The emotional state can influence decision-making under ambiguity. Cognitive bias tests (CBT) proved to be a promising indicator of the affective valence of animals in a context of farm animal welfare. Although it is well-known that humans can influence the intensity of fear and reactions of animals, research on cognitive bias often focusses on housing and management conditions and neglects the role of humans on emotional states of animals. The present study aimed at investigating whether humans can modulate the emotional state of weaned piglets. Fifty-four piglets received a chronic experience with humans: gentle (GEN), rough (ROU) or minimal contact (MIN). Simultaneously, they were individually trained on a go/no-go task to discriminate a positive auditory cue, associated with food reward in a trough, from a negative one, associated with punishments (e.g. water spray). Independently of the treatment (P = 0.82), 59% of piglets completed the training. Successfully trained piglets were then subjected to CBT, including ambiguous cues in presence or absence of a human observer. As hypothesized, GEN piglets showed a positive judgement bias, as shown by their higher percentage of go responses following an ambiguous cue compared to ROU (P = 0.03) and MIN (P = 0.02) piglets, whereas ROU and MIN piglets did not differ (P > 0.10). The presence of an observer during CBT did not modulate the percentage of go responses following an ambiguous cue (P > 0.10). However, regardless of the treatment, piglets spent less time in contact with the trough following positive cues during CBT in which the observer was present than absent (P < 0.0001). This study originally demonstrates that the nature of a chronic experience with humans can induce a judgement bias indicating that the emotional state of farm animals such as piglets can be affected by the way humans interact with them.

Low birth weight piglets are at high risk of mortality, because of the rapid depletion of their energy reserves after birth. At 3 h postpartum, 405 piglets weighing <1.1 kg were either dosed orally with 2 mL of (1) coconut oil (CO, 74 kJ/2 mL, n = 107 piglets), (2) commercial product (CP, 71 kJ/2 mL, n = 101 piglets), (3) water (W, 0 kJ/2 mL, n = 100 piglets) or (4) were sham-dosed (S, n = 97 piglets). Treatments were applied within litter (97 sows). Before treatment piglets were weighed, scored for vitality and blood glucose concentration (subset: CO = 45 piglets, CP = 38 piglets, W = 49 piglets and S = 44 piglets) and rectal temperature were measured. Rectal temperature was remeasured 1 h post-treatment (4 h postpartum). At 24 h post-treatment (27 h postpartum), vitality, weight and blood glucose were remeasured. Piglets were weighed on D5, D7, D10, D14, D21 and at weaning (27 ± 0.1 day old). Mortality rate and cause were recorded until 24h period post-treatment and until weaning. Data were analysed using Generalised Linear Mixed Models in SAS. There was no overall effect of treatment on any of the parameters measured. In conclusion, a single oral of fat-based energy supplement dose at birth did not improve growth, survival, rectal temperature or vitality of low birth weight piglets.

Nurse sow strategies are used to manage large litters on commercial pig farms. However, new-born piglets transferred to nurse sows in late lactation might be compromised in terms of growth and survival. We investigated the effects of two nurse sow strategies on piglet growth, suckling behaviour and sow nursing behaviour. At 1-day post-farrowing, the four heaviest piglets from large litters were transferred to a nurse sow either 21 (1STEP21, n=9 litters) or 7 (2STEP7, n=10 litters) days into lactation. The remainder of the litter remained with their mother and was either kept intact (remain intact (RI), n=10 litters) or had some piglets cross-fostered to equalise birth weights (remain equalised (RE), n=9 litters). The 7-day-old piglets from 2STEP7 were transferred onto a sow 21 days into lactation (2STEP21, n=10 litters). The growth of new-born piglets on 1STEP21 and 2STEP7 nurse sows was initially lower than in RI litters (F 3,33.8=4.61; P<0.01), but weaning weights did not significantly differ (F 4,32.7=0.78; P>0.5). After the 1st week of lactation, the weights and growth rates did not differ between treatments. Fighting behaviour during nursing bouts decreased over time. The frequency of fights was higher in 1STEP21 and 2STEP21 litters compared with RI litters (t 122=3.06 and t 123=3.00, respectively, P<0.05). The 2STEP21 litters had shorter nursing bouts than RI and 1STEP21 litters (t 107=−2.81 and t 81.7=2.8, respectively, P<0.05), which were more frequently terminated by 2STEP21 than RI sows (t 595=2.93; P<0.05). Transferring heaviest piglets from RI and RE litters to nurse sows reduced the percentage of teat changes during nursing bouts (RI: F 1,275=16.61; RE: F 1,308=43.59; P<0.001). In conclusion, nurse sow strategies do not appear to compromise piglet growth. However, new-born piglets transferred onto sows in late lactation experienced more competition at the udder, suggesting that the sows’ stage of lactation is of importance to how achievable nurse sow strategies are. Thus, the two-step nurse sow strategy is likely the best option (in relation to growth and suckling behaviour), as it minimises the difference between piglet age and sow stage of lactation.

Management strategies are needed to optimise the number of piglets weaned from hyper-prolific sows. Nurse sow strategies involve transferring supernumerary new-born piglets onto a sow whose own piglets are either weaned or fostered onto another sow. Such ‘nurse sows’ have extended lactations spent in farrowing crates, which could have negative implications for their welfare. This study used 47 sows, 20 of which farrowed large litters and had their biggest piglets fostered onto nurse sows which were either 1 week (2STEP7, n=9) or 3 weeks into lactation (1STEP21, n=10). Sows from which piglets were removed (R) were either left with the remainder of the litter intact (I) (remain intact (RI) sows, n=10), or had their litters equalised (E) for birth weight using piglets of the same age from non-experimental sows (remain equalised (RE) sows, n=9). Piglets from 2STEP7 were fostered onto another nurse sow which was 3 weeks into lactation (2STEP21, n=9). Back-fat thickness was measured at entry to the farrowing house, at fostering (nurse sows only) and weaning. Sows were scored for ease of locomotion and skin and claw lesions at entry to the farrowing house and weaning. Salivary cortisol samples were collected and tear staining was scored at 0900 h weekly from entry until weaning. Saliva samples were also taken at fostering. Data were analysed using GLMs with appropriate random and repeated factors, or non-parametric tests were applied where appropriate. Back-fat thickness decreased between entry and weaning for all sows (F 1,42=26.59, P<0.001) and tended to differ between treatments (F 4,16=2.91; P=0.06). At weaning RI sows had lower limb lesion scores than 2STEP7 and RE sows (χ 2 4=10.8, P<0.05). No treatment effects were detected on salivary cortisol concentrations (P>0.05) and all nurse sows had a higher salivary cortisol concentration at fostering, compared with the other days (F 10,426=3.47; P<0.05). Acute effects of fostering differed between nurse sow treatments (F 2,113=3.45, P<0.05); 2STEP7 sows had a higher salivary cortisol concentration than 1STEP21 and 2STEP21 sows on the day of fostering. 2STEP7 sows had a higher salivary cortisol concentration at fostering, compared with 1STEP21 and 2STEP21 sows. Tear staining scores were not influenced by treatment (P>0.05). In conclusion, no difference was detected between nurse sows and non-nurse sows in body condition or severity of lesions. Although some nurse sows experienced stress at fostering, no long-term effect of the nurse sow strategies was detected on stress levels compared with sows that raised their own litter.