Assessment of the Non All-in All-out System and the All-in All-out System for Sows and Their Suckling Piglets

The goal of this study was to investigate the behavior of sows and their piglets reared using the non-all-in all-out system and the all-in all-out system during a 3-week lactation period. The 24 Landrace×Yorkshire sows (range of parity: 2 to 4) and their litters (range of litter size: 8 to 13 piglets) for each treatment (8 sows/replication) were used in this experiment. The sows were housed in farrowing crates (0.6×2.1 m) located in pens (2.4×1.8 m) with totally perforated metal flooring. The crates were kept in an experimental room that was thermostatically controlled to approximately 23°C. A 250-W infrared heat lamp was turned on above the creep area during lactation. The sows and piglets were conventionally managed. The animals were recorded during a 24-h period at 1, 4, 7, 14 and 21 days postpartum. The videotapes were scanned every minute to obtain an instantaneous behavioral sample. In the behaviors of sows, lateral recumbency decreased and ventral recumbency increased from 1 day postpartum to 21 days postpartum. In comparison to the AIAOS, sows in the NAIAOS spent much less time on lateral recumbency (p<0.01), whereas they spent much more time on ventral recumbency and sitting at 4 days postpartum (p<0.05 and p<0.05). In the behaviors of piglets, lying increased and unsuccessful suckling decreased until 21 days postpartum. Compared to the AIAOS at 4 days postpartum, piglets in the NAIAOS spent less time lying (p<0.01), whereas they spent much more time walking and unsuccessful suckling (p<0.01 and p<0.05). These results suggest that the allin all-out system may be preferable to the non-all-in all-out system for promoting welfare. (Asian-Aust. J. Anim. Sci. 2006. Vol 19, No.


INTRODUCTION
In the current swine industry, the all-in all-out system (about 50% in Korea) has been used largely to prevent diseases, better synchronize the sows' return to estrus, and reduce manual labor.Sows are moved into a farrowing room about 7 days before parturition and are reared until weaning in the same room.According to previous studies (Jensen, 1988;Götz, 1991;Jensen et al., 1991;Lou and Hurnik, 1998), the rearing environment affects the behavior of the sows and their piglets.The sows' and piglets' behavior affects the mortality rate and growth of the piglets, which impacts the number of piglets available for market, and the time needed for them to reach market weight (Mahan and Lepine, 1991).Body damage, impaired growth, and disease are indicators of poor welfare (Broom, 1991).A piglet's mortality rate during the 3-week lactation period ranges from 14.2 (Fahmy and Bernard, 1971) to 23.5% (Dyck and Swierstra, 1987).The three main causes of death are stillbirth, crushing by the sow, and starvation (English and Morrison, 1984;Dyck and Swierstra, 1987).Of these, crushing by the sow and starvation are closely associated with communication between the sow and its piglet.If the sow hears the call of its piglet in the danger of being crushed, the sow can change its posture and the piglet will survive.Starvation occurs if the piglets fail to achieve regular and adequate suckling (Fahmy and Bernard, 1971;English and Morrison, 1984;Dyck and Swierstra, 1987).If piglets get inadequate milk after birth because they can not communicate with their sow due to other noise, they become hypothermic and weak and may be crushed by their sow (Fahmy and Bernard, 1971;Dyck and Swierstra, 1987).The piglet's weaning weight is affected by the sow's nursing frequency (Spinka et al., 1997).Some sows and their piglets housed in a farrowing room occasionally show nursingsuckling behaviors synchronized by mutual sound …stimuli (Walser, 1986;Kasanen and Algers, 2002).
The goal of this study was to investigate the behavior of sows and their piglets reared in the non-all-in all-out system (NAIAOS) and the all-in all-out system (AIAOS) during the first 3 weeks of lactation, and to evaluate the NAIAOS and the AIAOS from a behavioral standpoint.litter size: 8 to 13 piglets) were used in this experiment from the 1 day postpartum to weaning.The windowless farrowing house of the NAIAOS (n = 24 sows; 8 sows/replication) for 72 sows and their piglets was not divided into compartments, whereas that of the AIAOS (n = 24 sows; 8 sows/replication) was divided to raise 8 sows and their piglets in each room.The sows were housed in farrowing crates (0.6×2.1 m) located in pens (2.4×1.8 m) with totally perforated metal flooring.All sows were moved into the farrowing crates, which were kept in rooms that were insulated and thermostatically controlled to about 23°C.A 250-W infrared heat lamp was installed above the creep area and turned on until weaning.Some newborn piglets were cross-fostered immediately after parturition and all piglets had their teeth clipped and tail docked at 1 day postpartum.The male piglets were castrated at approximately 3 days postpartum.The sows were fed a standard ration of commercial concentrate twice daily, between 07:00 and 08:00 h and between 18:00 and 19:00 h, and they had unlimited access to water at all times (Table 1).The piglets were supplied creep feed from 10 days of age (Reese et al., 2001;Table 2).

Observations
On the days of observation, all piglets were marked on the back with a conspicuous number.The behavior of the animals was recorded during a 24-h period at 1, 4, 7, 14 and 21 days postpartum using invisible LED (light-emitting diode) lamps (950-nm wavelength), CCD (charge-coupled device) cameras, multiplexers, and time lapsed VCRs (videocassette recorder).Piglets that were ill or crushed  Relatively low or fast speed locomotion on the ground in which propulsive force derives from the action of legs, which included standing and sitting Successful suckling Successfully switching from teat massage and slow suckling movements to the rapid, regular suckling movements indicative of milk ingestion (Fraser, 1980) when at least 50% of the littermates did so (Milligan, 2001) Piglet Unsuccessful suckling Failed to switch from teat massage and slow suckling movements to the rapid, regular suckling movements indicative of milk ingestion when at least 50% of the littermates did so were excluded from this experiment.The videotapes were scanned every minute to obtain an instantaneous behavioral sample.The mutually exclusive behavioral categories recorded are shown in Table 3.

Statistical analysis
The number of times in which the animals engaged in each of the designated behavioral categories was counted for each observation day.These numbers were then converted into percentages.Data were analyzed using analysis of variance procedures, and were also tested for normality.The data were approximately normal and so were analyzed without transformation.The paired student t-test was used to compare the behaviors of sows and piglets reared using the NAIAOS and the AIAOS.The postures of piglets' standing and sitting were excluded from the analyses because these values were almost all zero.All analyses were performed using SAS (SAS Inst., Cary, NC, 2000).

Sow behaviors
In the NAIAOS, the frequency of lateral recumbency decreased rapidly from 1 day postpartum to 4 days postpartum (p<0.05;Table 4) and then was maintained at an almost constant level, while that of ventral recumbency increased rapidly from 1 day postpartum to 4 days postpartum (p<0.05) and then was maintained at an almost constant level.In the AIAOS, the lateral recumbency frequency decreased from 4 days postpartum to 21 days postpartum (p<0.05), while the ventral recumbency frequency was maintained at an almost constant level from 1 day postpartum to 7 days postpartum and then increased rapidly (p<0.05).The frequency of sitting in the NAIAOS increased gradually from 1 day postpartum to 21 days postpartum (p<0.05),whereas that in the AIAOS was maintained at an almost constant level during the experimental period.In both the NAIAOS and the AIAOS, the standing posture increased gradually up to 14 days postpartum, and then decreased (p<0.05).The frequency of feeding in both the NAIAOS and the AIAOS increased gradually from 1 day postpartum to 7 days postpartum (p<0.05), and then was maintained at an almost constant level.The drinking frequency in the NAIAOS increased gradually from 1 day postpartum to 21 days postpartum (p<0.05),whereas that in the AIAOS increased from 1 day postpartum to 4 days postpartum (p<0.05) and then was maintained at an almost constant level.The order of behavioral frequency in both the NAIAOS and the AIAOS during the experimental period was lateral recumbency, ventral recumbency, sitting, standing, feeding, and drinking.In comparison to the AIAOS at 4 days postpartum, sows in the NAIAOS spent much less time on lateral recumbency (p<0.01),whereas they spent much more time on ventral recumbency and sitting (p<0.05 and p<0.05).

Piglet behaviors
In the NAIAOS, the frequency of lying increased rapidly from 4 days postpartum to 7 days postpartum, decreased rapidly from 7 days postpartum to 14 days postpartum, and then increased again (p<0.05;Table 5).In the AIAOS, the frequency of lying increased gradually from 1 day postpartum to 7 days postpartum, decreased rapidly from 7 days postpartum to 14 days postpartum, and then increased again (p<0.05).The walking frequency in the NAIAOS increased rapidly from 7 days postpartum to 14 days postpartum and then decreased (p<0.05),whereas that in the AIAOS decreased from 1 day postpartum to 4 days postpartum, increased from 4 days postpartum to 14 days postpartum, and then decreased (p<0.05).The successful suckling frequency in the NAIAOS was maintained at almost constant levels during the experimental period, whereas that in the AIAOS increased from 1 day postpartum to 4 days postpartum, decreased from 4 days postpartum to 7 days postpartum (p<0.05), and then was maintained at almost constant levels.The frequency of unsuccessful suckling in both the NAIAOS and the AIAOS decreased gradually between 1 and 21 days postpartum (p<0.05).The order of behavioral frequency in both the NAIAOS and the AIAOS during the experimental period was lying, successful suckling, walking, and unsuccessful suckling.In comparison to the AIAOS at 4 days postpartum, piglets in the NAIAOS spent much less time lying (p<0.01),whereas they spent much more time walking and unsuccessful suckling (p<0.01 and p<0.05).

DISCUSSION
The results in both the NAIAOS and the AIAOS from this study showed that lateral recumbency decreased and ventral recumbency increased during the postpartum period.The study by Götz (1991), observing 3 sows at 1, 2 and 3 weeks postpartum in a similar experimental environment, showed that lateral recumbency decreased between the first week and the following 2 weeks, similar to the results of this study.The data on lateral recumbency and ventral recumbency may have reflected the recovery of the animal's body from parturition (Jeon et al., 2003).In the current study, the frequency of sitting and standing in both the NAIAOS and the AIAOS increased gradually from 1 day postpartum to 14 days postpartum.Jensen (1988) and Stangel and Jensen (1991) reported that, with the advance of lactation, foraging and locomotion increased in a seminatural enclosure, whereas lying down, both laterally and ventrally, decreased.The rise in activity after parturition closely related to the recovery of the animal's body from parturition (Jeon et al., 2003).The order of behavioral frequency at 1 and 4 days postpartum in both the NAIAOS and the AIAOS was lateral recumbency, ventral recumbency, sitting, standing, feeding, and drinking and this was similar to the results reported by Lewis and Hurnik (1985) and Lou and Hurnik (1998).In this study, the nursing frequency was a part of lateral recumbency frequency and the same as the successful suckling frequency of piglets.The nursing frequency in both the NAIAOS and the AIAOS was maintained at an almost constant level.The almost constant trend in nursing frequency between 1 and 3 weeks postpartum was in agreement with previous results (Götz, 1991) for sows in a similar experimental environment, but differed from the decreasing trend seen in seminaturally reared sows (Jensen, 1986(Jensen, , 1988)).This difference may be attributable to the more frequent nursing that occurred in response to the grunting sound stimuli from different sows in a room compared to a seminatural environment (Bate et al., 1999;Kasanen and Algers, 2002) or the difference in individual milk yields of the animals used in each study.Compared to the AIAOS at 4 days postpartum, sows in the NAIAOS spent much less time on lateral recumbency, whereas they spent much more time on ventral recumbency and sitting.The lower lateral recumbency frequency and higher ventral recumbency and sitting frequency of sows in the NAIAOS may be due to the response to the more frequent troublesome massage of piglets that occurred in response to the nursing-suckling sound stimuli from different sows and litters in a room compared to the AIAOS (Bate et al., 1999;Kasanen and Algers, 2002).This suggests that the AIAOS is preferable to the NAIAOS in facilitating rest for the sows.
In both the NAIAOS and the AIAOS, the frequency of lying down in piglets was the lowest on their first day postpartum and the unsuccessful suckling frequency was as high as the result reported by Milligan et al. (2001).This was because the piglets were not completely developed (Jeon et al., 2003).The transition from a nonsynchronous neo-suckling to a cyclical and synchronous suckling in piglets occurs within approximately 10 hours after the birth of the first piglet (Lewis and Hurnik, 1985;Castren et al., 1989), and their playful behaviors increase from about 3 days of age (Blackshaw et al., 1997).Activity, including standing, sitting, and moving about, increases with age (Fraser, 1978).As above, the frequency of walking trended to increase with age in both the NAIAOS and the AIAOS.Compared to the AIAOS at 4 days postpartum, piglets in the NAIAOS spent much less time lying down (or resting), whereas they spent much more time walking.This suggests that the AIAOS is preferable to the NAIAOS for allowing the piglets to rest.The lower lying down frequency and higher walking frequency of piglets in the NAIAOS may be due to the response to the more frequent nursing-suckling sound stimuli by different sows and litters in a room compared to the AIAOS (Bate et al., 1999;Kasanen and Algers, 2002).The order of behavioral frequency at 1 and 4 days postpartum was similar to previous results (Lewis and Hurnik, 1985;Lou and Hurnik, 1998).As with the sows, the piglets spent most of their time during the first 3 weeks of life lying down.In both the NAIAOS and the AIAOS, the successful suckling frequency was maintained at an almost constant level during the experimental period, whereas the unsuccessful suckling frequency decreased gradually as the piglets aged.The results were closely associated with the piglets' behavioral development and the sows' recovery from parturition (Jeon et al., 2003), and reflect the findings of Castren et al. (1989) and Newberry and Wood-Gush (1985).Unsuccessful suckling can cause starvation (Fahmy and Bernard, 1971;English and Morrison, 1984;Dyck and Swierstra, 1987), crushing by the sows (Fahmy and Bernard, 1971;Dyck and Swierstra, 1987), and lower weaning weight.Unsuccessful suckling is an indicator of poor welfare according to Broom (1991).If the grunt signal of the sow is masked by different sows' and litters' noises, the odds of unsuccessful suckling may be increased (Algers and Jensen, 1985;Lewis and Hurnik, 1986;Algers and Jensen, 1991;Algers, 1993;Kasanen and Algers, 2002).The unsuccessful suckling in the NAIAOS was higher than in the AIAOS at 4 days postpartum.No starvation occurred in this study.For reasons mentioned previously, it is believed that the AIAOS is more effective than the NAIAOS for reducing unsuccessful suckling.
In conclusion, since sows and piglets can show more frequently comfortable resting and less unsuccessful suckling in the AIAOS than in the NAIAOS, the AIAOS may be preferred over the NAIAOS for enhancing welfare.

Table 3 .
The mutually exclusive behavioral categories used for behavioral observation

Table 4 .
Percentage (±SD) of the behaviors of sows according to

Table 5 .
Percentage (±SD) of the behaviors of piglets according