Introduction

There are many variables that influence the two main components of overall herd reproductive performance- sow and boar fertility. Regardless of breeding management, i.e., natural service (NS) vs. artificial insemination (AI), broad management areas like environment, nutrition, lactation length or farrowing room management and health status are generally similar in both protocols. However, unlike natural service, successful matings that should culminate sow and boar fertility into the birth of live offspring are no longer the responsibility of the boar, but a person. Therefore, technician knowledge and experience in achieving successful matings is the most important variable in ensuring successful reproductive performance when using AI, assuming that all other factors between NS and AI are constant.

A successful mating could be defined as "Providing and ensuring that a sufficient number of live spermatozoa are deposited and retained in the female reproductive tract at a optimal time relative to ovulation". One may argue that live sperm are not necessarily fertile spermatozoa, but current methods to evaluate boar spermatozoa fertility are not good and, in fact, the ability to determine if the spermatozoa put into the female are actually living is truly a superior benefit of AI compared to NS. Generally speaking, most boars are pretty good at achieving a successful mating once they have learned what they are suppose to do. I think the same can be said for an AI technician as well, however, technicians are probably not as inept as the boar at determining when and how often to mate a sow or gilt. This proceeding will discuss these two important variables in depth and provide tips on ensuring that like an experienced boar, you to can perform a successful mating with AI.

Developing an Insemination Strategy

There are two factors involved in determining when to breed: How long do spermatozoa survive in the female tract and when does ovulation occur. The fertile life span of a spermatozoa population in the female with the ability to produce a pregnancy is estimated to be from 12 to 36 hours, even though motile sperm have been recovered 10 days following insemination. The fertile life span of the ovulated egg is approximately 8 hours. Therefore, once ovulation occurs, it is necessary for a viable population of spermatozoa to be present in the oviducts at this time. Sperm require a minimum of 4 h in the female before they acquire the ability to fertilize an egg, and thus insemination during ovulation is not necessarily good because post-ovulation insemination will increase the chances of abnormal fertilization (Hunter, 1988) and may interfere with uterine preparation for implantation (Rozeboom et al., 1997).

Good fertilization results (> 90%) in the majority of sows, can be achieved when a single insemination is performed during a 24 h period before ovulation (Waberski et al., 1994; Soede et al., 1995; Steverink et al., 1997). This optimal "AI time" is assuming that 1) an adequate number of sperm are inseminated and retained (> 1 billion and less than 20 ml of back flow; Steverink et al., 1997), and 2) Semen is relatively fresh (less than 38 h old; Waberski et al., 1994). This interesting finding would therefore suggest that all producers need to do is determine when ovulation is going to take place, and inseminate the sow. However, on average, ovulation takes place 35 to 45 h after onset of oestrus (standing response in presence of a boar) but the variability between sows is extremely large (10-85 hours; Weitze et al., 1994; Soede et al., 1995), thus making it difficult to accurately time a single insemination. Therefore, performing multiple inseminations throughout estrus is an easy answer to increase the chances that one insemination will be performed at an "optimal time". Here are four simple steps in establishing a AI protocol on the farm.

Step one: adjustments for time of ovulation:

Even though there is large variation in the time that ovulation occurs after estrus is first detected, ovulation consistently take place at a relatively fixed two-thirds of estrus length in most herds and females. Thus, if a female's estrus length is known, then we have a pretty good idea when ovulation will occur. There is some variation in ovulation time relative to estrus length, but this response can be verified using real time ultrasonography. Briefly, once estrus is detected transcutaneous flank ultrasonography (Weitze et al., 1989) can be performed using an ultrasound machine, preferably with a with a 5 MHz micro-convex probe (Universal Medical Systems, Bedford Hills, NY.) to detect the presence (pre-ovulation) or absence (post-ovulation) of tertiary follicles greater than 6 mm in diameter. When pre-ovulatory follicles are present, ultra sound should be repeated morning and afternoon until these follicles disappear. The absence of large follicles indicates that ovulation occurred. Females should then be check for estrus at the same interval until they are no longer in standing heat. Subtract 6 h from each time (ovulation and estrus length) and an estimation of ovulation relative to estrus length can be accurately established in this herd. The number of females required to accurately establish this relationship should be around 10 % of the total population of the breeding herd.

Step two: adjustments for wei:

It has been well established that estrus length is related to weaning-to-estrus-interval (WEI) (Rojkittikhun et al., 1992; Weitze et al., 1994; Kemp and Soede, 1996); sows with a short (3-4 days) WEI on average have a long estrus length and in contrast, sows with long WEI (>6 days) have shorter estrus lengths. Therefore on most farms, sows with a WEI of 6 days or more should be inseminated sooner after estrus is detected to make sure that the first insemination is before ovulation. To more accurately time inseminations, producers should attempt to determine the average length of estrus on their farms relative to WEI so that on their farm an efficient insemination protocol can be developed. An estimation of average estrus length can help in developing an efficient insemination strategy so that: 1) costly, unnecessary AI doses are not wasted, and 2) late inseminations that may negatively effect pregnancy establishment can be avoided (Rozeboom et al., 1997). An important consideration in developing a AI protocol on average estrus length is that farms cannot use estrual lengths derived from other farms or genetics. The duration of estrus can be variable across farms and is influenced by many factors such as: housing conditions, stress conditions, season, parity, genetic background (see review by Soede and Kemp, 1997), but within a farm, estrus lengths do remain fairly constant.

Step three: adjustments for insemination interval:

The goal for your operation should be to insure that at least one AI is performed within 24 h of ovulation. What we don't really know is if two or more inseminations performed during this time has an additive effect. Nevertheless, regardless of the interval between inseminations, the length of time that the female "stands to be mounted" by a boar will determine how many times to inseminate. Generally, females that are in strong estrus for 3 days will receive 3 inseminations. In these females, shortening the interval between the second and third insemination should in theory, help reduce the negative consequences of mistiming the last AI.

There is no ideal universal number or insemination interval. The ideal frequency and intervals between each insemination on each farm is influenced by semen age at AI, semen storage conditions, individual boar fertility and the composition of the AI dose (percentage of seminal plasma or neat semen used to make up an AI dose). Conservatively speaking, when freshly extended (<72 hours old) semen (3-4 billion motile cells) containing 10-12% seminal plasma is routinely inseminated by an experienced technician at 24-h intervals, a viable population of spermatozoa should be in the female at all times. However, some data would suggest that spermatozoa viability and fertility in the female tract decrease when any of these factors are reduced or changed (Flowers, 1994; Waberski et al., 1997; Rozeboom et al., 1999). Thus, when semen conditions fall outside these criteria, intervals probably should be reduced to 12-18 h. Perhaps the most important consideration to remember is that when estrus length increases, reproductive performance will improve with an insemination frequency increase (Flowers and Esbenshade, 1993). Finally, it's important to consider that gilt estrus behavior patterns differ from sows. Estrus lengths are generally shorter and often less pronounced in gilts and therefore, the first insemination should occur immediately following detection of estrus. A follow up insemination 24 hours later should follow only if the gilt is still in standing heat.

Step four: adjustments for heat detection frequency:

Increasing the frequency of estrus detection will provide for a more accurate determination of the true beginning and end of estrus. Estrus detection is a very labour intense and time-consuming procedure, and consequently, most operations do not check heat more than once per day. However, it may be cost-effective to heat check sows twice a day for 3 to 4 breeding periods in order to accurately establish an average estrus length relative to return-to-estrus intervals. This assessment will enable the farm to develop an efficient insemination protocol for either once or twice per day estrus detection schedules. If an operation chooses to switch back to only morning heat checks, the timing of the first insemination can be adjusted based on the expected length of estrus for that group of females. In most cases, a shorter interval between estrus detection and first insemination will be required when heat detection is performed once vs. twice per day since these female may have been in estrous for a longer period of time prior to detection.

Putting it all Together

Based on averages from summarized research, a hypothetical protocol is depicted in table 1 for females that were first detected in estrous in the morning of day 1(once or twice per day heat detection).

¹ Not all farms will this pattern of estrus length. Start with these intervals and perform estrus detection on about 4 breeding groups until females are no longer standing. If there are no differences in WEI and estrus lengths then all females can be treated the same. Important: Females that return to estrus after mating should also be checked as their estrus lengths may differ from weaned females.
² To determine average ovulation time, multiply average estrus length by .70 and assume 6 h for 2 X/day heat detection and 12 h for 1 X/day heat detection variation in this number.
³ Interval from estrus detection to first AI is shorter with less frequent estrus detection or in females with shorter estrus lengths. Interval between first and subsequent AI depends on semen quality issues (see text).
⁴ Second AI should be scheduled based on average ovulation time (within 12 h of ovulation).
⁵ Perform third or (fourth?) AI only in groups of females that have estrus lengths greater than 60 h (two days). Shorten interval regardless of semen quality to prevent late inseminations.

Heat Detection

Establishing a successful breeding program can only be done if accurate heat detection is frequently performed. To recognize estrus, technicians must be observant, diligent, and have a complete understanding of female cyclic behavior (Table 2). In addition to this perseverance and knowledge, additional measures must be taken to increase the efficiency of this process.

Male/female separation- Pheromones produced by boars are the most potent and effective inducer of the standing reflex in receptive females. The best method of heat detection is to commingle the sow and boar and observe if the female "stands to be mounted" (immobilisation response). This immobilisation response consists of strong, energy consuming muscle contractions that the female can only maintain for brief (5-10 minutes) time periods. At the end of this time period, most females require as long as 40 minutes to an hour to physical recover and respond in the same manner again. This interlude is called refraction, and therefore, when males and females are housed together or in close proximity, females that are in heat may not actually display a standing heat response when observed occasionally (once or twice per day). Thus, housing females and males separately until estrus detection is performed will elicit a stronger and more consistent response in females that are in heat.

Second service AI- In many AI programs, the boar is allowed to mate naturally with the female when she is first detected in estrous and a follow up service is performed artificially on the next day (Flowers, 1995). This procedure is an excellent way for producers to become proficient in AI techniques without sacrificing a drop in reproductive performance. When producers are confident in detecting estrus, the transition from these combination breedings to exclusive AI services occurs.

Direct Male/Female Contact- With exclusive AI, commingling the sow and boar can be frustrating because of inability to move the female away from the boar to perform an AI. Therefore, bringing the sow and boar in close proximity, allowing fence-line nose to nose contact and applying backpressure to the female should provide sufficient stimulation to detect estrus. For sows housed in crates, running a boar in front of sows while a breeding technician applies backpressure is a common and effective method of estrus detection. When the boar is directly in front of a female in a crate, she will move forward and assume the standing reflex. It is imperative in this situation that the boar be in direct nose-to-nose contact with the female because often non-estrual females in crates will "act" like they are in heat until the boar contacts them. Furthermore, heat checking females in small groups will aid in the accuracy of detection and prevent both over exposure (unwanted refraction) and lack of boar exposure (miss-interpreted heats) that often occurs when one or two personal attempt to heat check more females than they can give immediate attention to.

Take-Home Messag

Optimal reproductive performance in a swine operation generally occurs when an insemination strategy ensures that a sufficient population of live spermatozoa is placed in the female tract at an optimal time prior to ovulation. An insemination strategy to more accurately time inseminations can be established once within herd reproductive characteristics such as estrus lengths and return-to-estrus intervals are known. However, the successful development and employment of any insemination strategy is only as good as the accuracy of estrus detection performed on the farm. Like any newly adapted on farm procedure, accurate records should be keep and the procedure evaluated regularly so that changes can be made meet performance goals.

Suggested Reading

Carr, J. 1996. Effective heat detection. Proceedings of the 3rd Beyond 2000 Conference, "Focus on AI". pp. 20-23.

Flowers, W. L. and K. L. Esbenshade. 1993. Optimizing management of natural and artificial matings in swine. J. Reprod. Fertil. Suppl. 48:217-228.

Flowers, W. L. 1995. Characterization of the implementation and use of artificial insemination (AI) in North Carolina swine industry. J. Anim. Sci. (suppl. 1):6.

Flowers, W. L. 1996. Detection of Estrus. Pork Profitability Summit.

Hunter, R. H. F. 1988. The fallopian tubes. Their role in fertility and infertlility. Springer Verlag Berlin. Heidelberg, Germany.

Kemp, B. and Soede, N.M. 1996.. Weaning to oestrus interval in relation to timing of ovulation and fertilisation results in sows. J. Anim. Sci., 74: 944-949.

Rojkittikhun, T., Sterning, M., Rydmer, L. and Einarsson, S. 1992. Oestrous symptoms and plasma levels of oestradiol-17b in relation to the interval from weaning to oestrus in primiparous sows. Proc 12^th International Pig Veterinary Society, The Hague, The Netherlands, p. 485 (Abstract).

Rozeboom KJ, Troedsson MHT, Shurson GC, Hawton JD and Crabo BG. 1997 Late estrus or metestrus insemination subsequent to estrual inseminations decreased farrowing rate and litter size in swine. J. Anima. Sci. Journal of Animal Science 75:2323-2327.

Rozeboom, K. J., M. H. T. Troedsson, H. H. Hodson, G. C. Shurson, and B. G. Crabo. 1999. The importance of seminal plasma on the fertility of subsequent artificial inseminations in swine. (Accepted).

Soede, N.M., Wetzels, C.C.H., Zondag, W., de Koning, M.A.I. and Kemp, B. 1995. Effects of time of insemination relative to ovulation, as determined by ultrasonography, on fertilization rate and accessory sperm count in sows. J. Reprod. Fertil., 105: 135-140.

Soede, N.M. and Kemp, B. 1997. Expression of oestrus and timing of ovulation in pigs. J. Reprod. Fertil., Supplement 52: 91-103.

Steverink, D.W.B., Soede, N.M., Bouwman, E.G. and Kemp, B. 1997. Fertilisation results as influenced by insemination to ovulation interval and sperm dosage in sows. J. Reprod. Fertil. 111: 165-171.

Waberski, D., Weitze, K.F., Lietmann, C., Lubbert zur Lage, W., Bortolozzo, F.P., Willmen, T. and Petzoldt, R. (1994) 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 fertilisation. J. Reprod. Fertil. Supplement 52: 105-109.

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