Abstract
Tail biting is one of the biggest welfare problems in pigs. However, depending on the individuals involved (e.g., tail biter/victim), pigs seem to change their behaviour prior to tail biting events, which raises the possibility of early detection and thus prediction and prevention of tail biting. In this retrospective explorative study, we used datasets from four different studies with 9 trials of rearing (4 pens/trial with 24 pigs/pen) and fattening (8 pens/trial with 12 pigs/pen) that focused on the exploration behaviour of undocked pigs towards plant-based enrichment materials. From this dataset, we identified 8 pens from rearing (n = 192 pigs) and 6 pens from fattening (n = 72 pigs) in which individual tail biters were identified. From this dataset, we investigated whether any a priori behavioural changes in exploration or feeding could be identified with respect to tail biting. Furthermore, the effects of weight parameters from suckling to fattening were examined. Using linear mixed effects models, we found that exploration duration was linked to days prior to tail biting in rearing, depending on CatPig (category of pigs: biter, victim, neutral pig) (P = 0.001), in fattening independent of CatPig (P<0.0001), and by duration, amount and frequency of feed consumption in fattening (P<0.0001). Some weight parameters covaried with CatPig in rearing (weight-gain suckling: P = 0.0018; weaning weight: P = 0.019) and fattening (weaning weight: P = 0.07; start weight at fattening: P = 0.03; weight-gain rearing: P = 0.02). Suitable indicators for future early detection trials of tail biting could be exploration duration in rearing and fattening and feeding data in fattening. Moreover, weight parameters in rearing and fattening and exploration duration in rearing may be used to identify individual pigs that might become tail biters in an upcoming tail biting event. The retrospective explorative nature of our analysis revealed interesting patterns; however, further studies are needed to confirm our findings.