Consequences for Insemination Strategies

Bas Kemp, Dorothé W.B. Steverink and Nicoline M. Soede

Wageningen Institute of Animal Sciences, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen, The Netherlands

Introduction
The life span of eggs after ovulation and the life span of a sufficient number of sperm cells capable of fertilization within the oviduct limits the period in which inseminations can lead to successful fertilization, relative to ovulation (Hunter, 1994). In recent years (especially with the development of ultrasound techniques to study the timing and process of ovulation) more quantitative data have become available on the consequences of variation in insemination-to-ovulation interval on reproductive parameters. This review describes effects of variation in insemination-to-ovulation interval on fertilization in the sow. Also, factors that influence the limits of the optimal insemination-to-ovulation interval (in terms of fertilization) are discussed. Knowledge on the range of insemination-to-ovulation intervals that lead to successful fertilization is important because it will set the accuracy with which insemination should be timed relative to ovulation. Together with information on the possibilities to predict the time of ovulation during estrus this can lead to more knowledge of the effects of insemination strategies on reproductive performance. Therefore, this review also discusses possibilities to predict ovulation. In spontaneously ovulating sows, duration of ovulation ranges between 1 and 3 h (Soede et al., 1992); since this is a rather short period, we refer to it as ovulation moment rather than duration of ovulation in this review.

Effects of Insemination-to-Ovulation Interval on Partial Fertilization and Embryonic Development

To study effects of insemination-to-ovulation interval on subsequent fertilization, Soede et al. (1995a) inseminated sows at various times relative to ovulation (as detected with transrectal ultrasound) and slaughtered the animals at about 120 h after ovulation to study fertilization rate and embryonic development. Embryos were morphologically assessed and subsequently processed for counting the number of nuclei (to calculate the number of cell cycles) and spermatozoa bound to the zona pellucida (accessory sperm count). Oocytes were classified as unfertilized if the nuclei count was zero or one. Embryos with degenerated morphology and a low number of nuclei were classified as degenerated; the remaining embryos were considered to be normal. Fertilization rate was defined as the percentage of normally developing embryos relative to all embryos and oocytes recovered. Sows were assigned on the basis of fertilization rate to classes varying from no fertilized eggs at all to total fertilization of all ovulated eggs. The number of degenerated embryos was about 4% irrespective of interval between insemination and ovulation. In Figure 1, the percentage of sows within each class of fertilization rate is shown for different insemination-to-ovulation intervals. The percentage of sows with partial or no fertilization increases when insemination takes place more than 24 h before ovulation or directly after ovulation. In these sub-optimal periods a large variation between sows can be seen. Even when inseminating more than 40 h before ovulation, 15% of the sows showed a 100% fertilization.

Collective data from three experiments (Soede et al., 1995a, b and Steverink et al., 1997) show that litters with partial fertilization also have a slightly retarded embryo development and increased variation in embryonic development at day 5 of pregnancy (Kemp and Soede, 1997). Whether the effect on embryo development has an impact on embryonic mortality remains to be investigated.

The Limits of the Insemination-to-Ovulation Interval

Studies in which ultrasound was used to monitor the time of ovulation in relation to a large range of insemination-to-ovulation intervals (Waberski et al., 1994 b; Soede et al., 1995a, b) led to the conclusion that inseminations can be performed between 0 and 24 h before ovulation with no significant negative effects on fertilization rate. Nissen et al. (1997) found optimal results in terms of numbers of Day 16 embryos and farrowing rate and litter size in sows inseminated between 28 h before ovulation to 4 h after ovulation (as detected by transrectal ultrasound).

From the above mentioned literature it has become clear that fertilization results are dependent on insemination-to-ovulation interval. The general trend suggests that when fresh semen is used at an adequate dose (2-3 x 10⁹ sperm cells), the optimal insemination-to-ovulation interval is between 0 and 24 h before ovulation. It is, however, clear that some sows have good fertilization results outside these limits and others sows have bad fertilization results inside these limits. The semen factors and sow factors that may influence the optimal period for insemination relative to ovulation are listed below.

Semen Characteristics

Experiments of Soede et al. (1995a, b) have been performed using 3 x10⁹ sperm cells and Nissen et al. (1997) used 2 x10⁹ sperm cells. Steverink et al. (1997) studied whether the optimal period between insemination and ovulation is influenced by sperm dosage (using 1, 3, or 6 x10⁹ sperm cells). Results showed that the optimal period for insemination of 0 to 24 h was hardly influenced by the dosages as applied in the study. This indicates that the life span of a functional sperm reservoir is not very sensitive to variation in the number of sperm cells inseminated, at least within the range of 1 to 6 x10⁹ sperm cells.

In Soede et al. (1995a, b), semen age at insemination varied from 12 to 38 h (stored in BTS at 17^oC) and was not related to reproductive performance or accessory sperm count. However, Waberski et al. (1994b) showed that liquid storage of semen (BTS, 17^oC) for 48 to 87 h resulted in a significant reduction, from 80 to 54%, in fertilization rate in sows that were inseminated between 12 and 24 h before ovulation. Storage for 87 to 120 h did not result in a further reduction in fertilization rate (fertilization rate was on average 50%), but did result in a (small) reduction in fertilization rate in sows that were inseminated between 0 and 12 h before ovulation (73% compared to 85% in sows that were inseminated with semen younger than 87 h). Storage of semen therefore limits the life span of sperm cells in the female reproductive tract and consequently shortens the optimal period for insemination relative to ovulation. The extent of this effect of storage is probably dependent on several factors such as the type of extender, the quality of semen, the number of sperm cells and the 'quality' of the sows. For example, in gilts that were inseminated after ovulation, Waberski et al. (1994b) found a reduction in fertilization rate when semen was stored in BTS for more than 24 h and in Androhep for more than 48 h. Therefore, effects of storage of liquid semen on the optimal period of insemination relative to ovulation were dependent on the type of extender used.

In future, frozen boar semen may be available for practical use. However, Waberski et al. (1994a) showed that fertilization results using frozen boar semen are optimal in only a limited period (between 0 and 4 h before ovulation). For very specific reasons, frozen semen could be used in combination with hCG-timed ovulation. Nevertheless, new freezing processes may lead to the wider use of frozen boar semen.

Unfortunately, no information is available regarding the optimal period for natural mating in relation to ovulation. Since natural mating results in much more stimulation of the sow than insemination, affecting both sperm transport and ovulation (Soede, 1993), a positive effect of mating as compared to insemination may be anticipated. However, there are indications that under certain (socially limited) housing conditions of sows, mating behaviour of boars may also have a negative influence on fertility (Hemsworth et al., 1978; Soede, 1993), possibly related to fear of the 'unknown'.

Sow Characteristics

Not much is known about sow characteristics (e.g., parity, breed) that may influence the optimal period between insemination and ovulation. The sows in the study of Soede et al. (1995a, b) were from three different genetic lines that were grandparent lines of the fattening pigs; one sow line (selected mainly for fertility traits) and two boar-lines (selected mainly for growth and meat traits). In these genetic lines, the percentage of good embryos in the sows that were inseminated within 16 h after ovulation was 95 (n=16), 66 (n=24), and 61 (n=16), respectively (LSM, line: P=0.03, experiment: n.s.). These data suggest that oocyte survival, sperm transport and/or sperm capacitation may vary between different breeds and consequently may influence the optimal period for insemination relative to ovulation.

Estrus and Ovulation

Insemination more than 24 h before ovulation or insemination after ovulation result in a reduced fertilization rate and consequently a reduction in farrowing rate and litter size. Therefore, an important question for optimal estrus detection and insemination management is; which characteristics of estrus gives the best prediction of the time of ovulation? Factors that can be measured objectively, like changes in vaginal mucus conductivity (Stokhof et al., 1996) or changes in body temperature (Soede et al., 1997) around estrus are poor predictors of ovulation.

Up to the early nineties, sows were thought to ovulate at a relatively fixed time, 38 to 40 h after onset of estrus. However, studies in which ovulation was assessed with ultrasonography, have shown that ovulation takes place at very variable times after the onset of (spontaneous) estrus in German, Dutch, Danish and Swedish experiments (Weitze et al. 1994; Soede et al., 1992, 1995a, b; Nissen et al. 1997; Dalin et al. 1995, respectively). Although the average ovulation time is not very variable between the experiments (between 35 and 48 h), the ovulation time of individual sows varies between 10 and 85 h after onset of estrus. Therefore, although in practice, onset of estrus is the parameter used to assess insemination time, it is not a good predictor of the moment of ovulation.

However, when time of ovulation in these studies was expressed in terms of the interval between onset of standing heat in response to a boar, as a percentage of the total length of the estrus period, ovulation occurred between 64 and 72% of the way through the heat period. Furthermore, in these experiments, the duration of estrus explained as much as 50 to 60% of the variation in ovulation time after the onset of estrus (Figure 2). The variation in ovulation time between individual sows still seems quite large (between 39% and 133%), but this is primarily caused by those sows with a short duration of estrus. Thus, the relative timing of ovulation during estrus seems quite constant. Although Weitze et al. (1990) showed that insemination with seminal plasma at onset of estrus may advance ovulation by up to 14 h, the effects on the relative timing of ovulation were limited (ovulation at 59% of estrus), because the gilts concomitantly show a reduction in duration of estrus averaging 8 h. Further studies have shown that the advancement of ovulation is caused by estrogen and a peptide fraction in the seminal plasma that somehow result in a reduction in the interval from LH to ovulation (see review by Waberski, 1997).

From the above, it is clear that at the moment, the duration of boar-detected estrus seems the best predictor for ovulation time during estrus. Unfortunately, the duration of estrus is highly variable and gives only a retrospective estimate of the time of ovulation. Therefore, research should be aimed at finding better prospective estimators for the timing of ovulation. The ultimate prediction would be to be able to monitor the rise in LH levels in a simple way, but for pigs this may still be impracticable. Possibly, a combination of parameters (maybe both physical and behavioural) can give a better estimate of ovulation time than only one parameter. However, at the moment, no such combination of parameters is known.

Factors Affecting the Duration of Estrus

As already shown above, a large variation is found in the duration of estrus between sows and between experiments. Many factors may influence the duration of estrus like genetic background, stress, season, and differences in boar stimulation. Literature studying these factors is scarce and sometimes contradictory and therefore this field needs more attention (Soede and Kemp, 1997).

Studying field data from 52 Dutch farms collected for 3 to 18 months revealed significant and consistent (from month to month) differences in mean estrus duration varying from 31 to 65 h between individual farms (In Figure 3, data of 13 farms are presented). These data also showed that on all farms a consistent decrease of duration of estrus with an increase in weaning to estrus interval (WEI) up to approximately 6 days (Steverink et al., unpublished results). Similar effects have been reported elsewhere (Rojkittikhun et al., 1992; Weitze et al., 1994; Kemp and Soede, 1996). This influence is quite large; an increase in WEI of 3 days resulted in an average decrease in duration of estrus of 24 h (Kemp and Soede, 1996). The decrease in duration of estrus also resulted in a decrease in the interval from onset of estrus to ovulation (Weitze et al., 1994; Kemp and Soede, 1996). Kemp and Soede (1996) showed that the percentage of sows that ovulated within 32 h from onset of estrus increased from 21%, 24%, 50% to 72% for sows in estrus on days 3 to 6 after weaning, respectively.

Consequences of Variation in Timing of Ovulation After Onset of Estrus for the Insemination Strategy

As is clear from the outline above, sows show a lot of variation in timing in ovulation after onset of estrus, both between and within farms.

Reducing the variation in timing of ovulation can be done by control and synchronization of ovulation after weaning using hormones like PMSG, hCG, and GnRH (Brüssow et al., 1996). This allows fixed time AI. One should bear in mind, however, that the success rate of such systems is highly dependent on the management system used, that it might mask reproductive problems due to poor management on a farm, that these systems can give rise to ethical discussions, and that the use of exogenous hormones is costly.

Another way to deal with that variation in terms of an insemination strategy is to start inseminating at the first signs of heat and repeat inseminations every 24 h until the end of estrus. In addition to such an approach being time consuming and expensive, there is also a question whether such an intense insemination scheme could have negative effects. Soede et al. (1995b) showed that a second insemination within 4 h after ovulation (first insemination was performed before ovulation) did not affect fertilization rate in sows. Therefore, post-ovulatory inseminations do not seem to have a negative impact on the fertilization process. However, Rozeboom et al. (1997) concluded that late estrus or met-estrus inseminations after estrual inseminations decreased farrowing rate and litter size. De Winter et al. (1992) showed that females inoculated with bacteria during late estrus or met-estrus were more susceptible to uterine infections as compared to sows inoculated during early or mid estrus. It seems that very intensive insemination strategies can, therefore, give negative reproduction results.

A better approach could be to take in account farm data on duration of estrus (in relation to of weaning to estrus interval), to use this information to estimate the moment of ovulation and then pinpoint the insemination strategy based on that knowledge. This approach requires an accurate detection of the start and end of the estrus period on an individual sow basis. With such an approach probably one or two inseminations in a heat period could be sufficient. At our lab, Steverink is working on a simulation model including factors influencing the optimal insemination to ovulation interval and factors that can be used to predict ovulation. The simulation model can be used to study consequences of insemination strategies (including timing of insemination, age of semen, number of sperm cells used, etc.) on fertilization. Such analyses may further increase the efficiency with which semen is used for AI.

Summary

In sows, insemination between 0 and 24 h before ovulation results in high fertilization rates and consequently, a low number of rebreeders and a slightly higher litter size. Outside this optimal period for insemination, the percentage of sows with poor fertilization results increases, especially due to increased partial fertilization. Sow factors such as parity and breed, and semen factors such as quality and number of sperm cells, may influence the length of the period for insemination relative to ovulation in which optimal fertilization rates are found. However, in general, one may conclude that insemination should occur between 0 and 24 h before ovulation to result in optimal fertilization results.

In pig husbandry, timing of insemination is based on onset of estrus. Unfortunately, this does not accurately predict the time of ovulation, as ovulation time varies between 10 and more than 60 h after the onset of estrus. At the moment, the best 'predictor' for ovulation time is the duration of estrus: irrespective of the duration of estrus, ovulation takes place at a relatively fixed time 70% of the way through estrus. Therefore, knowledge of factors that influence the duration of estrus may assist in adjusting the insemination strategy. The duration of estrus varies considerably, both between farms and within farms. The following factors have been found to influence the duration of estrus (and consequently the optimal time for insemination): farm, boar-stimuli, stress, and weaning-to-estrus interval. An efficient insemination strategy could be achieved when taking in account factors influencing the optimal insemination-to-ovulation interval and factors that predict ovulation.

References

Brüssow, K.P., W. Jöchle and U. Hühn. (1996) Control of ovulation with a GnRH analog in gilts and sows. Theriogenology, 46: 925-934.

Dalin, A.M., T. Nanda, F. Hulten and S. Einarsson. (1995) Ovarian activity at naturally attained estrus in the sow: ultrasonographic and LH Study. Acta Veterinaria Scandinavia 36: 377-382.

De Winter, P.J.J., M. Verdonck, A. de Kruif, L.A. de Vriese and F. Haesebrouck. (1992) Endometritis and vaginal discharge in the sow. Animals Reproduction Science, 29: 51-58.

Hemsworth, P.H., R.G. Beilharz and W.J. Brown. (1978) The importance of the courting behaviour of the boar on the success of natural and artificial matings. Applied Animal Ethology 4: 341-347.

Hunter, R.H.F. (1994) Causes for failure of fertilization in domestic species. In: Embryonic mortality in domestic species. MT Zavy and RD Geisert (editors), CRC-press, Boca Raton, p 1-22.

Kemp, B. and N.M. Soede. (1996) Weaning to estrus interval in relation to timing of ovulation and fertilization results in sows. Journal of Animal Science 74: 944-949.

Kemp, B. and N.M. Soede. (1997) Timing of insemination relative to ovulation. Journal of Reproduction and Fertility Supplement 52: In press.

Nissen, A.K., N.M. Soede, P. Hyttel, M. Schmidt and L. D'Hoore. (1997) The influence of time of insemination relative to time of ovulation on farrowing frequency and litter size in sows, as investigated by ultrasonography. Theriogenology 47: 1571-1582.

Rojkittikhun, T., M. Sterning, L. Rydmer and S. Einarsson. (1992) Estrus symptoms and plasma levels of oestradiol-17? in relation to the interval from weaning to estrus in primiparous sows. Proceedings of the 12th Congress of the lnternational Pig Veterinary Society 485 (Abstract), The Hague, The Netherlands.

Rozeboom, K.J., M.H.T. Troedsson, G.C. Shurson, J.D. Hawton and B.G. Carbo. (1997) Late estrus or metestrus insemination after estrual inseminations decreases farrowing rate and litter size in swine. Journal of Animal Science, 75: 2323-2327.

Soede, N.M., B. Kemp and J.P.T.M. Noordhuizen. (1992) The duration of ovulation in pigs, studied by transrectal ultrasonography, is not related to early embryonic diversity. Theriogenology 38: 653-666.

Soede, N.M. (1993) Boar stimuli around insemination affect reproductive processes in pigs: a review. Animal Reproduction Science 32: 107-125.

Soede, N.M., C.C.H. Wetzels, W. Zondag, M.A.I. de Koning and B. Kemp. (1995a) Effects of time of insemination relative to ovulation, as determined by ultrasonography, on fertilization rate and accessory sperm count in sows. Journal of Reproduction and Fertility 104: 99-106.

Soede, N.M., C.C.H. Wetzels, W. Zondag, W. Hazeleger and B. Kemp. (1995b) Effects of a second insemination after ovulation on fertilization rate and accessory sperm count in sows. Journal of Reproduction and Fertility 105: 135-140.

Soede, N.M. and B. Kemp. (1997) Expression of oestrus and timing of ovulation in pigs. Journal of Reproduction and Fertility Supplement 52: In press.

Soede, N.M., W. Hazeleger, J. Broos and B. Kemp. (1997) Vaginal temperature is not related to the time of ovulation in sows. Animal Reproduction Science 47: 245-252.

Steverink, D.W.B., N.M. Soede, E.G. Bouwman and B. Kemp. (1997) Fertilization results as influenced by insemination-ovulation interval and sperm dose in sows. Journal of Reproduction and Fertility: In press.

Stokhof, S., N.M. Soede and B. Kemp. (1996) Vaginal mucus conductivity as measured by Walsmeta MKIV does not accurately predict the moment of ovulation or the optimum time for insemination in sows. Animal Reproduction Science 41: 305-310.

Waberski, D., K.F. Weitze, T. Gleumes, M. Schwartz, T. Willmen and R. Petzoldt. (1994a). Effects of time of insemination relative to ovulation on fertility with liquid and frozen semen. Theriogenology 42: 831-840.

Waberski, D., K.F. Weitze, C. Lietmann, W. Lubbert zur Lage, F.P. Bortolozzo, T. Willmen and R. Petzoldt. (1994b) The initial fertilizing capacity of long term stored liquid semen following pre- and postovulatory insemination. Theriogenology 41: 1367-1377.

Waberski, D. (1997) Effects of semen components on ovulation and fertilization. Journal of Reproduction and Fertility Supplement 52: In press.

Weitze, K.F., J.F. Lotz, A. Everwand, T. Willmen and D. Waberski. (1990) Interaction between inseminate, uterine and ovarian function in the sow II Investigations into the influencing of ovulation by the use of sperm free media. Reproduction in Domestic Animals, 25 197-204.

Weitze, K.F., H. Wagner-Rietschel, D. Waberski, L. Richter and J. Krieter. (1994) The onset of heat after weaning, heat duration, and ovulation as major factors in AI timing in sows. Reproduction in Domestic Animals 29: 433-443.