J. Anim Sci.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J. Anim Sci. 2007. 85:1758-1769. doi:10.2527/jas.2006-510
© 2007 American Society of Animal Science
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
jas.2006-510v1
85/7/1758    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Salak-Johnson, J. L.
Right arrow Articles by Curtis, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Salak-Johnson, J. L.
Right arrow Articles by Curtis, S. E.

ANIMAL PRODUCTION

Space allowance for dry, pregnant sows in pens: Body condition, skin lesions, and performance1

J. L. Salak-Johnson2, S. R. Niekamp, S. L. Rodriguez-Zas, M. Ellis and S. E. Curtis

Department of Animal Sciences, University of Illinois, Urbana 61801


   Abstract
Top
 Abstract
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Different floor space allowances for dry, pregnant sows in pens were evaluated to determine the impacts of space on sow performance, productivity, and body lesions during 2 consecutive farrowings. Treatment groups of 5 sows/pen were assigned to 1.4, 2.3, or 3.3 m2 of floor space/sow or of 5 sows in individual stalls (1.34 m2). The experiment consisted of 6 replications (blocks 1 to 6), and within each block measurements were recorded for 2 consecutive pregnancies and farrowings. A total of 152 sows were measured at 1 farrowing, and 65 of those sows were measured at the successive farrowing (n = 217 records). Performance traits were BCS, BW, backfat (BF), days until rebred, and proportion culled. Litter traits were number of piglets born alive, male:female ratio, and proportions of stillborn, mummified, or dead piglets after birth. Litter performance measures were mean piglet BW and gain and litter BW. Lesion scores were assessed for several body regions. There were treatment and parity effects and interactions for several traits. An interaction of space treatment and parity occurred for sow mean BW, d-110 BW, BF, litter size, and litter and piglet BW and gain, with most effects in parity 2, 3, and 4 sows. Space affected sow mean (P < 0.001) and d-110 BW (P < 0.05) and mean BF and adjusted BF (P < 0.001); sows in pens at ≤2.3 m2 of space had greater (P < 0.05) BW and BF depth, and primiparous sows had less (P < 0.05) BW than greater parity sows. Few differences were found for litter traits, except for litter size, litter weaning BW, and pig BW gain. Sows in pens at 3.3 m2 had the largest (P < 0.05) litters. Weaning BW was greater (P < 0.05) for litters born to sows that gestated in stalls. Sow BCS and lesion score were affected by floor space (P < 0.001) and parity (P < 0.01); BCS was reduced in sows in pens at 1.4 m2, and their lesion score was consistently greater (P < 0.05) than that of other space allowances. Primiparous sows often had the least (P < 0.05) lesion scores. For sows in pens, linear and quadratic responses were detected for many of the traits assessed. As floor space increased, BW and BF increased (P < 0.01), but as space decreased, lesion scores increased. Although the effects of gestation system were found and lesion scores often were greater as space decreased, differences in productivity traits were unremarkable with respect to sow welfare or performance compared with industry norms.

Key Words: body condition • group • lesion • stall • sow environment


   INTRODUCTION
Top
 Abstract
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
The typical accommodation for pregnant sows in the United States is the individual stall. This housing system is one of the most controversial issues now facing the swine industry. There is growing concern about sow housing systems, with emphasis being placed on sow welfare (Bracke et al., 2002aGo,bGo; McGlone et al., 2004aGo). Following European trends, use of individual gestation stalls has been banned in Florida and Arizona. These bans have meant that research on the welfare of sows in different gestation housing systems has gained national attention, and US producers must consider modifying their systems or finding an alternative.

Few studies have shown reproductive impairment among pregnant sows kept in groups compared with sows in stalls (Peltoniemi et al., 1999Go; Kongsted, 2005Go). Others have shown no differences in reproduction measures between sows housed in stalls or groups (Gjein and Larssen, 1995Go; Harris et al., 2006Go). Greater skin lesion scores have been found among sows kept in pens during gestation than those in stalls (Turner et al., 2000Go; Anil et al., 2003Go; Karlen et al., 2006Go).

Most studies of dry sows residing in loose-housing systems have ignored any possible effects of floor-space allowance. Consequently, neither any impact of grouping on dry-sow state of being (often called well-being) nor appropriate grouped-sow space allowance has been ascertained. A thorough understanding of the nature of dry sows when grouped must precede any such system being adopted; otherwise, sow performance, health, and state of being may be unintentionally reduced.

The primary objective of this study was to determine the impacts of floor-space allowance for dry sows in pens (group-size constant) on performance, body lesion score, and body condition score. Moreover, differential effects of keeping sows in individual stalls vs. pens on those traits were evaluated.


   MATERIALS AND METHODS
Top
 Abstract
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Animals, Housing, and Experimental Design
The University of Illinois Institutional Animal Care and Use Committee approved the protocol for this experiment. Crossbred primi- and multiparous sows (n = 217 records) derived from PIC genetic lines kept at the University of Illinois Swine Research Center were used to compare the effects of housing sows in pens or in individual stalls during pregnancy. For each group, balanced for BW and parity, 20 sows were randomly allotted to 1 of 4 treatments: 5 sows in a pen at 1.4 m2 per sow, 5 sows in a pen at 2.3 m2 per sow, 5 sows in a pen at 3.3 m2 per sow, or 5 sows in 5 respective individual stalls (each 2.12-m long x 0.61-m wide) at 1.34 m2 per sow. All pens were 4.12-m long, so pen widths for the respective group pens were 1.62, 2.80, and 3.98 m.

The experiment consisted of 6 replications (blocks 1 to 6), and within each block, measurements were recorded for 2 consecutive pregnancies and farrowings, with the exception that for 1 block, measurements were made only once because a majority of the sows in that block, regardless of treatment, were found to be open at d 25 postbreeding. The gestational time periods for the respective groups are presented in Table 1Go. If sows were culled from a group at the end of lactation, a new individual was added to that group to keep the subsequent group size constant.


View this table:
[in this window]
[in a new window]

  		Table 1. Approximate gestational periods for each sow group for consecutive farrowings
 
Before the study, all multiparous sows had been kept in individual stalls after their previous litters were weaned. All sows received AI within 24 h after the onset of estrus and again at 24 h later. All newly bred sows were kept in individual stalls until the experiment began. Pregnancy was diagnosed at d 25 postbreeding using an Aloka 500V ultrasound machine (Aloka Inc., Tokyo, Japan) for transabdominal examinations, with a 3.5-MHz, convex linear-array transducer, and then sows that were pregnant were moved to their respective assigned treatment locations. Sows remained in their respective gestation treatment groups until approximately d 110 of gestation, when they were moved to the farrowing facility. There, the sows were placed in individual farrowing stalls with Tenderfoot flooring (Tandem Products Inc., Minneapolis, MN), where they remained until the end of lactation. All litters were weaned at 17 to 19 d of age, and sows were returned immediately to the breeding area.

Sows were kept in a well-insulated, mechanically ventilated, closed house during the breeding and gestation periods. During winter months, the unit-heater thermostat located 1.5 m above the floor in the gestation room was set at 21 ± 1°C. Sows were individually fed a diet formulated to meet or exceed established nutrient allowances (NRC, 1998Go). During gestation, each sow was fed 2.5 kg/d of a corn-soy-based diet having a calculated composition (as-fed) of 12.5% CP and providing a calculated ME density of 3,300 kcal/kg. All sows were fed between 0700 and 0730 each day. Sows in individual stalls were individually fed. Sows in pens were fed as a group on the solid-floor section and were offered 2.5 kg per sow. Each pen and each stall was equipped with 1 nipple waterer. Lactating sows were fed ad libitum a corn-soy-based diet with a calculated composition (as-fed) of 16% CP and 3,426 kcal of ME/kg. Cross-fostering did occur in this study, following the standard operating procedures of the farm, and occurred rarely and was random.

Sow and Litter Performance
Multiple sow- and litter-related traits were recorded for a total of 217 farrowings. A total of 152 sows were studied in their respective first gestations with 37, 39, 38, and 38 sows in treatments having 1.4, 2.3, or 3.3 m2 of space or individual stalls, respectively. Data were collected on fewer sows (65) at the second farrowing. This reduced number of sows resulted from culling, and 1 entire group was excluded due to an extremely low conception rate independent of treatment. The 65 sows studied at the second farrowing were represented as 18, 16, 18, 13 sows in the 1.4, 2.3, or 3.3 m2, and stall treatments, respectively. Of the 217 farrowings, 79 sows were primiparous and were allocated to the 4 respective treatments at 21, 17, 18, and 23 sows.

Sow traits included BW and backfat (BF) at gestation d 25 and 110 and again at d 131 (end of lactation), days until rebred, and proportion culled. Sow BF was measured using a longitudinal ultrasound scan (Aloka model 500V B-mode with a 5011 probe, Corometrics Medical Systems, Wallingford, CT) cranial to the last rib each time at which BW was measured. Sow BW gain was adjusted for BF depth and for overall BW gain. Litter traits included number of pigs of each sex born alive, ratio of males:females within a litter, and proportions of piglets stillborn, mummified, or dead after live birth but before weaning, respectively. Litter and mean pig birth BW, weaning BW, and mean BW gain from birth-to-weaning were calculated.

Sow Body Condition, Health Descriptors, and Lesions
These measures were assessed every 3 d for the first 13 d of the experiment (phase 1), then weekly thereafter throughout the experimental period (phase 2). Using the visual-appraisal (sow’s rear aspect) method described by Coffey et al. (1999)Go, BCS (1 = lowest, 5 = greatest) was assessed. Multiple health descriptors were scored and recorded by an experienced evaluator across the 1 or 2 parities (total >3,000 records). Traits evaluated included hair coat condition, presence of dung, ability to stand, lameness, and various lesion scores.

Hair coat condition was scored based on appearance, rated 0 (normal) or 1 (poor). Normal referred to a hair coat in which hairs were evenly distributed, fine, appropriately oriented on the body, and carrying a sheen. Poor referred to a hair coat in which bald patches were present or hairs were unevenly distributed, coarse, woolly, crimped, or lying in a disoriented fashion, or appearing dull. Evaluation of degree of dunging across the entire body on both sides of the sow was 0 (no dung), 1 (some dung), or 2 (excessive dung). Sows also were assessed for the ability to stand up and move (lameness). Standing-up was based on the apparent ease with which a sow went from lying to standing, resulting in a rating scale of 0 (no difficulty), 1 (slow, stiff), 2 (difficulty), or 3 (unable). For sows in pens, lameness score was assigned once the sow was standing and began to walk, resulting in a rating scale of 0 (no lameness), 1 (slight lameness), or 2 (marked lameness). Lameness scores for sows in stalls were recorded similarly, with the exception that the sow moved only forward and backward in her stall, but on weighing d 0, 110, and 131 they did move along a corridor and were scored for lameness. The same individual recorded all scores.

Several lesion scores were made and recorded at the beginning of the experiment (d 25) and every 3 d for the first 2 wk (phase 1). After that, the lesion scores were recorded weekly throughout gestation and again at the end of lactation (phase 2). Figure 1Go identifies the body locations used to assess lesion scores. Lesion scores were made and recorded for each of the following regions: head, ears, neck, chest/breast, shoulders, back, udder, rear, vulva, perineum, legs, and hooves. The variable perivulva denoted the maximum score found at the perineum or the vulva. Sow lesion score was based on the presence or absence of an apparently new or old lesion in conjunction with severity of the wound. Scoring definitions were: 0 = normal (no lesions), 1 = dehairing, callus, balding; 2 = redness, swelling; 3 = swelling + callus, abscess; 4 = moderate wound, scabbed over scratch; 5 = marked wound, fresh scratch; and 6 = severe wound, open wound. Thus, a sow could receive a score ranging from 0 (normal, no lesions) to 6 (severe wound, open wound) for any location on any particular day. Sows also could be assigned combined scores for a given location. For example, a sow might have at a certain location: dehairing (1), redness + swelling (2), swelling + callus (3), and marked wound/fresh scratch (5), for a total combined lesion score of 11.


Figure 1
View larger version (11K):
[in this window]
[in a new window]

  		Figure 1. Schematic of body regions used to assess sow lesion scores.

 
Statistical Analysis
Linear mixed-effects models were used to analyze sow performance traits, except for the binary trait cull, and most health descriptors, except for the binary traits hair coat, standing up, lameness, and lesion scores for chest, breasts, vulva, perineum, perivulva, and front and hind hooves. These health descriptors were transformed into binary traits due to the distribution of scores, such that values equal to 0 were assigned a 0 and values other than zero were assigned a 1. Binary traits were analyzed using the {chi}2 and Fisher’s exact test in PROC FREQ (SAS Inst. Inc., Cary, NC). Results of these approaches were similar, so only those of {chi}2 tests are presented here.

The sow performance mixed-effects model included fixed effects of space (4 levels: pens at 1.4, 2.3, or 3.3 m2/sow or stall), parity (5 levels: 1 to ≥5), and the space x parity interaction. A random effect of group (6 levels: sow groups 1 through 6) was included in the model to account for potential environmental and management differences across groups. A repeated-measurements structure was used to account for measurements across parity within sow while accounting for heterogeneity of variance across parities and space. The experimental unit was sow because the measurements corresponding to space level stall were from a single sow or experimental unit per stall. The heterogeneity of variance specified in the model accommodates potential differences in variation across space levels including those due to different number of sows per enclosure (stall or pen).

Sow BW at postbreeding d 25, 110, and 131, respectively, were analyzed as 3 separate response variables and together (mean BW), and included the fixed effect of day at measurement in the model and using a repeated-measurements model for day at measurement with an unstructured variance-covariance matrix of order 1. Likewise, BCS was analyzed using a mixed-effects repeated-measurements model for day that included the fixed effect of day at measurement. Litter performance variables were adjusted for cross-fostering (number of piglets added to or removed from the litter). The health-indicator mixed-effects model used within respective phases included the fixed effects of space, parity, space x parity interaction, and linear and quadratic trends on body condition. The model also included a random effect of group and a repeated-measurement structure across day within sow while accounting for heterogeneity of variance across parities and space. Estimates and linear and quadratic orthogonal contrasts with respect to pen density were computed using PROC MIXED of SAS. The significance level was set at P ≤ 0.05, and trends were noted at P ≤ 0.10.


   RESULTS
Top
 Abstract
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Despite the numerous space x parity interactions for many of the traits assessed here, very few means among treatments within parity were significant; thus only those measures in which there was a significant interaction among space treatments within parity are reported here. When differences were detected, they were found most often among parity 2, 3, or 4 sows. Table 2Go shows the means and associated P-values for space x parity interaction for those traits that were different among parity 2, 3, and 4 sows across treatments. There was space x parity interaction for sow mean BW (P = 0.014) and BW at postbreeding d 110 (P = 0.03), but not for BW at d 25 or 131 or BW gain (data not shown). Parity 3 sows in pens at 1.4 m2 had less (P < 0.05) mean BW than did parity 3 sows in pens at 3.3 m2 (Table 2Go). Regardless of space, parity 4 sows in pens had greater (P < 0.05) BW than did parity 4 sows in stalls (Table 2Go). At gestation d 110, parity 2 and 3 sows in pens at 3.3 m2 had greater (P < 0.05) BW than did parity 2 and 3 sows in stalls. Parity 4 sows in pens at 2.3 m2 also had greater (P = 0.04) BW at d 110 than did parity 4 sows in stalls (Table 2Go). In general, sows in pens at 2.3-m2 floor space had deeper (P ≤ 0.05) mean BF at all parities than did sows in other space treatments (Table 2Go). Within respective parities, BF depth adjusted for sow BW was deeper (P ≤ 0.05) in sows that gestated in pens at 2.3 m2 than in those in pens at any other space allowance (Table 2Go). Only parity 2 sows in stalls had less (P < 0.001) BF depth when adjusted for BW than did those sows in pens, regardless of space allowance.


View this table:
[in this window]
[in a new window]

  		Table 2. Least squares means (±SE1) for significant treatment x parity interactions for performance, productivity, and litter-related traits for parity 2, 3, and 4 sows in pens at various floor-space allowances or individual stalls during gestation
 
Treatment x parity interaction occurred for a few litter-related traits, but there were few differences among space allowance within parity for these measures. Parity 2 and 3 sows in pens at 3.3 m2 floor space had larger (P ≤ 0.01) litters than did the same parity sows in pens at 2.3 m2 or stalls, and parity 4 sows in pens at 3.3 m2 also farrowed larger (P < 0.05) litters than did sows in stalls (Table 2Go). Parity 3 sows in pens at 1.4 m2 farrowed larger (P ≤ 0.01) litters than did parity 3 sows in pens at 2.3 m2 (Table 2Go). Among sows in pens, regardless of space allowance, there was no space x parity interaction for number of pigs weaned, but primiparous sows in pens at 3.3 m2 (9.5 ± 0.68) weaned more (P = 0.03) piglets than did sows in stalls (8.8 ± 0.25). Treatment x parity interaction also occurred for litter and piglet BW at weaning and BW gain. Litters from parity 3 sows that had gestated in stalls had greater (P = 0.05) BW at weaning than did litters from sows in pens at 1.4 m2 (Table 2Go). Litter BW at weaning was greater (P < 0.05) for parity 4 sows in pens at 1.4 m2 than did litters from parity 4 sows in pens at 3.3 m2 (Table 2Go). Piglets weaned from parity 4 sows in pens at 3.3 m2 also had less (P < 0.001) BW and BW gain from birth-to-weaning than did those piglets from sows that gestated in pens at 1.4 or 2.3 m2 or in stalls (Table 2Go).

For body condition and lesion scores and health descriptors, data were analyzed during first 2 wk (phase 1) and weekly throughout the duration of the study (phase 2); thus results are presented by phase. There was space x parity interaction for sow BCS in phases 1 and 2 (P < 0.001). During phase 1, parity 3 sows in pens at 2.3 or 3.3 m2 and stalls had greater (P ≤ 0.05) BCS (3.3 ± 0.18, 3.4 ± 0.18, and 3.6 ± 0.11, respectively) than did parity 3 sows in pens at 1.4 m2 (2.8 ± 0.10). Again during phase 2, parity 3 sows in pens at 1.4 m2 had less (P ≤ 0.01) BCS (2.3 ± 0.13) than did parity 3 sows in pens at 2.3 or 3.3 m2 and stalls (3.5 ± 0.09, 3.9 ± 0.12, and 3.4 ± 0.08, respectively).

There was space treatment x parity interaction for dung scores in phases 1 (P = 0.024) and 2 (P < 0.001). Despite this, all scores were <1 regardless of treatment or parity (data not shown). Interaction for lesion scores at ears (P ≤ 0.08), shoulders (P ≤ 0.01), and hind legs (P < 0.01) during both phases of the experiment were detected, but there were few effects among space allowance within parities, except for shoulder lesions.

Shoulder lesion scores were similar among pen space during both phases, except that parity 2 sows in pens at 1.4 m2 (4.1 ± 0.42) had greater (P < 0.05) shoulder lesion scores than did parity 2 sows in pens at 2.3 and 3.3 during phase 1 (2.6 ± 0.32, 2.5 ± 0.28, respectively). During phase 1, shoulder lesion scores were greater (P ≤ 0.001) among primiparous sows and parity 2 sows in pens at 1.4 m2 (2.8 ± 0.34, 4.1 ± 0.43; respectively) than for primiparous and parity 2 sows in stalls (0.80 ± 0.21, 0.78 ± 0.28; respectively). Primiparous sows in pens at 2.3 m2 had greater (P < 0.001) shoulder lesion scores (2.6 ± 0.28) than did sows in stalls. During phase 2, shoulder scores in primiparous sows (2.4 ± 0.23, 1.9 ± 0.23, 2.1 ± 0.20; respectively) and parity 2 sows (3.3 ± 0.24, 2.8 ± 0.21, 3.0 ± 0.23; respectively) in pens across floor space allowances had greater (P < 0.001) shoulder lesion scores than did those primiparous (0.48 ± 0.13) and parity 2 (0.69 ± 0.19) sows in stalls. Interaction for rear lesion scores was only detected during phase 2; parity 2 sows kept in pens (regardless of floor space) had greater (P < 0.001) rear lesion scores than did those sows in stalls (data not shown).

Days to rebreed, culling rate, various litter-related traits, and lesion at some regions of the body were not affected by gestation space treatment, parity, or space x parity interaction.

Treatment and Parity Effects: Sow-and Litter-Related Traits
Probability values and means for gestation space effects on sow and litter performance-related traits are presented in Table 3Go. Sow BW at d 110 (P = 0.03) and 131 (P = 0.08) were both affected by gestation space allowance, but neither BW gain nor loss was affected (Table 3Go). At d 110 of gestation, sows in pens at 2.3 or 3.3 m2 of floor space had greater (P ≤ 0.05) BW than those in pens at 1.4 m2 or in stalls (Table 3Go). Sow BW at d 131 tended (P = 0.08) to be affected by gestation space; sows in pens at 2.3 m2 floor space had greater (P ≤ 0.05) BW than did sows in pens at 1.4 m2 or stalls (Table 3Go). Mean BW for sows in pens at floor space of 2.3 and 3.3 m2 was greater (P < 0.05) than for sows in either pens at 1.4 m2 or stalls (Table 3Go). Sow BF depth also was affected by gestation space (Table 3Go). Sow mean BF and BF covariate-adjusted for BW were deepest (P < 0.001) in sows kept in pens at 2.3 m2 floor space compared with those in all other floor space treatment groups (Table 3Go).


View this table:
[in this window]
[in a new window]

  		Table 3. Main effects of treatment on productivity and litter-related traits for sows in pens at various floor-space allowances or individual stalls during gestation (least squares means ± SE1)
 
There was an overall space effect on litter size (P = 0.001; Table 3Go). Sows in pens at 3.3 m2 floor space farrowed more (P < 0.05) piglets per litter than did sows in any other treatment, but the number of pigs weaned was similar among treatments (Table 3Go). Litter weaning BW and piglet birth-to-weaning BW gains were affected by gestation floor space (Table 3Go). Litter BW at weaning was similar for litters from sows kept in pens regardless of floor-space allowance, but piglets from sows kept in stalls had greater (P = 0.02) BW at weaning than did piglets from sows that gestated in pens at 2.3 m2. Piglets from sows that gestated in pens at 1.4 m2 or stalls tended to gain more (P < 0.10) BW from birth-to-weaning than did those from sows kept in pens at 3.3 m2 of floor space (Table 3Go).

Parity also affected sow BW at d 25, 110, and 131 (P < 0.001). Primiparous sows had less BW (P ≤ 0.001) than did multiparous sows at all measured time-points (Table 4Go). Mean BF was less (P < 0.001) in primiparous sows, but when BF was adjusted for sow BW, BF depth was deeper in primiparous sows than in sows at all other parities (Table 4Go). Parity did not affect litter size in any floor space group.


View this table:
[in this window]
[in a new window]

  		Table 4. Parity effects on performance, productivity, and litter-related traits for sows in pens at various floor-space allowances or individual stalls during gestation (least squares means ± SE1)
 
Treatment and Parity Effects: Health Descriptors and Body and Lesion Scores
Probability values and means for space effects on health descriptors and body condition and lesion scores are presented in Table 5Go. Again, data were initially collected every 3 d for the first 2 wk (phase 1), then weekly thereafter for the duration of the study (phase 2).


View this table:
[in this window]
[in a new window]

  		Table 5. Main effects of treatment on health descriptors, BCS, and lesion score for sows in pens at various floor-space allowances or individual stalls during gestation (least squares means ± SE1) in phases 1 and 2
 
Sow BCS was affected by gestation floor space during both phases of the study (Table 5Go). Sows in pens at 2.3 or 3.3 m2 of floor space had greater (P ≤ 0.01) BCS (greater condition) than did sows at 1.4 m2 during phase 1 and 2. Sows in stalls had greater (P < 0.001) BCS than did those sows in pens at 1.4 m2 during phase 1 (Table 5Go). Sows in stalls also had greater (P < 0.001) BCS than did sows in any pen spacing during phase 2 (Table 5Go).

During both phases, body lesions at ears, neck, shoulders, rear, and hind legs were all affected by gestation space allowance (P < 0.001); sows in pens had greater body lesion scores (more lesions) than did sows in stalls (Table 5Go). Sows in pens at 1.4 m2 had more body lesion scores at shoulders and rear than did sows in pens at 2.3 and 3.3 m2 during both phases (P < 0.01; Table 5Go). Moreover, sows in pens at 1.4 and 3.3 m2 had greater (P < 0.01) hind leg lesion scores than did sows in pens at 2.3 m2. During phase 2, head and ear lesion scores were greater in sows in pens at 1.4 m2 than in those in other pen floor spacings or stalls (P < 0.01; Table 5Go). Also, sows in pens at 3.3 m2 had less body lesions (0.36 ± 0.08) at back than did sows in pens at 1.4 m2 (0.68 ± 0.08; P < 0.001).

Gestation space allowance also affected perivulva lesion score (P < 0.001). During phase 1, sows in pens at 1.4 m2 of floor space had more perivulva scores of 1 (27.2%) than did sows in pens at any other floor space allowances (16% combined) or stalls (6.0%). Those same sows also had greater 1 scores at the perivulva site during phase 2 (26.5%) than did sows in pens (19% combined) or in stalls (9.7%). Lesion scores on chest, breast, front legs, and udder were similar among floor space treatment groups during both phases (data not presented).

Dung scores differed during the first 2 wk on experiment (phase 1), being greater (P < 0.05) for sows in pens at 2.3 m2 and in stalls than for sows in pens at 1.4 m2 (Table 5Go). During phase 2, dung scores also were greater (P < 0.001) for sows in pens at 2.3 or 3.3 m2 or in stalls than for those in pens at 1.4 m2; however, the mean dung score for all floor spacings was low, at <1 (Table 5Go).

There was no space effect on a sow’s ability to stand up or lameness score during phase 1. But during phase 2, the ability of a sow to stand up (P = 0.03) and her lameness score (P < 0.001) were affected. Sows in pens at 3.3 m2 had more standing-up scores of 1 (9.4 %) than did sows in pens regardless of floor space (5.5%) or in stalls (6.6 %). Sows in pens at 3.3 m2 also received more lameness scores of 1 (19.2%) than did sows in pens at 1.4 (9.0%) or 2.3 m2 (11.8%) or in stalls (9.0%).

Hair coat was similar among all sows regardless of floor space during the first 2 wk of the study (phase 1). During phase 2, gestation space tended (P = 0.08) to affect hair coat score, with more sows in pens at 3.3 m2 of floor space (6.3%) having a score of 1 (poor) as against those in the other pen spacings (4.3%). Sows in stalls had greater frequency (7.2%) of hair coat score of 1 than did sows in pens at 1.4 or 2.3 m2.

Parity also influenced several health descriptors (Table 6Go). Sow BCS was affected by parity during study phase 1 and 2, being greater (P < 0.01) in parity ≥4 sows than in all other sows (Table 6Go). Parity also influenced lesion scores at head, ears, neck, shoulders, rear, and hind legs during both phases (P < 0.001; Table 6Go). In general, primiparous sows had reduced head, neck, shoulders, rear, and hind leg lesion scores throughout the experiment (P < 0.05). During phase 1, parity ≥4 sows had the greatest lesion scores at head, neck, and shoulders than did primiparous sows (P < 0.05; Table 6Go), whereas, during phase 2, parity ≥2 sows had greater lesion scores at head, neck, shoulders, rear, and perivulva than did primiparous sows. Dung score was influenced by parity during phases 1 (P ≤ 0.05) and 2 (P < 0.001), but hair coat condition was similar across parities. During phase 1, primiparous sows had greater (P = 0.01) dung scores than did parity 3 sows (Table 6Go). During phase 2, primiparous sows also had greater dung scores than did parity 2 (P = 0.02) or ≥5 sows (P < 0.001; Table 6Go).


View this table:
[in this window]
[in a new window]

  		Table 6. Main effects of parity on health descriptors, BCS, and lesion score for sows in pens at various floor-space allowances or individual stalls during gestation (least squares means ± SE1) in phases 1 and 2
 
Effects of Floor Space
Sow Performance and Litter Productivity.
There were significant linear and quadratic effects of floor space on sow mean BW, BF, BW-adjusted BF, and BW loss. Both sow mean BW (P < 0.005) and BF (P = 0.01) increased in linear fashion as floor space increased. Sow BW loss (d 110 to 131) tended (P = 0.07) to respond in a quadratic manner, with sows in pens at 2.3 m2 of floor space losing the least BW. Backfat (P < 0.001) and adjusted BF (P < 0.001) also responded in quadratic fashion, with sows in pens at 2.3 m2 having the most BF depth.

Several litter-related traits were associated linearly and quadratically with sow floor-space allowance during gestation. Litter size, litter weaning BW, piglet weaning BW, and birth-to-weaning BW gain had positive linear trends with sow floor space allowance (P < 0.05). Number of piglets weaned and male:female ratio were affected quadratically; litters from sows that gestated in pens at 2.3 m2 of floor space each weaning fewer pigs (P = 0.03) but having more male young (P = 0.04).

Body Condition and Lesion Scores.
There were linear and quadratic responses to floor space allowance for BCS during both phases of the experiment (P < 0.001). The BCS was affected in quadratic fashion during phases 1 (P = 0.02) and 2 (P < 0.001), with sows in pens kept at 2.3 m2 of floor space each having the greatest BCS.

During both study phases, there also were linear and quadratic responses for lesion scores in sows kept in pens at different floor space allowances. During phase 1, for sows kept in pens, as floor space allowance decreased, lesion score increased at shoulders (P = 0.001), rear (P < 0.001), and hind legs (P = 0.04). Lesion scores at the neck during phase 1 also responded in quadratic fashion to floor space (P = 0.04); sows in pens at 2.3 m2 had reduced scores. A similar linear relationship was apparent in phase 2 for lesion scores at head (P = 0.001), ears (P = 0.003), neck (P = 0.01), shoulders (P < 0.001), back (P < 0.001), and rear (P < 0.001). During phases 1 and 2, rear (P < 0.001; P < 0.0001) and hind leg (P < 0.001; P = 0.04) lesion scores also were associated quadratically; once again, sows kept in pens at 2.3 m2 of space had reduced scores. Shoulder lesions were reduced (P = 0.004) in sows kept at 2.3 m2 compared with sows in the other pen spacings.


   DISCUSSION
Top
 Abstract
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
The numbers of observations made in this experiment were relatively large. Hence, the extent of statistical significance levels tended to be more extreme than usual. Nevertheless, for the most part, treatment means for performance traits fell within respective ranges of commercial-industry norms (McGlone and Pond, 2003Go). Consequently, although floor space allowance affected performance of sows kept in group-pen systems during gestation; no optimal floor space allowance can be identified from results of this experiment alone. It is apparent that there are positive and negative aspects in all systems currently evaluated, and there is a need for improvement, regardless of the housing system.

One remarkable finding of this experiment was the lack of impact of the relatively high lesion scores in sows kept in pens, especially at 1.4 m2 of floor space, on sow productivity. Moreover, treatment x parity interactions were detected for many of the measured traits assessed in this study, although very few were detected among treatments within parity. Most such effects were found among parity 2, 3, and 4 sows.

Results of the research reported herein, first, confirm conclusions in a review of welfare issues related to sow-housing systems (Barnett et al., 2001Go) and a report of meta-analyses performed on scientific studies of performance of sows kept in group pens or individual stalls (McGlone et al., 2004aGo). In particular, in this experiment, the performance of dry sows kept in gestation stalls was often similar to that of sows kept in group pens. In particular, numbers of piglets born alive to and weaned from sows that gestated in stalls did not differ among treatments, and BW and BF means for sows in stalls were very close numerically to (and, in the practical sense, no different from) the means for sows in groups. However, few differences were found among treatments.

Reproductive performance of sows and productive performance of piglets were quite similar regardless of the experimental environment in which the sow resided during pregnancy. These findings are similar to those reported by others (Langendijk et al., 2000Go). Sows in pens at 3.3 m2 of floor space farrowed more piglets per litter, but the number of pigs weaned was similar among treatments. But, piglets from sows that gestated in stalls had greater BW at weaning than did piglets from sows that gestated in pens, especially at 2.3 m2, and tended to gain more BW from birth-to-weaning. Bates et al. (2003)Go found that piglets from sows that gestated in groups were heavier at birth but found no effect on the number of piglets born alive or weaned; however, piglets from sows that gestated in stalls were heavier at weaning.

Sow performance in this study agreed with the concept that what is considered to be acceptable animal state of being exists over a range of environmental conditions provided by the variety of keeping systems evaluated in the experiment; and although on the "welfare plateau" an improvement in environmental quality may well enhance animal state of being, a system located even at the lowest point of the "welfare plateau", by definition, will meet minimal ethical standards for acceptability (Curtis, 1985Go). Others have reported that sow performance is greater for sows kept in stalls than in groups (Peltoniemi et al., 1999Go; Kongsted, 2005Go; Karlen et al., 2006Go), whereas others have found greater performance among sows in groups (Bates et al., 2003Go).

Individual Stalls vs. Group Pens
At certain times during pregnancy, BW was greater in sows kept in pens, but BW of even those sows residing in stalls or in pens at 1.4 m2 floor space was within the acceptable range (McGlone and Pond, 2003Go). Also, BCS of sows in stalls averaged 3.7, which was greater than that of those sows gestating in pens. But, again, stall and pen sows alike all averaged "very good" BCS (Coffey et al., 1999Go). The differences in BCS that were found might be explained in a number of ways.

First, some sows in gestation stalls might have retained less dietary ME because they had less opportunity to huddle with one another on those occasions when effective environmental temperature in the sow micro-environment was relatively low (Verstegen and Curtis, 1988Go). Secondly, sows in pens might have accumulated more muscle mass during pregnancy (Marchant and Broom, 1996Go), and this might have resulted in an increase in BW but not necessarily in BCS. Finally, it is plausible that some systematic but unintentional bias entered the BCS evaluations due to observer vantage point. The BCS of sows in pens was evaluated while the sow was standing in a pen, whereas BCS of those in stalls was evaluated while the sow was standing in a stall. The different points of view might have introduced optical illusions. Thus, sow BCS was used as linear and quadratic trends in the statistical model for lesion scores.

However, only shoulder and rear lesion scores during phase 1 (d 0 to 13) of the study were influenced by sow BCS. This may imply that this discrepancy most likely cannot be explained by any relationship between sow BCS and her lesion score. Although sows in stalls on average had greater BCS, those sows also on average had less BF than did the sows in pens, a relationship opposite what would be predicted (Coffey et al., 1999Go).

Body lesion scores were consistently greater in sows kept in pens regardless of floor space compared with those sows kept in stalls. Others have reported similar results (Gjein and Larssen, 1995Go; Anil et al., 2003Go; Karlen et al., 2006Go). The greatest lesion scores, regardless of body region, were found in sows that gestated in pens at 1.4 m2. In sows in pens, the greatest lesion scores during the first 13 d of the experiment (phase 1) were recorded at the neck, shoulders, and rear, whereas during phase 2, lesion scores were greater at shoulders and rear only. Moreover, frequency of perivulva lesions was much greater in sows in pens at 1.4 m2 than in those in any other treatment group. Despite substantial lesion scores in the sows in pens at 1.4 m2 floor space, no major differences were found in overall performance among sows in groups at other floor space allowances or in stalls.

Space Allowance in Group Pens
Sows in pens at 1.4 m2 floor space each generally had less BCS, BW, and BF than did those kept in pens at 2.3 or 3.3 m2. Traits of even those sows kept in pens at the minimal floor-space allowance evaluated (1.4 m2/sow) were within normal acceptable ranges (McGlone and Pond, 2003Go). Moreover, although those same sows consistently had greater lesion scores throughout the study, most often their performance was similar to or tended to be greater than that of sows in the other pen treatments. For instance, piglets from sows kept during pregnancy in pens at 1.4 m2 of floor space tended to gain more BW from birth to weaning than did those from sows kept in gestation pens at 3.3 m2. However, the significant treatment differences found in litter weaning weight most likely reflected treatment differences in number of pigs weaned per litter. Overall, there was no systematic effect of space allowance on reproductive and productive performance of group-kept sows and their litters. Sows in pens at 3.3 m2 floor space did farrow larger litters, but number of pigs weaned was similar across pen treatments. Moreover, there were several productivity, BCS, and lesion score traits that responded linearly and quadratically to floor space allowance for sows in groups during gestation. Most often, as floor space per sow increased, values for a particular trait increased. Among the traits that responded in quadratic manner (e.g., lesion score), sows kept in pens at 2.3 m2 floor space had lesser lesion scores than did those in other pen treatments. Despite these differences in lesion scores, there were only limited differences on sow- and litter-related traits.

Parity
The several effects of parity on measured performance traits that were found here followed conventional trends (McGlone and Pond, 2003Go; McGlone et al., 2004aGo). It is of interest to note that there was little evidence that younger sows were at any disadvantage in a group pen during gestation. In fact, lesion scores were often greater in greater-parity sows than in primiparous sows, regardless of treatment. Two possible explanations exist for this finding; most of the agonistic activity within the group may have occurred between greater ranking individuals, and the greatest ranking individuals would more likely be attacked during rank-order challenges by midranking individuals. Sow BF depth adjusted for BW was actually deeper in primiparous sows than in greater-parity sows, perhaps because the older sows had lower BF than did primiparous because it was soon after their previous lactation. In general, BW and reproductive performance was similar among primiparous sows. Harris et al. (2006)Go reported no differences in reproductive performance or BW between gilts that gestated in stalls compared with gilts housed in small pens. Moreover, in the current study the differences that were detected across parities were most often found among parity 2, 3, and 4 sows.

Other
The decision to compose a penned group of 5 sows instead of some other number was based on the concept that smaller groups of pigs generally have both a more stable social hierarchy and more linear hierarchical relationships (Gonyou, 2001Go) and the empirical observation that, in dry sows in groups, social tension seems to be minimized and reproductive performance maximized in groups numbering around 5 (P. Dziuk, University of Illinois, Urbana, personal communication). That may have been the case in this study because in sows in pen treatments, body lesion scores increased substantially during the first 10 to 13 d of study, with scores most often peaking at d 10. Often, by d 13, lesion scores began to decline, regardless of floor space allowance, but they never returned to baseline level. Once lesion scores had plateaued, they remained stable throughout the experiment. Sows kept in pens at 1.4 m2 of floor space consistently had the greatest lesion scores. It is possible that, in the future, producers may want to pen sows in smaller groups based on increased understanding of the social dynamics of sows in groups.

Just as space allowance can affect sow performance traits (e.g., BCS, BW, and BF), it is possible that changes in the physical aspect of the gestation stall can do likewise (Curtis et al., 1989Go; Tauson, 1995Go; McGlone et al., 2004bGo). For example, a gestation stall design that allows a sow more freedom of movement and the opportunity to turn around, which in itself is 1 of the 5 freedoms, may improve sow well-being (McFarlane et al., 1988Go; Johnson et al., 1990Go). The physical exercise associated with sow movement also might result in more muscle and BW gain over the gestation period (Marchant and Broom, 1996Go). Another feature that may improve the well being of an individually kept sow is the size of the gestation stall. To accommodate the range of body sizes of pregnant sows on commercial farms, stall sizes for the majority of sows would need to increase, as would total floor space, according to findings reported by McGlone et al. (2004b)Go. Thus, it is apparent that there are advantages and disadvantages associated with all housing systems.

Although there are advantages and disadvantages associated with sows kept individually as against in groups during gestation, no housing system has been improved, developed, or optimized such that it ultimately improves the welfare of the gestating sow. Again, the research reported herein yielded no evidence that the experimental living environments evaluated had any marked effects on sow or pig performance. In contemporary, controlled comparisons, mainly groups of 5 sows at 1.4, 2.3, or 3.3 m2 of floor space per sow or individual stalls (1.34 m2 per sow), the reproductive and productive performance of sows and the preweaning performance of their piglets were similar over 2 successive parities of sows. Moreover, when sows were kept in any 1 of these 4 different gestation systems, the biologic performance of sows and piglets in all accommodations fell within acceptable industry norms. No one experimental system substantially improved sow well-being.


   Footnotes
 
1 The research herein reported was supported by the National Pork Board and the Illinois Agricultural Experiment Station. The authors gratefully acknowledge M. A. Sutherland, J. Shott, S. A. Cumber, and S. R. Tuftedal for technical assistance. Back

2 Corresponding author: johnso17{at}uiuc.edu

Received for publication July 28, 2006. Accepted for publication March 1, 2007.


   LITERATURE CITED
Top
 Abstract
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 


Anil, L., K. M. Bhend, S. K. Baidoo, R. Morrison, and J. Deen. 2003. Comparison of injuries in sows housed in gestation stalls vs. group pens with electronic sow feeders. J. Am. Vet. Med. Assoc. 223:1334–1338.[CrossRef][Medline]

Barnett, J. L., P. H. Hemsworth, G. M. Cronin, E. C. Jongman, and G. D. Hutson. 2001. A review of the welfare issues for sows and piglets in relation to housing. Aust. J. Agric. Res. 52:1–28.[CrossRef]

Bates, R. O., D. B. Edwards, and R. L. Korthals. 2003. Sow performance when housed either in groups with electronic sow feeders or stalls. Livest. Prod. Sci. 79:29–35.[CrossRef]

Bracke, M. B., J. H. Metz, B. M. Spruijt, and W. G. Schouten. 2002b. Decision support system for overall welfare assessment in pregnant sows B: Validation by expert opinion. J. Anim. Sci. 80:1835–1845.[Abstract/Free Full Text]

Bracke, M. B., B. M. Spruijt, J. H. Metz, and W. G. Schouten. 2002a. Decision support system for overall welfare assessment in pregnant sows A: model structure and weighting procedure. J. Anim. Sci. 80:1819–1834.[Abstract/Free Full Text]

Coffey, R. D., G. R. Parker, and K. M. Laurent. 1999. Assessing sow body condition. In Publ. ASC 158, University of Kentucky Cooperative Extension Service, Lexington, KY.

Curtis, S. E. 1985. What constitutes animal well-being? Pages 1–14 in Animal Stress. G. P. Moberg, ed. American Physiological Society, Bethesda, MD/Williams & Wilkins, Baltimore, MD.

Curtis, S. E., R. J. Hurst, T. M. Widowski, R. D. Shanks, A. H. Jensen, H. W. Gonyou, D. P. Bane, A. J. Muehling, and R. P. Kesler. 1989. Effects of sow-crate design on health and performance of sows and piglets. J. Anim. Sci. 67:80–93.[Abstract/Free Full Text]

Gjein, H., and R. B. Larssen. 1995. Housing of pregnant sows in loose and confined systems—A field study. 1. Vulva and body lesions, culling reasons and production results. Acta Vet. Scand. 36:185–200.[Medline]

Gonyou, H. W. 2001. The social behaviour of pigs. Pages 147–176 in Social Behaviour in Farm Animals. L. J. Keeling and H. W. Gonyou, ed. CABI Publ., Wallingford, UK.

Harris, M. J., E. A. Pajor, A. D. Sorrells, S. D. Eicher, B. T. Richert, and J. N. Marchant-Forde. 2006. Effects of stall or small group gestation housing on production, health and behavior of gilts. Livest. Sci. 102:171–179.[CrossRef]

Johnson, R. W., S. E. Curtis, R. K. Balsbaugh, and I. A. Taylor. 1990. Sow behavior in a hinged free-pivoting sided gestation system. J. Anim. Sci. 68(Suppl. 1):263–264.

Karlen, G. A. M., P. H. Hemsworth, H. W. Gonyou, E. Fabrega, A. D. Strom, and R. J. Smits. 2006. The welfare of gestating sows in conventional stalls and large groups on deep litter. Appl. Anim. Behav. Sci. doi:10.1016/j.applanim.2006.0.014

Kongsted, A. G. 2005. A review of the effect of energy intake on pregnancy rate and litter size-discussed in relation to group-housed non-lactating sows. Livest. Prod. Sci. 97:13–26.[CrossRef]

Langendijk, P., N. M. Soede, and B. Kemp. 2000. Effects of boar contact and housing conditions on estrus expression in weaned sows. J. Anim. Sci. 78:871–878.[Abstract/Free Full Text]

Marchant, J. N., and D. M. Broom. 1996. Effects of dry sow housing conditions on muscle weight and bone strength. Anim. Sci. 62:105–113.

McFarlane, J. M., K. E. Bøe, and S. E. Curtis. 1988. Turning and walking by gilts in modified gestation crates. J. Anim. Sci. 66:326–333.[Abstract/Free Full Text]

McGlone, J. J., and W. G. Pond. 2003. Pig Production: Biological Principles and Applications. Thomson Delmar Learning, Clifton Park, NY.

McGlone, J. J., B. Vines, A. C. Rudine, and P. DuBois. 2004b. The physical size of gestating sows. J. Anim. Sci. 82:2421–2427.[Abstract/Free Full Text]

McGlone, J. J., E. H. von Borell, J. Deen, A. K. Johnson, D. G. Levis, M. Meunier-Salaun, J. Morrow, D. Reeves, J. L. Salak-Johnson, and P. L. Sunberg. 2004a. Review: Compilation of the scientific literature comparing housing systems for gestation sows and gilts using measures of physiology, behavior, performance, and health. Prof. Anim. Sci. 20:105–117.

NRC. 1998. Nutritional Requirements of Swine. Natl. Acad. Press, Washington, DC.

Peltoniemi, O. A. T., R. J. Love, M. Heionen, V. Tuovinen, and H. Saloiemi. 1999. Seasonal and management effects on fertility of the sow: A descriptive study. Anim. Reprod. Sci. 55:47–61.[CrossRef][Medline]

Tauson, R. K. 1995. Comparative evaluation and development of housing systems for laying hens. Pages 83–93 in Animal Behavior and the Design of Livestock and Poultry Systems. Northeast Agricultural Engineering Service, Ithaca, NY.

Turner, S., M. Ewen, J. Rooke, and S. Edwards. 2000. The effect of space allowed on performance, aggression and immune competence of growing pigs housed on straw deep-litter at different group sizes. Livest. Prod. Sci. 66:47–55.[CrossRef]

Verstegen, M. W. A., and S. E. Curtis. 1988. Energetics of sows and gilts in gestation crates in the cold. J. Anim. Sci. 66:2865–2875.[Abstract/Free Full Text]



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
jas.2006-510v1
85/7/1758    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Salak-Johnson, J. L.
Right arrow Articles by Curtis, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Salak-Johnson, J. L.
Right arrow Articles by Curtis, S. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS