Group-lactation housing from 7 or 14 days post partum: effects on sow behaviour

Group-lactation housing may improve sow welfare by increasing the opportunity for the sow to move about and express social and maternal behaviours (van Nieuwamerongen et al. 2014). This study tested the hypothesis that group-housed lactating sows would (1) engage in less aggression and more positive social interactions when mixed at 7 rather than 14 days post partum; and (2) show better maternal behaviour (nursing behaviour and sow-piglet interactions) than sows housed in farrowing crates. One hundred and twelve sows (Large White × Landrace, PrimeGro™ Genetics, Corowa, NSW; Parity 1 to 7) and their litters were allocated to one of three treatments over four time replicates: (1) group lactation (GL) from 7 days post partum (GL₇, n = 48 sows); (2) GL from 14 days post partum (GL₁₄, n = 48 sows); or (3) farrowing crate (FC; n = 16 sows). All sows farrowed in standard farrowing crates, where FC sows remained until weaning. However, GL₇ and GL₁₄ sows and their litters were transferred from farrowing crates to GL pens (one pen of five sows at 8.4 m²/sow and one pen of seven sows at 8.1 m²/sow, per treatment and replicate) at 7 and 14 days post partum, respectively. All treatments were weaned at 28 days post partum. Treatments were balanced for sow parity, weight and litter size, and there were no treatment differences in litter weight and sex ratio, or within pen/crate variation in these variables. For GL pens, two focal sows per pen (one high and one low parity) were video recorded from 0700 to 1700 on the day after mixing (D2) and 2 days before weaning (pre-weaning, PW). Of the four FC sows per replicate, two were video recorded on the same days as GL₇ (FC₇) and two on the same days as GL₁₄ (FC₁₄). Data gathered continuously from video records included sow aggressive and nursing behaviours while sow time-budgets were observed using point sampling at 5-min intervals. Behaviours were analysed with LMM and GLMM models (SPSS v23.0, IBM, Armonk, NY, USA), with the main effects of housing (GL v. FC), litter age at mixing (7 v. 14 days) and observation day (D2, PW), as repeated-measures and controlling for pen and replicate as random factors (Table 1). There were no significant interactions between main effects.

Table 1.  Sow behaviour (per sow and pen/crate) in each housing treatment (Group lactation, GL; Farrowing crate, FC; Housing), relative to age of GL litter at mixing (7, 14; Age) and observation day (d 2 post-mixing, D2; 2 days pre-weaning, PW; Day)

Aggression (bites/knocks) between GL sows increased by nearly 40% from d 2 to 26; however, there was no effect of age at mixing on aggression. GL₁₄ sows engaged in more positive interactions with conspecifics (i.e. nosing sow) than GL₇ sows. Sows spent less time lying in GL pens compared to FC and interacted with piglets more frequently. Whilst there was no difference in number of nursings between GL and FC treatments, GL housing disrupted nursing behaviour, as evidenced by reduced proportion of successful nursing bouts, a tendency for increased sow terminated bouts and a longer inter-nursing interval, compared to FC sows. Sow aggression and disrupted nursing behaviour in GL may result in compromised welfare and growth of sows and piglets, and is being investigated.