Research shows gestational heat stress results in differences in behavior feed intake and growth and affects the nutrient composition of milk produced by gilts SONGQIUJU/ISTOCK/THINKSTOCK

Research shows gestational heat stress results in differences in behavior, feed intake and growth, and affects the nutrient composition of milk produced by gilts.

Emerging evidence expands industry’s understanding of the true impact of heat stress

When females gestate through the heat of summer, the environmental stress reduces productivity in several ways. Researchers and producers have known for decades that the pregnant gilt or sow is directly affected by heat stress — reducing farrowing rate and increasing stillborn rate among other detrimental consequences. What may come as a surprise, however, is that her unborn offspring are also affected by gestational heat stress.

Our most recent research has shown that some of the resulting impacts can be long-lasting, and even appear to be transgenerational. When considered on an individual basis, many of the physiological alterations described here would have minimal impact on productivity and profitability. But when the additive effects are considered, the economic losses associated with gestational heat stress are amplified and may justify more serious consideration of heat abatement strategies.

In order to precisely examine the effects of gestational heat stress, we subjected pregnant gilts to controlled heat stress in environmental chambers, resulting in two treatment groups: gestational thermoneutral (gTN) or gestational heat stress (gHS). Other than their time in the chambers, gilts from both treatments were managed as one contemporary group. Near the end of gestation, all gilts were moved to a thermoneutral farrowing room and managed as one group. Their piglets were weaned at about 21 days and grown out together in a wean-finish barn at the University of Missouri’s South Farm. All of the female offspring (F1 generation) were retained for further studies focused on reproductive, lactational and production parameters.

Prior to breeding

At the wean-finish barn, gilts from both treatment groups were penned together. While that precluded comparison of feed intake by treatment, we did evaluate behaviors. Cameras were used to record posture and location of each pig over several weeks of the growing period in July.

There were some subtle differences in behavior, but the most striking was the percentage of pigs standing at the feeder by treatment. Gilts that were from gTN dams spent significantly less time at the feeder than gHS gilts, especially at the beginning and end of the day (see Figure 1).

Barrow littermates of these gilts were housed individually in small pens at the same farm, and weekly measures of growth and feed intake were taken. As might be expected based on gilt behavior, gHS barrows ate significantly more feed than gTN barrows in both grower (6%) and finisher (10%) phases of production (see Figure 2).

Corresponding improvements in growth were not observed, indicating that the growth of barrows from the gHS treatment was not as efficient as growth of the gTN barrows.

Gilts (n=124) were transported to the TAREC research farm at Virginia Tech at the end of the “finishing phase.” Attainment of puberty and body weight at a constant age was similar for gTN and gHS females. Those females were subsequently inseminated and allowed to farrow, thus producing an F2 generation of piglets. It is important to note that half of the F2 piglets were born in the spring, while the other half were born in the summer, adding an additional facet for comparing performance.

Pregnancy and gestation

The length of gestation and weight gain during gestation was similar for gTN (114.8 ± 0.21 days; 105.46 ± 9.54 pounds) and gHS (115.0 ± 0.25 days; 101.69 ± 6.40 pounds) females. Number of piglets born, number born alive or stillborn did not differ statistically between treatments, although all numerically favored gTN offspring (see Figure 3).

Although the numbers of piglets born did not differ between treatment groups, gHS gilts tended to have lower piglet survival than their gTN counterparts (88.9 ± 0.02% versus 93.9 ± 0.02%; P=0.08). Interestingly, during lactation, body weight loss did not differ (49.99 ± 7.50 versus 41.87 ± 7.54 for gHS and gTN, respectively), even though gHS gilts tended to consume more feed during lactation (11.95 ± 0.25 versus 11.31 ± 0.26 pounds per day for gHS vs. gTN, respectively; P=0.07). This suggests the heat stress experienced by these females while they were fetuses in utero affected their long-term efficiency.

Milk production

Besides growing and reproducing, gilts are expected to produce milk to support growth of their suckling piglets. Milk production of the females from the F1 generation was measured at peak lactation using the weigh-suckle-weigh technique. While this technique has its advantages, it does leave room for error, making it difficult to detect small differences in milk production. Inherent difficulties in controlling piglet urination and defecation between weighings, as well as the disturbance stress associated with repeated piglet handling, undoubtedly reduces the accuracy of this technique.

With that said, results of the weigh-suckle-weigh analysis showed that gilts exposed to gTN and gHS produced similar amounts of milk. Interestingly, there was also no effect of season of farrowing on milk production, nor any interactions between gestational treatment and season of farrowing.

The lack of difference in milk production between the spring-farrowing and summer-farrowing gilts was somewhat surprising since previous studies have found that heat stress can decrease milk production by as much as 35% (Renaudeau and Noblet, 2001; reviewed in Black et al., 1993). These previous studies used different methods to measure milk production, and some used multiparous, rather than primiparous, sows. These variations between studies may explain the discrepancies in results.

Colostrum and milk nutrients

In addition to the amount of milk available to the offspring, the quality and nutrient content of the milk is also important for health and growth of the piglets. Gestational environment of the F1 generation did not affect milk fat percentage, milk solids nonfat (SNF) percentage, nor somatic cell count. Females that had been in utero in heat stress conditions produced milk with lower lactose content (5.21 ± 0.04% versus 5.10 ± 0.04% for gTN and gHS, respectively; P<0.05), while their milk tended to have a higher protein percentage than gTN gilts (5.24 ± 0.15% versus 5.61 ± 0.14% for gTN and gHS, respectively; P=0.07).

The decrease in milk lactose content, and tendency for greater milk protein content of gHS gilts, provides evidence that the metabolic state imparted by hyperthermic intrauterine development persists into adulthood to alter the milk nutrient content of lactating gilts.

In particular, it is known that lactose, which is the primary carbohydrate present in sow milk, is positively correlated with both milk yield and piglet growth through weaning (Noble et al., 2002). The reduced lactose present in the milk of gHS gilts was not associated with decreased milk yield or differences in weaning weight in the present study, although the lack of differences may have been related to previously described challenges with milk data collection.

For all treatments, milk fat was least at Day 21 compared to other observations, and milk lactose content was lower in the colostrum than in the transition (Day 7) or mature milk (Day 14 and Day 21). Milk SNF and milk protein were greatest in the colostrum, and declined from colostrum to Day 7 and from Day 7 to Day 14.

When compared across seasons, milk fat was higher in gilts that farrowed in summer compared to spring, and this was especially apparent at the Day 7 milk collection (Table 1). 

Contrary to milk fat, milk lactose was lower in the milk of summer-farrowing gilts compared to spring-farrowing gilts, particularly in colostrum samples (Table 2). Interestingly, milk SNF was greater in the milk of summer-farrowing gilts than spring-farrowing gilts, with the most distinct difference apparent in the colostrum samples (Table 3).

Growth, performance

Considering the observed differences in nutrient composition of the milk produced by the F1 generation, we were interested in measuring the growth and performance of the gilts and barrows of the F2 generation. These piglets, then, were the offspring of gTN (OgTN) dams or offspring of gHS (OgHS) dams. Any potential differences observed in this generation would be mediated through alterations in physiology imparted on their dams when their dams were fetuses or permanent genetic differences passed from generation to generation. We found that there were no treatment-related differences in postweaning growth owing to the dams’ gestational environment (OgHS versus OgTN) as assessed by average daily gain, body weight at 113.7 ± 0.4 days of age and adjusted days of age at a common market weight. There also was no difference in weight or age at marketing between OgHS and OgTN pigs.

When pigs were marketed, carcass characteristics were assessed (n=32 male OgHS, 31 female OgHS, 33 male OgTN and 33 female OgTN). Loin depth and loin eye area did not differ between pigs born to gHS or gTN dams. Interestingly, however, there was a statistical tendency for pigs born to gHS dams to have greater backfat than pigs born to gTN dams (Figure 4).

This difference is particularly pertinent because it is consistent with backfat differences observed in the barrows from the F1 generation and would make sense given the higher feed intake without corresponding increased daily gain in those pigs. Male offspring of the gHS dams also had the shortest carcass length of any of the treatment groups (Figure 5). Finally, there was also a tendency for pigs born to gHS dams to have greater dressing percentage than pigs born to gTN dams.

Conclusions and implications

Many producers are aware of the direct effects of heat stress on their breeding herd (delayed age at puberty, irregular estrous cycles, decreased conception rates and increased abortions, in particular), but may fail to realize that the growth and performance of these pigs affected by hyperthermia during intrauterine development will be impaired for many months after summer has ended. Evaluating the impact of heat stress on-farm is complicated because pigs suffering the effects of heat stress are around at different times of the year than those not.

The results of our experiments demonstrate that gestational heat stress results in differences in behavior, feed intake and composition of growth, and affects the nutrient composition of milk produced by primiparous F1 gilts. Numerical differences in litter size and piglet survival traits all favored the gTN environment, and failure to reach statistical significance may be the result of limited number of observations. Differences in carcass characteristics of the F2 generation of offspring were also apparent, suggesting that consequences are long-lasting and may be passed from generation to generation. This supports previous work completed by collaborators at Iowa State University, which showed that pigs exposed to hyperthermia during intrauterine development had greater daily adipose deposition at the expense of lean tissue accretion.

All data reported here are from first-parity females, and of course, we expect them to be more susceptible to suffering from high temperatures given they are also still growing. The economic impact of summer temperatures is likely underestimated for the commercial swine industry. There has been no consideration of the long-term effects on the breeding herd from gilts that were fetuses during the summer; there is little recognition of the impact on feed efficiency or carcass composition of market hogs that were fetuses in the summer.

Taken together, the results of these experiments emphasize the need for investigation into heat abatement strategies to alleviate the negative consequences of heat stress both for the effects we observe during the summer and those not evident until far into the future.

This work was made possible with financial support of the USDA-National Institute of Food and Agriculture, the pork checkoff and Agricultural Experiment Station funds from the University of Missouri, Virginia Tech and Iowa State University.

Others contributing to this research are Matt Lucy, University of Missouri; Lance Baumgard, Nick Gabler and Jason Ross, Iowa State University; Jeffrey Wiegert, North Carolina State University; Mark Estienne and Robert Rhoads, Virginia Tech.  

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.