New social phenotypes from automated computer vision systems for reducing aggression in pigs through selective breeding

Fund

AWF Research Fund

Grant

£24,972

Research Period

2022

Area of study

Breeding for Better Welfare

Description

Investigator(s): Prof Andrea Doeschl-Wilson

Project length: 6 months (due to start October 2023)

Aggression in commercial farms reduce welfare and performance of pigs. Despite numerous efforts to find a practical solution, this issue remains one of the priority welfare issues for pigs in the UK (®).  Genetic selection could provide a long-term solution to control aggression. However, due to the lack of reliable phenotypes, selection for this complex behaviour is difficult in practice.

In the ‘big data’ era, social network analysis (SNA) has gained much attention for studying social interactions underlying e.g. infectious disease spread or aggression in animal groups. Previous studies have shown that SNA provide novel “heritable” behavioural phenotypes that describe the direct and indirect role of each animal in pen level aggression and resulting skin lesions (®,®,®).These heritable phenotypes provide promising candidates for reducing harmful aggression through genetic selection. However, monitoring the behaviour of the animals at the farm level is laborious and costly when conducted by human experts. As an alternative, recent on-farm monitoring and automated capture systems provide exciting opportunities to generate informative data of social behaviour between animals.

The recipient's industry partner, PIC, has implemented these systems to track the positions and interactions of pigs within the pen in real time. In this project, they will assess the feasibility and scope of these systems to construct social networks, using SNA, and identify novel behavioural phenotypes to reduce harmful behaviour through selective breeding. The involvement of the industry partner will ensure swift uptake of the scientific findings by the pig industry to improve animal welfare in commercial farms.