The method of housing sows during the breeding-gestation phases of pork production has undergone considerable transition during the last 40 years.

The move to housing sows indoors had several major drivers:

  • A worker could more easily manage a larger number of sows (feeding, vaccinating, mating and moving individual animals);

  • A worker could more easily manage the environmental aspects needed by the sows;

  • More sows could be housed in a smaller area; and

  • Reproductive performance of the herd would be enhanced.

The move to individual gestation stalls made feeding easier, prevented sows from fighting during mixing and feeding, allowed for individual health care and vaccination, provided better control of the dunging area, prevented sows in estrus from excessively riding each other, and allowed more sows to be housed in a specific amount of space.

Although gestation stalls offer the benefits of controlled management, many perceive them to be a welfare problem. As a result, many pork producers are seriously considering their options for housing gestating sows.

Whether constructing a new gestation facility or remodeling an existing one, sow welfare is just one aspect of a total decision-making process that involves many other factors, including capital costs, anticipated performance, ease of management and operating cost.

Health and welfare of sows

People who want individual stalls banned believe that the requirements of the sow are freedom from malnutrition, thermal discomfort, physical discomfort, lack of exercise, injury, disease, suppression of all normal behaviors, fear and stress. Although there is disagreement as to how welfare should be assessed, scientists have assessed the well-being of gestating sows and gilts according to the animal's social behavior, health status, skin lesions, feet and leg soundness, locomotion, muscle weight and bone strength, concentration of cortisol, immune system function, heart rate and reproductive performance.

Some people perceive that the housing of sows as a group is more welfare-friendly. However, the housing of sows in groups can have welfare problems, such as: (1) aggression during mixing of animals; (2) aggression at time of feeding; (3) bullying by dominant animals; (4) injuries of feet, leg and back due to excessive riding of each other during estrus; (5) excessively high feed intake by dominant animals that results in fat sows; (6) excessively low feed intake by subordinate animals that results in thin sows; (7) vulva biting; and (8) wounds and scars from fighting.

The rate of injuries, body lesions, claw lesions and body condition scores can be influenced by feeding method, parity distribution within the pen, floor space per animal, use of bedding and physical properties of the floor.

Sows with low social status within a group have injuries on a greater number of body areas than sows with high social status one week after mixing into either static or dynamic groups. Also, sows with a low social status have a greater number of bodily injuries four weeks after mixing into static groups. Although injury scores have a wide fluctuation for sows housed in an individual stall or pen containing an electronic sow feeder, total injury scores have been reported to be higher for sows housed in pens than for sows housed in stalls.

Group housing of sows

It is important that producers realize there is more than just one group housing system. Some of the many factors involved with group housing are number of animals per pen; size of animals per pen; amount of space per animal; type of flooring (total slats, partial slats, solid concrete floor); use of mold-free bedding, method of feeding (mechanical or non-mechanical, floor feeding, individual feed drops, interval feeding, trickle feeding); use of feeding stalls (locked, unlocked, self-locking, computerized); geographic location; genetic composition of sows; temperament of animals; safety aspects for workers; and complexity of accomplishing work tasks (estrous detection, mating animals, moving animals, feeding animals).

Each group housing system has benefits and drawbacks. Therefore, producers must decide what they want to achieve, and then implement the design components that will most likely reach those goals.

Following is a discussion of numerous aspects of sow gestation management that must be considered:

Reproductive performance

The method of housing sows plays an important, but not exclusive, role in determining reproductive performance of sows. Genetics, health, environment, geographic location, worker skill, management procedures and other factors also impact reproductive performance.

Although a large number of studies have been published comparing sow performance in different housing systems, care must be taken when interpreting data generated from records gathered from several different farms. It is difficult to make absolute conclusions that one type of housing is better than another, because most farms have only one system and, therefore, do not test housing options under the same management and environment.

Table 1 on page 11 shows the influence type of gestation housing has on number of piglets born alive/litter for 19 sets of data. There is no clear and consistent pattern to identify which housing system is the best.

In studies comparing stalls vs. group housing, the percentage indicating a numeric increase in number of piglets born alive/litter was evenly split between sows housed in stalls and sows grouped in pens.

An analysis of records from 71 farms in northern Italy found that housing sows in individual stalls during the entire reproductive cycle (mating and gestation), compared to other housing systems, gave better performance in number of piglets born/litter, farrowing rate and number of piglets weaned/sow/year (Table 2).

The records also showed that grouping sows at 28 to 50 days after insemination produced more weaned pigs/sow/year compared to grouping sows at 14 to 28 days after insemination.

A Swedish study found the number of piglets born alive/litter was greater when sows were housed in stalls compared with sows housed in a group immediately after service and fed with an electronic sow feeder (Table 3). However, in the same study, Herd 2 found a numeric increase of 0.2 piglets born alive/litter for sows housed in groups compared with sows housed in stalls.

Feeding systems

Feed intake during gestation is restricted to prevent excessive body weight gain and fat deposition.

It is known that excessive feed intake during early gestation increases embryonic death in gilts, but not in multiparous sows.

Excessive feed intake during gestation decreases feed intake during lactation. Excessive underfeeding of gestating sows can reduce piglet birth weight, piglet viability and lower body fat reserves at farrowing and weaning.

Research has shown that food deprivation for 48 hours after ovulation is associated with changes in reproductive hormones, changes in metabolic hormones, a decrease in number of sperm cells transported to the sperm reservoir of the oviduct, a lower cleavage rate of embryos and a delayed transport of ova. Fasting of sows on Days 10 and 11 of gestation can also have detrimental effects on reproduction. Thus, the control of feed is a major consideration when designing and managing a gestation facility.

  • Electronic sow feeding (ESF) system

    The computerized feeding system allows sows to be loosely housed and fed individually. The computer can be used to change the total volume of feed each sow receives, and it can be adjusted to give each sow her entire meal in a single visit or during several smaller meals throughout the day.

    Aggressive physical acts, particularly vulva biting, do occur while sows are waiting for their turn to enter the feeder.

    The suggested number of animals per electronic feeder is 40 to 65 sows. The minimum space/sow ranges from 18 to 32 sq. ft. In general, bedding is not used with an electronic sow feeding system in the United States.

    There are essentially two management schemes for ESF:

    Option A uses a static group of 40-65 sows/pen with only one electronic feeding station. All animals in the group are in the same phase of production.

    Option B uses a dynamic group from about 80 up to 200 sows with two to five electronic feeding stations. Every week sows enter and leave the group; thus, the animals are in different productive phases. The introduction of bred sows to an existing group at one to eight days after mating has increased the incidence of bred sows returning to estrus by 10% and reduced litter size by 0.2 piglets/litter compared with introducing bred sows at 22 to 29 days after mating.

    Although the data in a Swedish study is confounded between type of group (dynamic or static) and feeding method (ESF or floor-fed), farrowing rate and litter size born alive was not different when sows were grouped four weeks after mating (Table 3, Herd 3). However, the percentage of sows removed from the group was two times greater in the dynamic housing method.

    Research in Canada found a larger number of piglets born/100 sows bred when sows were mixed more than 35 days after breeding, compared with mixing less than seven days after breeding (Table 4). Although the use of an ESF system helps ensure that sows receive the correct feed allowance, sows with low social ranking have lower bodyweights, higher injury levels, lower position in the feeding order and are displaced more often from the drinkers than high-ranking sows.

  • Feeding-resting stalls

    This feeding system allows the sows to freely roam in a large pen with other sows except when they are fed. The surface of the lying area can be total slats, partial slats, solid concrete with bedding or solid concrete without bedding. The sows enter body-length, individual feeding stations (one feeding stall/sow), where they are fed on the floor or in a trough that continues in front of all the stalls. The purpose of this system is to reduce aggression during feeding. Because each sow can randomly enter any feeding stall, individualized rationing is not possible.

    Body-length stalls are used to improve the welfare of the sows. The body-length feeding stalls are also used as a resting area.

    When bedding is used, the feeding stalls are placed on an 8- to 16-in.-high platform, depending on depth of bedding. The minimum amount of space provided/sow is 14.8 sq. ft.

  • Self-locking or manual-locking individual feeding stall

    Self-locking feeding stalls are designed so the rear opening is closed when a sow enters the stall (Figure 1). Manual-locking stalls allow for easier vaccination, estrous detection and artificial insemination, for example.

    Because each sow can randomly enter any of the feeding stalls, individualized rationing is not possible with this feeding system unless each sow is fed by hand.

  • Non-locking individual feeding stall

    Researchers have investigated the influence of the length of feeding stall partition (19.5 in. wide × 6.5 ft. long body-length stall, 19.5 in. wide × 15.6 in. long shoulder-length stall, or no partition) and type of food (wet or dry) on the amount of aggression, frequency of changing position at the trough and duration of time at feeding trough in groups of pregnant sows.

    When sows were provided dry feed, it was reported that increasing the length of partitions resulted in a significant reduction in the number of bites, total aggressive behaviors and displacement at the trough. Time spent at the trough increased.

    When sows were provided wet feed, the number of bites or duration of time feeding at the trough was not different between body-length and shoulder-length partitions.

    Top-ranking sows received less bites toward their head, shoulder and body, and were less frequently displaced at the trough than sows with a lower ranking when eating from a trough with no partition or shoulder partition. Vulva bites were greater when sows consumed either wet or dry feed from a feeding stall with a body partition, compared to a shoulder-feeding stall or a stall with no partitions. Individualized rationing is not possible with this feeding system.

  • Trickle feeding system

    Another method to possibly limit aggression and feed intake by dominant sows is the trickle feeding system. Sows are usually kept in stable groups (4 to 60 sows) and shoulder-length barriers (19 to 24 in. long) separate the feeding trough. An auger apparatus slowly delivers 0.2 to 0.4 lb. of food/minute over a period of approximately 15 to 30 minutes.

    In the ideal system, there is no incentive for sows to move away from the feeder to bully other sows. The slow rhythm of feed distribution encourages the sows to remain at the feed space for the duration of the feeding period. Because each sow can randomly enter any feeding space, individualized rationing is not possible with the trickle feeding system.

  • Floor feeding

    When feeding on the floor, the highest incidence of aggression occurs during the first 30 minutes after feed is delivered. As expected, dominant sows defend the center of the pile of feed. Subordinate sows quickly grab food at the edges and move only when forced to do so. Unequal feed intake between sows within the group has detrimental effects on body reserves, especially for the low-ranking sows.

Body weight gain is reportedly 40 to 50 lb. lower for low-ranking sows compared to high-ranking sows when floor-fed. Aggression over food during a single feeding is not totally eliminated by providing piles of feed at several locations within the feeding area (Figure 2).

Space requirement

Although the amount of space needed/sow or gilt is a critical factor, the optimal amount of space needed/sow or gilt when group-housed during gestation has not been adequately investigated. The suggested space requirement when sows or gilts are housed in groups in the United States appears in Table 5.

Group size

The optimal number of sows/pen and management procedures have not been adequately investigated. A wide range in number of sows/pen and management procedures are utilized.

With respect to reproductive performance in two research projects, farrowing rate and litter size were not different over the range of five to 40 sows or 12 to 28 sows/pen.

In reality, group size is often confounded with group stability because larger groups can usually only be operated on a dynamic basis. Because the number of pens and size of pens often cannot be easily changed, pork producers quite frequently add recently bred sows to a pen during the breeding phase and during the first 30 days of gestation. The variation in number of sows bred/week or group is a contributing factor to this problem.

Ease of management

Each type of breeding-gestation facility design has advantages and disadvantages, as previously noted. An important consideration when designing a breeding-gestation facility is the ease in performing estrous detection; artificial insemination, pregnancy detection; health procedures; moving of animals (width of alley, open gates cutting off alley, ease of working gate latch); and feeding and watering.

The ease to successfully artificially inseminate estrous females is particularly noteworthy. If weaned sows are housed in groups, a procedure must be implemented whereby estrous sows can be inseminated without being ridden by other sows during the insemination process.

Table 1. Influence of Gestation Housing System on Number of Piglets Born Alive/Litter
Group-housed indoors
Feeding stall
Location of study Stalls indoors Floor fed Locked Open Electronic feeder Trickle feeder
Nebraska 9.80 9.60
Netherlands 10.31 10.11
Sweden 11.03 10.88 11.13
Sweden 11.80 11.50
United Kingdom 10.77 10.70
Netherlands 10.70 10.90 11.00 10.70
Sweden 10.20 10.10 11.30
Sweden 10.40 10.20
Sweden 11.42 11.21 11.34
Texas 8.90 9.90
United Kingdom 10.20 10.50
Sweden 10.02 10.32
Minnesota (Parity 1) 9.80 10.50
Minnesota (Parity 2) 10.11 10.12
Denmark (Herd 1) 11.20 10.70
Denmark (Herd 2) 11.40 11.60
Denmark (Herd 3) 11.90 11.40
Kansas 9.77 9.77
Sweden
(gilts) 9.10 9.70
(sows) 11.8 10.8
Table 2. Influence of Housing Method on Reproductive Performance (71 farms in Northern Italy)
Method houseda
Item G S SG1 SG2 GS
Weaning to mating Group Stall Stall Stall Group
Stage of gestation
0 to 14 days Group Stall Stall Stall Group
14 to 28 days Group Stall Group Stall Stall
28 to 50 days Group Stall Group Group Stall
50 to 110 days Group Stall Group Group Stall
Farrowing rate, %
1997 76.28 77.71 69.60 72.68 70.03
1998 75.85 76.61 70.56 70.59 69.77
Average number of piglets born alive/litter
1997 9.89 10.24 9.49 9.78 10.14
1998 9.87 10.18 9.50 9.63 10.24
Number of piglets weaned/sow/year
1997 19.78 20.82 18.61 19.27 19.28
1998 19.47 20.66 18.06 18.38 18.92
aG is group housing for the entire time of mating and gestation.
S is stall housing for the entire time of mating and gestation.
SG1 is stall housing during mating and grouping during 14 to 28 days of gestation; group housing remaining period of gestation.
SG2 is stall housing during mating and grouping during 28 to 50 days of gestation; group housing remaining period of gestation.
GS is group housing during mating and housing in stalls 14 to 28 days of gestation; stall housing remaining period of gestation.
Table 3. Influence of Gestation Housing System on Sow Performance
Item Herd 1 Herd 2 Herd 3
Type of housing Dynamic group Dynamic group Stall Dynamic group Stall Dynamic group Static group
Feeding ESF1 ESF Individual ESF Individual ESF Floor
Type of flooring Partly slotted Deep litter Partly slotted Partly slotted Partly slotted Partly slotted Partly slotted
Time of mixing After service After service After service 4 wks after service 4 wks after service
No. litters 313 348 354 455 265 364 365
Sows removed, % 17 13 29 24 12
Farrowing rate, % 83 84 87 86 94 94 95
Liveborn per litter 10.7a 10.7a 11.3b 11.9 11.7 11.8 12.0
(Nielsen, 2003)
1ESF = electronic sow feeder; a,bMeans differ (p < .05)
Table 4. Influence of Gestation System on Live Piglets per 100 Sows Bred
Mixed pre-implantation
(< 7 days after mating)
Mixed pre-implantation
(> 35 days after mating)
Item Stalls Static Dynamic Static Dynamic
Gilts 763 666 678 734 763
1st parity 894 891 855 965 914
2nd parity 973 906 958 929 1,020
Mature sows 951 910 884 995 995
(Gonyou, 2004)
Table 5. Recommended Space Requirements for Sows and Gilts Housed in Groups (MWPS-43)
Animal Body weight, lb. Solid floor (sq. ft./head) Fully and partially slotted (sq. ft./head)
Breeding gilt 250 to 300 40 24
Breeding sow 300 to 500 48 30
Gestating gilt 250 to 300 20 14
Gestating sow 300 to 500 24 16

Fixed and Variable Cost

The National Pork Board recently released a CD, “Sow Housing Alternatives Calculator,” which contains spreadsheets to estimate the cost of building or remodeling a gestation facility.

The spreadsheets evaluate the production and financial implications of remodeling an existing individual-stall gestation facility to house sows in groups, building a new gestation facility to house sows as groups, and constructing a new hoop structure that houses sows and feeds them either indoors or outdoors.

The main input categories of the model include cost of building structure, cost of equipment, annual ownership cost and annual variable cost of gestation facility. The following annual ownership costs can be easily changed: labor, feed, utilities, veterinary and health supplies, semen, loan payment and depreciation on breeding stock. The user can enter known values or have the computer calculate values.

After the total annual ownership and variable costs are calculated, the user can change the reproductive performance values (farrowing rate, litter size, litters/sow/year) to estimate their effect on cost of the gestation phase/pig weaned.