Administration of P.G. 600 to sows at weaning and the time of ovulationas determined by transrectal ultrasound1

R. V. Knox*,2, S. L. Rodriguez-Zas*, G. M. Miller*, K. L. Willenburg*, and J. A. Robb†

*Department of Animal Sciences, University of Illinois, Urbana 61801 and†Department of Agriculture, Illinois State University, Normal 61790-5020

ABSTRACT: This study determined whether the in-terval from estrus to ovulation was altered by givingP.G. 600 to sows at weaning. Mixed-parity sows re-ceived P.G. 600 i.m. (n = 72) or no treatment (n = 65)at weaning (d 0). Beginning on d 0, sows were observedfor estrus twice daily. At the onset of estrus and thereaf-ter, ultrasound was performed twice daily to determinethe average size of the largest follicles and time of ovula-tion. Weaning age (20.1 ± 0.4 d) did not differ (P > 0.10)between treatments. More P.G. 600 sows expressed es-trus within 8 d (P < 0.01) than controls (94.4% vs 78.4%,respectively). Parity was associated with expression ofestrus (P < 0.02), with 78% of first-parity and 93% oflater-parity sows exhibiting estrus. However, no treat-ment × parity effect was observed (P > 0.10). The inter-val from weaning to estrus was reduced (P < 0.0001)by P.G. 600 compared with controls (3.8 ± 0.1 d vs4.9 ± 0.1 d). Follicle size at estrus was not affectedby treatment (P > 0.10). The percentage of sows thatovulated did not differ (P > 0.10) for P.G. 600 and controlsows (90.3% vs 81.5%, respectively). Time of ovulationafter estrus was not affected by treatment and averaged44.8 h. However, univariate analysis indicated that the

Key Words: Estrus, Farrowing, Litter Size, Ovulation, Sows

2001 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2001. 79:796–802

Introduction

A reduced reproductive performance by sows is asso-ciated with season (Hurtgen and Leman, 1980; Love etal., 1993; Xue et al., 1994), lactation length, and parity

1This research was supported in part by the Illinois Council onFood and Agricultural Research (C-FAR) and the Department of Agri-culture, Illinois State University. The authors thank Tammy Knox,Jodi Keubler, and Kevin Kalmer for their excellent technical assis-tance. The animal care and use committee of Illinois State Universityapproved the protocol for use of animals in this experiment.

2Correspondence: 360 Animal Sciences Laboratory, 1207 WestGregory Drive (phone: (217) 244-5177; fax: (217) 333-8286; E-mail:rknox@uiuc.edu).

Received June 21, 2000.Accepted November 28, 2000.

796

interval from weaning to estrus influenced the intervalfrom estrus to ovulation (r = 0.43, P < 0.0001). Further,multivariate analysis showed an effect of treatmenton the intervals from weaning to estrus, weaning toovulation (P < 0.0001), and estrus to ovulation (P <0.04). Within 4 d after weaning, 81% of the P.G. 600sows had expressed estrus compared with 33% of con-trols. However, this trend reversed for ovulation, withonly 35% of P.G. 600 sows ovulating by 36 h after estruscompared with 40% of controls. The estrus-to-ovulationinterval was also longer for control and P.G. 600 sowsexpressing estrus ≤ 3 d of weaning (45 h and 58 h,respectively) than for sows expressing estrus after 5 d(39 h and 32 h, respectively). Farrowing rate and littersize were not influenced by treatment. However, theinterval from last insemination to ovulation (P < 0.02)indicated that more sows farrowed (80%) when the lastinsemination occurred at ≤ 23 to ≥ 0 h before ovulationcompared with insemination ≥ 24 h before ovulation(55%). In summary, P.G. 600 enhanced the expressionof estrus and ovulation in weaned sows but, breedingprotocols may need to be optimized for time of ovulationbased on the interval from weaning to estrus.

(Hurtgen and Leman, 1981; Koketsu and Dial, 1997;Koketsu et al., 1997). The problem originates fromfewer sows returning to estrus, increased time to returnto estrus, and higher return rates after mating, whichcollectively result in lower farrowing rates. Becausesuppressed gonadotropin secretion and ovarian inactiv-ity may be involved, exogenous gonadotropins havebeen used to stimulate ovarian activity. Both PMSG(Britt, 1986; Sechin et al., 1999) and PMSG/hCG (Lan-caster et al., 1985; Bates et al., 1991; Kirkwood et al.,1998) have been used in sows at weaning. Generally,gonadotropins given at weaning improve the percent-age of sows returning to estrus and reduce the time toreturn to estrus. However, only a few of the studiesindicate an improvement (Lancaster et al., 1985; Britt,1986; Sechin et al., 1999), and others indicate no changeor even slight declines in farrowing rates and litter

P.G. 600 effects at weaning 797

sizes (Bates et al., 1991; Estienne and Hartsock, 1998;Kirkwood et al., 1998). Differences in results are notunexpected when considering differences in seasons,parity, genetics, housing, and management among thestudies. Despite these differences, most of the previousstudies show that gonadotropins induce more sows intoestrus when seasonal effects are most notable and theyinduce more primiparous sows into estrus, which tendto have greater incidences of reproductive failure thanmultiparous sows.

The reason for a greater induction of estrus, but withlittle or no improvement in farrowing rate, is somewhatperplexing. Therefore, this study was conducted to de-termine whether sow response to P.G. 600 at weaningwas related to follicle development, ovulation, or time ofovulation and whether the time of breeding influencedfarrowing rate and litter size in sows induced into es-trus with P.G. 600.

Materials and Methods

One hundred and thirty-seven DK44 sows (DekalbChoice Genetics, St. Louis, MO) of mixed parity wererandomly assigned to receive P.G. 600 (400 IU of eCGand 200 IU of hCG, n = 72) or no treatment (control, n =65) at weaning (d 0) across 16 weaning periods (months)over 1 yr. Average parity for control sows was 2.9 ± 0.2(parity 1, n = 17, and parities 2 through 6, n = 48) andfor sows receiving P.G 600 was 2.7 ± 0.2 (parity 1, n =20, and parities 2 through 8, n = 52). Sows receivedP.G. 600 or no injection (control) on the day of weaning(d 0). Immediately before use, lyophilized P.G. 600 wasmixed with 5 mL of sterile diluent supplied by the man-ufacturer and administered i.m. in the neck using a 38-mm 20-gauge needle. Sows were moved on d 0 to anestrus detection area where they were housed individu-ally in gestation crates and observed for estrus twicedaily at 0800 and 1600 using fence-line contact with amature boar for 15 min. Estrus detection continued at12-h intervals until the end of estrus. Estrus lengthwas classified as the interval from onset of standingestrus (0 h) until the last time the female would standfor the back-pressure test. At the onset of estrus andthereafter, transrectal ultrasound was performed twicedaily at 0700 and 1700 to determine the average sizeof the three largest follicles and the occurrence of ovula-tion after estrus onset. Ovulation was confirmed withan additional transrectal ovary scan at 12 h after theinitial observation of large follicle disappearance. Ul-trasound was performed using an Aloka 500 V witha 7.5-MHz linear transducer attached to a polyvinylchloride transrectal stabilizing rod. The technique fortransrectal ultrasound has been described previously(Knox and Althouse, 1999). Ovulation was classified ashaving not occurred when the number of large follicles(≥ 6.5 mm in diameter) present was representative ofthe ovulation rate for the line and age of the sow (range:approx. 10 to 22). Ovulation was classified as in prog-ress and not completed when more than half but notice-

ably less than the expected pool of ovulatory folliclesstill remained. Ovulation was considered completewhen no large follicles remained or fewer than fourlarge follicles were still present. Sows were bred at 12and 24 h after first observed in standing estrus usingsemen from a commercial supplier (Dekalb Choice Ge-netics) diluted in long-term extender at a concentrationof 3.0 × 109 sperm cells per 80 mL. The semen wasused within 3 to 4 d after collection. The proportionof experimental sows that expressed estrus and weremated, were observed for farrowing and for total num-ber of pigs born (alive and still-born).

Housing and Management

After weaning, sows were kept in an environmentallyregulated facility that maintained an air temperaturebetween 18 and 25°C and a lighting regimen of approxi-mately 8 to 10 h of light each day. Sows were fed oncedaily ∼2 kg of a standard 14% CP corn-soybean mealgestation diet containing 0.65% lysine, 0.75% Ca, 0.60%P, and 3,300 kcal ME/kg. All sows were given ad libitumaccess to water. Approximately 21 d after estrus, sowswere moved into pens in a modified open-front buildingwith an outdoor run, with 5 to 7 sows in each pen.Sows remained in pens until they were moved into anenvironmentally regulated confinement facility. Sowswere placed into gestation crates at 70 to 80 d of gesta-tion and subsequently into farrowing crates on d 110of gestation.

Statistical Analysis

The continuous variables studied were as follows:time of ovulation from onset of estrus (hours), folliclesize at estrus (millimeters), interval from weaning toestrus and to ovulation, and litter size (total numberof pigs born). The model for the continuous variablesincluded (where appropriate) treatment, month, parity,lactation length, interval from weaning to estrus, aver-age diameter of the three largest follicles at estrus,length of estrus, interval from time of the last insemina-tion to ovulation (≥ 24 h before, ≤ 23 to ≥ 12 h before,and ≤ 11 to ≥ 0 h before) and associated interactions.Univariate mixed effects linear model analyses (SASInst. Inc., Cary, NC) were performed with month as ablock random effect and the rest of the independentvariables as fixed effects. Variables that were nonsig-nificant were removed from the model. In considerationof the related nature of the variables studied, a multi-variate analysis (PROC GLM/MANOVA) of the effectof treatment on interval from weaning to estrus andfrom estrus to ovulation was conducted. This approachaccounted for the correlation between these two vari-ables while testing for the effect of treatment simulta-neously on both dependent variables.

Data that measure the length of time until the occur-rence of an event, such as days from weaning to ovula-tion, are called survival data. Analysis of survival data

Knox et al.798

Table 1. Least squares means ± SE for sowsadministered P.G. 600 or no treatment (control) at

weaning on the continuous variables: wean-to-estrusinterval, average diameter of the largest follicles at

estrus, interval from estrus to ovulation, lengthof estrus, interval from last insemination

to ovulation, and total pigs born

Treatment

Item Control P.G. 600 SE P-value

Animalsa 65 72Lactation lengthb 20.4 19.9 0.6 NSWean to estrusc 4.9 3.8 0.1 <0.0001Follicle size, mmd 7.7 7.7 0.2 NSOvulation houre 43.7 45.6 3.2 NSEstrus lengthf 56.9 54.1 4.5 NSLast AI before 13.3 15.7 1.9 NSovulationg

Total bornh 11.9 10.6 0.6 NS

NS = not significantly different.aNumber of sows; SE based on 65 sows.bDays from farrowing to weaning.cDays from weaning to estrus.dMean of three largest follicles at estrus.eHours from onset of estrus to ovulation.fHours from onset (0 h) to last standing response.gHours from last insemination to ovulation.hTotal number of pigs born alive and still-born.

requires special considerations, such as that the re-sponse cannot be negative, and this must be incorpo-rated into any analysis. Hence, in addition to linearunivariate and multivariate analysis, Kaplan-Meier’snonparametric time-dependent variable analysis(PROC LIFETEST) was conducted to gain informationon the effect of treatment on days from weaning toestrus, hours from estrus to ovulation, and days fromweaning to ovulation. The binary traits (occurrence ofestrus, ovulation, and farrowing) were analyzed usinga logistic regression (PROC LOGISTIC) with indepen-dent variables as previously described.

Results

The average weaning age (range: 10 to 34 d) was notdifferent (P > 0.10) between treatment groups (Table1). Likewise, there was no significant effect of monthor lactation length on the response variables, and henceboth variables were removed from the final modelsfitted.

More sows treated with P.G. 600 at weaning ex-pressed estrus within 8 d (P < 0.01) compared withcontrols (Table 2). Parity had a significant effect on thepercentage of sows expressing estrus (P < 0.02). Amongfirst-parity sows, 78% (60%, 89%, representing 95% con-fidence interval) expressed estrus compared with 93%(86%, 97%, representing 95% confidence interval) ofsecond- and later-parity sows. There was no treatment× parity interaction (P > 0.10), and means with 95%confidence limits indicated that P.G. 600 increased the

percentage of parity 1 (90%, 67 to 98%) and parity 2and greater sows (96%, 86 to 99%) that expressed estruscompared with parity 1 (59%, 35 to 79%) and parity 2and greater sows (85%, 72 to 93%) in the control group.The duration of estrus was influenced by the wean-to-estrus interval (P = 0.02) but not by treatment (Table1) or any other variable; the duration of estrus waspositively correlated with estrus-to-ovulation interval(r = 0.71, P < 0.0001).

Follicle size at estrus was not affected by treatment(Table 1). Follicle size at estrus was associated withparity (P < 0.05), with least square means for first-and later-parity sows equal to 7.5 ± 0.2 and 7.9 ± 0.1mm, respectively.

There was no significant effect of treatment on thepercentage of sows that ovulated (Table 2) or on theaverage interval from estrus to ovulation (Table 1). Uni-variate analysis indicated that time of ovulation afteronset of estrus was influenced by interval from weaningto estrus (P < 0.001). Further, multivariate analysis ofthe effect of treatment on the intervals weaning to es-trus and estrus to ovulation indicated that both inter-vals are negatively correlated (r = −0.43) and the Ho-telling-Lawley and Wilk test statistics suggested thatP.G. 600 treatment was associated (P < 0.0001) withsignificant variation of both intervals. Additionally, thetime of ovulation after onset of estrus was greater forcontrol and P.G. 600 sows expressing estrus ≤ 3 d ofweaning (45 h and 58 h, respectively) than for sowsexpressing estrus after 5 d (39 h and 32 h, respectively,Table 3).

The effect of treatment on the percentage of sowsexpressing estrus and ovulating after weaning was ex-amined using Kaplan-Meier’s nonparametric survivalanalysis. The log rank and Wilcoxon statistics test ho-mogeneity between groups and indicated a significanteffect of treatment on the interval from weaning toestrus (Figure 1; P < 0.001), on the interval from estrusto ovulation (Figure 2; P < 0.04), and on the intervalfrom weaning to ovulation (Figure 3; P < 0.001). Esti-mates from survival analysis indicate that 8% and 29%of control and P.G. 600 sows showed estrus 3 d afterweaning, respectively, and, by d 5 after estrus, 75% and97% of control and P.G. 600 sows showed estrus (P <0.0001; Figure 1). By contrast, 47% and 35% of controland treated sows ovulated 36 h after estrus, respec-tively, whereas 98% and 95% of control and P.G. 600sows ovulated 72 h after estrus (P < 0.04; Figure 2).When the overall period from weaning to ovulation isconsidered, 8% and 42% of control and P.G. 600 sowshad ovulated by 5.5 d after weaning, whereas 92% and99% of control and P.G. 600 sows had ovulated by d 7.5after estrus (P < 0.0001; Figure 3). Treatment was notsignificantly associated with the time of last artificialinsemination before ovulation, but P.G. 600 sowstended to be inseminated slightly more than 2 h earlierrelative to ovulation compared with controls (Table 1).

The farrowing rate was not influenced by P.G. 600(Table 2) but was associated with the time of the last

P.G. 600 effects at weaning 799

Table 2. Effect of P.G. 600 or no treatment (control) for sows at weaning on the binaryresponse variables: Percentage of sows expressing estrus,

ovulating, and farrowing

Treatment

Item Control P.G. 600 P

Animalsa 65 72Estrus, %b 78.4e 94.4 0.01

(66.7, 86.9)f (85.9, 97.9)Ovulate, %c 81.5 90.3 NS

(70.0, 89.3) (80.8, 95.4)Farrowing rate, %d 67.4 71.1 NS

(51.4, 80.2) (57.0, 82.1)

NS = not significantly diferent.aNumber of sows randomly assigned to treatment.bPercentage of sows expressing estrus within 8 d.cPercentage of treated sows ovulating.dPercentage of treated sows that farrowed.eLeast squares mean.fLower and upper 95% confidence interval.

insemination relative to ovulation (P < 0.02). More sowsfarrowed when the last insemination occurred at ≤ 23to ≥ 0 h before the time of ovulation than when insemi-nated ≥ 24 h before ovulation (Table 4). The significanceof time of the last insemination relative to ovulationobserved for farrowing rate was not evident in the totalnumber of pigs born (Table 4). Furthermore, P.G. 600treatment did not have a significant effect on numberof pigs born (P > 0.05, Table 1).

Discussion

Results of this study show that P.G. 600 administeredto sows on the day of weaning stimulates ovarian follicledevelopment and results in a greater incidence of estrusand a shortened interval from weaning to estrus com-pared with untreated sows. However, these benefitswere not reflected in improved farrowing rates and lit-ter sizes. In P.G. 600-treated sows, the interval fromweaning to estrus was shorter and the interval fromonset of estrus to ovulation was longer than for controlsows. Because of this treatment effect, more P.G. 600sows received their last insemination ≥ 24 h before the

Table 3. Least squares means (± SE) for time of ovulation (Ov) by interval fromweaning to estrus in sows given P.G. 600 or no treatment (control) at weaning

Treatment

Control P.G. 600

Interval, da Ov, hb nc Last AId Ov, hb nc Last AId

≤ 3 45.0 ± 7.9 4 + 57.6 ± 3.5 20 −4 46.9 ± 4.8 13 + 47.3 ± 2.8 35 +5 to 6 39.3 ± 2.7 34 + 32.0 ± 4.6 13 +

aInterval from weaning to estrus (in days).bInterval from estrus to ovulation (in hours). Least squares mean ± SE.cNumber of observations.dTime of last artificial insemination occurring ≤ 23 h to 0 h (+) or ≥ 24 h before ovulation (−).

time of ovulation. The breeding protocol in this studyemployed twice daily estrus detection with predeter-mined inseminations occurring at 12 and 24 h afterdetected estrus. This methodology may have resultedin a greater frequency of nonoptimal inseminationtimes for the P.G. 600-treated sows.

The postweaning period is associated with economicimportance because it affects measures of reproductiveefficiency, such as interval from weaning to estrus, non-return to estrus rate, open sow days, farrowing rate,and litter size. Reproductive failure at the time of wean-ing has been linked to parity, lactation length, andseason of the year (Koketsu and Dial, 1997). Much ofthe failure has been associated with a lack of ovarianactivity, which is thought to originate from reducedgonadotropin stimulation, increased prolactin levels, orreduced responsiveness of the ovaries (Britt, 1986; vande Wiel and Booman, 1993; Kermabon et al., 1995). Themechanisms for these influences remain unclear, butP.G. 600 given to sows at weaning induces an increasein estrogen, an LH surge, and improves return to estrusin multiparous sows (Saoulidis et al., 1995; Estienneand Hartsock, 1998) and primiparous sows (Bates et

Knox et al.800

Figure 1. Estimated cumulative percentage of sows ex-pressing estrus after weaning in response to P.G. 600 orno treatment (control) at time of weaning.

al., 1991; van de Wiel and Booman, 1993) and reducesthe interval from weaning to estrus in both primiparous(Bates et al, 1991; Kirkwood et al., 1998) and multipa-rous sows (Estienne and Hartsock, 1998; Lancaster etal., 1985). In this study, P.G. 600 treatment shortenedthe wean-to-estrus interval as well as increased theproportion of sows returning within 8 d of weaning. Theinterval from weaning to estrus, although only differingbetween treatments by 1 d, would have been muchgreater, especially for the control group, if the day toreturn to estrus included all sows that expressed estrusmore than 8 d after weaning. In the present study,parity affected return to estrus rate, with 78% of first-parity and 93% of later-parity sows exhibiting estrus,but parity was not associated with significant variationon the number of days from weaning to estrus. However,P.G. 600 improved the percentage of parity 1 and

Figure 2. Estimated cumulative percentage of sowsovulating after estrus in response to P.G. 600 or no treat-ment (control) at time of weaning.

Figure 3. Estimated cumulative percentage of sowsovulating after weaning in response to P.G. 600 or notreatment (control) at time of weaning.

greater parity sows that expressed estrus, suggestinga potential benefit for use on both groups of animals.

In the present study, we did not observe an effect oflactation length on measures of reproductive perfor-mance. However, in studies that have observed a de-cline in reproductive performance, the length of lacta-tion and the reproductive measures affected are notalways in agreement (Levis, 1997; Koketsu and Dial,1997). With very short lactation lengths of less than 10d, increases in the wean-to-service interval are ob-served. However, even with this short length of lacta-tion, farrowing rate and litter size did not differ fromsows weaned at 22 d. Another explanation for our in-ability to detect an effect of the length of lactation maybe the body condition of the sows at weaning. Althoughthis was not measured or reported in the present study,when compared with some other commercial lines ofgenetics, the sow line used for this study has been ob-served to be heavier at farrowing, consume more feedduring lactation, and lose less backfat during lactation(Moeller, 2000).

Table 4. Least squares means of farrowing rate andnumber of pigs born by interval from the time of

last insemination before ovulation (last AI) in sowsgiven P.G. 600 or no treatment at weaning

Last AIa nb Estimated farrowing, %c Total number bornd

≤ 11 to ≥ 0 h 28 79.1 (59.7, 90.6) 11.7 (10.3, 13.2)≤ 23 to ≥ 12 h 42 80.6 (65.2, 90.2) 11.5 (10.3, 12.7)≥ 24 h 54 54.8 (39.8, 68.9) 10.8 ( 9.4, 12.2)

P < 0.02 NS

aHours from last insemination before ovulation.bNumber of sows inseminated.c95% confidence interval limits.d95% confidence interval limits includes born alive and still-born.

P.G. 600 effects at weaning 801

Results of this study show that P.G. 600 inducedfollicle development, resulting in expression of estrusin 94% of P.G. 600-treated sows, and that these folliclesovulated in 90% of these cases. Follicle size was normaland was comparable to controls, and no incidence ofcystic follicles was detected. The follicle size at estruswas similar to that reported at estrus by Nissen et al.(1997). Although the significance remains unknown,parity-1 sows tended to have smaller follicles comparedwith later-parity sows, which could reflect lower natu-ral gonadotropin production and aid in explaining loweroverall return rates for these sows.

The average time of ovulation was not affected bytreatment, and sows ovulated approximately 45 h afterestrus. Our ability to observe a treatment effect mayhave been limited due to the 12-h frequency of ultra-sound observation. More frequent observation, at 4 to8 h intervals, has previously been used to characterizeovulation times in sows (Soede et al., 1995, Nissen etal., 1997) and may have been necessary to detect thiseffect of treatment. However, the observation frequencyused in the present study does provide information onthe effect of inseminations that occur within 12 h ofovulation. This is important and practical because moststudies have shown that inseminations occurring 0 to28 h before ovulation produce maximal fertilizationrates and litter sizes (Soede et al., 1995; Kemp andSoede, 1996; Nissen et al., 1997). Therefore, observationat 12-h intervals would provide an opportunity to opti-mize insemination times before ovulation.

P.G. 600 did influence the interval from weaning toestrus and from estrus to ovulation, which were relatedto one another. Relationships between wean-to-estrusinterval and ovulation hour have previously been re-ported. In those studies the relationships of wean-to-estrus interval and estrus-to-ovulation interval were r= 0.24 (Kemp and Soede, 1996) and 0.29 (Nissen et al.,1997), which are similar to those observed in our studyfor controls (r = 0.35) and P.G. 600-treated sows (r =0.47). Despite these significant relationships, it is im-portant to point out in the previous studies and in thepresent experiment that more than half of all sowsthat return to estrus early or late after weaning do notovulate late and early, respectively. Therefore, delayingthe insemination of sows that return early and insemi-nating late-returning sows early will fail to achieve thedesired results in more than half of the cases. However,in half the cases in which a relationship did occur, weobserved that the estrus-to-ovulation interval waslonger for both control and P.G. 600 sows expressingestrus ≤ 3 d of weaning (45 h and 58 h, respectively),compared with sows expressing estrus after 5 d (39 hand 32 h, respectively). Treatment also influenced thisinterval, and, by d 4 after weaning, 81% of P.G. 600sows had expressed estrus compared with only 33% ofcontrols. In contrast, by 36 h after estrus, only 35% ofP.G. 600 sows had ovulated compared with 40% forcontrols. This pattern continued at 48 h, during whichonly 75% of treated sows and 95% of control sows had

ovulated. Despite the delayed interval from estrus toovulation in P.G. 600-treated sows, P.G. 600 sows ovu-lated sooner after weaning than controls. By d 6.5 afterweaning, more than 85% of treated sows had ovulatedcompared with only 50% of controls. This could be ofimportance for scheduling timed inseminations because94% of treated sows ovulate in a 3-d period, comparedwith 3.5 d needed to achieve this proportion ovulatingin controls.

An examination of farrowing rate and litter size inresponse to P.G. 600 at weaning in the present studyshowed little or no effect compared with controls eventhough more animals expressed estrus, ovulated, andwere mated. A lack of effect of P.G. 600 has also beenobserved for farrowing rate and litter size by Kirkwoodet al. (1998) and Lancaster et al. (1985), and even slightdecreases were observed by Bates et al. (1991) and Es-tienne and Hartsock (1998). The reasons for the failureto improve these reproductive measures were not clear,but reduced farrowing rates and litter sizes have beenreported to occur in response to season of the year (Love,1978; Stork, 1979) and the effects of season combinedwith parity (Hurtgen and Leman, 1981), the effect ofparity with length of lactation (Svajgr et al., 1974), andsow housing (Hurtgen and Leman, 1980).

In this experiment, the lack of observed improvementin farrowing rate despite a 17% (94%/78%) increase inthe number of sows mated is presumed to be due to thenearly 30% of P.G. 600-treated females that did notreceive their last insemination within 24 h prior toovulation. Nissen et al. (1997), Soede et al. (1995), andWaberski et al. (1994) reported that, when single insem-inations do not occur within −24 to +4 h of ovulation (0h), fertilization rates, pregnancy rates, and farrowingrates are reduced. This appears to be the most likelyexplanation for the observations of the present studybecause sows not inseminated within ≤ 23 to ≥ 0 h rangebefore ovulation in our study tended to have reducedfarrowing rates. After weaning, 30% of all P.G. 600 sowsreturned to estrus within 2 to 3 d, and these animalsovulated at 55 and 65 h after estrus. Thus, none ofthese sows would have received their last inseminationwithin 24 h of ovulation on a schedule of 12- and 24-hinseminations after detection of estrus. The 30% of theanimals not inseminated could account for the failedfarrowing rate alone.

From a practical standpoint, Burke (1999) reportedthat, based on sow weaning schedules in the USA, 45%of on-farm semen demand is for Tuesday delivery and25% of demand is for Thursday delivery. In this typicalindustry scenario and in the present study, most sowsare weaned in groups on Thursdays. The bulk of semendelivery is scheduled for Tuesday; however, one otherdelivery day is also needed, either Friday or Saturday,which allows high-quality semen to be present for sowsthat return and must be mated in less than 5 d afterweaning. Previous reports (Dewey et al., 1994) showedthat most sows return to estrus within 5 to 7 d, however,in circumstances with known genetics, nutrition and

Knox et al.802

management situations, and longer lactations, manysows begin to return to estrus in 4 to 6 d. In the presentstudy, sows were weaned on Thursday and semen wasshipped within 1 d of collection for arrival on Saturday(1000) and on Tuesday (1000). Because many P.G. 600sows expressed estrus on Sunday and Monday, approxi-mately 51% of these sows were inseminated with sementhat was 72 to 96 h of age, whereas only 25% of controlsreceived semen of this age. Although it is not clearwhether in this situation this difference could have af-fected farrowing rate and litter size, inseminating sowswith semen ≤ 72 h of age would certainly be more desir-able than insemination with semen aged ≥ 72 h. Per-haps if this is involved in overall fertility, then semencollections and deliveries should be adjusted to accountfor earlier return to estrus with P.G. 600. We hypothe-size that the time of last insemination relative to thetime of ovulation may have masked an effect of age ofsemen when mated because sows that expressed estrusearlier after weaning would have tended to be insemi-nated with semen of older age after collection. Thisolder-aged semen could contribute to fewer eggs fertil-ized or more defective embryos resulting from the com-bination of aged semen with an increased estrus-to-ovulation interval in females that returned to estrusearlier after weaning.

Implications

This study indicates that P.G. 600 can provide sig-nificant advantages for inducing more sows to expressestrus and ovulate across different months of the year,when compared with no treatment. This effect is evi-dent in both primiparous and multiparous sows. How-ever, the time of breeding may need to be adjusted whenusing P.G. 600 at weaning in order to achieve improvedfarrowing rates and litter size because more P.G. 600sows express earlier and ovulate later compared withuntreated sows. Perhaps one way to account for thiseffect would be to breed sows based on the duration ofestrus because most sows ovulate approximately three-quarters of the way through estrus.

Literature Cited

Bates, R. O., B. N. Day, J. H. Britt, L. K. Clark, and M. A. Brauer.1991. Reproductive performance of sows treated with a combina-tion of pregnant mare’s serum gonadotropin and human chori-onic gonadotropin at weaning in the summer. J. Anim. Sci.69:894–898.

Britt, J. H. 1986. Improving sow productivity through managementduring gestation, lactation and after weaning. J. Anim. Sci.63:1288–1296.

Burke, P. 1999. In: L. A. Johnson and H. D. Guthrie (ed.) Proc.IV Int. Conf. on Boar Semen Preservation. Beltsville, MD. pp149–150.

Dewey, C. E., S. W. Martin, R. M. Friendship, and M. R. Wilson.1994. The effects of litter size of previous lactation length andprevious weaning-to-conception interval in Ontario swine. Prev.Vet. Med. 18:213–223.

Estienne, M. J., and T. G. Hartsock. 1998. Effect of exogenous gonado-tropins on the weaning-to-estrus interval in sows. Theriogenol-ogy 49:823–828.

Hurtgen, J. P., and A. D. Leman. 1980. Seasonal influence on thefertility of sows and gilts. J. Am. Vet. Med. Assoc. 177:631–635.

Hurtgen, J. P., and A. D. Leman. 1981. Effect of parity and seasonof farrowing on the subsequent farrowing interval of sows. Vet.Rec. 108:32–34.

Kemp, B., and N. M. Soede. 1996. Relationship of weaning-to-estrusinterval to timing of ovulation and fertilization in sows. J. Anim.Sci. 74:944–949.

Kermabon, A., A. Prunier, J. Djiane, and R. Salesse. 1995. Gonadotro-pins in lactating sows exposed to long or short days duringpregnancy and lactation: Serum concentrations and ovarian re-ceptors. Biol. Reprod. 53:1095–1102.

Kirkwood, R. N., F. X. Aherne, and G. R. Foxcroft. 1998. Effect ofgonadotropin at weaning on reproductive performance of primi-parous sows. Swine Health Prod. 6:51–55.

Knox, R. V., and G. C. Althouse. 1999. Visualizing the reproductivetract of the female pig using real-time ultrasonography. SwineHealth Prod. 7:207–215.

Koketsu, Y., and G. D. Dial. 1997. Factors influencing the postwean-ing reproductive performance of sows on commercial farms. The-riogenology 47:1445–1461.

Koketsu, Y., G. D. Dial, and V. L. King. 1997. Returns to serviceafter mating and removal of sows for reproductive reasons fromcommercial swine farms. Theriogenology 47:1347–1363.

Lancaster, R. T., G. R. Foxcroft, M. P. Boland, and S. Edwards.1985. Fertility of sows injected with exogenous estradiol and/orgonadotropins to control post-weaning estrus. Anim. Reprod.Sci. 8:365–373.

Levis, D. 1997. Effect of lactation length on sow reproductive perfor-mance. University of Nebraska Cooperative Extension EC 97-275-A, Lincoln.

Love, R. J. 1978. Definition of a seasonal infertility problem in pigs.Vet. Rec. 103:443–446.

Love, R. J., G. Evans, and C. Kluiec. 1993. Seasonal effects on fertilityin gilts and sows. J Reprod. Fertil. (Suppl. 1) 48:191–206.

Moeller, S. 2000. Genetic, parity effects on productivity traits. Na-tional Hog Farmer 30:22–28.

Nissen, A. K., N. M. Soede, P. Hyttel, M. Schmidt, and L. D’Hoore.1997. The influence of time of insemination relative to time ofovulation on farrowing frequency and litter size in sows, asinvestigated by ultrasonography. Theriogenology 47:1571–1582.

Saoulidis, K. J., S. K. Kyriakis, S. Samouilidis, A. Tsinas, and K.Chizios. 1995. Ein beitrag zur brunstauslosung und erhohterfruchtbarkeit bei zuchtsauen durch die verabreichung gonado-troper hormone (PMSG und HCG). Tieraerztl. Umsch. 50:13–16.

Sechin, A., J. C. Deschamps, T. Lucia, Jr., J. A. G. Aleixo, and V.Bordignon. 1999. Effect of equine chorionic gonadotropin onweaning-to-first service interval and litter size in female swine.Theriogenology 51:1175–1182.

Soede, N. M., C. C. H. Wetzels, W. Zondag, W. Hazeleger, and B.Kemp. 1995. Effects of a second insemination after ovulation onfertilization rate and accessory sperm count in sows. J. Reprod.Fertil. 105:135–140.

Stork, M. G. 1979. Seasonal reproductive inefficiency in large pigbreeding units in Britain. Vet. Rec. 104:49–52.

Svajgr, A. J., V. W. Hays, G. L. Cromwell, and R. H. Dutt. 1974.Effect of lactation duration on reproductive performance of sows.J. Anim. Sci. 38:100–105.

Van de Wiel, D. F. M., and P. Booman. 1993. Post-weaning anoestrusin primiparous sows: LH patterns and effects of gonadotropininjection and boar exposure. Vet. Q. 15:162–166.

Waberski, D., K. F. Weitze, T. Gleumes, M. Schwarz, T. Willmen,and R. Petzoldt. 1994. Effect of insemination relative to ovulationon fertility with liquid and frozen boar semen. Theriogenology42:831–840.

Xue, J., G. D. Dial, W. E. Marsh, and P. R. Davies. 1994. Multiplemanifestations of season on reproductive performance of com-mercial swine. J. Am. Vet. Med. Assoc. 204:1486–1489.