Drawing pig feeding patterns: Sensor-driven algorithms for individual welfare monitoring

Summary

Modern sensor technologies have created avenues to monitor the welfare of individual pigs throughout their lifespans. Individual pig behaviour can be measured continuously, but to detect behaviours indicative of welfare problems reliably we need a good understanding of the behavioural variation over time and between pigs. In this thesis, individual feeding patterns of pigs were analysed in detail, to describe and distinguish feeding behaviour of pigs in conventional and welfare-compromised situations. It was found that pigs applied individual feeding strategies dependent on their physical characteristics, activity and social rank. During welfare issues, most pigs did not change their behaviour, though almost all pigs fed less during heat stress. Often, these responses were dependent on pigs’ feeding strategies. This large variation in feeding behaviour and in responses to welfare issues hamper reliable identification of welfare issues using sensors. This thesis highlights alternative pathways to achieve continuous welfare monitoring of individual pigs.