Will AVMA See the Light?
Sows Should Not Be Confined to Crates

Pregnant sows forced into uncomfortable positions by their crate "homes" at one of the nations' largest pig factories, a supplier to a major US fast food chain. Diane Halverson/AWI

It has been estimated that about 70% of the almost six million breeding sows in the US spend three-quarters of their adult lives confined in narrow, two foot by six and one-half foot gestation crates or stalls, and the other one-quarter in equally narrow farrowing crates, constructed to limit their mobility in the presence of their piglets.

As a consequence of their confinement, and despite being given preventative doses of antibiotics and laxatives in their feed, crate-housed sows live fewer years and are subject to more maladies, including osteoporosis, lameness, muscle deterioration, mastitis and constipation, than their counterparts on humane farms. Industry scientists have estimated yearly sow death rates on some of the largest factory farms, which use crates, at a stunning 20% of the farm's herd.

This is the compelling background against which the ethical appropriateness of housing breeding pigs in crates must be evaluated.

In 2002, the American Veterinary Medical Association (AVMA) went on record supporting the use of gestation crates. In response to the furor created over this untenable position, the AVMA decided to reassess its stance and appointed a task force to conduct a review of the current scientific literature with a view to recommending an appropriate position.

The November 1, 2003 issue of the Journal of the American Veterinary Medical Association (JAVMA), referring to "the heightened interest in the welfare aspects of housing for pregnant sows," recommended for its readers' edification a "scientific article comparing injuries sustained by pregnant sows in individual versus group housing" by Anil, et al. To AWI's chagrin the study featured in JAVMA was so poorly designed it ensured that crates appeared to be better for gestating pigs than housing them in groups.

The study, which supposedly compared group-housing to crate-housing, assessed the welfare of sows solely by tallying injury scores to quantify and compare pain in the two systems (sows single-housed in crates will not have the means or opportunity to injure each other). Additional parameters of welfare, such as bone loss, lameness, and incidence of mastitis, which also cause pain, should have been used; this would have been less obviously biased in favor of crate housing.

Even the authors acknowledged that the feeding system they chose for the group-housed sows, a single electronic sow feeder (ESF), had been cited for causing increased aggression and injuries (Van Putten, et al., 1990). In a 1988 article in the scientific journal Applied Animal Behaviour Science, Dr. Van Putten described ESF systems as examples of "farming beyond the ability of pigs to adapt," because they require pigs, who normally forage and eat together as a social group, to line up and take turns entering the feeder. "Obviously," said Van Putten, "the remarkable improvements in knowledge, obtained by applied ethological research, have not been taken into account in drafting concepts for computer controlled housing systems.... After all, it is an ethical point: either we choose to continue working against the nature of farm animals or, if we accept the introduction of a new era in pig farming, we welcome the opportunity to work with the animals by meeting their needs."

Anil and her cohorts listed options that might have reduced sow injuries in their group housing system such as providing a separate enclosure or solid walls for the ESF so sows outside the feeder could not see the sow inside; feeding a high fiber diet that might reduce appetite and aggression; and enriching the environment. However, they concluded that the "practicality and scientific value of these options are not yet known." This statement highlights another shortcoming of their research; they did not build on and extend the work of other scientists whose research has demonstrated the practicality and scientific value of those options.

For example, Professor Peter Brooks, University of Plymouth, has described scientific research undertaken to minimize competition and fighting among sows in ESF systems. He recommended the very options that Anil, et al. listed, but dismissed as unproved: providing protection around the system for the eating sow, making bulk materials such as corn and grass silage continuously available to the sows and enriching the environment with straw bedding to satisfy sows' hunger and permit a wide range of behavioral activities.

Dr. Ingvar Ekesbo has described the Swedish deep-bedded group housing systems (see "A Successful System for Housing Pregnant Sows in Groups," page 6), enriched with straw and equipped with individual feeding stalls that allow sows to eat at the same time. Contrary to the claim by Anil, et al. that individual feeding facilities are expensive for producers, these systems are cost-effective and provide good welfare. Deep straw beds save on labor costs for cleaning. They compost and provide warmth in winter. Individual feeding stalls provide an alternative lying area for sows, who like to get away from the straw beds when the weather is warm, and serve as a restraining area when the farmer needs to administer medical treatment or wants to close in the sows to clean the pens.

Anil and her colleagues contended that fighting is a permanent feature of dynamic groups, yet Swedish farmers learned ways to promote peaceful relationships in dynamic sow groups, such as housing new sows together where they form stable subgroups before farmers introduce them into established sow groups. In Anil's experiment sows remained in gestation crates for 10 days before they were introduced to the established group. Rather than entering as a stable subgroup, new sows entered the established sow group as separate individuals, increasing the likelihood of conflicts and injury.

Unfortunately, the authors of the JAVMA article do not appear to have had sufficient knowledge of scientific and practical advances in group sow housing to design a system that could provide a fair and unbiased comparison between individual and group housing. Instead, their study repeated what is known from earlier studies: sows housed in groups with a single ESF on fully slatted floors without environmental enrichment have high injury rates.

It cannot be concluded from Anil, et al. that it would be unwise or premature to support a resolution banning gestation crates. Effective alternatives to crate housing of pregnant sows exist and render crate housing of sows obsolete, as well as morally objectionable.

AWI urges the AVMA membership and task force not to accept studies on their face value but to scrutinize carefully the authors' assumptions, methodology, and command of the scientific literature. This is particularly critical on such a politically charged issue as sow housing, which calls into question a clearly cruel method of housing with scientifically dubious origins that nevertheless has been embraced and fiercely defended by the pork industry.