RESUMO
The aim of this study was to analyse spontaneous play behaviour in litters of domestic pigs (Sus scrofa) for sources of variation at individual and litter levels and to relate variation in play to measures of pre and postnatal development. Seven litters of commercially bred piglets (n = 70) were born (farrowed) within a penning system (PigSAFE) that provided opportunities for the performance of spontaneous play behaviours. Individual behaviour was scored based on an established play ethogram for 2 days per week over the 3 week study period. We found strong evidence of litter differences in play behaviour (F(6,63) = 27.30, p < 0.001). Of the variance in total play, 50% was attributable to differences between litters with a lesser proportion (11%) to between piglets within litters. We found similar evidence of litter differences when we analysed the separate play categories (e.g. for locomotor play: F(6,63) = 27.50, p < 0.001). For social and locomotor play the variance was partitioned in a broadly similar way to total play; however for object play the variance was distributed with a more even balance across and within litters. In terms of explanatory factors we found little evidence that at the litter level differences in play were associated with differences in general activity. Of the prenatal factors measured, we found that birth weight was positively associated with total play and the play categories (e.g. with total play: F(1,64) = 12.8, p < 0.001). We also found that postnatal piglet growth up to weaning (as a percentage of birth weight) had a significant positive association with total play and the play categories (e.g. with object play: F(1,66) = 20.55, p < 0.001). As found in other studies, on average males engaged in more social play (e.g. non-injurious play fighting: F(1,63) = 39.8, p < 0.001). Males also initiated more play bouts on average than females (F(1,62) = 4.41, p = 0.040). We conclude that the study of differences between litters and individuals provides a robust approach to understanding factors potentially influencing play behaviour in the pig. This work also provides support for the use of play as a welfare indicator in pre-weaned piglets as the litter differences in play we observed were associated positively with physical development.