Feed is the largest cost of pork production in Canada and therefore improving feed efficiency (FE) is essential to boost the Canadian pig producer profitability. Additionally, improving FE increases industry competitiveness, decreases demand on global feed resources, and complements environmental sustainability.
Feed conversion ratio (FCR) and residual feed intake (RFI) are common indicators for FE in pigs. Currently, FCR is used for lower feed intake (FI) selection to reach a fixed market weight; however, FCR as a ratio is not statistically preferable and might increase the body weight instead of reducing FI.
Residual feed intake (RFI) is defined as the difference between observed FI and expected FI considering requirements for growth and maintenance. Previous research indicated that RFI has moderate heritability and genomic selection for RFI is feasible.
Selective breeding for RFI has been shown to be a promising tool to reduce feed cost. Selection for reduced RFI has also provided decreased environmental impact. Therefore, it is vital to examine the possibility of using RFI in comparison with FCR as an indicator for FE in Canadian swine industry.
Machine learning have been applied in many fields including agriculture and livestock industry. Genomic selection using machine learning has been successfully used in swine globally but not in the Canadian swine industry. This project will combine traditional genetics, state-of-art genomics and machine learning approaches to improve feed efficiency in the Canadian swine industry.
