High-throughput digital phenotyping and quantitative genetics of feed efficiency in Atlantic salmon

Summary

Feed efficiency (FE) quantifies how well fish convert available energy into useful products. It is increasingly recognised as a key trait in Atlantic salmon and other aquaculture species, given that feed costs make up 30 – 70 % of production expenses and 75 % of environmental impacts. Selective breeding for higher FE could be an effective strategy to lower feed costs, improve nutrient use, and promote environmental sustainability. However, direct selection demands accurate, large-scale phenotyping of FE-related traits at the individual level, which is challenging in complex aquaculture environments. An alternative is indirect selection through correlated traits, such as growth and fat content. Yet, the potential benefits depend on genetic variation and correlations with these production traits, which are rarely studied in salmon. Moreover, growth, body composition, and FE in salmon are dynamic, influenced by developmental stages and environmental conditions. Therefore, these complex but interconnected traits need to be studied across different life stages to explore opportunities for improving FE. This PhD project aims to develop and validate digital phenotyping techniques for rapid, repeated, and non-invasive assessment of various production traits, particularly individual feed intake (IFI), FE, and whole-body fat (WBF) content. The objective is to investigate the genetic basis and explore the potential for genetic improvement of FE-related traits. The data utilised in this thesis originate from large FE trials conducted across three life phases on 2800 Atlantic salmon from 35 full-sib families of the MOWI genetics breeding programme in Norway. The datasets include IFI, Feed conversion ratio (FCR), phenotypic residual feed intake (pRFI), genetic residual feed intake (gRFI), WBF, average daily gain (ADG), Body weight (BW), spectral recordings, and genotyped data.