry. The aims of this PhD thesis were to design an ozonation system for RAS, based on the ozone demand of the specific system, where the water quality is improved without compromising the fish welfare, and then, to be able to monitor the ozone within the system with a novel, accurate and real-time measuring method relying on the fluorescence sensitivity towards ozone. Aquatic dissolved organic matter has numerous fluorescent properties and it is highly reactive with ozone. Water samples were collected from several aquatic facilities around Denmark and then were analysed with fluorescence spectroscopy to determine the fluorescent character. RAS samples were subjected to ozonation to investigate the responsiveness towards ozone. Similar fluorescence components were present in all studied RASs, with different reactivity suggesting that a florescence based sensor could be used as an indirect ozone dosage determination tool in water, since the fluorescence intensities and dissolved organic matter degradation by ozone were well correlated. Furthermore, potential applications of such a sensor were proposed. To design an optimal ozonation system, the water from the specific facility should be analysed in terms of ozone demand and ozone lifetime. Each facility is unique since the process water content of solutes, the operational conditions, and the water treatment units, the fish species and stocking density vary greatly, resulting in different water matrix. Therefore, the ozone reactions in this particular water should be investigated prior to installation, to ensure that the amount of ozone required to improve water quality is sufficient and it will be degraded long before it reaches the biofilters or the culture tanks. Having experimentally determined the ozone demand and kinetics, the predicted ozone dosages were applied in pilot-scale RASs to verify the effect of ozone on water quality of continuous operated freshwater RASs. Several water parameters were investigated including non-volatile organic matter, chemical oxygen demand, biological oxygen demand, ammonia, nitrate and nitrite levels, particles number and size, and microbial activity. Fluorescence organic matter was analysed by fluorescence Excitation Emission Matrix (EEM) spectroscopy coupled with Parallel Factor analysis (PARAFAC) for a more accurate identification of the organic matter. The overall water quality was significantly improved upon ozonation, proportionally to ozone dosage applied, suggesting that the predicted ozone dosages matched with the needs of the water. During ozonation no fish mortality was observed. The water matrix has huge influence in the design of an ozonation system. The determination of ozone in seawater samples is more complicated than in freshwater. In seawater, where bromide is naturally present, brome-oxides are formed in presence of ozone, which are toxic to fish. Thus, it is vital to be able to determine the critical point between the ozone dosage required to improve water quality and avoiding the formation of brominated by-products. Attempts to measure fast ozone in brominated water were made setting the basis for a modified analytical method. Further investigations are needed to increase the accuracy and to verify the breakpoint between optimal ozonation and brominated by-product formation inhibition in seawater RAS. In conclusion, this PhD study elucidated that when ozone is properly implemented in a RAS, having taken into consideration the ozone demand and the lifetime of ozone for the system of interest, the water quality of a RAS will be remarkably improved. To determine the ozone demand and lifetime, fluorescence spectroscopy was used, since it was found to be a good indicator of organic matter accumulation in RAS and highly sensitive towards ozone treatment. The detailed analysis of the fluorescence dissolved organic matter contained in RAS water, revealed four independently varying fractions with different reactivity and responsiveness to ozone, suggesting that a fluorescence based sensor targeting a specific wavelength transition could be used to determine indirectly the ozone concentration in water.