Fighting of gilts after mixing is associated with early removals, altered litter sex ratio and lower piglet survival

Aggressive behaviour can compromise the welfare of group-housed gilts and sows and affect reproductive performance (Spoolder et al. 2009), but the extent of individual aggression is frequently unknown. Lesions resulting from fighting indicate the intensity and extent of aggressive encounters between sows (Bunter and Boardman 2015). The aim of this study was to investigate if lesions resulting from fighting amongst gilts were associated with subsequent reproductive outcomes for group-housed sows.

Gilts (n = 3238) scored for the extent of lesions resulting from fighting at 24 hours after mixing post-selection and again before farrowing (n = 1929) were used in this study. Lesion scoring and their grouping were described in Bunter and Boardman (2015). Gilts were also scored for pre-farrowing condition and locomotion. Condition scores represented under- to over-conditioned sows (scored: –1, 0, 1) while locomotion was scored on a four point scale, from 0 (normal) to 3 (very poor). Gilts removed from the herd after selection without a farrowing event were identified (n = 881), and farrowed sows were recorded for litter size and average piglet birth weight. Data were also available from a smaller subset of litters (n = 915) at the time of analyses, to investigate sex-ratio of live born piglets within litters and piglet survival until weaning. The association between lesion score categories (Bunter and Boardman 2015) and removals without a farrowing event was assessed using logistic regression, submitting score groups (anterior, posterior, or whole body) separately to the analysis, after accounting for selection date (61 levels) and breed (two levels). Implications of lesion scores as covariates for reproductive traits were examined using linear models with ungrouped scores (SAS^®; USA).

The extent of fight lesions 24 hours after mixing was highly associated (P < 0.001) with selection date and breed, but not gilt weight. Breed differences in lesion scores were no longer evident for sows rescored before farrowing. Relative to other score groups, there was an increased tendency for selected gilts with high anterior lesion scores (group 3) recorded after selection to be removed from the herd without a farrowing event (31.1 vs 26.3%, P = 0.026), but removals for a specific reason (e.g. feet and leg problems, stale, not in pig) were not statistically significant. There were no significant associations between lesion scores of gilts after selection and their pre-farrowing condition or locomotion scores, or between lesion scores (either after selection or before farrowing) with sow reproductive traits, such as litter size or average piglet birth weight. In contrast, sows with more fight lesions scored before farrowing had reduced pre-farrowing condition score (P = 0.005), poorer locomotion scores (P < 0.001) and a slightly shorter lactation length (P = 0.004). Higher anterior, but not posterior, lesion scores recorded on gilts after selection were also linearly associated (P = 0.041) with an increasing ratio of female : male piglets. This equates to a maximum change in sex ratio of 4.2% across a seven-score range in lesions (0 to 30+ fight lesions), and implies that engagement in fighting had physiological consequences that directly or indirectly affected the sex-ratio of offspring born almost 6 months later. A sex-ratio biased towards females has been repeatedly demonstrated in guinea pigs subjected to an unstable social environment and was accompanied by reduced maternal androgens, which also affect fertility (Kemme et al. 2009). Increasing posterior (P = 0.019), whole body (P = 0.021) or to a lesser extent anterior lesion scores (P = 0.057) recorded on gilts 24 hours after mixing were also associated with a decreased proportion of piglets that survived from birth until weaning.

Overall, lesion scores resulting from fighting explained little of the variation in gilt removals (R² < 0.5%) or reproductive traits (typically R² < 1–2%), even when associations identified were statistically significant. Therefore, individual variation in engagement in fighting is just one of many, frequently unidentified factors contributing to variation in sow wastage and reproductive outcomes under group housing. Moreover, lesion scores are non-specific descriptors for individual behaviours and (or) stress relating to aggression at the time of scoring, and thus may have limited predictive utility.