Optimisation of algorithms to predict biomass and size distribution of fish reared in cages, using the Aquanetix production management software

• Autor(a) principal: Soares, Diogo Alexandre Marques
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Texto Completo: http://hdl.handle.net/10400.1/12361
• topic: Sparus aurata; Dicentrarchus labrax; Aquacultura em jaulas; Modelo de alimentação; Biomassa

This Thesis aims to optimise the algorithms used to estimate actual biomass and weight distribution in gilthead sea bream (Sparus aurata) and European sea bass (Dicentrarchus labrax) cages by the Aquanetix Software. For this, we first try to understand the practical functioning of a fish farm that uses cages and how can the used procedures affect the collection of data or its veracity. Then, we use the data collected by the company and the observations made on the field in order to attempt the optimization of the algorithms that estimate biomass and weight distribution of the Aquanetix Software. The data parameters analysed were the moving average of the estimated biomass, mortality, density, number of fish and mean weight. Two time periods were tested for the moving average of the estimated biomass, at fourteen and thirty days prior to the first harvest. Between these two periods, the one at thirty days seemed to provide the better biomass estimation. Mortality and density showed to have no apparent influence in the deviations found between the biomass estimations and the total biomass harvested. The number of fish was found to be overestimated in the majority of the studied cages (n=7), with the exception of only cage 109. The mean weight was found to be underestimated in the majority of the studied cages, with the exception of only cage 03. At the end, all proposed goals were achieved. In conclusion, every cage of sea bream studied (n=4) shows an under estimation of the mean weight of fish at first harvest, which in turn leads to an underestimation of the biomass. This suggests that every sea bream cage is currently being under fed, most likely, due to a fault on the feeding model which is probably overestimating the specific feeding rate (SFR) for this species. Alterations should be made to the feeding model in order to resolve this unbalance. The results for the sea bass cages were shown to be more inconclusive since all of the studied cages for this species (n=3) appear to have no common reason to explain the errors found for the estimation of their biomass.