The University of Minnesota's swine research facility at Waseca, MN, compares sows housed in individual stalls vs. pens, but the jury is still out.

The coordinator of sow housing research at one of the largest swine research facilities in the United States has concluded that both types of confinement — individual stalls and group pens — can adequately accommodate the needs of sows in gestation.

Samuel K. Baidoo joined the University of Minnesota's Southern Research and Outreach Center at Waseca, MN in 1999. “We decided to compare the impact of two housing systems — individual stalls and group pens — on sow longevity and productivity in the new swine research facility completed in 2001,” explains Baidoo.

The project modification and sow herd expansion includes about 800 sows — 400 sows housed in individual gestation stalls and six pens of 60 sows each (360) with one electronic sow feeder/pen (Osborne Industries, Inc.).

There are eight farrowing rooms with 16 farrowing units/room. Groups of 64 sows are washed, weighed and probed for backfat, before farrowing and at weaning, to determine changes in weight and backfat during lactation. Litters are weaned every two weeks at 18 days of age and 500-600 pigs are fed out in nearby wean-to-finish facilities, Baidoo explains. Surplus pigs are fed out by area producers and in finishing facilities at the University of Minnesota-Morris.

After farrowing, sows are returned to their respective gestation housing systems. “It is a very interesting dynamic system in the sense that the difference between stalls and group housing is more of a perception than an issue of sow productivity or animal care,” he says.

One focus at the southern Minnesota research facility has been on longevity, comparing sow survival in the two systems. “There are sows in both systems that have been here since 2001, and some of these sows are close to their 14th parity,” Baidoo notes. Data is for 2001 to mid-2007.

Figure 1 for the parity distribution by sow gestation housing system shows the retention of sows within the herd over time. The overall average parity for individual stalls is 4.5 parities, and for group-housed sows is 4.2 parities. At 1.98 years average sow productive lifetime, sow longevity slightly favors individual stalls, compared to group-housed sows at 1.84 years. For example, the 1.98 years was computed from the average number of years the sows were productive on the farm. One parity = 115 days (gestation) + 28 days (farrowing room from Day 109 of gestation to weaning) + 5 days (wean-to-estrus interval) + 22 days (repeats) = 170 days; 365 days/170 days = 2.14 parities/year. Average parity from 1 to 12 = 4.24. Number of productive years within the herd (4.2 parities/2.14 parities/yr) = 1.98 years for sows in individual stalls.

Baidoo and fellow scientist Roger Walker agree that sows shouldn't be culled just because of their age, provided they are still productive, healthy and in good body condition.

Older sows are used to the system and present fewer management challenges. “They are easier to take care of because they respond to management very well,” says Baidoo.

Production Comparisons

“The overall performance of sows in individual stalls was generally as good if not better than performance in group-penned sows,” says Baidoo, associate professor of swine nutrition and management.

When Baidoo pooled litter data for 2001-2007 (Table 1), he found that average total born was 11.5 pigs for sows in the electronic sow feeder (ESF) pen system and 11.6 pigs for sows housed in stalls. Similarly, born alive averaged 10.4 pigs in pens vs. 10.6 in stalls.

More pronounced differences in sow performance appear in Table 2, which reveals a greater total loss of backfat during lactation in sows from individual stalls vs. pens.

Noticeably, farrowing rate for the period averaged 84.5% for sows in the ESF system vs. 88.8% for sows in stalls, says Baidoo.

The impact of gestation housing by parity on average daily lactation feed intake and stillborn rate were virtually identical for both sow groups.

Both housing types recorded wean-to-estrus intervals of about five days.

There was no difference in preweaning mortality rates for the two groups of sows.

Dynamic Group System

Baidoo says there are some valid reasons why performance is slightly lower in the group setting. “Our group system is a dynamic system in the sense that two groups of sows are mixed during gestation.” Groups of sows are added at two different times, as opposed to a static system, where no sows are mixed.

Group-housed sows are bred by artificial insemination in stalls, then returned to pens with pregnant sows five days later. This mixing results in altercations between sows. This is a social behavior of sows to establish hierarchy. Baidoo suspects that is having an impact on litter size.

“What is interesting is that the fighting is not between the newly introduced sow group and the old, established sows in the pens. It's just between the newly introduced sows, because to them everybody is a stranger,” he observes.

The altercations normally last for about 24 to 48 hours. But once that is done, group-housed sows seldom fight.

Walker says his experience has shown that if sows can “tough it out” through two parities in pens, their longevity can equal that of sows in stalls.

Baidoo would prefer to keep group-housed sows in stalls for at least 24-28 days post-breeding to ensure proper embryo implantation, but there isn't adequate room because of the two housing systems. A modification of the system would include a set of individual stalls for breeding and housing sows for 24-28 days, before moving them into a pen system.

Baidoo says groups of 10-15 sows can be moved from pens to farrowing rooms and back without problems, compared to sows always housed in individual stalls.

Management Differences

While the sows seem to perform equally well in stalls or pens, management is another matter. The barn staff prefers individual stalls for treatment because it is easier to find and treat sows. In pens, they have to search through 60 sows to single out one for treatment.

One advantage for groups with the Osborne system is the hand-held, point-and-scan data logger, which reads the sow's ear tag and provides her history.

Baidoo says electronics help, but it doesn't ease the management load. “In fact, it may seem like the group housing system is easy to manage, but it demands more management skills and staff need to pay closer attention to details,” he notes.

Staff will check the computer often to see if a pregnant sow in the pens is sick or has lost the electronic ear tag — “obvious reasons why they won't go to the feeder,” he adds.

Sows that lose their tags cannot be fed because the computer has nothing to read. “That is a problem,” he assures.

Baidoo reports the Osborne electronic sow feeder is a very durable and reliable feeding system that in some ways mirrors individual stalls in feeding sows to body condition. The feeder dispenses feed slowly, allowing the sow to eat all or part of its daily ration at one time. If a sow leaves the feeding station before consuming her daily ration, she can return to eat the balance, because the computer knows she hasn't finished her daily allocation.

A nearby water dispenser allows the sow to enjoy liquid feed, which encourages sows to finish their feed at one setting, Baidoo suggests.

Animal Care

Judging which housing system provides the best care for sows is not such an easy task.

It's well known that stalls provide a controlled environment where individual sows are protected from others in the group, where pregnancy, health and well-being can be maintained.

That compares to a group setting where fights can produce scratches, bruises, feet and leg injuries. Vulva biting occurs as sows queue up to go through the feeding station, says Baidoo.

One thing for sure: older sows and young gilts are never mixed in a pen. Gilts from isolation are bred in stalls; one half of the group stays in individual stalls and the other half is moved into training pens until Day 109 of gestation, when they move to farrowing rooms. Training of gilts to use the ESF systems takes about two to three days. Some gilts learn quickly, while others are a challenge. But eventually, they learn how to use the system for feed. No more training is required as long as the sow stays in the breeding herd.

Newly weaned gilts join their cohorts, half being assimilated in group housing in pens and the other half housed in individual stalls, Baidoo points out.

Visitors Welcome

The Waseca facility offers tours to the public and other researchers. Visitors usually prefer to see sows housed in groups because they can walk freely, socialize and lie next to each other, Baidoo explains.

In contrast, when sows are in individual stalls, they appear to be bored and stressed. “This perception that sows in stalls are always under stress is difficult to understand,” Baidoo declares.

During their studies, sows were tested for cortisol levels. Cortisol is the hormone that indicates stress. Higher cortisol levels indicate higher levels of stress.

“Interestingly, sows in the ESF systems have high cortisol levels almost all of the time,” says Baidoo. He observes the high-stress levels are produced by the constant interaction group-housed sows have with others in the pen.

Housing Studies Continue

Baidoo says the comparison of stalls vs. pens will continue. Changes may include widening gestation stalls to accommodate the large sows in the system. It's common to have 600-lb., Parity 6 sows in the research herd.

Pens may be divided in half so sow groups don't have to be mixed, hopefully reducing the fighting.

From the seven years of comparing stalls to pens in gestation, Baidoo is convinced that productivity is not an issue. Choosing a housing type is a 100% animal care issue.

“But I don't believe that we are ready in the pork industry to go to a 100% group-housing system. We need the individual stalls for breeding and to let the embryos settle after breeding for at least 24-28 days,” he states.

What's more likely to happen is there will be modifications of individual stalls and group housing systems to meet the needs of specific systems.

Table 1. Performance of Sows in Stalls and Pens with ESF
Criteria ESF Stalls
No. of Litters 1,593 1,646
Sow backfat, in.
Farrowing .78 .74
Weaning .70 .67
Loss during lactation .08 .07
Sow body weight, lb.
Day 109 of gestation 519.2 495.0
Weaning 496.3 488.0
Loss during lactation 26.0 26.4
Wean-to-estrus interval, days 5.2 5.1
Farrowing rate, % 84.5 88.8
Table 2. Performance of Sows in Stalls and Pens with ESF
Criteria ESF Stalls
No. of Litters 1,562 1,622
Total born 11.5 11.6
Born alive 10.4 10.6
Pig birth weight (lb.) 3.5 3.5
Pig weaning weight (lb.)* 14.1 13.9
Pigs weaned 9.5 9.6
Preweaning mortality, % 9.1 9.0
* Average weaning age = 18 days