Sow nutrition and fertility: how feeding shapes pig reproduction

Why the authors revisited sow nutrition and fertility#

Pig producers had long worked on the assumption that a gilt needed to reach a certain weight and a certain degree of fatness before she would cycle, and that feeding sows generously in pregnancy was simply good practice. By 1991, a body of research across North America, the United Kingdom and Western Australia had begun to complicate that picture. Aherne, Williams and Head set out to draw these threads together, covering the whole reproductive sequence from puberty through ovulation, conception, embryo survival, gestation and lactation, in a paper published as part of the Recent Advances in Animal Nutrition proceedings.

Their starting point is that pigs can store substantial reserves of protein, glycogen and fat, and that the body appears to use the state of those reserves, together with the pace of their mobilisation, as a signal to the reproductive system. The paper proposes that nutrient intake changes the levels of hormones and substrates such as growth hormone, insulin and insulin-like growth factor 1 in the blood, and that these in turn act on the hypothalamus, the pituitary gland and the ovary. The practical question behind the review is how a producer's feeding decisions, at each stage of a sow's reproductive life, translate into fertility outcomes.

How the evidence was assembled#

This is a review rather than a single experiment, so it does not describe one consistent set of animals or methods. Instead the authors bring together results from a number of separate trials, several conducted at the University of Alberta and the University of Western Australia, alongside published work by other researchers such as Britt, Hughes and Pearce, King, and den Hartog and Van Kempen.

The studies cited vary a good deal in design. Some manipulated feed intake during the rearing period and tracked the age and condition of gilts at puberty. Others altered energy or protein levels in early, mid or late gestation and measured ovulation rate, embryo numbers or embryo survival at slaughter on fixed days of pregnancy. Still others compared different feeding levels during lactation and recorded weaning-to-mating intervals and subsequent litter performance. Reading these trials together lets the authors test whether a consistent pattern emerges across independent groups, which is the main value of a review of this kind, though it also means that no single trial in the paper can carry the full weight of any one conclusion on its own.

Puberty: growth and metabolic state matter more than fat#

The paper argues that fatness itself is not a necessary trigger for puberty. Price and colleagues found that gilts slaughtered at 109 kg showed similar body composition whether or not they had already come into season by that point, and in a population of 349 gilts, Young and colleagues found a normal spread of fatness at puberty rather than the skewed pattern expected if a fat threshold controlled the event.

Data attributed to unpublished work by Magowan reinforce the point. Gilts fed freely from 50 kg reached puberty at 172 days, weighing 108 kg with 17.1 mm of backfat. Gilts held at 50 kg liveweight until 270 days and then fully fed did eventually reach puberty, 65 days later at 335 days, but with just 5.3 mm of backfat. A group held at 80 kg until 270 days and then released to full feeding reached puberty with 6.7 mm of fat, and gilts restricted to 85 percent of ad libitum intake from 50 kg reached puberty at 238 days with 9.6 mm of fat.

The authors conclude that prolonged underfeeding delays puberty rather than preventing it, and that underfed gilts cycle at lighter weights and lower fatness than well-fed animals. They suggest protein reserves, or whatever the body uses as a correlate of them, may matter more than fat itself, and that restricting feed intake to 60 to 85 percent of ad libitum during the 20 to 50 kg growing period delays puberty by about 10 to 14 days.

Growth rate, mature size and the puberty paradox#

Beltranena and colleagues reported a negative quadratic relationship between lifetime growth rate and age at puberty: below about 500 g per day, faster growth brought puberty forward; between 500 and 650 g per day growth rate had little effect; and above 650 g per day further increases in growth rate appeared to push puberty later rather than earlier.

The paper resolves this apparent contradiction by invoking mature body size. Williams and colleagues estimated the mature liveweight of a modern sow, at the point of zero nitrogen retention, at approximately 340 kg, and King suggested gilts reach puberty at around 30 percent of their mature weight. In the Price study, 90 gilts of mixed genotype were slaughtered at 109 kg, and 36 percent had reached puberty by then. Among those that did, the fastest growers reached puberty earliest, yet growth rate was significantly greater overall in the group that had not cycled by 109 kg. The authors explain this by proposing that the fastest-growing gilts simply have a larger mature size, so that 30 percent of their mature weight lies beyond 109 kg, while slower-growing gilts with smaller mature size pass that threshold earlier.

Ovulation rate and the effect of flushing#

The review confirms that raising feed or energy intake during rearing increases pubertal ovulation rate, and that short-term high-level feeding around the first oestrous cycle, commonly called flushing, increases ovulation rate at the second cycle. Beltranena and colleagues showed this is not a true superovulation but rather a correction of the otherwise low ovulation rates seen in gilts that had been feed-restricted.

The likely pathway runs through increased plasma insulin and insulin-like growth factor 1, along with increases in follicle stimulating hormone, luteinizing hormone and the pulse frequency of luteinizing hormone. The authors cite work suggesting the mechanism is a reduction in the atresia, or loss, of developing follicles rather than an increase in how many follicles are recruited in the first place. Because the rise in ovulation rate was not significantly correlated with any particular change in body weight or fatness between the first and second cycles, the authors favour a short-term metabolic effect of realimentation over a structural change in body condition.

Conception, embryo survival and the gestation feeding balance#

A review of 26 experiments by den Hartog and Van Kempen found no significant effect, beneficial or harmful, of feed intake before breeding or across the cycle leading up to mating on conception rate. Embryo survival is a more delicate story. High-level feeding during rearing or just before mating was associated with increased embryo mortality in the trials summarised in the paper, although because ovulation rate also rose, the final number of embryos present at 25 days of gestation was not consistently affected.

A summary of 12 experiments compiled by Hughes, with pre-mating feed intake standardised, found embryo survival of 77 and 82 percent under high and low early-gestation feeding respectively, a gap the authors judge unlikely to be statistically meaningful. Separate trials at the University of Alberta found no significant effect of energy or protein level fed from three days after mating to day 28 of gestation on either ovulation rate or embryo survival, though gilts with higher plasma progesterone around day 9 tended toward better and less variable embryo survival. A further Alberta trial, reported by Baidoo, found that gestation feeding level made no difference to embryo numbers in sows well fed during the previous lactation, but that sows underfed in both lactation and early gestation had the fewest embryos, the greatest embryo loss, and lower plasma luteinizing hormone.

Mammary development, late pregnancy and lactation feed intake#

The paper flags 75 to 105 days of gestation as a critical period for mammary gland development in gilts. Weldon and colleagues reported that high dietary energy during this window reduced the total DNA content of mammary parenchymal cells, implying fewer milk-secreting cells and potentially less milk. Unpublished work from the University of Western Australia, credited to Head, found that sows managed to be fat rather than lean at the same parturition weight had about half as many milk secretory cells and produced less milk.

In late gestation, Cole reported that 39.6 MJ DE per day was needed from day 90 of pregnancy to hold backfat steady, consistent with Close and colleagues, who found sows fed 21 MJ DE per day mobilised fat from day 87 and could lose up to 4.8 kg of fat, representing 20 percent of reserves, by the end of pregnancy. Once lactation begins, the paper describes voluntary feed intake as often inadequate for milk production, citing an average 28-day intake of 4.27 kg for gilts and 4.90 kg for sows even under ad libitum feeding, and Lynch's estimate that around 40 percent of first-litter gilts, 50 percent of second-parity sows and 70 percent or more of older sows fail to meet recommended intakes, so they draw on body protein and fat instead.

What this means for feeding programs#

Williams and Mullan's work, cited in the review, suggests that once body fat passes about one third of body weight, voluntary energy intake during lactation starts to fall, and a regression equation is offered for estimating total body fat from liveweight and P2 backfat. The authors note this implies sows between 120 and 180 kg at farrowing can carry up to 25 mm of P2 backfat before fatness itself begins to depress appetite. A set of declining target weight gains across the first six pregnancies is offered as one practical approach, aimed at keeping body condition stable over a sow's working life rather than chasing maximum gain in any single pregnancy, a theme also present in the same authors' catalogued work on early feeding for lifetime performance of pigs.

Several of the later trials discussed, including work by Spicer and Aherne and by Anderson and colleagues, found that when first-litter sows ate enough during lactation to hold their weight, the level of reserves at farrowing made no difference to later reproductive performance, which pushes against the idea that some tissue loss is an unavoidable part of breeding. We would treat the puberty and flushing findings as reasonably well supported, given they recur across several independent data sets summarised in the paper, but the embryo survival results rest on smaller trials with differences the authors themselves describe as not clearly significant, so they are better read as leads for further work than as settled rules for feeding gilts in early pregnancy. This paper itself sits among the pig and ruminant work gathered in the RAAN conference proceedings collection within the broader Livestock Library research index.

Sources and further reading#

• Trove library search: find a library that holds the paper
• Feedipedia animal feed database: an open-access database of animal feeds
• Dairy Australia: dairy industry research body

Questions#

— The Livestock Library team