Fish abundance estimation with imaging sonar in semi-intensive aquaculture ponds

To know the abundance of fishes and their size distribution in the semi-intensive rearing systems in traditional ponds is an aspect key to plan and manage efficiently the sales lots. Usually this information is obtained by means of sampling which mandatory supposes a direct catch and stressful and time consuming management of fishes. Therefore, in this work we propose the use of non-invasive procedures based on multibeam sonars or imaging sonars to count and size the fishes in the ponds. For that, we use a commercial technology portable-fixed multibeam imaging sonar and estimate the abundance in ponds of a gilt-head seabream (Sparus aurata) fishfarm from sonar image analysis and adapting statistical methodologies traditionally applied for bird abundance estimation. Additionally, a simulation software was developed to emulate the fish aggregation contained in the rearing ponds. This computer program allows the calculation of an abundance correction factor which depends on the transducer beam size in relation to the pond size. The results indicate that the estimation is as accurate as the obtained by the fishfarm manager using traditional sampling methods and additionally it is possible to obtain a realistic function of the size distribution which allows estimate the biomass by size contained in the rearing ponds.This work was supported by KTTSeaDrones project (0622_KTTSEADRONES_5_E), cofunded by the European Regional Development Fund, ERDF, through the Interreg V-A Spain-Portugal program (POCTEP) 2014–2020. We would like to express our gratitude to Rafael Rodríguez Sierra (Manager of ‘Salinas del Astur’) for his willingness to carry out all the experiments of the KTTSeaDrones project at the ‘Salinas del Astur’ facilities. Funding for open access charge: Universidad de Huelva/ CBUA