2.1 Animals and Housing
The work was performed on a commercial farm, not associated with an institution that has an ethics committee. However, this work followed accepted industry and welfare standards of the EU and Spain. The commercial farm was located near Arbeca (Lleida), Spain.
Two consecutive weaning groups of 50 Landrace x Large White sows were used in this study (N = 100 weaned sows). After weaning, sows were housed in conventional breeding crates (2 m x 0.6 m) and were fed 2.8 kg of a commercial sow diet once a day. All sows had free access to water. Sows used in this study were selected based on the criteria of parity (1 – 7), acceptable body condition score (8 mm minimum P2 backfat thickness), previous lactation length (26.3 ± 1.6), and recent prolificacy (litter size = 14.35 ± 0.35). Unhealthy sows, sows with corpus hemorrhagicum (CH) or corpus luteum (CL) by the second day post-weaning, and sows that showed no ovulation after 192 hours post-weaning were excluded from the study. No hormonal or other treatments were used during this study to induce estrus or ovulation. To avoid cross-contamination between the different stimulation methods, all sows within the same weaning group were allocated to the same treatment group which then was changed randomly over time.
- 2 Estrus detection
Sows were checked for estrus behaviors daily at 09:00 and 15:00 starting from the first day post-weaning until sows ceased showing the standing reflex. Boar contact was performed with a trained, adult boar for the live-boar control group (BG). Estrus detection was conducted by a worker leading a boar along the corridor in front of the sows. The boar stood a few minutes in front of a group of 3 sows ensuring nose to nose contact. At the same time, another worker was doing the BPT. The BPT consisted of applying pressure to the back of the sows trying to mimic the tactile stimuli of a boar and evoke the standing reflex .
Estrus detection for the PG was conducted by spraying 4 mL of BB on the snout of each sow and conducting the BPT from 15 seconds after spraying. Three sows were sprayed, then each sow in this group was tested one by one. An audio record of boar courting grunts was played from the beginning of the estrus detection session to stimulate all sows to stand up as well as until the end of the estrus detection session. The live mature boar was not present during the BPT in this group.
The standing reflex was defined as when a sow was motionless with rigid limbs during the BPT . A sow was considered in estrus if she showed the standing reflex during the BPT regardless of the presence of other signs of estrus such as vulva reddening and swelling or mucus discharge. The onset of estrus was defined as the time between two consecutive BPT where sows showed the standing reflex on the second but not the first BPT. The end of estrus was defined as the time between two consecutive BPT where sows showed the standing reflex on the first but not the second BPT. The wean-to-estrus interval (WEI) was the time in hours from weaning until the onset of estrus. Estrus duration in hours was calculated by subtracting the WEI from the end of estrus.
Transrectal ultrasonography was used to determine the time of ovulation. A portable ultrasound (Esaote Mylab Delta, Chagrin Falls, OH, USA) coupled with a microconvex probe was used. Ultrasound images were collected at a frequency of 8.6 MHz. Data interpretation was conducted following Kauffold  methods. The first ultrasound image was collected at 48 hours post-weaning to assess sow reproductive tract and detect a possible lactational estrus. Sows with visible CH or CL at this time were excluded from the study. After 48 h post-weaning, transrectal ultrasonography was performed on each sow every 24 hours until 96 hours post-weaning then every 12 hours until 192 hours post-weaning until detection of ovulation. Ovulation was determined by the presence of CH or CL in the presence or absence of pre-ovulatory follicles. The ovulation time was defined as the time between two transrectal ultrasonography where the first one showed only preovulatory follicles (> 6 mm) and the second one showed CH with no more than one pre-ovulatory follicle. In case that more than one preovulatory follicles were visible with a CH, the ovulation time was the time the image was taken. Follicles were measured vertical to the screen. The weaning to ovulation interval (WOI) was defined as the time in hours from weaning to ovulation. The estrus to ovulation interval (EOI) was defined as the time in hours from the onset of estrus to ovulation. Follicular size at onset of estrus was estimated by linear regression.
2.5 Insemination, reproductive and prolificacy performance
The breeding protocol used in this study was a non-conventional method adapted to the farm needs. Table 1 shows the breeding protocol followed in relative to the onset of estrus. This variable time of insemination was based on the results of previous studies [5, 14, 15] to ensure sows were inseminated 0 to 24 hours before ovulation. The technique used was a conventional insemination with a dose of 80 ml and a minimum of 27.5 million spermatozoid per milliliter provided by the local insemination center. At 25 days post insemination, ultrasound was performed on each sow to confirm pregnancy. Pregnancy was confirmed by the presence of embryonic vesicles. The potential breeding rate was calculated by dividing the number of sows bred over the number of ovulating sows. The percentage of sows inseminated twice was calculated by dividing the number of sows that received two inseminations by the total number of sows bred. The pregnancy rate was defined as the number of sows found pregnant by ultrasound at 25 d over the number of sows bred . Litter size, number of piglets born alive, and stillborn were recorded at farrowing.
2.5 Statistical analysis
The experimental design was a completely random design with two treatments (BG vs PG) with 100 weaned sows enrolled in the study. Each measure was evaluated for adherence to the assumptions of parametric analyses. Non-parametric statistics were used when ANOVA assumptions were not met. The Fisher’s exact test was conducted to evaluate differences in ovulation rate, breeding rate, number of sows inseminated twice, and pregnancy rate. The Mann-Whitney test was used to determine differences in sow reproductive parameters between the BG and the PG. All statistical analyses were conducted in SAS statistical software (SAS version 9.4; SAS Inst., Inc., Cary, NC, USA). A statistically significant difference was declared at a p-value ≤ 0.05.