The Challenge of Breeding Herd Management: Making the Best of a Bad Job

George R. Foxcroft Alberta Pork Research Centre / Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Alta, T6G 2P5, Canada.


The title to this paper may seem rather pessimistic. However, I've been in swine reproduction research for long enough to know that the challenges we face in breeding herd management are largely created by others! Despite the emergence of specific dam lines that have been selected for reproductive merit, the overpowering effect of ongoing aggressive selection for lean growth performance and feed conversion efficiency, has left us with important management problems in the nutrition of the replacement gilt and primiparous sow.

Our data suggest that body growth in the gilt is now proceeding more rapidly than the process of sexual maturation. Hence we have to deal with gilts that are getting heavier at sexual maturity under most management situations. This has consequences for lifetime maintenance costs in the breeding herd. Also, because gilts are at market weight before they are considered part of the breeding herd, there is a financial penalty to retaining gilts in the gilt pool, unless they are to be bred. Thus, however difficult it may be to induce estrus in these gilts and to have them conceive at the desired time, there is a great reluctance to cull them. In a recent presentation to swine veterinarians I was asked whether treatment with exogenous gonadoptrophins was efficacious for inducing anestrous gilts to start cycling. My response was that the logical choice was to cull the gilt, and that treatment of an apparently subfertile animal might, in all likelihood, require treatment of the same animal as a subfertile, weaned sow. This of course raises the question as to whether the early appearance of pubertal estrus is indeed related to lifetime fertility. I wish we had a clearer answer to this question. However I believe we should try and develop a view of gilt management that at least allows this possibility. This well be discussed in the first section of this paper.

In relation to the lactating and weaned sow, the most deleterious effect of selection for lean growth performance appears to be the reduction in voluntary feed intake during the first lactation. There are claims that with very aggressive feed management it is possible to prevent loss of body condition in some genotypes during the first lactation. However, most herds experience the problem of lactational catabolism and an associated reduction in fertility in the weaned first parity sow. In early weaning situations, the problem is further compounded by weaning sows in a relatively infertile period after farrowing. Thus, management decisions made in favour of better growth responses in the weaned pig, work against the achievement of good fertility in the breeding sow. The underlying basis of this conflict needs to be clearly understood if we are to develop flexible management systems for the lactating and weaned sow. These systems must also account for different genotypes in different management situations. Ultimately, it may be just as important to match the nutrition and genotype of the gilt and sow to the pattern of management, as it will be to match the nutrition and genotype of slaughter generation pigs to the end product required by the packer. The second section of this paper will therefore briefly review the important issues involved in developing better management of the lactating and weaned sow.

Future management of replacement gilts / Growth rate and time of first mating

In earlier studies with the Camborough genotype it was demonstrated that the minimum age at puberty was around 160 days and that this could be achieved at body weights of around 90 kg if gilts were fed appropriately. However, if potential replacement gilts are managed in the same way as finished pigs, then body weights at the onset of first estrus will be higher. With optimal use of boar stimulation it is probable that a considerable proportion of gilts can be induced to cycle even earlier than the herd average, perhaps as early as 140 days of age. If this can be achieved, then producers have the option of selecting gilts as replacements on the basis of early onset of estrus. Later responding gilts will still have been identified before they reach market weight and can be culled without penalty.

With this type of aggressive management, it will be also be possible to consider gilt synchronization procedures and other management techniques that may enhance first litter size. All of this can be achieved before target breeding weights are reached. Therefore the timed introduction of replacement gilts can be associated with a minimum number of non-productive days. As the non-productive days of replacement gilts can account for 30% or more of all the non production days of the breeding herd, this is extremely important to breeding herd efficiency.

Optimal use of boars in gilt management.

  • There is no replacement for direct boar contact as a stimulus for onset of puberty.

It has been shown consistently that direct access of one or more gilts to the salivary pheromones of a mature boar is the only effective stimulus for puberty onset. In this situation the boar is delivering a "primer" pheromone in his saliva, which is actually a modified male steroid hormone. In a direct contact situation, the gilts will actually solicit the boar and several gilts at one time can gain access to the salivary signal. Thus, even 10 to 15 minutes per day of direct contact between a group of 6 to 10 gilts and a mature boar will be effective for inducing puberty. It is not difficult to appreciate that even if gilts are allowed good fence-line contact with a boar, only one or two gilts at a time can gain access to the head of the boar, and this assumes that the boar is willing to stand at the fence line for long periods to provide this service! It has also been shown that some other component of boar presence is also contributing to the induction of puberty and this is clearly removed in the fence line situation.

The management of the gilt/boar interaction is greatly facilitated by the use of vasectomized boars of an appropriate size. The animals can essentially be left unsupervised for the period of daily stimulation. If breeding occurs this is an advantage.

  • Several studies have shown that breeding at the pubertal estrus improves conception rate and litter size born when gilts are bred with fertile boars or AI at second estrus.

Thus in our standard gilt management program, groups of 6 gilts are stimulated with a rotation of three vasectomized boars on a daily basis, and as far as possible all gilts are bred at first standing heat. In addition to improving gilt fertility, this also help to maintain high libido in the boars and prevents them from becoming overly aggressive when they are with the gilts. Gilts can then be grouped according to the time of first recorded estrus. This already starts to create potential breeding groups that can be efficiently observed for second estrus and bred when required.

Heat synchronization

If a large enough gilt pool is maintained, then gilts will probably be available to meet most of the breeding targets set. However, without other means of heat synchrony, it is virtually impossible to meet all breeding targets without expanding the size of the gilt pool and thus incurring many extra non productive days. Thus consideration can be given to possible management techniques that will produce heat synchronization.

  • Feeding of the oral progestagen, allyl trenbolone (Regumate) for 14 - 18 days in randomly cyclic gilts will result in gilts coming into their next heat over a 4 to 6 day period after the last day of feeding.

If the stage of the estrous cycle is known, the number of days that Regumate needs to be fed can be considerably reduced, without loss of efficacy. The fertility of the gilts is not affected by this treatment and indeed may be improved. We find this particularly true for AI use, because fresh semen can be ordered in a very predictable way and we can concentrate on gilt breeding over a concentrated period of time.

  • Some degree of synchrony can be achieved with the use of prostaglandin F2a (PGF), the luteolysin used extensively in cattle.

However, the corpora lutea of the pig are only sensitive to PGF from day 12 of the cycle. At most, estrus can therefore be advanced by some 5 days. Nevertheless, this may still be useful in bringing a number of gilts into a tighter breeding group.

  • Some producers are reported to use fertile breeding and prostaglandin-induced abortion as a means of achieving synchronization.

This may produce ethical concerns, but the advantage is that gilts can be bred randomly and then breeding groups created by treating any gilts in the early stages of pregnancy (days 12 to 35) with a single injection of PGF.

  • An alternative strategy under consideration is the combined use of exogenous gonadotropins, like PG600, to induce ovulation in gilts, followed by estrogen treatment about 11 days later to induce a pseudopregnancy. This could then also be terminated at any stage with PGF.

Thus management systems can be developed to induce onset of heat at younger ages and at lighter weights. This will improve the efficiency of replacement gilt management and will be particularly applicable on large units where replacement gilt management can be centralized and managed by specialized staff. Gilt rearing facilities should be properly designed to allow effective use of the boar, which I see as an indispensable part of the program. In the future, I envisage that careful management of the health status of cooperative enterprises will allow for gilts to be bred to specified breeding targets in a central facility and then distributed to farrow-to- weaning units as pregnant females. These gilts will already have been selected on the basis on sexual precocity and acceptable breeding performance at first estrus. Furthermore, by having been selected into the gilt replacement program at an early age, their nutrition will have been managed to reduce weight at breeding and therefore reduce the lifetime maintenance costs of the breeding herd.

  • If really efficient gilt pool management can be achieved, and the fertility of the early weaned, first parity sow becomes an increasing problem, then we might again want to look at the concept of single cycle gilt production as a serious option.

Future management of the lactating and weaned sow

In seeking improved management options for early weaned sows we must be very aware that management decisions in large breeding herds need to take account of work practices. Management systems should facilitate those aspects of good animal husbandry that are critical to optimal fertility. In this respect it may be as important to ensure that weaned sows can be bred over a specific period of days, than to reduce the average weanling to estrous interval by one day. In other words, management systems have to work in practice as well as on the computer screen! We must also accept that the first parity sow is inherently more susceptible to the challenges of lactation.

  • Therefore a major focus on the management of the first parity female will pay the greatest dividends in improving breeding herd performance.

To achieve optimal fertility in the weaned sow we need to overcome,

  1. The inherent infertility of the sow in the early period after farrowing
  2. Negative effects of lactational catabolism when feed intake does not meet the demands of milk production

The best way to achieve these objectives may be to,

  1. Wean the first parity sow early in lactation but then block the return to estrus in some controlled way.
  2. Use more innovative management of the nutrition of the gestating/lactating first parity sow to improve her fertility after weaning.

To justify these suggestions I will briefly review information on three aspects of the biology of the sow that are relevant to early weaning systems.

  • Effects of time after farrowing on the function of the reproductive system
  • Suckling as the primary block to reproduction during lactation
  • Effects of metabolic state during lactation and after weaning on sow fertility

This review will draw on information accumulated on the biology of the lactating and weaned sow over the last 30 years, and the results of research at the University of Alberta that has looked at the interaction between the metabolic state of the sow during lactation and her fertility after weaning.

Effects of time after farrowing on the reproductive system

  • There is an unavoidable period of infertility in the early postpartum period

Pregnancy in large mammals requires massive adaptation on the part of the mother to provide for the needs of the fetus. This long term commitment terminates in the dramatic changes associated with parturition. Not least of these, are the hormonal changes that control the delivery of the young, and the onset of milk production and nursing behaviour. In the domestic pig, and in women, the metabolic demands of lactation, on top of the earlier metabolic demands of gestation, are associated with a deep lactational anestrus. This allows the mother to use her body reserves to attend to the nurturing of her young which are born in a relatively immature state. This period also allows the reproductive system to become fully functional again. In the sow it has been possible to record this recovery in different ways.

Uterine involution

This term describes the return of the uterus to its non-pregnant state. Although information on uterine involution in the sow is not extensive, the data available suggest that involution is finally complete around 21 days postpartum. Both genotype and suckling affect the rate of uterine involution. As suckling has been shown to be a critical stimulus or rapid uterine involution in the pig and comparable species, we must assume that early weaning in the sow (say at less than 10 days) will delay uterine involution.

  • Early weaning studies clearly indicate that complete uterine involution is not essential for the next pregnancy to be established. However, it is generally accepted that incomplete uterine involution is probably one factor that contributes to poor embryonic survival and hence reduced litter size born in early weaned sows.

Decovery of the brain and pituitary

As the sow comes into estrus, the secretion of the pituitary gonadotrophic hormones, Luteinizing Hormone (LH) and Follicle Stimulating Hormone (FSH) is responsible for the growth of ovulatory follicles and the number of ovulations that occur (ovulation rate). A major "surge" in LH is also needed to actually cause ovulation. Early in lactation, the secretion of LH and FSH is limited and this is another factor contributing to the infertility of the early weaned sow. Thus, not only the uterus, but also the central components of the reproductive system (the brain and pituitary) need some time to recover after farrowing.

  • The need for the reproductive system to recover from the effects of pregnancy partly explains the reduction in sow fertility as weaning age falls below 14 days.

Unfortunately, there is little evidence that any of these limitations in the early postpartum period can be rectified with hormonal or any other treatments. Indeed, early weaning is likely to delay recovery of the reproductive system after farrowing.

Suckling as the primary block to reproductive activity in lactation

Active LH and FSH secretion can be observed in sows immediately after farrowing and this period has therefore been described as the "Hypergonadotrophic" phase of lactation. If the litter remains with the sow, this uninhibited period of LH secretion is eventually suppressed around day 3 of lactation, as the inhibitory effects of suckling become established at the level of the brain and pituitary. This would then represent the "Hypogonadotrophic" phase. We have clearly demonstrated the importance of suckling as the primary inhibitor of LH secretion. If the litter is removed from the sow at farrowing, active LH secretion is maintained and ovarian follicular development occurs. This can be associated with observed estrus. However, in many cases the pattern of follicular development is abnormal and ovulation does not occur. As lactation progresses, there appears to be a gradual increase in LH secretion (a "Recovery" phase), as long as the sow does not become seriously catabolic.

The pattern of follicular development observed during lactation reflects the pattern of gonadotrophin secretion. A number of large ovarian follicles may be present immediately after farrowing, but a week later follicular development is minimal. Then, as lactation progresses, there is a gradual increase in the number of medium to large sized follicles.

Partial removal of the inhibitory effects of suckling by reducing the size of the litter suckling (split or partial weaning) only produces a transient increase in LH secretion; however, there are lasting effects on ovarian development.

  • This suggests that manipulation of litter size before weaning can be an effective management tool for increasing ovarian development and sow fertility.

Because the gonadotrophin response to split weaning is so transient, it has been suggested that split weaning should only occur for 2 to 3 days before final weaning. However, we believe this recommendation needs further evaluation, and any period of reduced suckling input may give beneficial results.

In the sow, therefore, a lack of ovarian activity in early lactation appears to be almost entirely suckling dependent and the suckling-induced inhibition of LH secretion in early lactation is the primary cause of lactational anestrus. Unfortunately, although we understand something of the mechanism by which suckling acts on the brain to inhibit LH secretion, it has not been possible to develop a practical treatment to reliably induce a fertile estrus during lactation. Even if we could, the detrimental effects of poor metabolic state in the lactating, primiparous sow would probably still result in reduced ovarian responses to such treatments and reduced fertility.

An immediate increase in LH secretion is invariably observed in response to weaning and is the stimulus for ovarian follicular development. The magnitude of the increase in LH secretion after weaning has not been consistently related to the weaning-to-estrous interval or to ovulation rate at first post-weaning estrus. However, a number of studies have reported a relationship between the pattern of LH secretion during lactation and fertility after weaning. These data suggest that the pattern of hormone secretion during lactation can have a lasting effect on the ovary.

  • Evidence that changes in LH and FSH secretion in late lactation and immediately after weaning affect sow fertility, may explain the efficacy of treating sows with low doses of exogenous gonadotrophins (products like PG600) at weaning.

Overall, therefore,

  • Suckling is the primary inhibitor of the reproductive system during lactation. Therefore, the only effective way to bring the sow into estrus is to find some way of blocking, or reducing, the effects of suckling.
  • As suckling intensity affects the pattern of hormone secretion during lactation, and this in turn may predetermine the fertility of the sow after weaning, manipulating suckling intensity may be important in achieving the overall goal of maximum pigs/sow/year.
  • Because of the inhibitory effects of suckling, on top of the effects of postpartum recovery, there is a period in early lactation when the reproductive system is very inactive. Weaning during this period will result in low fertility. Unfortunately this period coincides with that selected as optimal for SEW systems!

Effects of metabolic state during lactation and after weaning on sow fertility

  • Loss of body condition in lactation can have very negative effects on sow fertility after weaning.

Nutrition and metabolic state during lactation produce major effects on ovarian follicular development. It is not surprising therefore, that nutritionally-mediated differences in sow fertility after weaning are associated with differences in LH secretion in lactation. Recent studies determined whether the pattern of change in the body condition and metabolic state of the sows, as well as the absolute amount of body tissue loss, affected subsequent fertility. Notwithstanding the dominant inhibitory effect of suckling, differences in the pattern of feed intake during lactation produced significant effects on episodic LH secretion during lactation and on post- weaning fertility (see Table 1).

Table 1. Effects of Pattern on Feed Intake Over a 28-day Lactation on Postweaning Fertility in Primiparous Sows Bred at First Postweaning Estrus.

Group AA Group AR Group RA
Weight loss in lactation (kg) 11.0a 21.12b 24.75b
Backfat loss in lactation (mm) 2.19c 4.61b 5.38b
Ovulation rate 19.86c 15.44b 15.43b
Embryo survival (%) to day 28 87.53a 64.43b 86.50a
Weaning to estrous interval (days) 3.7a 5.1b 5.6b

Superscripts b and a denote treatment differences at P<.002 and c denotes differences at <.05. (L.J. Zak, PhD Thesis, University of Alberta, 1997)

  • These results clearly indicate that the pattern of metabolic change during lactation can affect different components of sow productivity (ovulation rate and embryonic survival, and thus potential litter size; weaning to service interval and thus non-productive days).

In addition to our own research, work from the University of Minnesota also led to the hypothesis that restricted feed intake and catabolism in early lactation, and the associated reduction in LH secretion, had lasting consequences for the fertility of the sow after weaning. The results of this group led to the further suggestion that reductions in feed intake in the first or second week of lactation has a more profound effect on sow fertility than reductions in feed intake in later lactation. However, the extensive studies of our own group on the endocrine mechanisms mediating effects of nutrition on ovarian function in the gilt, led us to conclude that a period of increased catabolism immediately before weaning in the sow would have the most negative impact on subsequent fertility. Also, we know that effects of metabolic state on subsequent fertility involve both central (gonadotrophin-mediated), as well as local ovarian, effects. Other relatively short-term effects of nutritional state on ovarian function and fertility can be found in work on nutritional flushing in the gilt and in the studies on the influence of energy and insulin treatment on the final stages of ovarian follicular development before ovulation.

  • Relating these results to the sow therefore, we suggest that increasing catabolism in late lactation, which will carry over into the period after weaning when ovarian development occurs, will be most detrimental to fertility.

By considering the overall mechanisms mediating the effects of nutrition during lactation on postweaning fertility, we have attempted to reconcile these apparently conflicting ideas. it has long been established that one potential mechanism by which nutrition affected ovulation rate was by changing the number of maturing follicles available for recruitment into the ovulatory population. Thus in the sow, similar mechanisms will affect the size of the pool of follicles at the time of weaning. Additionally, the sensitivity of the ovary to gonadotrophin stimulation immediately after weaning, and hence the number of follicles that are finally ovulated, will be affected by both the previous and current metabolic status of the sow. Thus, both before and after weaning, nutritionally-mediated effects acting centrally (gonadotrophin mediated) and at a local ovarian level may affect sow fertility. For example, reported associations between nutritionally mediated differences in insulin status during lactation and subsequent fertility likely involve direct ovarian effects. The "imprinting" of follicles in early lactation may have other, lasting, consequences for follicle development at weaning. We have shown that the quality of follicles that grow and ovulate is important for normal oocyte maturation. Differences in follicle quality at weaning, due to different metabolic states during lactation, may therefore affect the quality of the oocytes ovulated at the postweaning estrus, and be an important factor determining subsequent embryonic survival. In our latest studies in the lactating, primiparous sow, we have obtained data to support this suggestion.

  • Therefore, different mechanisms exist by which changes in metabolic state at all stages of lactation affect the fertility of the sow after weaning.
  • Although the negative effects of increased catabolism in early lactation on embryonic survival after mating could be reversed by feeding to appetite immediately before weaning, the effects on ovulation rate were not reversed.
  • Further study of the effects of metabolism on different components of the sow's fertility will help us to develop innovative ways to improve fertility.

In the context of very early weaned sows, there is an additional question: Are follicles that ovulate after weaning already affected by the metabolic state of the sow during late gestation? It has been proposed that the final growth of follicles in the pig to ovulatory size covers a period of about 19 days. Thus, as many conventionally managed sows become catabolic in late gestation, negative effects on fertility of early weaned sows may already occur before farrowing!

The data in Table 1 illustrate another important feature of recent studies of the primiparous lactating sow. Even when fed to appetite throughout lactation, these sows still lost 11 kg of bodyweight and 2 mm of backfat over a 28-day lactation. The poor appetite of many primiparous sows therefore results in tissue catabolism during lactation, whereas higher parity sows are able to satisfy the demands of lactation through adequate feed intake.

  • Differences in metabolic state of sows of different parities underlies the need to consider different management strategies for primiparous, as compared to multiparous, sows.

Thus it is imperative to understand the characteristics of individual genotypes in the farm environment in order to make informed decisions about the management of the early weaned sow.

Possible management strategies for the lactating and weaned sow

First and foremost we must define the response of particular genotypes to particular patterns of management; in the context of current trends in the industry, their response to weaning early weaning systems. Only then can we make sensible suggestions about strategies to improve weaned sow fertility. We can consider a number of possible situations, in each case assuming that we are dealing with primiparous sows weaned at less than 14 day (i.e. the worst case scenario). Although these scenarios may not relate directly to the management of most herds in western Canada, they focus on the type of problems that may be seen in first parity sows weaned after even longer lactations. The following represent the situations that might be encountered:

Scenario 1: Sows with few problems of adequate voluntary feed intake during lactation, and hence with lactational catabolism, but with a relatively short weaning-to-estrous interval.

Scenario 2: Sows with few problems of voluntary feed intake and lactational catabolism, but with a longer and more variable weaning-to-estrous interval.

Scenario 3: Sows that become catabolic in lactation, but nevertheless have a short weaning-to-estrous interval. (University of Alberta type sows)

Scenario 4: Sows that become catabolic in lactation and but show a longer and more variable weaning-to-estrous interval. (University of Minnesota type sows)

It is likely that in all genotypes, early weaned sows will show some increase in the weaning-to- estrous interval and that this will be more variable. Additionally, most primiparous sows will show reduced fertility in terms of conception rate and litter size born, and overall productivity of the sow in terms of pigs/sow/year will fall. Depending on the genotype and its response to early weaning, this may be associated with increased early embryonic mortality after weaning, a decrease in ovulation rate, or both.

In relation to the earlier sections of this paper, we believe that by weaning at 14 days or less, a part of the problem will be caused by the short postpartum interval. On top of this there may also be problems of the sow becoming catabolic by the second week of lactation

  • Our thinking therefore is that a delay in the return to estrus is required to ensure improved fertility after weaning early. We predict that the extra non-productive days and a slight decease in litters/sow/year will be more than offset by an increase in litter size born.
  • Despite the impressive response of primiparous sows to "skip-a-heat" breeding, we do not believe that the 21 non-productive days incurred will be economically acceptable.
  • If early weaned sows show a prompt and fairly synchronous return to after early weaning, then a shorter "skip-a-heat" could be achieved by using prostaglandin F2a treatment on day 12 or 13 after first estrus to regress to corpus luteum and shorten the first estrous cycle by 3 to 5 day. This would also likely result in some degree of estrus synchronization at second heat.
  • If sows show an extended, but predictable, weaning-to-estrous interval, then the best option may be to simply breed at this estrus and accept that fertility may be reduced compared to higher parity sows.
  • If there is either an unacceptable increase in the weaning-to-estrous interval, or problems with variability, then treatment with PG 600 at, or on the day of weaning, can be effective in increasing the number of sows bred by a specific target day.
  • If an inconsistent pattern of weaning-to-service intervals disrupts breeding programs, or if a delay to first breeding is needed to enhance overall fertility, then another strategy is to feed sows the oral progestagen Regumate from the day of weaning until 5 days before sows are required to be bred. This will block the onset of estrus and provide good heat synchrony after withdrawal.
  • Another experimental treatment being tried would achieve the same objective. In this protocol sows are induced to ovulate immediately after farrowing by injecting HCG during the "hypergonadotrophic" phase described earlier. If effective, this would result in an induced estrous cycle during which estrus is blocked by the progesterone produced by the sow's own corpora lutea. If no other treatment is given, sows should return to estrus around day 22 after farrowing, and can be weaned at any time before this. An earlier, synchronous return to estrous should also result from PGF treatment of these sows from 12 days after initial HCG treatment.
  • In all early weaned situations producers should consider the potential advantages of enhancing fertility with split weaning. The number of days over which to reduce litter size before final weaning needs evaluation in each management system.
  • Finally, producers should always try and design their breeding barns to take advantage of the stimulatory effect of boars in reducing the weaning-to-estrous interval and improving fertility. As the boars are likely present anyway, it makes good sense to design facilities to take advantage of this extra contribution to breeding herd performance.