There have been important developments in the measurement of farm animal welfare in recent years. Measuring animal welfare is one thing, implementing a farm animal welfare assessment system another. The implementation of such a system occurs in an environment that is influenced by economic, political, technological and socio-cultural factors which interact with each other. This creates enormous complexity, generates a huge number of different potential ‘futures’, and makes the eventual impact that the system will have on the welfare of farm animals uncertain. This article draws upon strategic management literature to apply scenario analysis as a technique to help understand the variance of the uncertainty associated with the implementation of an animal welfare assessment scheme. Specifically, it develops two extreme scenarios based on a theoretical European-wide implementation: one scenario in which all uncertain factors influence the implementation of the assessment system in a negative way, and one scenario in which all these factors have positive impacts. These scenarios provide insight into the variance of possible futures in which the system may have to function. Although consumers are an important stakeholder group, their role in creating uncertainty for the system may be overestimated; it is apparent that the roles of companies, brands and certification organisations deserve significant attention, as well as any relevant institutional structure.

There is continuing debate about the space needs and requirements of broiler chickens, The aims of this study were to measure the amount of floor area a six-week-old broiler occupies for different behaviours and to use the obtained results in two models to estimate the number of birds that can be kept per m2 in large flocks simulating different levels of behavioural synchronisation. Photographs were taken of overhead projections of broilers (2.468 kg on average) kept in floor pens of 1 m2 with either eight (low density) or 16 birds (high density) per pen. Individual body space was measured from these photographs for seven behaviours. Posture and density affected body space of the behaviours idle, drinking, and ground pecking. The first model, computing space needed per bird performing a behaviour in relation to flock size, showed that 15.3-15.7 birds m−2 (37.8-38.7 kg m−2) can be housed maximally, based on low density measurements and 18.5-19.4 birds m−2 (45.7-47.9 kg m−2) based on high density measurements. The second model, computing stocking density based on synchronisation of behaviour and body space, showed that 13.7-15.9 birds m−2 (33.8-39.2 kg m−2) can be housed maximally based on low density measurements and 15.4-18.6 birds m−2 (38.0-45.9 kg m−2) based on high density measurements. Results based on high density measurements implied that birds are compressed. Given the restrictions of a limited number of behaviours and no inclusion of movement and social interactions in the models of this study, stocking density in large flocks should not exceed 16 birds m−2 (39.4 kg) because that would lead to compression of birds which will suppress opportunities for behavioural expression and therefore impair welfare.

This study assessed the effects of increasing dietary fibre levels in concentrate rations and providing access to straw in racks on the welfare of pregnant sows housed in small static groups. In a 2 × 2 factorial design experiment, 128 Large White × Landrace pregnant sows were offered one of two diets: (i) High fibre diet with 9% crude fibre, or (ii) Control diet with 4.5% CF, and one of two levels of access to a foraging substrate: (i) access to straw in racks or (ii) no straw. The study was replicated eight times using groups of four sows, and treatment periods lasted four weeks. Sows were housed in pens with voluntary cubicles and a slatted exercise area and were offered a wet diet twice a day. Back-fat levels were measured before sows were mixed into groups at 28 days post partum, and four weeks later. Aggressive interactions were recorded on the day of mixing, and injury scores were recorded one week post mixing. Scan sampling was used to collect data on general activity, posture and location of the sows, and on sham-chewing and bar-biting behaviours across the treatment period. In addition, detailed focal observations were carried out on all sows across the treatment period. Straw usage was also recorded. There were no treatment effects on changes in back-fat levels over the treatment period. Treatments had no effect on post-mixing aggression or on injury scores. However, focal observations showed that sows with access to straw were involved in fewer bouts of head-thrusting over the treatment period. Control diet sows spent more time inactive than sows on the high fibre diet, however high fibre diet sows spent more time lying with eyes closed than sows on the control diet. Sows on the high fibre diet with access to straw showed less sham-chewing and bar-biting behaviour than sows in other treatments. These results show that although a diet containing 9% crude fibre promoted resting behaviour, it was necessary to combine it with access to straw to reduce stereotypic behaviour of sows in small static groups.

Tail biting is a major problem in modern pig (Sus scrofa) production and results in a reduction of animal welfare and productive performance. Biting behaviour has been shown to be decreased by the use of enrichment objects. In this study, 108 pigs housed in a room with 12 pens were observed and a sequence of seven different enrichment materials was tested. Gilts and barrows were housed together and received a new enrichment object each week starting from three different points in the fattening period, ie 20, 40 or 70 kg bodyweight. Toy-contact and biting-penmate behaviour were observed during one hour at day of introduction and five days later. A continuous sequence of seven enrichment objects reduced biting-penmate behaviour and the number of wounds compared to providing only a single toy (chain). This study also confirmed that not every object was feasible as an enrichment object for growing pigs. Generally, the highest toy contact was observed together with the highest biting-penmate behaviour. Most toy-contact and biting-penmate behaviour was observed between 20 and 40 kg bodyweight and was decreasing over age. Providing a sequence of toys for the first time induced toy-contact behaviour while reducing bitingpenmate behaviour but decreased after applying the same sequence for the second or third time. The presence of a single chain or a sequence of different toys had no effect on growth and feed conversion. The ideal sequence should maintain toy-contact behaviour without competition in order to avoid biting-penmate behaviour and reduced animal welfare.

The use of laboratory animals in scientific procedures in the United Kingdom is highly regulated and, in the case of non-human primates (NHPs), it is intended that they are only used in scientific studies when no other animal species would provide a suitable model and when no alternative methods are available. Additionally, there must be a high likelihood that the research will lead to important medical advances. The number of NHPs used in scientific procedures each year in the UK are relatively small (in 2010 less than 0.1% of the 3.6 million animals used) however, it is thought that the welfare cost to NHPs in a laboratory environment may be greater than for other animals because of their sentience and social nature. In 2006, the report of a working group chaired by Sir David Weatherall, recommended that the major funding bodies of non-human primate research in the UK should systematically review the outcome of all of their research using NHPs over a ten-year period.