Ranking for fight lesion scores is not consistent over time

Aggressive behaviours such as fighting compromise the welfare of group-housed gilts and sows. The extent of lesions that result from fighting is a simple measure of the aggression received by individual animals within a group, and is a potential measure of individual behaviour (Turner et al. 2006). However, group dynamics and individual behaviour can change over time. The hypothesis investigated in this study was that for individual gilts regrouped within parity over time, post-selection and pre-farrowing lesion scores might not be a consistent indicator of aggressive behaviours.

A subset of gilts (n = 3,238) selected at 170 ± 3.3 (mean ± SD) days of age, housed in temporary groups of 20–40 gilts/pen, were scored (0–3) for fight lesions on each quadrant of the body separately, 24 hours after mixing post-selection. Each scoring increment represented an additional five lesions. Total scores for the anterior or posterior regions (0–6) or over the whole body (0–12) were accumulated. Non-zero values were re-scaled to create 0–3 score categories, representing a range of 0–30+ lesions for anterior and posterior scores, or 0–60+ lesions over the whole body. Gilts were subsequently allocated to single parity groups of up to 10 sows post-mating for their gestation period, and rescored (n = 1,929 at 342 ± 15.1 days of age) for fight lesions (0–3) over the whole body upon transfer to farrowing accommodation using the same scoring increment. The range in lesion count across scores was therefore relatively lower pre-farrowing, from 0–15+. Sows removed from groups prior to transfer were not scored pre-farrowing. Associations between anterior and posterior lesion score categories recorded post-selection, and between post-selection and pre-farrowing lesion scores, were examined using a Chi-square test.

With regard to gilts, 5.5% had no lesions 24 hours after mixing whereas 28.7% of sows had no lesions pre-farrowing, indicating a large change in social dynamics between these time points and the visible evidence of fighting amongst sows (Table 1). A reduction in lesion scores was expected, since the development of a social hierarchy within a stable gestating group should reduce measures of antagonistic interactions between sows (Arey 1999). However, the presence of lesions pre-farrowing suggests that some source(s) of motivation for agonistic behaviour within gestation groups (e.g. competition for food) were present. The percentage of sows with high lesion scores post-selection was higher for anterior compared to posterior scores (15.3% vs 4.1%), indicating most fight injuries were received on the front of the sow. There was an association (P < 0.0001) between anterior and posterior scores. Sows without anterior lesions were unlikely to have posterior lesions (<3% of sows) and sows actively engaged in fighting (high anterior scores) also had high posterior lesion scores. In contrast, there was no association (P > 0.05) between lesion scores of gilts recorded post-selection with their lesion scores pre-farrowing. Engagement in fighting post-mixing as gilts was not a predictor of individual engagement in fighting within a new group at a later stage. This is consistent with the results of Tönepöhl et al. (2013), who found no relationship between a sow’s behaviour for initiating aggression with their own lesion scores recorded 10 weeks later.

Table 1.  The percentage distribution of anterior and posterior lesion score groups recorded post-selection (n = 3,238) and their associations with a pre-farrowing score of 0 (from n = 1,595)