Movement change detected by optical flow precedes, but does not predict, tail-biting in pigs

Abstract

Videos of pigs were analysed using an optical flow algorithm to measure group movement in the four days leading up to and including an outbreak of tail-biting. 8 pens in which tail-biting occurred (Outbreak) were retrospectively matched with 8 similar pens in which no tail-biting occurred over the lifetime of the pigs (NoOutbreak) and videos analysed for the same four days. Results showed an increase in movement over the four days in both groups and no difference between Outbreak pens and their matched NoOutbreak pens. On the contrary, Outbreak and NoOutbreak matched pens showed similar changes in behaviour, suggesting that the pigs were influencing each other or responding to common external circumstances. Such a tendency to “match” the behaviour of pigs in neighbouring pens makes it difficult to identify which pens will show tail biting.