TY - JOUR
AU1 - Rault, J.-L.
AU2 - Morrison, R. S.
AU3 - Hansen, C. F.
AU4 - Hansen, L. U.
AU5 - Hemsworth, P. H.
AB - Abstract This project compared the effects of grouping sows after weaning or within 2 d after insemination on sexual behavior, aggression, injuries, stress, and mating success. At weaning (d 0), 360 sows were housed in groups of 10 sows at 4.4 m2 per sow (group weaned [GpW]) or individual stalls (stall weaned [StW]), with 18 groups per treatment. Six days after weaning (d 6), 7 inseminated GpW sows were moved to pens at 2.1 m2 per sow and keeping acquainted sows, and simultaneously, groups of 7 inseminated StW sows were mixed at 2.1 m2 per sow. Group-weaned sows showed greater variation in the onset of estrus (P = 0.02) but not in the length of estrus compared to StW sows (P = 0.21), with 7% fewer GpW sows inseminated within 5 d of weaning (P = 0.05). Group-weaned sows showed lower sexual receptivity scores, showing less spontaneous standing during boar exposure and partly compensating by a greater response to the back-pressure test (both P < 0.01). The GpW treatment also showed greater variability in sows inseminated twice within 6 d of weaning, with 3 out of 18 pens having only 5 mated sows out of 10. Mixing after weaning resulted in higher levels of stress than mixing after insemination, with GpW sows having higher plasma cortisol concentration than StW sows on d 1 (P < 0.001) but no treatment differences on d 7 in cortisol concentration or aggression at feeding (P = 0.48). Group-weaned sows experienced greater weight loss during the first week postweaning (P = 0.05). Anogenital sniffing in GpW sows was frequently observed from d 2 to 5, but mounting and flank nosing increased on d 4 and 5. Frequency of sexual behavior initiated by GpW sows tended to correlate with weight loss (P = 0.08), and sexual behavior received correlated positively with cortisol concentration at d 1 (P = 0.005). In conclusion, sows housed in groups at weaning and regrouped after insemination experienced higher stress than sows housed in individual stalls at weaning and mixed in groups after insemination. This resulted in lower mating success within 5 d of weaning, which in turn increased between-week variability. The lower sexual receptivity in sows grouped at weaning may be due to suppressed estrus-related behaviour, with ovulation occurring, or delayed ovulation beyond d 6. Further research is needed to identify underlying mechanisms to reduce variability, manage aggression and sexual behavior, and optimize estrus detection in group-housed weaning systems. INTRODUCTION A critical time for sow reproduction and productivity is the period around mating. Sows are particularly sensitive to stressors during that period (Knox et al., 2014), hence the reason why the use of mating stalls is still allowed in many countries from a few days to a few weeks after insemination. It is generally agreed that stress can impair reproduction: Turner et al. (2005), in a review of the literature, concluded that prolonged stress and sustained elevation of cortisol can disrupt reproductive processes in female pigs. Hence, mixing shortly after insemination may cause a stress response and affect fertilization and implantation success. The general consensus in the scientific literature is that most pregnancy losses occurs within the first 30 d of gestation, as the embryonic period is a sensitive time when a range of environmental, social, genetic, nutritional, hormonal, and biochemical factors interact with each other, all having significant influences on conception rate and ultimate litter size (Ashworth and Pickard, 1998). Mixing sows directly after weaning could also affect sexual behavior and hence their return to estrus and reproductive success. However, the effects of the housing system around estrus and insemination on sow welfare and reproductive performance have received little research attention. Furthermore, little is known about the impact of stress on sexual behavior in particular. This project examined the effects of grouping sows after weaning or after insemination on sexual behavior, aggression, injuries, stress, and mating success, with the aim of determining the need for more thorough investigation of temporal effects and mechanisms. We hypothesized that sows housed in groups after weaning would experience higher stress from aggression and sexual behavior between weaning and insemination but that group-weaned sows will experience lower stress after insemination compared to sows housed in stalls from weaning to insemination and mixed after insemination. MATERIALS AND METHODS The project was approved by the Rivalea Animal Ethics Committee in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (National Health and Medical Research Council, 2013) and conducted in September corresponding to the spring season in the southern hemisphere. Animals and Housing Treatments On the day of weaning (d 0), 360 Landrace × Large White sows were allocated by parity and estimated BW (based on visual observation) to 1 of 2 treatments: group-weaned (GpW) treatment or stall-weaned (StW) treatment in groups of 10 sows over 3 time replicates, 1 wk apart. Sows with gait scores of 2 or 3 (i.e., lame) on d 0 were not included. We obtained 18 groups per treatment: 6 groups of parity 1 sows and 3 groups each of parities 3, 4, 5, and 6. Out of 36 groups, 22 groups constituted single parities and 10 groups contained 1 or 2 sows of the parity above or beyond, whereas 4 groups were parities 5 and 6 when BW differences prevailed over parity ranking to constitute uniform groups balanced across treatments. For the GpW treatment, 10 sows, unacquainted with each other, were mixed into a pen with 4.4 m2 per sow. Pen design characteristics consisted of a 6.3-by-7-m floor space, including feeding stalls, with 10 full-body-length feeding stalls on the front side of the pen, a 1.5 m strip of slatted floor and the rest as solid concrete floor, and 2 nipple drinkers on each side of the pen. The floor area of the pens was cleaned between each replicate. Sows were fed in the water trough located in front of the stalls with 2.7 kg per sow of a dry commercial pelleted diet dropped into the trough 3 times a day (approximately 0730, 1130, and 1430 h) for a total of approximately 8.1 kg of dry feed per sow daily, hence close to ad libitum feed access from d 1 to 6 (DE: 13 MJ/kg DM; CP: 13.3%; and Lys content: 0.5%). Group-weaned sows were manually locked in the stalls at the first feeding of the day to allow for the subsequent mating check and artificial insemination if in estrus. Sows were not locked in at the other 2 feeding times or at other times of the day. For the StW treatment, sows were individually housed in 2.2-by-0.6-m stalls equipped with a nipple drinker and horizontal bars. Stall-weaned sows were located across the alleyway facing the GpW pens and fed the same diet delivered in a manner identical to the GpW treatment. Stall-weaned sows were managed in “groups” of 10 sows (10 adjacent stalls) to control for the number and management of sows in GpW pens, and therefore the term “group” rather than pen will be used hereafter. Boar Exposure, Estrus Detection, and Artificial Insemination On d 1, a 2-yr-old Large White × Landrace boar was allowed to roam freely down the hallway between the GpW and StW treatments for 1.5 h in the morning with all GpW sows locked in the stalls at the first feeding to provide similar boar fence line stimulation between treatments. From d 2 to 6, all sows were locked in the stalls at feeding and 2 boars in rolling carts were moved down the hallway once a day around 0800 h, with a pause of 30 s in front of each sow. The side of presentation of each boar along the alleyway (GpW or StW treatment side) was reversed each day to balance for boar effect. Daily estrus detection was conducted by 2 stockpeople whose identity varied within and between replicates. Sexual receptivity was scored using a 3-point system specifically developed for this study. A score of 2 was assigned if the sow showed a spontaneous standing response (immobile stance, arched back, and cocked ears) on presentation of the trolley boar. Thereafter, the stockperson performed a “back-pressure test” on each sow by pressing with both hands on the back of the sow (Signoret and Du Mesnil du Buisson, 1961), and a score of 1 was assigned if the sow showed a standing response (immobile stance, arched back, and cocked ears) to the back-pressure test for at least 10 s. A score of 0 was attributed if the sow did not show a spontaneous standing response on presentation of the boar or a standing response to the back-pressure test for at least 10 s. Only sows that scored 1 or 2 were artificially inseminated, using commercial pooled semen. All GpW sows remained locked in the stall for the duration of estrus detection and insemination for the pen. Following the first insemination, sows were subsequently inseminated the following day, irrespective of their receptivity score. Receptivity scores were still recorded for each individual sow from d 2 to 6. No boar was present in the building other than the 2 experimental boars housed approximately 20 m away from experimental sows. Postmating Period Six days after weaning (d 6), 7 sows out of 10 were selected from each group (GpW pen or groups of 10 StW sows) with the inclusion criterion of having been inseminated twice by d 6 (in other words, a first insemination within 5 d after weaning) and the exclusion criterion of lame sows with a gait score of 2 or 3 or a first standing response on d 6, which equated to a single insemination before moving. If more than 7 sows had been inseminated twice, the 3 focal sows selected on d 1 for blood sampling and sows having received their second insemination within 24 h before mixing were selected in priority and the remaining excess sows were excluded from the rest of the experiment at random. If fewer than 7 sows had been inseminated twice, the maximum number of sows inseminated twice was kept but no other sows were added to maintain groups of acquainted sows. The sows were moved to a different building in the morning, 50 m away, and placed in 5.03-by-3.05-m pens with 2.1 m2 per sow including shoulder feeding stalls. Groups originating from GpW treatment or StW treatment were alternated within rows of 12 pens. Pen design characteristics consisted of 10 shoulder stalls with 5 shoulder feeding stalls on either side of the pen, with a third of the area as slatted floor and the rest as concrete floor and 2 nipple drinkers. Sows were fed, in the troughs in the shoulder stalls on either side of the pen, 2.7 kg of dry feed per sow once daily around 0700 h from d 7. After a pregnancy check by ultrasonography 5 wk postinsemination, the ration was reduced to 2.4 kg per sow. The floor area of the pens was cleaned between each replicate. Sows remained in these pens and in their original groups until approximately d 110 of gestation when they were moved to individual farrowing crates. Measurements Weight and Gait Score. All sows were individually weighed at weaning on d 0 and the 7 sows from each group that remained in the postmating period were weighed again on d 6. After weighing, sows were gait scored when walking to their new pen to assess the incidence of lameness. The gait score scale ranged from 0 to 3 and was scored visually by a single observer in all replicates with 0 = normal gait; 1 = irregular gait indicated by a visible degree of difficulty in walking but still using all 4 legs and a swagger of caudal body while walking and shortened stride; 2 = severely lame indicated by a visible reluctance to bear weight on the affected limb; or 3 = no weight bearing on affected limb or total recumbency. Skin Injuries. On d 1 and 7, the following day after weaning and postmating mixing, respectively, each sow was individually assessed for skin injuries by 1 of 3 trained persons at approximately 0900 h, as described by Karlen et al. (2007). Skin injuries were categorized into fresh injuries (scratches, abrasions, cuts, and abscesses) or partially healed or old injuries. Each side of the sow's body was divided into 21 areas for injury data collection (Karlen et al., 2007). The number and the type of skin injuries were recorded, and from these records, the numbers of fresh and total injuries (fresh and old injuries) were collated for each sow on each observation day. Physiology. On d 1, blood samples were collected via jugular venipuncture from 3 focal randomly chosen sows within each group around 1200 h. Blood samples were collected within 2 min of restraint by snaring. On d 7, blood samples were collected from the same 3 focal sows from each pen if these remained in the group postmating. If a focal sow had been discarded, a new focal sow was randomly chosen for d 7. Each 5-mL blood sample was collected in lithium heparin tubes (10-mL lithium-heparinized tubes; BD Vacutainer, North Ryde, NSW, Australia) and subsequently stored on ice before being centrifuged for 10 min at 1,912 × g at 4°C, after which the plasma fraction was transferred to microtubes for long-term storage at –20°C. Plasma concentrations of cortisol were quantified using a commercial RIA kit (Diasorin Ltd., North Ryde, NSW, Australia). The sensitivity of the assay was 1.35 ng/mL. Mean intra-assay CV for low (8 ng/mL) and high (60 ng/mL) plasma samples were 3.2 and 6.8%, respectively. Pregnancy and Farrowing. Regular checks for return to estrus were conducted daily from 3 wk after insemination as well as a pregnancy test using ultrasonography 5 wk after insemination. Sows that returned to estrus, those that tested negative at the pregnancy test, and those with injury or in poor health were removed from treatment pens and not replaced in the groups by other sows. The total number of piglets born and the number of piglets born alive and stillborns and mummified piglets were recorded by stockpeople. Aggressive Behavior. Aggressive behaviors were recorded through plastic infrared dome color closed-circuit television cameras (2.8–12 mm, model 700TVL; Electrogear, Lansvale, NSW, Australia) mounted on the ceiling above each of the GpW treatment pens for d 0 and 1 and by placing Go-Pro cameras (model Hero3 white edition; GoPro Inc., San Mateo, CA) on the feeder line on d 7 for both GpW and StW treatment pens, which were all in identical mixing pens. Continuous observation from these recordings for d 0 and 1 were conducted only on the GpW sows for 3 h after introducing all sows to the pen on d 0 and for 3 h after releasing the sows from the stalls after feeding and boar exposure on d 1. The identity of the initiator and the receiver involved in each interaction was recorded. The frequency of aggressive behavior was recorded as a whole, without attempting to differentiate between different types of aggressive behavior (e.g., pressing, knock, bites), using an ethogram from Hemsworth et al. (2013; Table 1) but with the observation period adapted to this system because sows were locked in for the first feeding and therefore observations around aggression at feeding were not possible. If the interaction stopped for more than 5 s, any new behavior displayed was considered to be part of a new interaction. Aggressive behavioral observations on d 7 were collected for both the GpW and the StW treatment, using the same ethogram as for d 0 and 1 (Table 1). However, aggressive behaviors on d 7 were observed continuously for the first 30 min after the first feeding of being introduced to the pen, recording all interactions delivered and received for each sow, following a previously validated method (Hemsworth et al., 2013), but observing only the first feeding bouts rather than all feeding bouts on the first feeding day following mixing. From the aggressive behavior data on d 0, 1, and 7, the aggression index was calculated for each sow as the ratio of aggression delivered to the total number of aggressive interactions (i.e., aggression delivered/(aggression delivered + aggression received)), varying from 0 to 1. Sows were then classified as “dominant” if they delivered more aggression than they received (aggression index > 0.5), “subdominant” if they received more aggression than they delivered (aggression index > 0.05 and ≤0.5), and “submissive” if they delivered very little or no aggression (aggression index ≤ 0.05). This aggression index is similar to that devised by Mendl et al. (1992). Two observers recorded aggressive behavior with an interobserver reliability above 90% based on frequency of observations. Table 1. Ethogram of behaviors classified as aggressive behavior, adapted from Hemsworth et al. (2013)1 Behavior  Description  Parallel pressing  Pigs stand side by side and push with shoulders against each other, throwing the head against the neck or head of the other  Inverse parallel pressing  Pigs face front to front and then push their shoulders against each other, throwing the head against the neck and flanks of the other.  Head to body knocking  A rapid thrust upward or sideways with the head or snout against any part of the body behind the ears. Most of the knocks are performed against the front half of the receiver. The performer's mouth is shut.  Head to head knocking  A rapid thrust upward or sideways with the head or snout against the neck, head, or ears of the receiver. The performer's mouth is shut.  Bite  A pig delivers a knock with the head against the head, neck, or body of the other pig with the mouth open.  Behavior  Description  Parallel pressing  Pigs stand side by side and push with shoulders against each other, throwing the head against the neck or head of the other  Inverse parallel pressing  Pigs face front to front and then push their shoulders against each other, throwing the head against the neck and flanks of the other.  Head to body knocking  A rapid thrust upward or sideways with the head or snout against any part of the body behind the ears. Most of the knocks are performed against the front half of the receiver. The performer's mouth is shut.  Head to head knocking  A rapid thrust upward or sideways with the head or snout against the neck, head, or ears of the receiver. The performer's mouth is shut.  Bite  A pig delivers a knock with the head against the head, neck, or body of the other pig with the mouth open.  1An interruption of more than 5 s was considered a new bout. View Large Table 1. Ethogram of behaviors classified as aggressive behavior, adapted from Hemsworth et al. (2013)1 Behavior  Description  Parallel pressing  Pigs stand side by side and push with shoulders against each other, throwing the head against the neck or head of the other  Inverse parallel pressing  Pigs face front to front and then push their shoulders against each other, throwing the head against the neck and flanks of the other.  Head to body knocking  A rapid thrust upward or sideways with the head or snout against any part of the body behind the ears. Most of the knocks are performed against the front half of the receiver. The performer's mouth is shut.  Head to head knocking  A rapid thrust upward or sideways with the head or snout against the neck, head, or ears of the receiver. The performer's mouth is shut.  Bite  A pig delivers a knock with the head against the head, neck, or body of the other pig with the mouth open.  Behavior  Description  Parallel pressing  Pigs stand side by side and push with shoulders against each other, throwing the head against the neck or head of the other  Inverse parallel pressing  Pigs face front to front and then push their shoulders against each other, throwing the head against the neck and flanks of the other.  Head to body knocking  A rapid thrust upward or sideways with the head or snout against any part of the body behind the ears. Most of the knocks are performed against the front half of the receiver. The performer's mouth is shut.  Head to head knocking  A rapid thrust upward or sideways with the head or snout against the neck, head, or ears of the receiver. The performer's mouth is shut.  Bite  A pig delivers a knock with the head against the head, neck, or body of the other pig with the mouth open.  1An interruption of more than 5 s was considered a new bout. View Large Sexual Behavior. Sexual behaviors were recorded through the plastic IR dome color CCTV cameras mounted on the ceiling above each of the GpW treatment pen. Sexual behavioral observations were conducted daily from d 2 to 5 using one–zero sampling during 5 min intervals every 30 min from 0730 to 1730 h daily. The ethogram specifically developed for this study consisted of the behavior displayed by the initiator of the sexual interaction—flank nosing, attempt to mount, unsuccessful mount, mounting, and anogenital sniffing—and the behavior displayed by the receiver—standing posture, flee to pen, flee to stall, aggressive, and no reaction (Table 2). The identity of the initiator and the receiver involved in each interaction were also recorded. If the interaction stopped for more than 5 s, any new behavior displayed was considered to be part of a new interaction. Three observers recorded sexual behavior with an interobserver reliability above 90% based on frequency of observations. Table 2. Ethogram used for sexual behavior recording1 Behavior – initiator  Description  Flank nosing  Repeated contact with the flat tip of the nose to the flank area of another sow, where “contact” is the snout touching or appearing to touch, that is, within 5 cm of the flank, and “repeated” is 2 or more sequential contacts or head movements  Anogenital sniffing  Flat tip of the snout of the sow is orientated toward, at the same height as, and within 1 head distance of the vulva of another sow  Mount  Two legs on the back of another sow  Attempt to mount  The initiator is standing with the head or 1 leg on the back of another sow  Unsuccessful mount  Slipped on the floor, or attempt to mount bout <2 s  Behavior – initiator  Description  Flank nosing  Repeated contact with the flat tip of the nose to the flank area of another sow, where “contact” is the snout touching or appearing to touch, that is, within 5 cm of the flank, and “repeated” is 2 or more sequential contacts or head movements  Anogenital sniffing  Flat tip of the snout of the sow is orientated toward, at the same height as, and within 1 head distance of the vulva of another sow  Mount  Two legs on the back of another sow  Attempt to mount  The initiator is standing with the head or 1 leg on the back of another sow  Unsuccessful mount  Slipped on the floor, or attempt to mount bout <2 s  Behavior – receiver  Description  Standing posture  Immobile and stand upright for >5 s, immobile stance, arched back, and cocked ears  Flee to pen  Movement of 2 steps or more by the receiving sow away from the initiator to another location in the pen  Flee to stall  Movement of 2 steps or more by the receiving sow away from the initiator to a stall (including from 1 stall to a new stall)  Aggressive  Bite, knock or push  No reaction  No obvious behavioral reaction to the initiator  Behavior – receiver  Description  Standing posture  Immobile and stand upright for >5 s, immobile stance, arched back, and cocked ears  Flee to pen  Movement of 2 steps or more by the receiving sow away from the initiator to another location in the pen  Flee to stall  Movement of 2 steps or more by the receiving sow away from the initiator to a stall (including from 1 stall to a new stall)  Aggressive  Bite, knock or push  No reaction  No obvious behavioral reaction to the initiator  1Each behavior was scored as a 1 if it occurred within the sexual interaction bout observed or as a 0 if it did not occur. Behaviors were not mutually exclusive. An interruption of more than 5 s was considered a new bout. View Large Table 2. Ethogram used for sexual behavior recording1 Behavior – initiator  Description  Flank nosing  Repeated contact with the flat tip of the nose to the flank area of another sow, where “contact” is the snout touching or appearing to touch, that is, within 5 cm of the flank, and “repeated” is 2 or more sequential contacts or head movements  Anogenital sniffing  Flat tip of the snout of the sow is orientated toward, at the same height as, and within 1 head distance of the vulva of another sow  Mount  Two legs on the back of another sow  Attempt to mount  The initiator is standing with the head or 1 leg on the back of another sow  Unsuccessful mount  Slipped on the floor, or attempt to mount bout <2 s  Behavior – initiator  Description  Flank nosing  Repeated contact with the flat tip of the nose to the flank area of another sow, where “contact” is the snout touching or appearing to touch, that is, within 5 cm of the flank, and “repeated” is 2 or more sequential contacts or head movements  Anogenital sniffing  Flat tip of the snout of the sow is orientated toward, at the same height as, and within 1 head distance of the vulva of another sow  Mount  Two legs on the back of another sow  Attempt to mount  The initiator is standing with the head or 1 leg on the back of another sow  Unsuccessful mount  Slipped on the floor, or attempt to mount bout <2 s  Behavior – receiver  Description  Standing posture  Immobile and stand upright for >5 s, immobile stance, arched back, and cocked ears  Flee to pen  Movement of 2 steps or more by the receiving sow away from the initiator to another location in the pen  Flee to stall  Movement of 2 steps or more by the receiving sow away from the initiator to a stall (including from 1 stall to a new stall)  Aggressive  Bite, knock or push  No reaction  No obvious behavioral reaction to the initiator  Behavior – receiver  Description  Standing posture  Immobile and stand upright for >5 s, immobile stance, arched back, and cocked ears  Flee to pen  Movement of 2 steps or more by the receiving sow away from the initiator to another location in the pen  Flee to stall  Movement of 2 steps or more by the receiving sow away from the initiator to a stall (including from 1 stall to a new stall)  Aggressive  Bite, knock or push  No reaction  No obvious behavioral reaction to the initiator  1Each behavior was scored as a 1 if it occurred within the sexual interaction bout observed or as a 0 if it did not occur. Behaviors were not mutually exclusive. An interruption of more than 5 s was considered a new bout. View Large Statistical Analyses All data were checked for normality and homogeneity of variance, and transformations were applied as necessary (logarithmic or square root transformation). Continuous data were analyzed using a mixed model (PROC MIXED; SAS Inst. Inc., Cary, NC). The model included the fixed effects of treatment (GpW vs. StW), time replicate (1 to 3), and parity (1 and 3 to 6) and age of the previous litter at weaning (ranging from 25 to 34 d of age) if significant. Group nested within treatment and time replicate was included as a random effect in the model to control for the fact that sows where allocated to groups (equivalent to pens for GpW sows and groups of 10 stalls for StW sows) and groups where allocated to replicates as a blocking factor, with the group as experimental unit. However, for aggression at d 0 and 1 and sexual behavior, the fixed effect of treatment was removed from the model because these data were collected only for GpW sows and the group nested within replicate was included as a fixed effect instead of as a random effect. Furthermore, sexual behavior were pooled by day and analyzed as repeated measures across d 2 to 5 using sow nested within group as the repeated subject. When significant differences (P < 0.05) were detected, Tukey–Kramer adjustments were used to account for the number of pairwise comparisons. If the data were not normally distributed and could not be transformed to fit normality, data were analyzed using the Kruskal–Wallis nonparametric test of SAS (PROC NPAR1WAY; stillborn piglets, and mummified piglets) or the Welch's t test if the homogeneity of variance assumption was violated (wean-to-mate interval). For categorical data, the χ2 test of SAS was used (PROC FREQ) and Fisher's test was used if some of the cells had a frequency of 5 or less. Pearson correlations of SAS (PROC CORR) were conducted between variables of interest. Data are presented as least square means ± SE unless otherwise stated. RESULTS Wean-to-Mate Interval and Sexual Receptivity Test Overall, 89.4% of the sows showed a standing response to the sexual receptivity test with a score of 1 (response to the back-pressure test in the presence of the boar) or 2 (spontaneous response to the initial presence of the boar) between d 2 and 6, with 19 sows in each of the GpW and StW treatment failing to respond (χ2 test P = 1.00). Nine StW sows responded to the sexual receptivity test as soon as d 2 (scores of 1 or 2) whereas only 1 GpW sow responded on d 2 (Fisher's test P = 0.02; Table 3), but treatments did not differ on d 3 or 4 (Fisher's test P = 0.12 and χ2 test P = 0.27, respectively). More StW sows responded on d 5 whereas more GpW sows responded on d 6 (χ2 tests P = 0.002 and P < 0.001, respectively). The overall length of estrus, by summing the number of days that each sow scored 1 or 2 to the receptivity test, did not differ between GpW and StW sows (2.09 ± 0.07 and 2.23 ± 0.07 d, respectively; P = 0.21). The length of the spontaneous standing to boar exposure (score 2) was shorter in GpW sows than in StW sows (1.49 ± 0.08 and 1.90 ± 0.07 d, respectively; P < 0.001). However, GpW sows had a longer length of response to the back-pressure test (score 1) than StW sows (0.56 ± 0.06 and 0.31 ± 0.06 d, respectively; P = 0.003). Duration of estrus or wean-to-mate interval, based on receptivity scores, did not differ according to the aggression index at d 1 (χ2 tests P > 0.1) and was not correlated with cortisol on d 1, aggression on d 0 or 1, or fresh lesions on d 1 (all P > 0.1). Table 3. Sexual receptivity scores per treatment from d 2 to 6, shown as number of sows scored as 1 or 2 for group-weaned (GpW; n = 180) and stall-weaned (StW; n = 180) treatments Score frequency  d 2  d 3  d 4  d 5  d 6  Overall  StW = 1  7  4  29  8  8  56  StW = 2  2  7  74  147  112  342  GpW = 1  1  5  31  25  39  101  GpW = 2  0  1  60  120  87  268  Overall StW  9  11  103  155  120  398  Overall GpW  1  6  91  145  126  369  P-value between overall scores  P = 0.02  P = 0.12  P = 0.27  P = 0.002  P < 0.001    Score frequency  d 2  d 3  d 4  d 5  d 6  Overall  StW = 1  7  4  29  8  8  56  StW = 2  2  7  74  147  112  342  GpW = 1  1  5  31  25  39  101  GpW = 2  0  1  60  120  87  268  Overall StW  9  11  103  155  120  398  Overall GpW  1  6  91  145  126  369  P-value between overall scores  P = 0.02  P = 0.12  P = 0.27  P = 0.002  P < 0.001    View Large Table 3. Sexual receptivity scores per treatment from d 2 to 6, shown as number of sows scored as 1 or 2 for group-weaned (GpW; n = 180) and stall-weaned (StW; n = 180) treatments Score frequency  d 2  d 3  d 4  d 5  d 6  Overall  StW = 1  7  4  29  8  8  56  StW = 2  2  7  74  147  112  342  GpW = 1  1  5  31  25  39  101  GpW = 2  0  1  60  120  87  268  Overall StW  9  11  103  155  120  398  Overall GpW  1  6  91  145  126  369  P-value between overall scores  P = 0.02  P = 0.12  P = 0.27  P = 0.002  P < 0.001    Score frequency  d 2  d 3  d 4  d 5  d 6  Overall  StW = 1  7  4  29  8  8  56  StW = 2  2  7  74  147  112  342  GpW = 1  1  5  31  25  39  101  GpW = 2  0  1  60  120  87  268  Overall StW  9  11  103  155  120  398  Overall GpW  1  6  91  145  126  369  P-value between overall scores  P = 0.02  P = 0.12  P = 0.27  P = 0.002  P < 0.001    View Large Seven percent fewer GpW sows were inseminated within 5 d after weaning than StW sows (81.7 and 88.9%, respectively; χ2 test P = 0.05). When considering all sows, the wean-to-mate interval was longer by 1 d in GpW sows compared to StW sows (Welch's t test: 6.10 ± 0.37 vs. 5.08 ± 0.23 d; P = 0.02). However, the wean-to-mate interval (mating being based on the first insemination, corresponding to the onset of estrus) contained numerous outliers, extending from 2 to 28 d, with about two-thirds of these outliers being GpW sows and one-third StW sows and the CV for the wean-to-mate interval being greater for the GpW treatment than the StW treatment (CV: 0.63 and 0.37, respectively; P = 0.02). When considering the sows mated within 20 d of weaning (n = 342 sows), which could be considered to include delayed ovulation but exclude undetected ovulation (e.g., “silent” estrus; n = 12 sows with a wean-to-mate interval between 20 and 28 d, with 9 GpW sows and 3 StW sows), the wean-to-mate interval was not different between GpW and StW sows (5.10 ± 0.22 and 4.82 ± 0.21 d, respectively; P = 0.36). Because only sows that had received 2 inseminations by d 6 continued on the experiment and a decision was made not to introduce unacquainted sows into the group if fewer than 7 sows had been inseminated, 2 GpW groups consisted of 6 sows at d 6 and 3 GpW groups consisted of 5 sows out of the initial groups of 10 sows, whereas only 1 StW group consisted of 6 sows at d 6. This translated into 243 sows remaining on trial from d 6 until farrowing: 118 GpW sows and 125 StW sows. Body Weight and Gait Score According to the random selection of sows, treatments did not differ in BW on d 0 (n = 360; mean ± SE: 246.2 ± 2.2; P = 0.62). For the sows that remained on trial on d 6, GpW sows were 3 kg lighter in BW compared to StW sows (241.2 ± 1.1 and 244.3 ± 1.1 kg, respectively; n = 243; P = 0.05), accounting for initial weight on d 0 in the model (P < 0.001). Consequently, GpW sows lost more weight over that first week after weaning in comparison to StW sows (n = 243; BW d 6 to 0: –2.81 ± 1.12 and +0.32 ± 1.09 kg, respectively; P = 0.05). Very few poor gait scores were recorded, with 4 GpW sows and 2 StW sows scoring 1 on d 0 and the rest scoring 0 (n = 360; Fisher's test P = 0.41). Treatments did not differ in gait score on d 6, with 9 GpW sows scoring 1 and 1 GpW sow scoring 2 whereas 4 StW sows scored 1 (n = 243; Fisher's test P = 0.18). Aggressive Behavior, Skin Injuries, and Cortisol Aggression Day 0. Aggression between GpW sows over the first 3 h after mixing differed between groups (P < 0.001), with a mean of 9.2 bouts of aggression per sow over these 3 h and a range from 4 to 15 bouts per sow among groups. The aggression index on d 0, that is, whether a sow classified as a dominant, subdominant, or submissive, did not differ between groups (χ2 tests P > 0.05), with overall 30% GpW sows classified as dominant, 7% GpW sows as subdominant, and 63% GpW sows as submissive. Aggression on d 0 did not correlate with the wean-to-mate interval, length of estrus, or weight gain (P > 0.1; Table 4). Table 4. Pearson correlations between variables of interest collected between weaning and mating in group-weaned sows (r-values)   Aggression delivered d 01  Aggression received d 0  Aggression delivered d 12  Aggression received d 12  Plasma cortisol d 1  Fresh skin injuries d 1  Weight gain d 6–0  Aggression delivered d 01                Aggression received d 0  0.66***              Aggression delivered d 12  0.29***  0.14†            Aggression received d 1l  0.01  0.05  0.64***          Plasma cortisol d 1  –0.28*  –0.18  –0.01  0.20        Fresh skin injuries d 1  0.18*  0.19**  0.02  –0.01  0.20*      Weight gain d 6–03  –0.03  0.02  –0.02  –0.02  –0.18†  –0.13*      Aggression delivered d 01  Aggression received d 0  Aggression delivered d 12  Aggression received d 12  Plasma cortisol d 1  Fresh skin injuries d 1  Weight gain d 6–0  Aggression delivered d 01                Aggression received d 0  0.66***              Aggression delivered d 12  0.29***  0.14†            Aggression received d 1l  0.01  0.05  0.64***          Plasma cortisol d 1  –0.28*  –0.18  –0.01  0.20        Fresh skin injuries d 1  0.18*  0.19**  0.02  –0.01  0.20*      Weight gain d 6–03  –0.03  0.02  –0.02  –0.02  –0.18†  –0.13*    1Data were analyzed using the square root transformation. 2Data were analyzed using the logarithmic transformation (x + 1). 3n = 118 sows. †P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001. View Large Table 4. Pearson correlations between variables of interest collected between weaning and mating in group-weaned sows (r-values)   Aggression delivered d 01  Aggression received d 0  Aggression delivered d 12  Aggression received d 12  Plasma cortisol d 1  Fresh skin injuries d 1  Weight gain d 6–0  Aggression delivered d 01                Aggression received d 0  0.66***              Aggression delivered d 12  0.29***  0.14†            Aggression received d 1l  0.01  0.05  0.64***          Plasma cortisol d 1  –0.28*  –0.18  –0.01  0.20        Fresh skin injuries d 1  0.18*  0.19**  0.02  –0.01  0.20*      Weight gain d 6–03  –0.03  0.02  –0.02  –0.02  –0.18†  –0.13*      Aggression delivered d 01  Aggression received d 0  Aggression delivered d 12  Aggression received d 12  Plasma cortisol d 1  Fresh skin injuries d 1  Weight gain d 6–0  Aggression delivered d 01                Aggression received d 0  0.66***              Aggression delivered d 12  0.29***  0.14†            Aggression received d 1l  0.01  0.05  0.64***          Plasma cortisol d 1  –0.28*  –0.18  –0.01  0.20        Fresh skin injuries d 1  0.18*  0.19**  0.02  –0.01  0.20*      Weight gain d 6–03  –0.03  0.02  –0.02  –0.02  –0.18†  –0.13*    1Data were analyzed using the square root transformation. 2Data were analyzed using the logarithmic transformation (x + 1). 3n = 118 sows. †P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001. View Large Aggression Day 1. Aggression between GpW sows over the first 3 h after release from the stall after first feeding on the day following mixing tended to differ between groups (P = 0.08), with a mean of 2.4 bouts of aggression per sow over these 3 h and a range from 1 to 4 bouts per sow among groups. The aggression index on d 1 did not differ between groups (χ2 tests P > 0.1), with overall 29% GpW sows classified as dominant, 34% GpW sows as subdominant, and 37% GpW sows as submissive. Aggression delivered correlated between d 0 and 1 (P < 0.001; Table 4) but not aggression received between d 0 and 1 (P = 0.52). Aggression on d 1 did not correlate with the wean-to-mate interval, length of estrus, or weight gain (P > 0.1; Table 4). Aggression Day 7. Group-weaned and StW groups did not differ in aggression over the 30 min after first feeding on d 7 (P = 0.37; Table 5) or in the aggression index at d 7 (χ2 test P = 0.90; Table 5). Table 5. Aggressive behavior on d 7, cortisol concentrations on d 1 and 7, and skin injuries for d 1 and 7 for group-weaned (GpW; n = 118) and stall-weaned (StW; n = 125) treatments (least square means ± SEM) Variables  GpW  StW  SEM  P-value  Aggression delivered over the first 30 min after first feeding on d 7, frequency per sow1  8.3 (4.9–13.5)  11.3 (6.8–18.3)    P = 0.37  Aggression index d 7: dominant/subdominant/submissive, %  33/41/26  36/38/26    P = 0.90  Cortisol, ng/mL2      d 1  27.4  19.4  2.1  P < 0.001      d 7  11.8  13.8  2.8  P = 0.48  Fresh skin injuries, number      d 1  16.5  5.9  2.1  P < 0.001      d 7  7.1  24.1  3.1  P < 0.001  Total skin injuries, number      d 1  17.0  6.5  2.2  P < 0.001      d 7  30.2  31.0  3.0  P < 0.001  Variables  GpW  StW  SEM  P-value  Aggression delivered over the first 30 min after first feeding on d 7, frequency per sow1  8.3 (4.9–13.5)  11.3 (6.8–18.3)    P = 0.37  Aggression index d 7: dominant/subdominant/submissive, %  33/41/26  36/38/26    P = 0.90  Cortisol, ng/mL2      d 1  27.4  19.4  2.1  P < 0.001      d 7  11.8  13.8  2.8  P = 0.48  Fresh skin injuries, number      d 1  16.5  5.9  2.1  P < 0.001      d 7  7.1  24.1  3.1  P < 0.001  Total skin injuries, number      d 1  17.0  6.5  2.2  P < 0.001      d 7  30.2  31.0  3.0  P < 0.001  1Data were analyzed using the logarithmic transformation log(x + 1) and is presented as back-transformed with the 95% Confidence Interval in parenthesis. 2n = 54 sows per treatment. View Large Table 5. Aggressive behavior on d 7, cortisol concentrations on d 1 and 7, and skin injuries for d 1 and 7 for group-weaned (GpW; n = 118) and stall-weaned (StW; n = 125) treatments (least square means ± SEM) Variables  GpW  StW  SEM  P-value  Aggression delivered over the first 30 min after first feeding on d 7, frequency per sow1  8.3 (4.9–13.5)  11.3 (6.8–18.3)    P = 0.37  Aggression index d 7: dominant/subdominant/submissive, %  33/41/26  36/38/26    P = 0.90  Cortisol, ng/mL2      d 1  27.4  19.4  2.1  P < 0.001      d 7  11.8  13.8  2.8  P = 0.48  Fresh skin injuries, number      d 1  16.5  5.9  2.1  P < 0.001      d 7  7.1  24.1  3.1  P < 0.001  Total skin injuries, number      d 1  17.0  6.5  2.2  P < 0.001      d 7  30.2  31.0  3.0  P < 0.001  Variables  GpW  StW  SEM  P-value  Aggression delivered over the first 30 min after first feeding on d 7, frequency per sow1  8.3 (4.9–13.5)  11.3 (6.8–18.3)    P = 0.37  Aggression index d 7: dominant/subdominant/submissive, %  33/41/26  36/38/26    P = 0.90  Cortisol, ng/mL2      d 1  27.4  19.4  2.1  P < 0.001      d 7  11.8  13.8  2.8  P = 0.48  Fresh skin injuries, number      d 1  16.5  5.9  2.1  P < 0.001      d 7  7.1  24.1  3.1  P < 0.001  Total skin injuries, number      d 1  17.0  6.5  2.2  P < 0.001      d 7  30.2  31.0  3.0  P < 0.001  1Data were analyzed using the logarithmic transformation log(x + 1) and is presented as back-transformed with the 95% Confidence Interval in parenthesis. 2n = 54 sows per treatment. View Large Plasma Cortisol Concentration. Group-weaned sows had higher cortisol concentrations than StW sows on d 1, the day following weaning (P < 0.001; Table 5), and parity had an effect (P = 0.002), with parity 1 sows having lower cortisol concentrations than parity 4 and 6 sows (Tukey post hoc tests P = 0.002 and P = 0.05, respectively). Cortisol concentration on d 1 correlated negatively with aggression delivered on d 0 (P = 0.05; Table 4) but did not correlate with aggression received on d 0 or delivered or received on d 1 (P > 0.1). Cortisol concentration on d 1 tended to negatively correlate with weight gain (P = 0.10; Table 4) but did not correlate with the wean-to-mate interval or length of estrus (P > 0.1). Group-weaned sows and StW sows did not differ in cortisol concentrations on d 7, the day following grouping for the StW sows or regrouping for GpW sows (P = 0.48). Cortisol concentration on d 7 correlated negatively with aggression delivered on d 7 (P = 0.02; Table 6). Table 6. Pearson correlations between variables of interest collected after mating for stall-weaned sows previously housed in stalls or group-weaned sows (n = 243; r-values)   Plasma cortisol d 1  Fresh skin injuries d 1  Plasma cortisol d 7  Fresh skin injuries d 7  Aggression delivered d 71  Aggression received d 71  Plasma cortisol d 12              Fresh skin injuries d 1  0.20*            Plasma cortisol d 7  –0.10  –0.14          Fresh skin injuries d 7  –0.16  –0.09  0.29**        Aggression delivered d 71  –0.20†  0.01  –0.23*  0.00      Aggression received d 71  –0.16  0.00  –0.18†  –0.01  0.63***      Plasma cortisol d 1  Fresh skin injuries d 1  Plasma cortisol d 7  Fresh skin injuries d 7  Aggression delivered d 71  Aggression received d 71  Plasma cortisol d 12              Fresh skin injuries d 1  0.20*            Plasma cortisol d 7  –0.10  –0.14          Fresh skin injuries d 7  –0.16  –0.09  0.29**        Aggression delivered d 71  –0.20†  0.01  –0.23*  0.00      Aggression received d 71  –0.16  0.00  –0.18†  –0.01  0.63***    1Data were analyzed using the logarithmic transformation (x + 1). 2n = 78 sows. †P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001. View Large Table 6. Pearson correlations between variables of interest collected after mating for stall-weaned sows previously housed in stalls or group-weaned sows (n = 243; r-values)   Plasma cortisol d 1  Fresh skin injuries d 1  Plasma cortisol d 7  Fresh skin injuries d 7  Aggression delivered d 71  Aggression received d 71  Plasma cortisol d 12              Fresh skin injuries d 1  0.20*            Plasma cortisol d 7  –0.10  –0.14          Fresh skin injuries d 7  –0.16  –0.09  0.29**        Aggression delivered d 71  –0.20†  0.01  –0.23*  0.00      Aggression received d 71  –0.16  0.00  –0.18†  –0.01  0.63***      Plasma cortisol d 1  Fresh skin injuries d 1  Plasma cortisol d 7  Fresh skin injuries d 7  Aggression delivered d 71  Aggression received d 71  Plasma cortisol d 12              Fresh skin injuries d 1  0.20*            Plasma cortisol d 7  –0.10  –0.14          Fresh skin injuries d 7  –0.16  –0.09  0.29**        Aggression delivered d 71  –0.20†  0.01  –0.23*  0.00      Aggression received d 71  –0.16  0.00  –0.18†  –0.01  0.63***    1Data were analyzed using the logarithmic transformation (x + 1). 2n = 78 sows. †P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001. View Large Seventy-two percent (n = 78) of the focal sows remained on trial beyond d 6, with a treatment × day interaction (P = 0.003); GpW sows had higher cortisol concentrations than StW sows on d 1 (Tukey post hoc test P < 0.001), but there was no difference on d 7 (Tukey post hoc test P = 0.42), with both GpW and StW sows showing lower cortisol concentrations on d 7 than on d 1 (Tukey post hoc tests: P < 0.001 and P = 0.04, respectively). Skin Injuries. Group-weaned sows had higher total and fresh skin injuries than StW sows on d 1 (P < 0.001), the day following weaning. Total and fresh skin injuries on d 1 were positively correlated with cortisol concentration on d 1 (both r = 0.20, P = 0.04; Table 5) and aggression delivered and received on d 0 (P = 0.02 and P = 0.01) and negatively correlated with weight gain (both r = –0.13, P = 0.04). However, total and fresh injuries did not correlate with the wean-to-mate interval or the length of estrus (P > 0.1). On d 7, StW sows weaned in stalls and recently mixed had higher fresh skin injuries than GpW sows (P < 0.001; Table 5). Treatments did not differ for total (i.e., fresh and old) skin injuries on d 7 (P = 0.78). Total and fresh skin injuries at d 7 were positively correlated with cortisol concentration at d 7 (r = 0.20, P = 0.04, and r = 0.30, P = 0.003, respectively; Table 6) but did not correlate with aggression delivered or received on d 7 (P > 0.1). Sexual Behavior The number of sexual interactions between GpW sows varied according to the day (P < 0.001), increasing from d 2 and 3 to d 4, and remained high on d 5 (Tukey post hoc tests P < 0.001; Table 7). Sexual interactions initiated also varied according to the aggression index based on d 7 (n = 118; P < 0.001), with subdominant sows initiating less sexual behavior than dominant and submissive sows (0.22 ± 0.04 bouts, 0.47 ± 0.04 bouts, and 0.39 ± 0.05 bouts, respectively; Tukey post hoc tests P < 0.001 and P = 0.04, respectively). Table 7. Sexual behavior between group-weaned sows by day (n = 180; least square means ± SEM) Variables  d 2  d 3  d 4  d 5  SEM  P-value  Sexual interactions, average number of bouts observed per sow per 5 min scan      Sexual behavior  0.19a  0.27a  0.59c  0.41b  0.03  P < 0.001  Specific sexual behavior initiated, % of bouts1      Flank nosing  10.4a  11.6a  11.6a  21.5b  2.4  P < 0.001      Anogenital sniffing  88.5a  85.0a  82.6a  71.5b  0.5  P < 0.001      Mounts  0.1a  0.3a  1.7b  1.1ab  0.5  P = 0.008      Attempts to mount  0.9a  3.0ab  3.3ab  3.8b  1.0  P = 0.04      Unsuccessful mounts  0.2a  0.2a  0.8ab  2.0b  0.5  P < 0.001      All forms of mounts2  1.1a  3.5ab  5.8b  7.0b  1.5  P < 0.001  Specific sexual behavior received, % of bouts1      No reaction  62.4a  56.8ab  51.4b  54.9ab  3.3  P = 0.01      Standing posture  0.1  0.5  0.7  0.1  0.3  P = 0.06      Flee to pen  4.7a  6.7ab  9.0b  11.4b  1.8  P = 0.002      Flee to stall  0.4  1.1  1.5  1.3  0.5  P = 0.18      Aggression  0.7  0.8  1.9  2.8  0.9  P = 0.09  Variables  d 2  d 3  d 4  d 5  SEM  P-value  Sexual interactions, average number of bouts observed per sow per 5 min scan      Sexual behavior  0.19a  0.27a  0.59c  0.41b  0.03  P < 0.001  Specific sexual behavior initiated, % of bouts1      Flank nosing  10.4a  11.6a  11.6a  21.5b  2.4  P < 0.001      Anogenital sniffing  88.5a  85.0a  82.6a  71.5b  0.5  P < 0.001      Mounts  0.1a  0.3a  1.7b  1.1ab  0.5  P = 0.008      Attempts to mount  0.9a  3.0ab  3.3ab  3.8b  1.0  P = 0.04      Unsuccessful mounts  0.2a  0.2a  0.8ab  2.0b  0.5  P < 0.001      All forms of mounts2  1.1a  3.5ab  5.8b  7.0b  1.5  P < 0.001  Specific sexual behavior received, % of bouts1      No reaction  62.4a  56.8ab  51.4b  54.9ab  3.3  P = 0.01      Standing posture  0.1  0.5  0.7  0.1  0.3  P = 0.06      Flee to pen  4.7a  6.7ab  9.0b  11.4b  1.8  P = 0.002      Flee to stall  0.4  1.1  1.5  1.3  0.5  P = 0.18      Aggression  0.7  0.8  1.9  2.8  0.9  P = 0.09  a,bWithin a row, means without a common superscript differ (P < 0.05). 1Specific sexual behaviors were not mutually exclusive; hence, they could occur in the same bout of sexual behavior initiated or received. 2Mounts, attempts to mount, and unsuccessful mounts combined. View Large Table 7. Sexual behavior between group-weaned sows by day (n = 180; least square means ± SEM) Variables  d 2  d 3  d 4  d 5  SEM  P-value  Sexual interactions, average number of bouts observed per sow per 5 min scan      Sexual behavior  0.19a  0.27a  0.59c  0.41b  0.03  P < 0.001  Specific sexual behavior initiated, % of bouts1      Flank nosing  10.4a  11.6a  11.6a  21.5b  2.4  P < 0.001      Anogenital sniffing  88.5a  85.0a  82.6a  71.5b  0.5  P < 0.001      Mounts  0.1a  0.3a  1.7b  1.1ab  0.5  P = 0.008      Attempts to mount  0.9a  3.0ab  3.3ab  3.8b  1.0  P = 0.04      Unsuccessful mounts  0.2a  0.2a  0.8ab  2.0b  0.5  P < 0.001      All forms of mounts2  1.1a  3.5ab  5.8b  7.0b  1.5  P < 0.001  Specific sexual behavior received, % of bouts1      No reaction  62.4a  56.8ab  51.4b  54.9ab  3.3  P = 0.01      Standing posture  0.1  0.5  0.7  0.1  0.3  P = 0.06      Flee to pen  4.7a  6.7ab  9.0b  11.4b  1.8  P = 0.002      Flee to stall  0.4  1.1  1.5  1.3  0.5  P = 0.18      Aggression  0.7  0.8  1.9  2.8  0.9  P = 0.09  Variables  d 2  d 3  d 4  d 5  SEM  P-value  Sexual interactions, average number of bouts observed per sow per 5 min scan      Sexual behavior  0.19a  0.27a  0.59c  0.41b  0.03  P < 0.001  Specific sexual behavior initiated, % of bouts1      Flank nosing  10.4a  11.6a  11.6a  21.5b  2.4  P < 0.001      Anogenital sniffing  88.5a  85.0a  82.6a  71.5b  0.5  P < 0.001      Mounts  0.1a  0.3a  1.7b  1.1ab  0.5  P = 0.008      Attempts to mount  0.9a  3.0ab  3.3ab  3.8b  1.0  P = 0.04      Unsuccessful mounts  0.2a  0.2a  0.8ab  2.0b  0.5  P < 0.001      All forms of mounts2  1.1a  3.5ab  5.8b  7.0b  1.5  P < 0.001  Specific sexual behavior received, % of bouts1      No reaction  62.4a  56.8ab  51.4b  54.9ab  3.3  P = 0.01      Standing posture  0.1  0.5  0.7  0.1  0.3  P = 0.06      Flee to pen  4.7a  6.7ab  9.0b  11.4b  1.8  P = 0.002      Flee to stall  0.4  1.1  1.5  1.3  0.5  P = 0.18      Aggression  0.7  0.8  1.9  2.8  0.9  P = 0.09  a,bWithin a row, means without a common superscript differ (P < 0.05). 1Specific sexual behaviors were not mutually exclusive; hence, they could occur in the same bout of sexual behavior initiated or received. 2Mounts, attempts to mount, and unsuccessful mounts combined. View Large Specific sexual behavior initiated varied according to the day (Table 7). Anogenital sniffing was the most common behavior displayed but it was reduced on d 5, whereas flank nosing became more frequent. Mounting, attempts to mount, and unsuccessful mounts were relatively infrequent, but when all these forms of mounting behavior were considered together, all forms of mounts increased progressively from d 2 to 5. In terms of specific sexual behavior received, that is, the reaction from the recipient sow, sows generally showed no reaction to the sexual behavior interaction and rarely adopted a standing posture. Sows avoided sexual behavior interactions more on d 4 and 5 compared to d 2, by fleeing to another area of the pen. Relating sexual behavior to reproductive performance, sexual behavior both initiated and received were positively correlated with the length of estrus, as measured by the sexual receptivity test (r = 0.19, P = 0.01, and r = 0.18, P = 0.02), and to the spontaneous standing response to boar exposure (r = 0.20, P = 0.008, and r = 0.27, P < 0.001). Sexual behavior initiated was also negatively correlated with the wean-to-mate interval (r = –0.20, P = 0.008). There was a trend for sexual behavior initiated to correlate negatively with weight gain (r = –0.16, P = 0.08). Sexual behavior received was positively correlated with cortisol concentration on d 1 (r = 0.39, P = 0.005). Conception, Farrowing, and Litter Characteristics Following the 243 sows that remained on trial after d 6, conception rate as the percentage of sows confirmed pregnant based on ultrasonography and return to estrus 5 wk after insemination did not differ between GpW and StW sows (positive ultrasonography result, returns to estrus, and missing values were 114, 3, and 1 sows and 116, 4, and 5 sows, respectively; χ2 test P = 0.73). The insemination-to-farrow interval did not differ between treatments (both 116.9 ± 0.2 d; P = 0.90). Farrowing rate did not differ between GpW and StW sows, calculated using the sows that were mated twice and remained on trial after d 6 (95.76 and 90.40%, respectively; χ2 test P = 0.10). The total number of piglets born or born alive did not differ between treatments (means ± SE: total piglets born: 12.7 ± 0.2 pigs; piglets born alive: 11.7 ± 0.2 pigs; stillborn pigs: 0.8 ± 0.1; and mummified pigs: 0.2 ± 0.0 pigs; P > 0.1). DISCUSSION Mating Success Ninety percent of the sows displayed estrus-related behavior within the first week after weaning, with no overall difference between group-weaned sows and individually stall-weaned sows. However, group-weaned sows in our study displayed fewer spontaneous standing responses to boar exposure and partly compensated by more responses to the back-pressure test performed by a stockperson in presence of the boar. England and Spurr (1969) reported no difference in detection rate between group-housed or stall-housed sows. Yet different methods of estrus detection can influence the likelihood of successful detection (Kemp et al., 2005). Langendijk et al. (2000) found that group-weaning delayed estrus detection by the back-pressure test but not the spontaneous standing response to the boar, whereas we found the opposite. This discrepancy may be due to the smaller sample size for the spontaneous-estrus test in Langendijk et al. (2000). Indeed, our findings concur with the idea that the incidence of “silent estrus” may be higher with less intensive estrus detection protocols, because spontaneous standing to the boar provides less intense stimulation than the back-pressure test in presence of the boar (Kemp et al., 2005). This lower sexual receptivity highlights that efficient estrus detection methods may be relatively more important for group-weaning systems that stall-weaning systems. In agreement with Langendijk et al. (2000), the variation in the wean-to-mate interval was greater for group-weaned sows than stall-weaned sows, with 7% fewer group-weaned sows inseminated twice by d 6. Testing every 8 h, Langendijk et al. (2000) reported that group-weaning postponed the onset of estrus by 10 h in comparison to stall housing, but others found no effect (England and Spurr, 1969; Sommer, 1980) or even positive effects (Hemsworth et al., 1982; Pearce and Pearce, 1992), although the onset of estrus in these last 2 studies occurred more than 10 d after weaning. This variation in the onset of estrus could have 2 origins: there may be greater variation in the temporal pattern of estrus in group-weaned sows with delayed ovulation or group-housed sows may have higher incidence of “silent” estrus, that is, ovulation occurring while estrus behavior is suppressed. Our present study design does not allow us to directly answer this crucial question because we did not determine the occurrence of ovulation or test receptivity after d 6. Nevertheless, three-quarters of the sows excluded on d 6 but later inseminated fulfilled the delayed ovulation criteria (wean-to-mate interval between 6 and 19 d) whereas the remaining quarter fulfilled the undetected ovulation or silent estrus explanation (wean-to-mate interval ≥ 20 d). Insemination rate by d 6 was also more variable between group-weaned pens than between groups of stall-weaned sows. The greater variation in insemination rate found in this study and by others (Pedersen et al., 1993; Langendijk et al., 2000) suggests that the sexual behavior of sows may be susceptible to social stress. It is well recognized that reproductive performance may be affected by stress (Kongsted, 2004; Pedersen, 2007) but direct evidence remained scarce to date (Turner et al., 2005). Stress and Aggression A combination of stress indices were collected in this study: aggressive behavior, skin injury, and plasma cortisol concentration. As expected, fresh skin injuries occurred following mixing, on d 1 for group-weaned sows and on d 7 for the mixed stall-weaned sows. The effect on cortisol concentration was similar on d 1, but mixing of stall-weaned sows after insemination did not result in a cortisol increase. In fact, both treatments showed a reduction in cortisol concentration between d 1 and 7 and no difference in the frequency of aggression at feeding on d 7. These results suggest that sows may be more responsive to stressors, such as mixing, after weaning than after insemination. It is important to note that most studies examined the effects of mixing after 4 wk of pregnancy in order not to disturb embryo implantation (for review see Spoolder et al., 2009), whereas in the present study, grouped and stall-weaned sows were mixed within 48 h of the first insemination. This attenuation in cortisol response and aggression on d 7 compared to d 1 may also result from the fact that group-weaned sows had to cope with a greater combination of stressors on d 1, namely grouping in addition to weaning and new housing accommodation. Unfortunately, straight comparisons between the relative levels of aggression on d 1 and 7 are not valid because the feeding system and the timing and duration of behavioral observations varied to adjust for management differences. Although group-weaned sows were kept with acquainted sows on d 7, the similarities in aggression and cortisol concentration to simultaneously grouped stall-housed sows suggest that altering space allowance, group composition, or feeding systems postmating represent a stressful situation, although the fewer fresh skin injuries suggested less severe physical injuries at regrouping. Skin injuries on d 1 negatively correlated with weight gain, and cortisol concentration tended to negatively correlate with weight gain, suggesting that stressed sows performed poorly over the first week postweaning. However, we found no direct evidence that this stress affected sexual responsiveness or the wean-to-mate interval. Pedersen et al. (1993) reported that aggression received was positively correlated with wean-to-estrus interval and negatively with estrus duration, but a later study reported no effect on wean-to-estrus interval or estrus duration, although subordinate sows showed reduced sexual proceptivity and receptivity (Pedersen et al., 2003). Sexual Behavior Sexual behaviors were the subject of detailed observations in the group-weaned treatment. Sexual behavior peaked on d 4 and remained high on d 5, the last day of observation. Sexual behavior occurs throughout the day and the night (Pedersen, 2007), and different methods of observation may lead to differences in observed prevalence, although our method was designed to capture most of the daily pattern. Most sexual behaviors consisted of anogenital sniffing, but flank nosing and mounting behaviors increased on d 4 and 5, occurring in about 20 and 7% of the interactions, respectively. Although less frequent, mounting is a potentially injurious behavior. Sows that were the target of these sexual behaviors showed no reaction in about half of the interactions. They rarely adopted a standing posture, which suggests that the initiator was the sow in estrus, not the recipient. Sows fled from sexual behavior interactions as they became more frequent, but aggressive responses were rare. The frequency of sexual behavior initiated tended to correlate with lower weight gain, suggesting that sexual activity results in higher energy expenditure or stress. The frequency of sexual behavior received was correlated with cortisol concentration on d 1. Cortisol concentration at d 1 obviously was not affected by those sexual behaviors received later in the week, but both measures may relate to a common variable, such as the type of sows targeted by both mixing stress and sexual behavior. Sexual behavior may be used to ascertain dominance status, as suggested by Pedersen (2007). Nevertheless, our data did not support the hypothesis that recipient sows are chosen based on dominance status. Implications Group-weaned sows showed greater weight loss during the first week after weaning, which is likely linked to the higher stress levels experienced, due to either aggression, sexual behavior, or increased activity. This is unlikely to be related to competition for food as group-weaned sows had access to full-length feeding stalls with ad libitum access to feed. Unfortunately, weight and gait scores were not collected on sows discarded from the trial on d 6, particularly the 10% that did not show estrus-related behavior, which could have been most affected. Grouping sows after weaning or after insemination did not have any significant long-term effects on conception rate, returns, farrowing rate, or litter characteristics at farrowing. However, the sample size was relatively small to detect effects on farrowing performances (Hemsworth et al., 1982). Boar stimulation can stimulate estrus, depending on frequency, intensity, or length of exposure (Hemsworth et al., 1982; Langendijk et al., 2000; Kemp et al., 2005), but continuous boar contact can suppress estrus behavior (Hemsworth and Hansen, 1990; Tilbrook and Hemsworth, 1990; Knox et al., 2004). Boar stimulation in the present study was provided once daily for 30 s, with group-weaned sows locked in the feeding stalls. A majority of studies used the back-pressure test as a measure of receptivity, but only a few studies include parameters of proceptivity (defined as the behavior toward males that stimulate sexual activity), although proceptivity may serve to stimulate fertilization (de Jonge et al., 1994; Pedersen, 2007). Proceptivity could ease management by removing the need to lock in group-weaned sow for boar exposure if they willingly approach while in estrus. The effects of leaving group-weaned sows freely interacting with the boar (i.e., proceptivity) and different frequencies or lengths of boar exposure require further research. A choice was made in the present experiment to offer 4.4 m2 per sow in the mixing pen, keeping sows in static groups of same parities, with ad libitum feed, full length feeding stalls, and standardizing boar exposure for 30 s daily while locked in the feeding stalls for mating. A number of factors may exacerbate the effects on the behavior and stress of sows weaned in groups, such as mixed parities, dynamic grouping, absence of protection at feeding, restricted feed, or seasonal infertility. Conclusions Sows housed in groups at weaning and regrouped after insemination experienced higher stress than sows housed in individual stalls at weaning and housed in groups after insemination. Sexual receptivity was reduced, which could be due to fewer sows ovulating or suppressed estrus-related behavior. Notwithstanding, weaning sows in groups resulted in lower mating success within 5 d of weaning, which in turn increased between-week variability in sows mated. Clearly, further research is needed to identify underlying mechanisms to reduce variability, manage aggression and sexual behavior, optimize estrus detection, and assess reproductive performance in group-housed weaning systems. LITERATURE CITED Ashworth C. J. Pickard A. R. 1998. Embryo survival and prolificacy. In: Wiseman J. Varley M. A. Chadwick J. P. editors, Progress in pig science.  Nottingham Univ. Press,, UK. p. 303– 325. de Jonge F. H. Mekking P. Abbott K. Wiepkema P. R. 1994. Proceptive and receptive aspects of estrus behavior in gilts. Behav. Processes  31: 157– 166. Google Scholar CrossRef Search ADS PubMed  England D. C. Spurr D. T. 1969. Litter size of swine confined during gestation. J. Anim. Sci.  28: 220– 223. Google Scholar CrossRef Search ADS PubMed  Hemsworth P. H. Hansen C. 1990. The effects of continuous boar contact on the estrus detection rate of weaned sows. Appl. Anim. Behav. Sci.  28: 281– 285. Google Scholar CrossRef Search ADS   Hemsworth P. H. Rice M. Nash J. Giri K. Butler K. L. Tilbrook A. J. Morrison R. S. 2013. Effects of group size and floor space allowance on grouped sows: Aggression, stress, skin injuries, and reproductive performance. J. Anim. Sci.  91: 4953– 4964. Google Scholar CrossRef Search ADS PubMed  Hemsworth P. H. Salden N. T. C. Hoogerbrugge A. 1982. The influence of the post-weaning social environment on the weaning to mating interval of the sow. Anim. Prod.  35: 41– 48. Karlen G. A. Hemsworth P. H. Gonyou H. W. Fabrega E. Strom D. Smits R. J. 2007. The welfare of gestating sows in conventional stalls and large groups on deep litter. Appl. Anim. Behav. Sci.  105: 87– 101. Google Scholar CrossRef Search ADS   Kemp B. Soede N. M. Langendijk P. 2005. Effects of boar contact and housing conditions on estrus expression in sows. Theriogenology  63: 643– 656. Google Scholar CrossRef Search ADS PubMed  Knox R. Salak-Johnson J. Hopgood M. Greiner L. Connor J. 2014. Effect of day of mixing gestating sows on measures of reproductive performance and animal welfare. J. Anim. Sci.  92: 1698– 1707. Google Scholar CrossRef Search ADS PubMed  Knox R. V. Breen S. M. Willenburg K. L. Roth S. Miller G. M. Ruggiero K. M. Rodriguez-Zas L. 2004. Effects of housing system and boar exposure on estrus expression in weaned sows. J. Anim. Sci.  82: 3088– 3093. Google Scholar CrossRef Search ADS PubMed  Kongsted A. G 2004. Stress and fear as possible mediators of reproduction problems in group housed sows: A review. Acta Agric. Scand. A  54: 58– 66. Langendijk P. Soede N. M. Kemp B. 2000. Effects of boar contact and housing conditions on estrus expression in weaned sows. J. Anim. Sci.  78: 871– 878. Google Scholar CrossRef Search ADS PubMed  Mendl M. Zanella A. J. Broom D. M. 1992. Physiological and reproductive correlates of behavioral strategies in female domestic pigs. Anim. Behav.  44: 1107– 1121. Google Scholar CrossRef Search ADS   National Health and Medical Research Council ( 2013) Australian code for the care and use of animals for scientific purposes, 8th edition. Canberra: National Health and Medical Research Council. Pearce G. P. Pearce A. N. 1992. Contact with a sow in estrus or a mature boar stimulates the onset of estrus in weaned sows. Vet. Rec.  130: 5– 9. Google Scholar CrossRef Search ADS PubMed  Pedersen L. J. 2007. Sexual behavior in female pigs. Horm. Behav.  52: 64– 69. Google Scholar CrossRef Search ADS PubMed  Pedersen L. J. Heiskanen T. Damm B. I. 2003. Sexual motivation in relation to social rank in pair-housed sows. Anim. Reprod. Sci.  75: 39– 53. Google Scholar CrossRef Search ADS PubMed  Pedersen L. J. Rojkittikhun T. Einarsson S. Edqvist L. E. 1993. Post-weaning grouped sows: Effects of aggression on hormonal patterns and estrus behavior. Appl. Anim. Behav. Sci.  38: 25– 39. Google Scholar CrossRef Search ADS   Signoret J. P. Du Mesnil du Buisson F. 1961. Study of the behavior of the sow during estrus. In: Proc. 4th Int. Congr. Anim. Repro., The Hague, The Netherlands. p. 171– 175. Sommer B. 1980. Sow in individual pens and group housing-oestrous behaviour, parturition, fertility and damage to limbs. Anim. Breed. Absr.  48, 619. Spoolder H. A. M. Geudeke M. J. Van der Peet-Schwering C. M. C. Soede N. M. 2009. Group housing of sows in early pregnancy: A review of success and risk factors. Livest. Sci.  125: 1– 14. Google Scholar CrossRef Search ADS   Tilbrook A. J. Hemsworth P. H. 1990. Detection of estrus in gilts housed adjacent or opposite boars or exposed to exogenous boar stimuli. Appl. Anim. Behav. Sci.  28: 233– 245. Google Scholar CrossRef Search ADS   Turner A. I. Hemsworth P. H. Tilbrook A. J. 2005. Susceptibility of reproduction in female pigs to impairment by stress or elevation of cortisol. Domest. Anim. Endocrinol.  29: 398– 410. Google Scholar CrossRef Search ADS PubMed  Footnotes 1 This work was partially funded by a grant from the Australian Pork Cooperative Research Centre. We thank Rebecca Woodhouse, Maxine Rice, the Rivalea Research and Innovation team, and the staff from Rivalea for their help with conducting this project. American Society of Animal Science
TI - Effects of group housing after weaning on sow welfare and sexual behavior
JF - Journal of Animal Science
DO - 10.2527/jas.2014-8238
DA - 2014-12-01
UR - https://www.deepdyve.com/lp/oxford-university-press/effects-of-group-housing-after-weaning-on-sow-welfare-and-sexual-3JNKBZH7AI
SP - 5683
EP - 5692
VL - 92
IS - 12
DP - DeepDyve
ER -