| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GROWTH, PHYSIOLOGY, AND REPRODUCTION |
,4
* Section Foetal and Perinatal Biology, Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, the Netherlands;
and
Research Group Animal Husbandry, Animal Sciences Group of Wageningen University and Research Centre, Lelystad, the Netherlands; and
and
Research Group Animal Welfare, Animal Sciences Group of Wageningen University and Research Centre, Lelystad, the Netherlands
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: group housing • maternal effect • pig • social stress
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). These interactions are known to be stressful (Tsuma et al., 1996, Otten et al., 1997; Ruis et al., 2002).
The dominance rank of sows is known to affect litter size, with middle-ranking sows having smaller litters than high- and low-ranking sows (Nicholson et al., 1993). Litters from submissive sows weighed less at birth, and submissive sows had greater cortisol concentrations during gestation (Mendl et al., 1992; Nicholson et al., 1993; Zanella et al., 1998).
Elevated maternal cortisol concentrations or stress during gestation also result in reduced birth weights of the offspring (e.g., in rhesus monkeys: Schneider et al., 1999; rodents: Lesage et al., 2004; and pigs: Kranendonk et al., 2006a). In the latter study, piglets from sows with elevated cortisol concentrations during gestation displayed an increase in novelty-induced locomotion and a decrease in the salivary cortisol response to ACTH (Kranendonk et al., 2006a,b).
To our knowledge, no study to date has focused on the effects of the social rank of pregnant sows during group housing on development and behavior of their offspring. Hence, the primary aim of the present study was to determine whether the social rank of group-housed sows during gestation affected BW gain and litter characteristics, or BW, behavioral response to novelty, and slaughter characteristics of their offspring. We hypothesized that dominant sows would have reduced cortisol concentrations during gestation and give birth to heavier piglets than submissive sows and that growth, behavior, and slaughter characteristics of their offspring would differ from those of offspring from submissive sows.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Animals and Housing
The present experiment was performed with 5 batches of 3 groups of pregnant sows and gilts [166 Dalland animals (Topigs, Helvoirt, the Netherlands) in total], with a mean parity of 4.2 (range 1 to 9); hereafter all animals are referred to as sows] that were kept at the experimental pig farm of ID-Lelystad, the Netherlands. Groups consisted of on average 11.3 sows (range 7 to 14), and all sows participated in the experiment during one complete pregnancy only.
According to the farm’s standard procedure, sows were assigned to groups based on their backfat thickness before AI (
12; >12 but <15; or 
15 mm), which was measured at 5 cm from the midline over the last rib by use of a Lean-O-Meater (Renco Co., Minneapolis, MN). The nulliparous sows (n = 3 to 10 gilts per group, with 36 gilts total) were assigned to the 12-mm group. The experimental farm had a management system in which sows were inseminated every 3 wk. Batches of inseminated sows were separated in time by 3 or 6 wk, and the 3 groups from 1 batch were housed in the same room. The study was conducted from February 2004 (AI of batch 1) until April 2005 (slaughter of offspring from batch 5).
Sows were housed in a room with 18 farrowing pens before this experiment (farrowing pens were 1.7 x 2.6 m, with a solid piglet creep area and slatted, synthetic floor). Three to four days after AI, all sows of 1 group were simultaneously introduced into the group pen (6.00 x 3.15-m, with a solid concrete lying area, and a 3.80 x 3.50-m, slatted concrete dunging area). Two group pens were situated in 1 room, and during the first 4 wk after AI a boar was present in that same room, but in a third, separate boar pen. Four weeks after AI, pregnancy was diagnosed, and pregnant sows were transferred to another room without a boar (6.00 x 3.35-m pens with a solid concrete lying area and a 5.35 x 3.50-m slatted concrete dunging area; 4 pens per room with one pen empty), where they were kept with the same sows as during the first 4 wk of gestation. Nonpregnant sows were removed from the group, were inseminated again, and were placed in the next AI group that could be part of the experiment.
Sows were fed a commercially available sow diet according to normal husbandry procedures (2.5 to 3.2 kg·sow–1·d–1, depending on the stage of gestation; 8.3 MJ of NE·kg–1, 13.9% CP, and 0.71% lysine, as-fed basis). Water was available ad libitum from one water dispenser that was located in the feeding/dunging area of the pen. Mean room temperature during gestation was 20.1°C, and lights were on from 0600 until 2300 in the AI room and from 0700 until 1700 in the gestation room.
One week before the expected farrowing date, sows were relocated to farrowing pens (same design as described above). Here, sows were fed 3.2 to 6.5 kg·sow–1·d–1 of a commercially available, lactating sow diet (9.6 MJ of NE·kg–1, 15.6% CP, and 0.81% lysine, as-fed basis); water was available ad libitum. Mean room temperature during lactation was 21.5°C, and lights were on from 0700 until 1700.
Piglets were kept with their own mother until weaning at a mean age of 27 d (range of 23 to 31 d) and, only when the number of piglets exceeded the number of teats of a sow, surplus piglets were randomly selected and placed with another sow (social rank of the foster sow was not known at the time of transfer). Cross-fostered piglets were not used for behavioral testing. After weaning, piglets were transferred to rearing pens (1.20 x 2.90 m, 60% solid concrete floor, and 40% slatted metal floor) with their littermates, and at 10 wk of age they were transferred to fattening-pens (of varying sizes) in which they could be mixed with other litters. From 10 d of age onward, piglets were provided with creep feed (11.4 MJ of NE·kg–1, 20.4% CP, and 1.5% lysine, as-fed basis) and water ad libitum until weaning. Around the time of weaning, they were gradually adjusted to a weaner diet (ad libitum; 10.4 MJ of NE·kg–1, 17.5% CP, and 1.1% lysine, as-fed basis), which they received during the first 2 wk after weaning. From 6 wk of age until slaughter, they received a fattening pig diet ad libitum (9.7 MJ of NE·kg–1, 16.1% CP, and 0.9% lysine, as-fed basis).
Feeding Station
Sows were fed individually using an electronic sow feeding system (Fitmix, Mannebeck, Schüttorf, Germany). Each pen contained 1 sow feeding system, and sows were able to displace another sow from the feeding station (Figure 1). Each sow had an ear transponder to enable individual recognition by the feeding station, so that each sow received the ration she was entitled to have, by opening the feeding tube as long as she still had food left for the day. Sows were able to gather their ration within 24 h by multiple visits from 0600 onward. The feeding station registered every visit of each sow to the station by registering the sow number, the moment of arriving and leaving, and the amount of food provided.
|
It was presumed that dominant sows would be better able to maintain their position at the feeding station. These sows would be the first to visit the feeding station, could eat their ration in a relatively small number of visits, and the mean duration of their feeding visits would be relatively long compared with submissive sows (Tanida et al., 1993; Andersen et al., 1999). Preliminary behavioral observations demonstrated that sows often displaced other sows from the feeding station, and the displacing sow seemed to be able to gather a little amount of food left in the feeding tube. When the displacing sow had already finished her daily ration, she was registered as having a nonfeeding visit, though she may still have consumed some food.
Based on these preliminary observations, and using the data categories defined above, visit success (%) was calculated as: [number of nonfeeding visits/(number of nonfeeding visits + number of feeding visits)] x 100, and displacement success (%) was calculated as: [number of times a sow succeeded another sow within 2 s/(number of times a sow succeeded another sow within 2 s + number of times a sow was succeeded by another sow within 2 s)] x 100.
Agonistic Interactions of Sows
Agonistic interactions were studied in the first 2 batches (6 groups, n = 57 animals) to determine the relationship between social rank and the data from the feeding station. If correlations were highly significant, we assumed that the data from the feeding station could then be used to determine the social rank of all sows.
Cameras were mounted above the pens (one camera covered the lying area of the pen, another covered the feeding and dunging area), and behavior was videotaped with time-lapse video recorders. Sows were color-marked on their backs to allow individual identification. Behavioral observations were made 4 times during gestation: the first 4 d after group formation; the first 3 d after transfer to the gestation pen; 3 d during the 11th wk of gestation; and 3 d during the last week of group housing, before sows were relocated to the farrowing crates. Behavior was observed using behavioral sampling and continuous recording (Martin and Bateson, 1993) during the entire light period.
Agonistic interactions were defined as sows performing head-knocks, bites, displacements, and fights (Erhard et al., 1997). The sow that retreated at the end of the agonistic interaction was registered as having lost the interaction. For each sow, using the data from all 4 observational periods, the total numbers of interactions won and lost were registered and were used to calculate the percentage of agonistic interactions won [(interactions won/total number of interactions) x 100%]. In addition, using 2-way comparisons, the number of animals within the group a sow dominated (a sow dominated another sow when she won more agonistic interactions from that sow than she lost from that sow) was used to calculate the percentage of sows defeated {[number of animals an individual sow dominated/ (group size –1)] x 100%; Mendl et al., 1992}.
Classification of Sows
The behavioral data from batch 1 and 2 were correlated with data from the feeding station using Spearman rank correlations. These correlations were calculated using the residual terms of the regression models to adjust for possible differences among batches and groups. For the first 2 batches, data from the feeding station were significantly correlated with the percentage of sows defeated and the percentage of agonistic interactions won (P < 0.01; Table 1, Figure 2). The percentage of agonistic interactions won showed the greatest correlation with displacement success at the feeding station (rs = 0.90, P < 0.001; Figure 2B). This allowed us to determine the number of displacements performed and received for all sows, and to calculate the displacement success for all animals.
|
|
Because of health problems (high fever; not associated to aggression), 2 sows (one HSR and one LSR sow) were removed from the herd after farrowing. Litter data and offspring behavior and performance data from these animals were not included in the analyses.
Sow, Litter, and Piglet Characteristics
Body weight of the sows was determined before AI, before transfer to the farrowing crates (1 wk before the expected parturition date), and after weaning. In addition, backfat thickness of the sows was measured before transfer to the farrowing crates and after weaning of the piglets. Gestation length was calculated at parturition, with the day of AI defined as d 0. At parturition, the total number of piglets born, number born alive, number born dead, and number of mummified fetuses were recorded (placentas were not thoroughly examined, only well visible mummified fetuses were recorded). Within 24 h of birth, piglets were sexed and ear-tagged for identification. According to standard procedures, all piglets received an i.m. injection containing 200 mg of iron (1 mL per piglet; Prevan 200, Eurovet Animal Health B.V., Bladel, the Netherlands) at 3 d of age. No teeth clipping, tail docking, or castration was performed with the piglets. Body weight of the piglets was determined at birth, at 14 d of age, and at weaning (4 wk of age). Neonatal mortality was defined as piglets dying from birth until 3 d of age, and preweaning mortality was defined as as piglets dying from 4 d of age until weaning.
A number of piglets were sold for other experiments; therefore, the total number of piglets slaughtered was 228 and 211 for HSR male and HSR female piglets, respectively, and 410 and 336 for LSR male and LSR female piglets, respectively. Because piglets were not castrated, males were slaughtered earlier than females to avoid boar taint. Male offspring were slaughtered at a BW of approximately 100 kg, and female offspring were slaughtered at approximately 110 kg. Age at slaughter and slaughter weight were recorded, and the percentages of lean meat and fat were determined at 5 cm from the midline over the last rib, following common practice, with use of a Hennessy Grading Probe (Hennessy Grading Systems Ltd., Auckland, New Zealand). Personnel determining when the piglets had to be slaughtered or collecting slaughter data were unaware of the classification of the mothers of these piglets.
Novel Object Test
The novel object (NO) test was performed to study whether social rank of the sow would affect behavior of her individual piglets in an unfamiliar environment and in response to a NO (Wemelsfelder et al., 2000; Kanitz et al., 2004). In total, 282 piglets were tested (50 to 66 piglets per batch), of which 108 piglets belonged to HSR mothers and 172 to LSR mothers.
At the mean age of 20 d (range 16 to 24 d), 2 piglets per litter (1 gilt and 1 boar) were selected based on their BW (BW at 14 d of age within 1 SD of the mean BW of the litter) and health. The 2 selected piglets were taken out of their home pen together with a littermate that was not used for the test, placed in a cart, and brought to a room adjacent to the testing room, where they were allowed to adjust to the new environment for approximately 10 min. Then, 1 piglet was taken from the cart, carried to the testing room, and fitted with an elastic band around the chest to facilitate movement-tracking by use of the Ethovision 2.3 system (Noldus, Wageningen, the Netherlands).
The piglet was then placed into a test arena, which measured 3 x 3 m and was surrounded by 1-m-high solid wooden walls. A camera was mounted above the arena, and the entire test (10 min) was recorded on video. After 5 min, a NO (a red 25-L water container) was lowered from the ceiling to approximately 15 cm above the floor. The latency to touch the NO was recorded, as well as the percentage of time and number of times spent within 50 cm of the NO. The piglet was left with the NO for another 5 min and was then returned to its 2 littermates. Behavior was scored immediately (number of vocalizations, latency to touch the NO), or later (walking and running) using The Observer 4.1 software (Noldus).
Maternal Saliva Collection and Salivary Cortisol Analysis
During the 2nd, 7th, and 13th wk after AI, saliva samples were taken between 1200 and 1300 to determine salivary cortisol concentrations. Saliva was collected by allowing the animals to chew on two cotton buds (Hartmann, Nijmegen, the Netherlands). When these were thoroughly moistened, the buds were centrifuged, and saliva was collected and stored at –20°C until further analysis. The salivary cortisol concentration was determined in accordance to the method validated by Ruis and colleagues (1997), using a solid-phase RIA kit (Coat-A-Count Cortisol TKCO, Diagnostic Products Corporation, Apeldoorn, the Netherlands) modified for pig salivary cortisol. Samples were assayed in duplicate. The minimum detection limit of the assay was 0.4 ng/mL, and the interassay CV was 8.8% (n = 3).
Statistical Analyses
All data are presented as least squares means ± SEM. Results were considered statistically significant when P 0.05, and as tending to differ when 0.05 < P 0.10. All statistical analyses were performed with Genstat 7.1 (Payne, 2003, VSN International, Oxford, UK).
Continuous data were analyzed using a linear mixed model with random sow within group effects. Components of variance were estimated by REML. Percentages and counts were analyzed with a generalized linear mixed model with random sow within group effects, and components of variance were estimated by iterative reweighted REML (Engel and Keen, 1994). All models included main and interaction effects for the factors classification and batch (for litter characteristics) or classification and sex (for data on piglet level, with batch also included as a factor). Significance tests were performed with the exact F-test (ANOVA), an approximate F-test (GLM), or the Wald test (
2 for linear mixed model and generalized linear mixed model).
Sow and Litter Characteristics.
For the sow and litter characteristics (e.g., sow BW, parity, gestation length, litter size), sow was the experimental unit, and data were analyzed with group nested in batch as random factor. Sow characteristics were, where appropriate, corrected by covariate analysis for parity, BW before AI, and litter size. Litter size was added to the model as a covariate when analyzing gestation length and other litter data (sex ratio of the litter, percentage of piglets born alive, mummified, and dead before weaning). Maternal salivary cortisol concentrations were analyzed with parity and litter size as covariates.
Piglet Characteristics.
Piglet data (e.g., birth weight, weaning weight, behavioral data from the NO test) were analyzed within sow (before and after weaning, to indicate that piglets from one litter were offspring from one sow) and by group nested within batch as a random factor. Hence, we corrected for the fact that one sow produced several piglets in the experiment. Piglet birth weights were corrected by covariate analysis for gestation length, BW gain of the sow during gestation, parity, and litter size. Body weights of the piglets at 14 d of age and at weaning were analyzed using litter size, age, birth weight, BW loss of the sow during lactation, and parity of the sow as covariates. Piglet behavior during the NO test was analyzed with age, litter size, and parity as covariates. Slaughter characteristics were, where appropriate, analyzed with age at slaughter, slaughter weight, BW of the sow before AI, birth weight, and parity of the sow as covariates. Interactions between classification and sex, or main sex effects, were not significant, and they are therefore not mentioned in the results.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Week 2 and 7 salivary cortisol concentrations did not differ between HSR and LSR sows, but in wk 13, HSR sows tended to have lower salivary cortisol concentrations than LSR sows (Table 2). None of the litter characteristics differed between HSR and LSR litters (Table 3).
|
|
). At birth and at 14 d of age, HSR piglets tended (P = 0.09 and 0.06, respectively) to weigh more than LSR piglets, and at weaning, they weighed significantly more (Table 3). Male piglets weighed more than female piglets at birth (1,403 ± 17 g vs. 1,342 ± 18 g; P < 0.001), at 14 d of age (4.5 ± 0.1 kg vs. 4.4 ± 0.1 kg; P = 0.002), and at weaning (7.6 ± 0.1 kg vs. 7.5 ± 0.1 kg; P = 0.04).
Novel Object Test
Both before and after introduction of the NO, HSR piglets spent more time walking and running than LSR piglets (Table 4). There was a classification x sex interaction effect for the number of vocalizations before introduction of the NO, with LSR males vocalizing less than HSR males, HSR females, and LSR females (Table 4). After introduction of the NO, all LSR piglets vocalized less than HSR piglets (Table 4), and male piglets vocalized less than female piglets (117 ± 9 versus 167 ± 12, respectively; P < 0.001).
|
), and shorter in male compared with female piglets (respectively, 59.6 ± 3.3 vs. 69.8 ± 2.9 s; P < 0.01). The number of times piglets approached the NO did not differ between HSR and LSR piglets (Table 4), and males approached the NO more often than females (7.1 ± 0.8 versus 4.9 ± 0.6, respectively; P < 0.005). A classification x sex interaction effect (P = 0.003) was observed for the percentage of time spent near the NO, with LSR males spending less time near the NO than LSR females and HSR piglets (P < 0.01, Table 4).
Slaughter Characteristics
The HSR and LSR offspring were slaughtered at the same age (Table 5). As expected, male offspring were slaughtered at a younger age than female offspring (144.7 ± 0.7 vs. 156.4 ± 0.7 d, respectively; P < 0.001). A classification x sex interaction effect was observed for slaughtered weight (P = 0.002): HSR males and LSR males had lower slaughtered weights than female piglets (P < 0.01), and HSR females had lower slaughtered weights than LSR females (P < 0.05, Table 5). The percentage of lean meat was greater in HSR compared with LSR piglets, but the percentage of fat did not differ (Table 5).
|
|
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Displacement Success at the Feeding Station
Social rank was determined by assessing the success of the sow at displacing other sows from the feeding station because in 6 groups of sows the percentage of agonistic interactions won was highly correlated with displacement success at the feeding station. This is in line with previous studies that found high correlations between feeding behavior and dominance rank in pregnant sows (Tanida et al., 1993; Andersen et al., 1999). We therefore concluded that displacement data provided by the feeding station used in the present study were suitable for determining social rank in all animals.
Sow Characteristics
Mendl et al. (1992) and Jarvis et al. (2006) reported elevated salivary cortisol concentrations in submissive compared with dominant or control gilts. Therefore, in the present study, HSR sows were expected to have lower cortisol concentrations than LSR sows. However, maternal salivary cortisol concentrations only tended to be lower in HSR compared with LSR sows at 13 wk of gestation, and did not differ at 2 and 7 wk of gestation. Therefore, it seems unlikely that elevated maternal cortisol concentrations explain the differences between mothers and piglets from the HSR and LSR groups. Differences between our study and the studies of Mendl et al. (1992) and Jarvis et al. (2006) may be due to differences in management of the sows. Because only 3 saliva samples were collected throughout gestation in the present study, it may also be that this (frequency, timing) did not allow detection of differences between groups. Nevertheless, salivary cortisol concentrations of both groups observed in the present study were considerably greater than those from individually housed sows of previous studies using the same RIA-method (animals in crates, sampled at the same time of day: Kranendonk et al., 2005; animals in pens: Kranendonk et al., 2006a). It seems that group housing with the present feeding system results in greater cortisol concentrations. Firstly, this may be related to more activity in group housing systems compared with individual housing. Elevated cortisol concentrations and increased activity were also observed in piglets in an enriched housing condition compared with a barren housing condition (De Jong et al., 1998). Secondly, the large number of (agonistic) interactions, provoked by the present feeding station, may have caused elevated cortisol concentrations, irrespective of winning or losing, or of being the initiator or the receiver of the interaction.
High social rank sows gained more BW during gestation than LSR sows and subsequently lost more BW and fat during lactation. The increased BW gain can be explained, at least partly, by the fact that HSR sows often displaced other sows from the feeding station, enabling them to obtain little amounts of feed from the ration of the displaced sow. Increased BW gain during gestation in high-ranking sows has been reported earlier (Brouns and Edwards, 1994).
Litter and Piglet Characteristics
Litters from HSR and LSR sows did not differ in size or total weight (all piglets or live born piglets). This is in line with the study of Jarvis et al. (2006), who reported that submissive gilts did not differ from control gilts in reproductive performance or piglet birth weights. In the present study, HSR piglets tended to weigh more at birth than LSR piglets, which is in line with the study of Nowachowicz et al (1999), who did not find significant differences in birth weights from offspring of dominant and submissive sows, although the way of classification in their experiment is not quite clear. However, Mendl et al. (1992) assigned group-housed gilts to 3 groups: dominant, intermediate, and submissive gilts, and reported that litter weights were greater in dominant than in intermediate gilts, with submissive gilts in between these 2. This suggests that the relation among these 3 social rank classes and litter weight in their study was not linear, whereas it was linear in the present study (calculated with displacement success on a continuous scale, data not shown). Differences in methodology probably underlie differences between the 2 studies.
In the present study, LSR sows lost less BW (absolute and relative, data not shown) and less backfat during lactation than HSR sows. Although this was accounted for in the analysis, their offspring still had significantly lower weaning weights than offspring from HSR sows. This indicates that not only BW loss of the sow, but also social rank of the mother during gestation, affects the BW of her offspring. Body weights after weaning were also lower in offspring from gilts that were introduced to sows during gestation (Jarvis et al., 2006). These effects could be caused by several factors, like differences in feed intake, activity, or metabolism.
In the present study, lean meat percentage at slaughter was less in LSR piglets compared with HSR piglets. It may be that during gestation, testosterone concentrations were greater in HSR sows. This may have affected testosterone concentrations and, ultimately, the percentage of lean meat in their offspring. This would be in line with previous observations that androgen concentrations early in life are associated with carcass leanness at slaughter (e.g., Sinclair et al., 2001). Alternatively, a study in sheep has demonstrated that a significant reduction in maternal feed intake during gestation reduced the number of muscle fibers in their lambs (Fahey et al., 2005). Though similar observations have been found in some pig studies (see Rehfeldt, et al., 2004), results are inconclusive. It may be that the reduced BW gain of LSR sows during gestation in the present study has similar effects on muscle development in the piglets.
Interestingly, HSR and LSR offspring also differed in their behavior during NO test. High social rank offspring spent more time on locomotion and on presence near the NO than LSR offspring, and had a shorter latency time to touch the NO. In addition, HSR males vocalized more than LSR males when exposed to the new environment, and after introduction of the NO, male and female HSR piglets vocalized more than LSR piglets. Furthermore, male LSR piglets spent less time near the NO than male HSR piglets. It may be that HSR piglets were more active than LSR piglets or that locomotion and vocalizations are inhibited in LSR offspring, or enhanced in HSR offspring, due to a differences in anxiety. These results also demonstrate that effects of the social rank of the sow during gestation on novelty-induced behavior depend on the sex of the offspring. This is in line with several studies in other species that have reported that maternal stress or elevated glucocorticoid concentrations during gestation may affect male and female offspring differently (McCormick et al., 1995; Kaiser and Sachser, 2001; Kaiser et al., 2003). The underlying mechanisms of the alterations in novelty-induced behavior, and the sex differences, are not clear. Yet, our results indicate that offspring can be affected not only by elevated maternal cortisol concentrations during gestation (Kranendonk et al., 2006a,b) but also by the social rank of their mother during gestation.
This study indicates that when food supply is limited and the feeding position can be monopolized, sows will have regular agonistic interactions to acquire little amounts of extra food. Social rank of the sow during gestation is highly related to competition around the feeding station and affects BW gain and loss of the sow. In addition, it affects growth, novelty-induced behavior, and the percentage of lean meat at slaughter of the offspring. These factors are of commercial importance for a pig farmer because payment for weaned piglets is based on BW of the piglets, and for slaughtered piglets on slaughtered weight and lean meat percentage. From a welfare point of view, a large proportion of the sows (the LSR sows) will benefit from reducing competition around the feeding station. Furthermore, differences in feed intake may partly underlie observed differences. To improve equal feed intake in all animals, reduction of competition around the feeding station may be beneficial. This could be accomplished by a number of measures, e.g. 1) preventing leftovers in the feeding tube; 2) protecting the position of the sow at the feeding tube; 3) adding an additional feeding station to the pen; 4) supplying distraction materials.
According to European Union wellfare regulations, pregnant sows in the European Union must be housed in social groups from 2013 onwards because the sows will then be able to perform more natural behavior compared with crated sows. However, it is important that pig farmers pay attention to reducing competition around the feeding area, which may reduce aggression among the sows and minimize differences between HSR and LSR sows. Separating HSR and LSR sows will not solve these dominance differences because new hierarchies with dominant and submissive sows will be formed within HSR and LSR groups.
|
Footnotes |
|---|
2 The authors thank personnel of the experimental farm for taking care of the animals, and in particular Hans Kooijman for helpful discussions on the personalities of the sows used in the present study. Bas Engel and Willem Buist are acknowledged for their statistical support, and Janneke Allaart and Kim van Gaalen for their practical assistance. Furthermore, we thank Marcel Taverne, Victor Wiegant, Dinand Ekkel, and Eduard Mulder for critical comments on earlier versions of the manuscript.
4 Present affiliation: Netherlands Institute for Fisheries Research, IJmuiden, the Netherlands.
3 Corresponding author: g.kranendonk{at}vet.uu.nl
Received for publication February 8, 2006. Accepted for publication June 29, 2006.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
ski, and J. Kapelaska. 1999. Influence of electronically controlled individual feeding on behavior and reproductive performance of pregnant sows. J. Anim. Feed Sci. 8:45–49.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
