Relationship between vaginal mucus conductivity and time of ovulation in weaned sows

Abstract

This study investigated whether changes in the vaginal electrical resistance (VER) of vaginal mucus of weaned sows during the first 7 d post-weaning are associated with time of ovulation. Time of ovulation was determined by ovarian ultrasound carried out from 91 to 146 h after weaning and at different seasons. Vaginal electrical resistance was measured at 20, 44, 68, 91, 96, 102, 115, 120, 126, 140, 146, and 164 h post-weaning and was found to decrease between 120 h and 31 h before ovulation and then increase until 40 to 50 h after ovulation. Duration and timing of the nadir was affected by the season (P < 0.01). Estrus was observed from day 4 after the lowest VER values. Ovulation occurred between late day 5 and late day 6, while VER values were still increasing. Ovulation was earlier in lower parity sows (P < 0.001). Compared to 0 h (ovulation time), VER was significantly lower from 50 to 5 h before ovulation in autumn and from 40 to 21 h in winter, but such differences were not seen in spring. Lowest VER value was not correlated with time of ovulation. It was concluded that VER increases before ovulation and, although this increase is influenced by the season, it cannot be used to accurately predict ovulation in weaned sows.

Most artificial insemination (AI) protocols for swine rely on estrus detection in order to estimate time of ovulation. A high variability in the wean-to-estrus and estrus-to-ovulation intervals challenges the efficiency of AI, however, which decreases the fertility of the breeding herd (1). In weaned sows, common estrus synchronization protocols involve the use of exogenous gonadotrophins, such as equine chorionic gonadotrophin (eCG) or the combination of 400 IU eCG and 200 IU hCG (human chorionic gonadotrophin) (PG600) (2). The estrous response to hormonal treatments is affected by parity and season (3), however, which often results in lower farrowing rates and litter sizes (2,4).

Previous studies have reported the use of vaginal electrical resistance (VER) for predicting the estrous response in sows (5–7), given that changes in electrical conductivity of vaginal mucus seem to be related to changes in sexual hormone levels during the estrous cycle (8). In pigs (5), cattle (9), sheep (10), buffaloes (11), and rats (12), VER was observed to decrease during proestrus, followed by a gradual increase during estrus until ovulation. However, the use of VER to predict the time of ovulation has not been assessed in weaned sows. We hypothesized that sows ovulating sooner during the estrous period will show an earlier increase in VER levels than their delayed counterparts and that such an increase may be used to predict the time of ovulation. In order to test this hypothesis, we established 2 different steps: i) profile VER values between day 1 and day 7 after weaning across different seasons and parities, and ii) determine the relationship between VER and time of ovulation in weaned sows.

While transrectal real-time ultrasonography (RTU) accurately determines the time of ovulation in sows, it is relatively invasive and requires some degree of expertise. If VER can be used to predict ovulation in sows, it would represent a user-friendly, non-invasive, simple, and inexpensive method of determining the appropriate timing of insemination. With proper timing of insemination relative to ovulation, only a single insemination may be required to achieve pregnancy.

The objective of this study was to determine whether measurement of VER can be used as a predictor of time of ovulation in weaned sows and to determine possible seasonal variations.

This study was approved by the University of Guelph Animal Care Committee and was conducted on a 700-sow farrow-to-finish facility near Guelph, Ontario. To examine the relationship between VER and ovulation time, 76 mixed-parity Landrace-Yorkshire sows were used in 6 consecutive replicates.

The following groups were formed at weaning: Group 1 — 22 sows weaned June 4, parity range 1 to 12; Group 2 — 12 sows weaned June 10, parity range 1 to 11; Group 3 — 12 sows weaned June 16, parity range 1 to 9; Group 4 — 10 sows weaned October 1, parity range 1 to 12; Group 5 — 10 sows weaned October 8, parity range 1 to 12; and Group 6 — 10 sows weaned February 9, parity range 2 to 10. The groups were further divided according to the season of the year: 46 sows (Groups 1, 2, and 3) were evaluated in spring (March 20 to June 21); 20 sows (Groups 4 and 5) were evaluated in autumn (September 23 to December 21); and 10 sows (Group 6) were evaluated in winter (December 21 to March 20).

Weaned sows were housed individually in breeding stalls (allowing 1.3 m² per sow) and fed daily 2.5 kg of a corn-soybean diet formulated to provide 14.2 MJ ME/kg, 15% crude protein, and 0.61% lysine. Water was available at all times. All sows received 5 min of fence-line boar contact daily for 7 d post-weaning to facilitate estrus detection. Exposure to boars started on the day after weaning to stimulate ovarian activity and to determine the onset of estrus. Boar exposure was maintained until day 7 post-weaning to determine the end of the estrous period.

Ovulation time was determined by transrectal real-time ultrasonography (RTU) at 91, 102, 115, 126, 140, and 146 h post-weaning, using an Aloka SSD 500 (Aloka; Wallingford, Connecticut, USA) with a 7.5 MHz linear array transducer for visualization of the ovaries. Day of weaning was Day 0 and 2.00 pm on weaning day was regarded as 0 h. The disappearance of follicles > 6.5 mm in diameter signaled the start of ovulation. The schedule for VER readings was: Reading 1 (Day 1): 20 h post-weaning (at 10 am); Reading 2 (Day 2): 44 h post-weaning (at 10 am); Reading 3 (Day 3): 68 h post-weaning (at 10 am); Reading 4 (Day 4, morning reading): 91 h post-weaning (at 9 am); Reading 5 (Day 4; afternoon, 1 reading): 96 h post-weaning (at 2 pm); Reading 6 (Day 4; afternoon, 2 readings): 102 h post-weaning (at 8 pm); Reading 7 (Day 5, morning reading): 115 h post-weaning (at 9 am); Reading 8 (Day 5, afternoon, 1 reading): 120 h post-weaning (at 2 pm); Reading 9 (Day 5, afternoon, 2 readings): 126 h post-weaning (at 8 pm); Reading 10 (Day 6, morning reading): 140 h post-weaning (at 10 am); Reading 11 (Day 6, afternoon reading): 146 h post-weaning (at 4 pm); Reading 12 (Day 7): 164 h post-weaning (at 10 am).

The VER was determined using a cylindrical vaginal probe (Draminski; Electronics in Agriculture, Olsztyn, Poland) with 2 ring electrodes on the terminal pole. The resistance was obtained by measuring the voltage developed through the vaginal mucus (ohms; Ω) in response to an alternating electric current excitation at a frequency of 8 mA. The probe was inserted into the vaginal tract until about a quarter of the length of the probe remained outside the vulva. Readings were taken in triplicates by the same operator in order to obtain a mean value.

Data analysis consisted of 2 steps: first, we assessed whether there were significant changes in VER values post-weaning and then we evaluated whether VER changes were associated with the ovulation time of each animal to test the efficacy for prediction. To assess overall changes in post-weaning VER, we normalized the data by grouping VER values of each animal in 10-h intervals before and after their time of ovulation. When 2 or more readings fell within the same interval, the arithmetic mean was calculated. This method allowed overall changes in post-weaning VER to be analyzed independently of the ovulation time (Figures 1a, b, and c). To evaluate whether VER changes were actually associated with the ovulation time and could therefore be used as a predictor for it, we allotted all sows to 4 treatment groups based on their known time of ovulation, i.e., 102, 115, 126, and 140, and then we tested for differences in VER among the 4 groups at each time point between weaning and ovulation, using either the ovulation intervals calculated before or the hours at which VER was measured post-weaning as time points (Figures 2a and b, respectively). We also tested whether there was a correlation between the lowest VER value within each treatment group and the ovulation time.

Figure 1.

Evolution of vaginal electrical resistance (VER) mean values (± SE) before and after the observed time of ovulation in weaned sows in autumn (a), winter (b), and spring (c). * Values significantly different from time 0 at P < 0.05.

Figure 2.

Evolution of vaginal electrical resistance (VER) mean values (± SE) during the first 164 h post-weaning (a) and before and after the observed time of ovulation (b) grouped by sow ovulation category (groups 102, 115, 126, and 140 for sows ovulating 102, 115, 126, and 140 h post-weaning, respectively). * Values significantly different from time 0 at P < 0.05.

Data were analyzed by analysis of variance (ANOVA) using a linear mixed model with repeated measurements in SAS 9.2 (PROC MIXED; SAS Institute, Cary, North Carolina, USA). The linear model included time (h or interval), season, ovulation group, and their meaningful interactions as fixed effects, block as random effect, sow nested in block as repeated measurement, and parity as covariate. Normality of the residuals and presence of outliers were assessed by PROC UNIVARIATE (SAS 9.2) using the Shapiro-Wilk test, Q-Q plots, and externally studentized residuals. When necessary, data were power-transformed by a parameter, ϕ, whose optimal value was estimated using the maximum likelihood (ML) method (13). Preplanned comparisons and linear, quadratic, and cubic trends were computed using orthogonal contrasts obtained by PROC IML (SAS 9.2) and P-values were calculated using Student’s t-tests. Data are presented as least square means ± standard error (SE). Significant effects were considered at P < 0.05. The relationship between the time at lowest VER value and time of ovulation per sow was analyzed using a correlation analysis in SAS (PROC CORR). Pearson correlation coefficient (R) was considered significant at P ≤ 0.05.

All 76 sows monitored in the study showed estrous behavior before day 5 post-weaning: 8 sows (10.5%) at 91 h; 57 (75%) at 102 h; and 11 (14.5%) at 120 h post-weaning. Likewise, all sows in the study ovulated by day 6 post-weaning: 14 sows (18.4%) at 102 h; 12 (15.8%) at 115 h; 46 (60.5%) at 126 h; and 4 (5.3%) at 140 h post-weaning. Average parity of sows was 4.2, 4.2, 4.6, 4.7, 6.0, and 6.5 for groups 1 to 6, respectively. There was a significant effect of parity on the time of ovulation, with younger sows ovulating earlier than their older counterparts (parity 4.21 ± 0.21 at 102 h; 4.06 ± 0.23 at 115 h; 5.30 ± 0.12 at 126 h; and 5.25 ± 0.4 at 140 h post-weaning; P < 0.001).

The VER values showed a cubic trend throughout the study (P = 0.0001); they decreased between 120 and 31 h before ovulation and then increased until 40 to 50 h after ovulation. The maximum mean value measured was 46.5 ± 4.6 Ω at 50 h after ovulation and the minimum was 27.2 ± 2.7 Ω at 30 h before ovulation. Compared to 0 h (ovulation time), VER was significantly lower between 50 and 5 h before ovulation in autumn (P < 0.01; Figure 1a) and between 40 and 21 h before ovulation in winter (P < 0.01; Figure 1b). There were no differences in VER values before ovulation in spring (Figure 1c). Electrical resistance values were not significantly different among sows ovulating at 102, 115, 126, and 140 h at any time point before ovulation (Figure 2a). Electrical resistance at 0 h (ovulation time) was significantly higher for sows ovulating at 126 h, however, than for those ovulating at 115 and 102 h (36.3 ± 0.82, 31.8 ± 1.52, and 31.0 ± 1.54 for 126, 115, and 102 h post-weaning, respectively; P = 0.01) (Figure 2b). There was no relationship between time of lowest VER and time of ovulation (data not shown).

Timing of sperm deposition relative to ovulation is crucial to achieve good fertility rates in weaned sows, especially if using a single timed-AI protocol or reduced sperm numbers per dose. In the present study, we assessed whether VER provided a reliable tool for predicting time of ovulation in weaned sows. Numerous studies have shown that VER fluctuations reflect changes in steroid and gonadotrophic hormone levels (6,8,14). Consequently, vaginal probes have been used during the periestrus period or throughout the estrous cycle for non-invasive monitoring of VER in cyclic gilts and sows in order to detect estrus. To our knowledge, however, there is no information about the potential use of VER to detect ovulation.

Results from this experiment show significant VER changes in weaned sows before ovulation, but do not support our hypothesis that preovulatory VER levels will differ between sows having an early or late ovulation. Patterns of VER in the present study did not differ from previous experiments, with a gradual decrease post-weaning, followed by an increase before ovulation. The decrease in VER levels has been associated with the peak of estradiol that occurs before the luteotropin hormone (LH) surge, while the electrical resistance increased gradually around the LH peak (8). Changes in VER levels in the current study took place over a 4- to 5-d period, while the time difference among sows with an early and late ovulation was less than 30 h. As such, it is likely that changes in VER at different ovulation times are too small to be detected accurately, especially since VER does not seem to be directly associated with ovulation itself, but rather to changes in sexual hormone levels (8,6).

In addition to the lack of periovulatory changes, VER values were highly variable among sows at any given time point. This variability is likely due to the effect of external factors on the sow, such as season, parity, or wean-to-estrus interval (WEI), that act directly at the vaginal level and also through changes in the level of sexual hormones (6,15). A previous study indicated that shorter WEIs were associated with lower VER values earlier in the estrous period (6). Although we did not observe this effect in the present study, we found a significant seasonal effect on VER that could contribute to the high variability among our sows and also to the lack of correlation observed between time at the lowest VER and time of ovulation.

In addition to external factors, the electrical resistance of the vaginal mucus in the sow may also be affected by internal factors such as mucus composition, density, or the sow’s body temperature. Indeed, changes in the mucin content of the vaginal mucus in women influenced the mucus viscoelasticity during the menstrual cycle (16) and changes in the electrolytic composition [sodium, potassium, and chloride ions (Na⁺, K⁺, and Cl⁻)] of the cervical mucus has been associated with changes in estrogen levels (17). It is therefore concluded that, while changes in VER occur before and after ovulation, differences in VER cannot be used to accurately predict the time of ovulation in sows.