Modelling nutrient dynamics and waste valorisation in a fish-biofloc system

In this PhD project tools to improve the understanding of a Nile tilapia cultured in a biofloc system, that aims to enhance water quality through the addition of an external carbon source, were developed. Although biofloc culture systems potentially reduce nutrient losses in aquaculture, knowledge of the nutrient flows and accumulations in the system is not yet well-developed. The main objective of this thesis was to investigate how to simulate the nutrient dynamics and provide management strategy insights given the complexity of a biofloc system. This is performed by combining field observations (Chapter 2) with modelling approaches (Chapters 3, 4, and 5). In Chapter 2, Nile tilapia (Oreochromis niloticus) was cultured in outdoor experimental ponds in Indonesia for eight weeks. During the culture period, the effects of replacing starch-rich ingredients (Control-diet) with non-starch polysaccharides (NSP) rich ingredients (High-NSP-diet) on fish growth, water quality, and nutrient budgets were investigated. Diets were isonitrogenous but the High-NSP-diet contained three times more fibre, as carbon source, than the Control-diet. The results show the potential of High-NSP-diet containing lower quality, relatively cheaper, and locally available ingredients for biofloc culture. Fish growth, biofloc growth, and nutrient dynamics data from Chapter 2 were used to develop the nutrient dynamics models of a fish-biofloc system in Chapters 3 and 4. In Chapter 3, a mathematical model to understand the dynamics and flows of nutrients (carbon (C), nitrogen (N), and phosphorus (P)) in a biofloc-Nile tilapia-rearing system was developed. Process variables included are nutrients, feed, fish, biofloc, periphyton, and water volume. Except for biofloc, the behaviour of the process variables fit the observations with a normalized root mean square error (NRMSE) of less than 30%. Although the High-NSP-diet resulted in more organic waste than the Control-diet, the amount of unutilised C and P were similar between diets. This indicates the ability of biofloc and periphyton to assimilate more waste, especially carbon, in the High-NSP diet. In Chapter 4, the dynamics of biofloc were predicted by combining a fish growth model with the Activated Sludge Model No 1, which was initially developed for wastewater treatment systems. In this model, heterotrophic and autotrophic bacteria utilise the organic matter for growth and are partly consumed by fish. The model was fit to the measurements of fish, biofloc, and nitrogen in the Control-diet datasets with a NRMSE of 3% - 34%. To our knowledge, our study is the first that demonstrates the dynamics of nutrients, fish and biofloc in a fish-biofloc system. Chapter 5 consolidates current knowledge on fish growth, food web and nutrient dynamics from 29 farm-scale models. This knowledge can be used, for instance, to understand the carrying capacity of an aquaculture system and improve the feeding strategy. The majority of the reviewed models considered phytoplankton but neglected bacteria dynamics. Finally, Chapter 6 reflects on the aforementioned main findings of the thesis and how it contributes to sustainable aquaculture and smart aquaculture technology.