Modeling carbon dioxide dynamics in a land based fish farm

Aquaculture is one of the world's fastest growing food production sectors, and it is becoming a more important contributor to global food supply and economic growth which appears to be the most obvious choice for increasing the seafood products. Rainbow trout is one of the most widely farmed fish species in the EU and Italy, accounting for 21% of total EU production in 2018. This thesis presents the development of a dynamic model for carbon dioxide concentration within raceway of rainbow trout farm as a step toward the implementation of Precision Fish Farming. The dynamic model's structure and, as a result, the equations used to simulate carbon dioxide excretion by fish respiration, carbon dioxide concentration in the input water as well as air-water exchange rate of carbon dioxide, were derived from previous literatures and chemical equations. The model was calibrated on a comprehensive dataset collected at a trout farm located in Trentino-Alto Adige, Northern Italy during the H2020 project GAIN – Green Aquaculture INtensification in Europe. Water temperature, pH and dissolved oxygen concentrations in farm influent and effluent were recorded hourly, fish biomass and fish size distribution were estimated daily using non-invasive monitoring device for fifty days in July and August 2020. The model parameters were derived from the available literature. The results suggest that the model could predict the short term carbon dioxide evolution based on real-time hourly pH data and low frequency alkalinity measurement and therefore could be used for developing an early warning system for fish farmers.