NCSU Extension Swine Husbandry 2002
A more printable version of Swine News in Adobe Acrobat.

September, 2002 . Volume 25, Number 08

The Science of Odor



In last month’s edition of Swine News, the physiology associated with the establishment and maintenance of pregnancy in swine was reviewed. Figure 1 summarizes the chronological sequence of these events and their association with normal management practices. This month’s article illustrates how this information can be used in conjunction with performance data to help pinpoint when problems are likely to occur during a female’s reproductive cycle.

Critical Analysis of Reproductive Performance

Four pieces of information are needed to analyze reproductive performance:

  • Farrowing rate.
  • Litter size (number of pigs born alive).
  • Return interval of non-pregnant females.
  • Subset groupings.

Subset groupings refer to the ability to determine averages for any of the first three measures for specific subsets of animals, i.e., sows bred in summer versus winter; first parity versus older sows, etc. The importance of being able to conduct subset analyses will be discussed later. Table 1 contains 5 situations in which some aspect of reproductive performance is poor.

Situation 1-Poor Farrowing Rate with Regular Return Intervals and Poor Number Born Alive

Poor farrowing rate and number born alive, coupled with regular return intervals, can be caused by one or more of the following:

  • Poor fertilization rates.
  • High embryonic mortality during the first two weeks of pregnancy.
  • Low ovulation rates.

In sows that did not conceive, five viable embryos, the minimum needed to signal pregnancy, were not present on day 12 of pregnancy. Thus, the sows never received the first pregnancy signal, pregnancy was terminated, and they returned to estrus at regular intervals. Physiologically, it is similar to never being bred. Obviously, poor fertilization or high embryonic losses after fertilization create situations in which the numbers of viable embryos are low and reduce the likelihood of having enough embryos for pregnancy to continue. However, it is important to recognize that low ovulation rate creates a similar situation. Normal fertilization and embryonic survival rates in swine are 90 percent and 60 percent, respectively. Consequently, an ovulation rate of 10 ova (5 viable embryos / 0.90 / 0.60) is needed to produce 5 viable embryos. Anything less would yield an inadequate number of embryos needed for a successful, first pregnancy signal. The problems in the sows that did conceive and eventually farrow are similar in nature, but reduced in magnitude, compared with their counterparts that did not conceive. Both pregnancy signals occurred, so the minimum number of embryos were present, but litter size at birth was reduced due to losses during the first two weeks or lack of enough eggs being released during ovulation.

From a management perspective, the interval to scrutinize for potential problems in this situation is from farrowing through the first two weeks post-breeding. The most common causes of fertilization failure are poor semen quality and poor breeding management. Consequently, microscopic evaluations of semen should be performed along with a thorough review of estrus detection procedures and breeding regimens. Low ovulation rates and high embryonic mortality are often associated with short lactation lengths or poor feed intake during lactation. In the case of short lactation lengths, it is often difficult to determine what is too short. A good rule of thumb is anything less than 14 days. In addition, it is important to remember that, physiologically, reduction in the suckling intensity during the first two weeks of lactation by partial weaning can have the same physi-ological effects as early weaning. Finally, the presence of any social, physical, or environmental stresses during the first two weeks post-breeding should be investigated. Common ones include rough handling during movement of animals, high ambient temperatures, and competition for food and space among sows in pens.

Situation 2-Poor Farrowing Rate with Irregular Return Intervals and Poor Number Born Alive

Poor farrowing rates and small litters at birth, in conjunc-tion with irregular return intervals, present a situation that is very similar to the first one described, with one important exception, namely, that sows that do not farrow return to estrus at irregular intervals. Physiologically, this indicates there were enough embryos to produce the first pregnancy signal at day 12, but not the second signal, which occurs between days 17 and 28. This is why the return intervals of sows are greater than 28 days. The sows were pregnant during the first two weeks after breeding, but not during weeks three or four.

Low ovulation rates or high embryonic mortality between days 12 and 28 are the most likely reasons for this situation. Conversely, it is unlikely that fertilization failure is a contrib-uting factor. If it were, then embryonic numbers would be low before day 12, not after, and non-pregnant animals would be expected to exhibit regular returns to estrus (as in Situation 1). Consequently, management practices during lactation and during the third and fourth weeks of pregnancy should be examined. In many operations, sows are often moved from the breeding area to gestation barns by 30 days post-breeding. If sows are moved during the third or fourth week of pregnancy, the manner in which they are moved should be examined closely since their movement corre-sponds with a potential cause of poor performance in this situation - high embryonic mortality after day 12.

Importance of Subset Analyses

The first two scenarios dealt with situations in which both farrowing rate and number of pigs born alive were poor. The next three will address situations in which one measure of reproductive performance is good and the other is bad. Whenever, this occurs, there usually is a smaller group or "subset" of animals in which performance is bad. This certainly would be the case in a situation in which the farrowing rate was low, but the number of pigs born alive was good. Obviously, the sows that farrowed had normal litters, but a group or subset of females that did not farrow creates the poor farrowing rate problem. As a result, one of the first steps that should be taken when farrowing rate and the number born alive are moving in opposite directions is to examine the average performance of different groups of animals in the herd. As mentioned earlier, logical comparisons are things such as sows that were bred in the summer versus other seasons or first and second parity sows with older females. Another type of subset analysis that can be very useful, especially for investigating litter size problems, is called a frequency distribution or diagram. This type of analysis shows the number or frequency of sows that had a specific number of pigs born alive.

Figure 2 provides an example of a frequency diagram for litter size. On the horizontal or x-axis, categories of litter sizes are listed, i.e., 4 pigs per litter, 5 pigs per litter, etc. On the vertical or y-axis, the number of sows in the herd or group (frequency) that had litters of each size is listed. For example in herd A, 14 sows had a litter of 13 pigs born alive, but in herd B, there were only 6. What is most interesting about herds A and B is that they both have the same average for number of pigs born alive, 9.5. However, in herd A, the problem appears to be the result of a group or subset of sows that had very small litters of 3, 4, and 5 pigs. The rest of the herd appears to be very good, with litter sizes between 11 and 17. In contrast, in herd B, the problem appears to affect the majority of the herd because most sows have litter sizes of 10 pigs or less. In this particular example, investigations into the cause of low numbers of pigs born alive in these two herds would take different paths. In herd A, identification of the subset of sows should be the focal point, whereas in herd B, efforts should be directed toward an examination of standard management practices that affect the entire herd.

Situation 3-Good Farrowing Rate and Poor Number Born Alive

From a physiological perspective, when farrowing rate is good, this means that there were adequate numbers of embryos present on day 12 and between days 17 and 28 for both pregnancy signals to occur. Consequently, a logical assumption is that management during lactation, rebreeding, and the first month of pregnancy probably is good. The low number of pigs born alive most likely results from things that cause a high incidence of embryonic or fetal death after day 28. In this situation, the numbers of mummified fetuses and the numbers of stillborn pigs should be examined. If the number of mummified fetuses is high, then the problem likely occurred after day 50 of gestation. In contrast, if the number of stillborn pigs is high, then the problem probably occurred during the last week of pregnancy or during farrowing itself. As discussed in the previous section, subset analyses can be very helpful in troubleshooting situations in which farrowing rates are good, but litter size is not.

Low ovulation rates could create a situation in which the farrowing rate is good but the litter size is poor. However, the ovulation rate can’t be too low because sufficient embryos were present for both pregnancy signals to occur in most sows. In cases where low ovulation rate is involved, the cause usually is not related to short lactation lengths or an incomplete recovery of the sow’s reproductive system after weaning. If it were, then farrowing rates should also be low due to problems with uterine involution. Instead, the cause of low ovulation rate is more likely to be poor feed intake during lactation.

Situation 4-Poor Farrowing Rates with Regular Returns to Estrus and Good Number of Pigs Born Alive

Whenever numbers of pigs born alive is good and farrowing rate is poor, the problem nearly always is related to poor management in a group or subset of sows. Obviously, in the sows that did farrow, everything from a physiological standpoint worked well because they had large numbers of pigs. It is the sows that did not farrow that are causing the low farrowing rate. Consequently, the length of time it takes non-pregnant sows to return to estrus is a key piece of information. In the present situation in which sows show regular returns, the problem involves inadequate breeding regimens or high embryonic mortality during the first two weeks after breeding. Causes of high embryonic mortality during this period were discussed in Situation 1 and should be investigated.

If the problem seems to be breeding regimens, the timing of matings should be examined. The occurrence of ovulation relative to estrus varies considerably among and within herds. It also can vary from one estrus to another in the same sow. This natural variation in the timing of ovulation is one reason why sows normally are bred more than once during estrus. Skipping matings early in estrus and perform-ing additional ones late in estrus can create situations in which farrowing rate is poor. In both cases, ovulation has already occurred when the mating is administered, which either results in fertilization of degenerating eggs (skipped matings) or initiation of an unwanted inflammatory reaction by the uterus (late matings).

Situation 5-Poor Farrowing Rate with Irregular Returns to Estrus and Good Number Born Alive

The observation that non-pregnant sows return to estrus by 28 days or longer after being bred indicates that the first pregnancy signal on day 12 occurred without any problems. In other words, they were pregnant on day 12. However, some time after the first 2 weeks, pregnancy was terminated in a subset of animals. Possible causes of high embryonic mortality between days 12 and 28 should be examined. In addition, it is possible that a high incidence of fetal death later in pregnancy could also be involved. If this were the case, then the proportion of sows diagnosed as pseudopreg-nant or "not-in-pig" should also increase. As was discussed last month, once the second pregnancy signal occurs, sows usually maintain pregnancy to term unless something causes them to abort. If they lose all their viable fetuses after day 28, then they still act physiologically as if they are pregnant. Situations that cause high levels of fetal death after day 28 of pregnancy fortunately are rare. Consumption of mycotox-ins such as zearalenone and several reproductive diseases such as leptospirosis and parvovirus are two that have been linked to high incidences of fetal death and pseudopreg-nancy in sows. These situations can be identified with feed and serological analyses, respectively.


Reproductive problems in sows are the direct result of things that go wrong with the physiology of pregnancy. Because these processes occur in a coordinated and chronological sequence, examining farrowing rate, number of pigs born alive, and the length of time it takes non-pregnant sows to return to estrus can be used to estimate when during the reproductive cycle the problems are occurring. This information, in turn, can help producers identify problems and develop solutions. Table 2 shows locations of potential problems.

—Billy Flowers

Number of accesses to this file:
Last modified August 28, 2002.