Trade-offs between litter size and offspring fitness in domestic pigs subjected to different genetic selection pressures

Artificial selection of the domestic pig (Sus scrofa domesticus) offers a useful model for investigating changes in behaviour associated with reproductive trade-offs between litter size and fitness of offspring. The aim of this study was to evaluate effects of litter size on teat stimulation, sibling competition, and pre-weaning survival and growth in three populations of domestic pigs subjected to different selection pressures (a maternal line selected for high reproductive investment, a paternal line selected for meat production traits, and a crossbred line). We predicted that, with increasing litter size, piglets would spend more time in udder massage, be less likely to gain access to a teat during milk letdown and, if surviving to weaning, have lower, more variable body weights. We also predicted that maternal line sows would wean more piglets of higher weight, despite larger litter sizes, than paternal line sows. Sows (maternal line, n = 12, paternal line, n = 12, crossbred line, n = 14) were loose-housed with their litters in individual farrowing pens. We collected data on piglet behaviour during nursings at 1 day of age, when sibling competition was expected to be most intense. Piglets were weaned at 35 days of age, when they were weighed and cumulative mortality was calculated. As predicted, piglets in larger litters spent more time in pre- and post-letdown udder massage (P = 0.050 and P < 0.001, respectively). In larger litters, more piglets survived to weaning (P = 0.002), but at a cost of a lower proportion of nursings with letdown (P < 0.001), longer nursing intervals on average (P = 0.018), more piglets without a functional teat at letdown (P < 0.001), an increased risk of mortality due to starvation (P < 0.001) and crushing (P = 0.002), and lower (P = 0.039), more variable (P = 0.002) body weights at weaning. In the maternal line, nursing intervals lengthened with increasing litter size (litter size × breed: P < 0.001) despite more post-letdown udder massage (P < 0.001), and mortality due to crushing rose with increasing litter size (P < 0.001), without differential increments in number, weight or weight uniformity of weaned piglets with increasing litter size between breeds (litter size × breed: P > 0.1). Our results suggest that further artificial selection for larger litters in maternal lines will be unsustainable because increments in the number of piglets weaned have increasing costs (e.g. sibling competition, mortality, uneven growth) that compromise piglet welfare and fitness.