Sow aggression in groups: predicting and implications for sow welfare

© 2014 Dr. Megan VerdonReducing aggression between commercial group housed sows is one of the main challenges facing animal welfare scientists and pork producers alike, particularly because of the risk sow aggression poses to sow welfare and productivity. However, individual variations in the behavioural and physiological responses of sows to the same stressor means that the welfare of some sows may be more compromised than others, an effect that can be overlooked when assessing welfare at a group level. This thesis comprises two parts that examined the relationships between individual sow aggressive characteristics, group composition and sow welfare. Part one of this thesis contains two experimental chapters, in which sexually mature female pigs were classified based on their aggressive behaviour on the day after mixing into three aggressive strategies (Submissive, Subdominant, and Dominant). The first experimental chapter showed that individual female pigs vary consistently in their aggressive characteristics, both during and between gestations. Further, the pattern of aggression delivered by Dominant females, and the pattern of aggression received by each classification of female, reduced with subsequent feeding bouts on a day. The second chapter examined the relationship between the individual aggressive strategy and welfare. It was found that once a hierarchy is established, Dominant females have a reduced risk of receiving aggression, and hence reduced injuries. They are also more likely to gain greater weight and experience less stress. While Subdominant and Submissive animals are comparable in terms of aggression received, live weight gain and reproduction, Subdominant females are more likely to experience greater stress but Submissive females have more injuries later in gestation. In the second part of this thesis, two experiments were conducted: examining predictive tests of individual sow aggression in groups, and using the best of these predictive tests to manipulate the group composition and examine the consequences on sow welfare. Predictive tests conducted prior to mixing showed that the latency and intensity of interaction with a social stimulus were negatively correlated with sow aggressive characteristics after mixing, but not with gilts. Further, 85% of sows that made contact with the social stimulus in fewer than 3s delivered aggression on the day after mixing. However, manipulating the group composition based on predicted sow aggressiveness to form groups of high or mixed aggression had no effect on aggression delivered or received, injuries, cortisol concentrations, reproductive performance, or non-reproductive removals in the group as a whole. From these results it is concluded that variation exists in groups of sows between individual group housed sows, and between groups of sows, in terms of their aggression and consequently welfare and productivity. Submissive sows are likely to benefit from increased access to resources such as space, food and water. Furthermore, behavioural responses of female pigs to a social stimulus before mixing is related to aggressive behaviour on the day after mixing, particularly in the second gestation, but manipulating the group composition had no effect on sow behaviour, welfare or productivity of the group as a whole. The aggressive behaviour of female pigs appears flexible, changing with both social experience and group composition. This raises questions on the suitability of gilts or young pigs as a model for the behaviour of the sow