Comparison of single, fixed-time artificial insemination in weaned sows using 2 different protocols to synchronize ovulation

Abstract

The present study evaluated 2 fixed-time artificial insemination (FTAI) techniques in sows in 2 herds. At weaning, sows were assigned to 1 of 3 treatment groups: Group 1 received intramuscular injections of 600 IU of equine chorionic gonadotropin followed 3 days later by 5 mg of porcine luteinizing hormone; Group 2 received a 200-μg intravaginal dose of triptorelin acetate 4 days post-weaning; and Group 3 were heat checked daily and double-mated when observed in heat. Groups 1 and 2 were bred once at a fixed-time independent of estrous behavior. Time of ovulation was monitored by ultrasound in a subset of sows from each group. Both FTAI techniques resulted in sows farrowing within short intervals, leading to the weaning of pigs that tended to be older and heavier compared with controls. The use of FTAI, however, was associated with a trend to reduced reproductive performance compared with controls in 1 herd.

Introduction

An important aspect of swine production is reproductive efficiency. Maximizing sow productivity and fully utilizing available farrowing crates require that the sows return to estrus promptly after weaning (1). It is especially true for producers operating batch farrowing systems, that the variation in time from weaning to rebreeding be consistent and predictable. The wean-to-estrus interval, the duration of estrus, and the estrus-to-ovulation interval affect the sow-to-sow variation with regard to the timing of conception (2,3). In addition, the onset of estrus is not a reliable predictor for time of ovulation (3). Since optimal fertility is achieved by insemination during the 24-hour period prior to ovulation, it is general practice to inseminate sows every 24 h while they remain in standing heat, which usually involves at least 2 inseminations per estrus period (4,5). If insemination is performed too early or too late relative to ovulation, it will result in decreased fertility (6,7). With the help of hormonal treatment protocols, ovulation can be manipulated so that it is predictable, allowing for fixed-time artificial insemination (FTAI) to be performed (8). As the name suggests, single FTAI allows for a single insemination to be performed at a specific predetermined time resulting in reproductive performance that is comparable to the traditional multiple inseminations (9). When an FTAI protocol is followed, all weaned sows are bred at a specific time without determining the time of onset of signs of standing heat or judging the quality of the signs of estrus. An FTAI protocol may result in labor savings because it eliminates the need to heat check sows daily from weaning until breeding. In addition, the use of FTAI synchronizes breeding among a batch of sows which should result in sows farrowing within a short time period, thus creating a uniform group of similar aged pigs at weaning. This will also allow for more efficient use of labor at farrowing and during the first few days of neonatal care (10).

At present, there are at least 2 options available to Canadian pork producers interested in FTAI. One protocol requires the injection of equine chorionic gonadotropin (eCG) at weaning followed by an injection of porcine luteinizing hormone (pLH) 3 d later (4), and the second protocol involves an intravaginal dose of a gonadotropin-releasing hormone (GnRH) agonist, triptorelin acetate, 4 d after weaning (10). The objectives of this study were to compare the reproductive outcomes of weaned sows bred by 2 different FTAI protocols and a control group bred by conventional means (detection of estrus and using multiple matings), and to determine if FTAI protocols result in farrowings within a short time span to facilitate efficient allocation of human resources and uniform weaning weights.

Materials and methods

This research trial was approved by the Animal Care Committee of the University of Guelph, in accordance with the guidelines set forward by the Canadian Council of Animal Care.

This experiment was conducted on weaned Landrace × Yorkshire sows on 2 farms. Both farms weaned pigs at approximately 3 wk of age and it was predetermined that any sows with a lactation length shorter than 18 d would be excluded from the trial. Farm 1 is a 300-sow operation which practices monthly batch farrowing, uses purchased pooled-semen from a commercial boar stud, and historically has good farrowing rates (> 80%). Farm 2 is a 700-sow unit that performs weekly farrowing, uses fresh pooled-semen collected on-farm, and historically achieves a farrowing rate below the industry norm (< 80%). On each farm, systematic random sampling was used to assign sows at weaning to 1 of 3 treatment groups. Sows in Group 1 were injected intramuscularly (IM) with 600 IU of eCG (Pregnecol; Vétoquinol, Lavaltrie, Quebec) at weaning and 80 h later were injected IM with 5 mg of pLH (Lutropin; Vétoquinol). A single insemination of approximately 3 billion sperm was performed 36 h after administration of pLH. Sows in Group 2 received a 2-mL intravaginal dose of triptorelin acetate (OvuGel; 100 μg triptorelin per mL; Elanco, Guelph, Ontario) 96 h following weaning to induce ovulation. A single insemination of approximately 3 billion sperm was performed 24 h after the administration of triptorelin acetate. Sows in Group 3 served as the control group and were handled in a conventional manner post-weaning, using boar exposure for estrus detection twice daily. Sows in Group 3 received an insemination of approximately 3 billion sperm when discovered in strong standing heat and a second insemination 24 h later if the sow was still in standing heat. In all cases, inseminations of fresh, pooled-semen were performed by the staff on each farm in the presence of a boar to stimulate estrous behavior. On both farms a foam-tipped insemination rod was used to deliver semen into the cervix. For sows in Groups 1 and 2, the estrous behavior was not assessed, but rather, breeding was carried out at a fixed-time.

Sample size was calculated to determine a difference in the mean weaning weight of piglets among the treatment groups (11). Based on historic herd records, the average weaning weight of a pig in the control group was presumed to be 5.5 kg [standard deviation (SD) = 1.0 kg] and in the FTAI groups it was estimated to be 6.0 kg (SD = 0.9 kg). Therefore, in order to detect such differences with 80% power and 95% confidence, 58 sows were required in each group. Data recorded included: sow identification, parity, treatment group, batch number, treatment date, breeding date(s), farrowing date, room and pen number, piglet birth weights, total number of piglets born, litter characteristics, weaning date, litter size, and piglet weight at weaning.

Additionally, in a subset of 10 sows per treatment group on Farm 2, the time of ovulation was monitored by transabdominal ultrasonography (Honda HS-1600 scanner; Honda Electronics, Tokyo, Japan) to confirm the time of ovulation (12). Ultrasound examination was conducted at 8-hour intervals starting on Day 4 following weaning and at 6-hour intervals on Days 5 and 6 until ovulation was complete.

Data analysis

Data were entered into Microsoft Excel for PC 2007 (Microsoft, Redmond, Washington, USA), imported into Stata (Stata/SE 14.2 for Mac; Stata Corp, College Station, Texas, USA), and validated before analysis commenced. Significance was determined at P < 0.05. Descriptive statistics such as means and standard deviations were calculated. The proportion of pigs with a weaning weight of < 4.5 kg in each group was also determined on each farm to further examine the variation in piglet size.

Differences in the likelihood of farrowing, litter size, and piglet weaning weight between the group assignments were modelled separately by farm. The likelihood of farrowing (1 = yes, 0 = no) was compared using a logistic regression model. Litter size (count) representing the total born was analyzed using a multi-level mixed effects Poisson regression model with batch modelled as a random effect. Weaning weights of piglets (continuous) were compared using multi-level mixed effects linear regression models. Pen and room were modelled as random effects and batch was modelled as a fixed effect. The time of ovulation in sows was reviewed and the ranges from the time the first sow farrowed in the batch until the time the last sow farrowed in each treatment group were measured.

Results

Farm 1

The numbers of sows assigned to each treatment were 61, 61, and 66 for Groups 1, 2 and 3, respectively. All sows in Group 3 on each farm were mated twice at 24 h intervals. Descriptive statistics for performance parameters are presented in Table 1. After controlling for parity, the regression model indicated that Group 3 sows had, on average, 1.14 times more piglets (total) born per sow than did Group 2 sows (P = 0.02). After controlling for the other variables in the model, the regression model showed that the piglets in Groups 1 and 2 were heavier at weaning than piglets in Group 3 (P < 0.05). The range of farrowing dates, from the date the first sow farrowed in the batch until the date the last sow farrowed in each treatment group, revealed that Group 1 sows were typically the first sows that farrowed in each batch and all sows within this group farrowed in the least number of subsequent days compared to sows in Groups 2 and 3. The time from the first until the last sow farrowing was on average 3.15 d, 3.75 d, and 4.50 d for sows in Groups 1, 2, and 3, respectively. Among batches, the percentages of pigs that were small (< 4.5 kg) at weaning were 9%, 15%, and 12% for Groups 1, 2, and 3, respectively.

Table 1.

Production performance, represented by means ± standard deviation, of sows receiving single fixed-time artificial insemination (Groups 1 and 2) versus conventional double-mating during a natural estrus (Group 3) on Farm 1 and Farm 2.

Farm 1 (300-sow farm with 4-week batch farrowing and historically good performance)
	aGroup 1 (n = 61)	bGroup 2 (n = 61)	cGroup 3 (n = 66)
Farrowing rate (%)	80.3	75.4	88.7
Litter size (total born)	13.8 ± 4.1	13.2 ± 3.7	14.6 ± 3.9
Birth weight (kg)	1.6 ± 0.4	1.5 ± 0.4	1.6 ± 0.4
Weaning weight (kg)	6.4 ± 1.3	6.2 ± 1.5	6.2 ± 1.4
Age at weaning (days)	18.0 ± 1.4	17.7 ± 1.7	17.5 ± 1.4
Farm 2 (700-sow farm with weekly farrowing and historically low farrowing rate)
	Group 1 (n = 62)	Group 2 (n = 55)	Group 3 (n = 53)
Farrowing rate (%)	74.2	76.4	77.4
Litter size (total born)	12.7 ± 3.7	12.4 ± 4.1	12.6 ± 2.7
Birth weight (kg)	1.6 ± 0.4	1.7 ± 0.4	1.7 ± 0.4
Weaning weight (kg)	6.3 ± 1.6	6.2 ± 1.4	6.2 ± 1.5
Age at weaning (days)	19.0 ± 1.3	18.0 ± 1.5	17.9 ± 1.4

^aGroup 1: IM injection of 600 IU eCG at weaning followed by 5 mg pLH given IM 84 h later, then single breeding 36 h later.

^bGroup 2: A 2-mL intra-vaginal dose of triptorelin acetate 4 d after weaning and then a single insemination 24 h later.

^cGroup 3: Weaning is followed by boar exposure and daily heat checks, bred when sow first observed in heat and 24 h later if still in heat.

Farm 2

The numbers of sows assigned to each treatment group were 62, 55, and 53 for Groups 1, 2, and 3, respectively. Descriptive statistics for performance parameters are presented in Table 1. The ranges, from the time the first sow farrowed in the batch until the time the last sow farrowed in each treatment group, revealed that Group 1 sows were typically the first sows that farrowed in each batch and all sows within this group farrowed in the least number of subsequent days compared to sows in Groups 2 and 3. The times from the first until the last sow farrowing were on average 3.25 d, 4.0 d, and 4.25 d for sows in Groups 1, 2, and 3, respectively. The regression models indicated that the treatments had no association with the likelihood of farrowing, litter size, or weaning weights on Farm 2. Among batches, the percentages of pigs that were small (< 4.5 kg) at weaning were 12%, 10%, and 12%, for Groups 1, 2 and 3, respectively.

The ultrasound assessment on Farm 2 showed that 6 of 10 sows in Group 3 did not ovulate until 24 to 30 h after the first breeding, whereas the FTAI sows ovulated between 6 h and 24 h post-breeding.

Discussion

In the present study, there is evidence that FTAI resulted in poorer reproductive performance than conventional breeding on Farm 1, with Group 3 sows (controls) tending to have a higher farrowing rate compared to sows in Groups 1 and 2. The FTAI farrowing rates appear to be low for most modern farms but this is not a consistent finding in the literature. There are studies which found that farrowing rates were either similar among FTAI and control groups, as found on Farm 2, or higher for FTAI groups. For instance, a study by Cassar et al (4) reported that farrowing rates were notably higher in eCG-pLH treated sows inseminated once at a fixed-time compared to control sows that were inseminated at least twice. A study by Zak et al (13) found that an IM injection of pLH at the onset of estrus in weaned sows followed by a double FTAI produced a farrowing rate comparable to controls. It has been reported that farrowing rates are much lower if sows are not in standing heat when bred using a FTAI protocol (14). Unfortunately, the quality of the estrous behavior at the time of the breeding was not recorded in this trial, but if some of the sows were not in good standing heat at the time of the FTAI, this might partly explain why the farrowing rates in the FTAI groups were low on both farms. On Farm 2, in which the control sows recorded a farrowing rate of < 80%, similar to the FTAI groups, the explanation for the low performance may be the same, in that the herdsman may have bred some control and FTAI sows that were not in strong standing heat and at the optimum time for insemination. One of the potential advantages of FTAI programs is that breeding would become less reliant on the stockman’s experience and ability to determine signs of estrus.

In addition to farrowing rate, litter size is an important aspect of reproductive performance and sow productivity. On Farm 1, there was evidence that at least in the case of triptorelin acetate treatment (Group 2), the litter size was reduced compared with controls. The eCG-pLH treatment did not have a significant effect on litter size at either farm, which is consistent with previous literature (4,5,15).

Ultrasound imaging on a subset of sows on Farm 2 revealed that the timing of insemination for sows in Groups 1 and 2 was within 24 h before ovulation, with ovulation occurring within 12 h of insemination for sows in Group 1, as expected based on previous studies (8). Sows in Group 2 finished ovulating between 6 to 24 h after breeding, making it appropriate to use a single insemination. This is consistent with the findings of a study by Knox et al (8), in which they evaluated the use of a single intrauterine insemination using 1.5 or 2.5 billion sperm administered at 22, 26, or 30 h following triptorelin treatment, given 96 h post-weaning in sows. The average interval from administration of triptorelin to ovulation was 42.2 ± 0.4 h and ovulation occurred within a 24-hour period in 88% of sows. The sows in Group 3 completed ovulating 24 to 30 h after the first breeding, or up to 6 h after the second breeding. This range of ovulation times in Group 3 confirmed why at least 2 inseminations are typically required to achieve desired conception rates. This also highlights the ability of the hormonal treatments to narrow the window of ovulation times within a group of sows, allowing for FTAI. Although these results indicate that the timing was reasonable, ultrasonography was performed on Farm 2 only. Future studies should examine whether these results vary from farm to farm.

An objective of using FTAI is to reduce labor by tightening the breeding window so that instead of breeding sows over a period of several days they are bred in a single day. A consequence should be that all sows farrow close together, improving the efficiency of labor in the farrowing room by making cross-fostering, scheduling supervision of farrowing, and the processing of litters easier. Commonly, due to the variation in the wean-to-estrus interval and gestation length, the farrowing period in a batch of sows occurs over the span of 10 d (10). In the present study, the first sow in the batch to farrow on both Farm 1 and Farm 2 was generally in Group 1, the same group that required the least number of subsequent farrowing days. Compared to Group 3, Group 1 sows finished farrowing 1 d earlier. Due to the controlled breeding, it was predicted that sows in Groups 1 and 2 would farrow closely together and this would contrast with a wider range of farrowing dates for Group 3 sows. Surprisingly Group 3 sows farrowed over a relatively narrow range of days as well. Perhaps FTAI protocols would prove to be better suited for herds with greater variation in the weaning to breeding interval.

When sows farrow close together, the lactation length will be similar among them as well as the age and size of their piglets at weaning (10). On both farms, the FTAI groups had slightly older pigs at weaning compared to Group 3. In addition to increased age at weaning, the weaning weights of pigs in Groups 1 and 2 were slightly heavier than those of pigs in Group 3 on Farm 1 based on models that controlled for various factors including birth weight. A previous study found no difference between the weaning weights of pigs in control and triptorelin acetate groups (14). Presumably, on farms with consistently short weaning to breeding intervals there would be very little difference between FTAI and natural breeding with respect to weaning age and weight. An increased weight at weaning has been shown to better prepare pigs for more successful performance in the later stages of production (15). In particular, heavier weights at weaning have been shown to result in more rapid growth post-weaning (15). Uniform age and size at weaning is desired because it reduces the likelihood of pigs that are too young and small being moved into the nursery and mixed with older and larger animals. The results of this trial, which was conducted on 2 farms, show that farm differences do exist and are important to consider when applying FTAI protocols. Perhaps the use of FTAI provides more benefit when the herd reproductive performance is below average, particularly if the timing of insemination and observing heat signs are problems for the herdsman. Fixed-time artificial insemination protocols allow the reproductive performance to be less dependent on the stockperson’s ability to detect estrus and to determine the best time to inseminate because timing is dictated by the FTAI protocol. Since FTAI protocols create a tight group which farrows together, the monthly batch farrowing farm was expected to benefit the most from either protocol. However, in this study the farm using batch farrowing recorded that control sows were bred within a narrow time frame and reproductive performance was excellent in this group as well. Based on the ultrasound results on Farm 2, it was presumed that the timing of inseminations was not an issue, but it is possible that the hormone treatments resulted in the recruitment of immature follicles causing the lower farrowing rates and smaller litters observed on Farm 1. Future studies need to investigate the cause of the farm to farm differences observed in this trial.

Significant benefits of FTAI are the labor savings related to heat checking and double mating. It also enables producers to eliminate weekend inseminations and plan how much semen to have available on a particular day, so semen storage time and semen wastage can be reduced (10). This study demonstrated that there is a farm to farm variation as to whether FTAI produces reproductive performance comparable to conventional double mating or lower farrowing rates and litter size. Therefore, caution needs to be used when contemplating a change to FTAI because results may vary depending on herd characteristics and management factors.