Russell and Burch's Three Rs (replacement, reduction and refinement) remain the cornerstone for principles guiding humane experimental research. However, the concept of refinement has evolved considerably since its first inception and there have been numerous interpretations, some of which are regressive from the original definition. In this paper we examine the interpretations of refinement, and propose a harmonised progressive definition that is in line with changes in animal ethics and animal welfare science. Our definition should be applied to all aspects of refinement: those related to housing, husbandry and care, techniques used in scientific procedures, periprocedural care, health and welfare monitoring, and experimental design. We argue not only that the concept should include the avoidance or minimisation of adverse effects experienced at any time during the life of an animal destined for use in a laboratory, but also that it should be applied to the founder animals. Furthermore, we take a proactive stance and argue that refinement should include enhancing well-being through environmental enrichment. The acceptance and application of this new definition by legislative authorities and in guidelines would represent a significant step forward for animal welfare.

Although genetic selection for certain characteristics has compromised the well-being of domestic animals, selective breeding could and probably should be used to improve welfare. Three of the major behavioural problems facing the poultry industry are fear, feather pecking and social stress. However, these and many other behavioural traits respond readily to genetic selection. The present paper reports the results of selective breeding studies in which underlying fearfulness, sociality, feather pecking, adrenocortical responsiveness, and growth rate were manipulated in Japanese quail, Coturnix japonica, or in chickens, Gallus gallus domesticus. The apparent benefits of selection for appropriate levels of these characteristics are discussed, as are the ethical issues involved. The ultimate aims of the selection programmes and the studies described here are to improve the birds ‘ welfare and productivity by increasing their ability to interact successfully with their physical environment, with one another, and with human beings.

Animal welfare assessment at group level is a scientific discipline that is rapidly developing. The interest in welfare assessment systems is based on an ethical concern for the welfare of farm animals. The scientific community plays an important role in delivering an appropriate repeatable, valid and feasible framework for these assessments. Consideration of the potential applications of these techniques is important for deciding upon the requirements of specific assessment systems. This paper provides a brief overview of the different types of applications, which can be categorised broadly into research, legislative requirements (non-voluntary), certification systems (voluntary) and advisory/management tools. These applications may have various goals: quantification of welfare, provision of welfare assurance or welfare management. Assessment systems vary in many characteristics, such as whether they are animal- or resource-based, and whether they are based on single or integrated scores. Different applications will require different elements of these features.

Obtaining reliable welfare outcome measures from commercial farms can be challenging. We developed a training programme to train observers to score injuries of the tarsal joint, carpal joint and neck on dairy cows as part of an on-farm study. Twelve trainees were trained using protocols and photographs in a classroom session and on-farm visits. Continued repeatability checking was carried out during a refresher and mid-way assessment. Two trainers were used as the reference standard to which all trainees were compared. The study demonstrated that methods of scoring tarsal joint, carpal joint and neck injury can be learned by trainees from different backgrounds and high repeatability can be achieved and maintained at a very large regional or national level. Successful learning of injury scoring is dependent on protocols with strong definitions and photographs as well as repetitive training sessions. Additionally, continued repeatability checks are essential to ensure the reference standard continues to be met. This training programme can be used as a model to successfully train on-farm assessors.

Interview questionnaires were administered to the general public in central Scotland and northern England during summer 2007 to investigate consumer awareness of UK dairy production methods, welfare issues and recognition of ‘quality assurance’ product logos. Fifty percent of respondents gave UK dairy animal welfare a positive rating. Recognition of individual quality assurance logos was poor and 75% of respondents stated that they did not intentionally seek to buy products with any of the logos. Respondents’ perceptions of good dairy welfare included: appropriate feeding, good stockmanship, plenty of space, freedom to roam/free range and environmental cleanliness. Half of respondents felt they were poorly informed about food production and the majority of respondents (68%) would like more information on food production. Respondents believed that information on animal welfare provided by veterinarians and farmers would be reliable. Most respondents (93%) said they would pay more for good dairy welfare. The findings show that the general public are interested in animal welfare but could be better informed on dairy animal production and welfare. Veterinarians and farmers may have a potentially important role in providing this information with increasing demand for higher welfare provenance products potentially helping to improve animal welfare.

As animal industry and animal advocacy groups debate how farm animals should be treated, little research has focused on the attitudes of consumers in the United States. This study utilises results of a representative telephone survey to measure consumer attitudes towards farm animal welfare, and investigates how these attitudes vary across individuals. The survey finds that consumers desire high standards of animal care, even if it raises food prices and involves government regulation. Support is particularly strong from females, Democrats, and residents of the Northeastern United States. To provide high standards of animal care, consumers as a whole perceive allowing animals to exhibit natural behaviours and exercise outdoors to be more important than protection from other animals, shelter, socialisation, and comfortable bedding. Consumers vary in their perceptions though, and are divided into three classes: Naturalists, Price Seekers, and Basic Welfarists. Naturalists place great importance on allowing animals to exhibit natural behaviours and exercise outdoors, and comprise 46% of the sample. Price Seekers, comprising 14% of the sample, are primarily concerned with low prices. Basic Welfarists make up 40% of the respondents, and value animal welfare but perceive it can be achieved by simply providing food, water, and treatment for injury and disease. This last group perceives amenities, such as access to outdoors and ability to exhibit natural behaviours, unimportant for the well-being of farm animals.

Humans interact with fish in a number of ways and the question of whether fish have the capacity to perceive pain and to suffer has recently attracted considerable attention in both scientific and public fora. Only very recently have neuroanatomical studies revealed that teleost fish possess similar pain-processing receptors to higher vertebrates. Research has also shown that fish neurophysiology and behaviour are altered in response to noxious stimulation. In the light of this evidence, and in combination with work illustrating the cognitive capacities of fish, it seems appropriate to respond to a recently published critique (Rose 2002) in which it is argued that it is not possible for fish to experience fear or pain and that, therefore, they cannot suffer. Whilst we agree with the author that fish are unlikely to perceive pain in the same way that humans do, we believe that currently available evidence indicates that fish have the capacity for pain perception and suffering. As such, it would seem timely to reflect on the implications of fish pain and suffering, and to consider what steps can be taken to ensure the welfare of the fish that we exploit.

To use behaviours as indicators of stress it is important to understand their underlying causation. For a prey animal in the wild, such as a sheep, behavioural responses have evolved to evade detection and capture by predators. The behavioural responses of the wild ancestors of domestic sheep to the threat of predation are characterised predominantly by vigilance, flocking, flight to cover and behavioural inhibition once refuge has been reached. Some limited defensive behaviours are seen, mainly in females with young against small predators. Vigilance and flight distance are affected by the animal's assessment of risk and are influenced by the environment, social group size, age, sex and reproductive condition, as well as by previous experience with potential predators. Under conditions of stress, domestic sheep show similar behavioural reactions to wild sheep, although the threshold at which they are elicited may be elevated. This is particularly evident when comparing less selected hill breeds with more highly selected lowland breeds, and suggests that a continuum of responsiveness exists between wild and feral sheep, through hill breeds to the lowland sheep breeds. However, this may be confounded by the previous experience of the breeds, particularly their familiarity with humans. Behavioural and neurobiological evidence suggests that, although the behavioural response to predators (vigilance, flight) is innate, the stimuli that elicit this behavioural pattern may have a learned component. Since vigilance and flight distances are affected by the animal's perception of threat, they may be useful indices of stress in sheep and, as graded responses, give some indication of the level of threat experienced by the sheep. Thus they may indicate the amount of fear or distress experienced by the sheep and hence have the potential to be used in the assessment of welfare states.

Since reintroduction programmes involve moving animals from captive or wild environments and releasing them into novel environments, there are sure to be a number of challenges to the welfare of the individuals involved. Behavioural theory can help us develop reintroductions that are better for both the welfare of the individual and the conservation of populations. In addition to modifying captive environments to prepare animals for release to the wild, it is possible to modify the animals’ experience in the post-release environment. For releases to be more successful, they need to better accommodate the ecological and psychological needs of individuals. A better understanding of sensory ecology — how animals acquire and respond to information in their environment — is needed to develop new, more successful management strategies for reintroductions. Sensory ecology integrates ecological and psychological processes, calling for better synergy among researchers with divergent backgrounds in conservation and animal welfare science. This integrative approach leads to new topics of investigation in reintroduction biology, including more careful consideration of post-release stress and the role of social support. Reintroductions are essentially exercises in ‘forced’ dispersal; thus, an especially promising avenue of research is the role of proximate mechanisms governing dispersal and habitat selection decisions. Reintroduction biologists have much to gain from the study of mechanism because mechanisms, unlike function or adaptive value, can be manipulated to enhance conservation and welfare goals.

To be able to assess animal welfare the researcher must presuppose a number of background assumptions that cannot be tested by means of ordinary empirical data collection. In order to substantiate these assumptions two sorts of inferences have to be relied upon, which the authors designate by the terms ‘analogies’ and ‘homologies’. Analogies are evaluative, philosophical reflections by means of which it is made clear what provisions or states constitute the welfare of humans and other animals. By means of analogies it may, for example be argued that animal welfare consists of subjective experiences such as pain, boredom, pleasure and expectation. Also by means of analogies the relative ‘weight’ of these states can be decided. Homologies are part of theoretical science. They serve to clarify how the relevant experiences are linked to measurable anatomical, physiological and behavioural parameters.

An account is given of the steps which have to be taken to give a full answer to a question concerning the welfare of animals. In the account only farm animals are mentioned, but the same steps, of course, also have to be taken to answer questions concerning the welfare of other kinds of animals be they companion, laboratory, zoo or wild. Eight steps are described, and it is argued that both analogies and homologies are needed at very fundamental levels. Therefore, if animal welfare science is to provide relevant, rational and reliable answers to questions concerning animal welfare, it must be an interdisciplinary inquiry involving philosophical reflections and theoretical biology.

Expansion of the meat inspection process to incorporate animal-based welfare measurements could contribute towards significant improvements in pig (Sus scrofa domesticus) welfare and farm profitability. This study aimed to determine the prevalence of different welfare-related lesions on the carcase and their relationship with carcase condemnations (CC) and carcase weight (CW). The financial implications of losses associated with CC and CW reductions related to the welfare lesions were also estimated. Data on tail lesions, loin bruising and bursitis, CW and condemnation/trimming outcome (and associated weights) were collected for 3,537 slaughter pigs (mean [± SEM] carcase weight: 79.2 [± 8.82] kg). Overall, 72.5% of pigs had detectable tail lesions, whilst 16.0 and 44.0% were affected by severe loin bruising and hind limb bursitis, respectively. There were 2.5% of study carcases condemned and a further 3.3% were trimmed. The primary cause of CC was abscessation. While tail lesion severity did not increase the risk of abscessation, it was significantly associated with CC. Male pigs had a higher risk of tail lesions and of CC. The financial loss to producers associated with CC and trimmings was estimated at €1.10 per study pig. CW was reduced by up to 12 kg in cases of severe tail lesions. However, even mild lesions were associated with a significant reduction in CW of 1.2 kg. The value of the loss in potential CW associated with tail lesions was €0.59 per study pig. Combined with losses attributable to CC and trimmings this represented a loss of 43% of the profit margin per pig, at the time of the study, attributable to tail biting. These findings illustrate the magnitude of the impact of tail biting on pig welfare and on profitability of the pig industry. They also emphasise the potential contribution that the inclusion of welfare parameters at meat inspection could make to pig producers in informing herd health and welfare management plans.

Our view of which individuals should be the subjects of our moral actions is expanding to include more people and more species. Animal welfare is the subject of rapidly increasing concern in most countries in the world, and this concern is resulting in changes in the ways in which animal users keep and treat animals. Ethical decisions about whether the killing of an animal is justifiable should be considered separately from those about how poor welfare can be and still be acceptable. The term ‘euthanasia’ should be restricted to killing an animal for its own benefit. Quality of life (QoL) in humans is generally taken to include: physical condition and any impairment of this resulting from injury or disease; capacity to function; perception of functioning; and satisfaction with functioning in relation to what is believed possible. If the welfare of an individual is its state as regards its attempts to cope with its environment, then welfare is essentially the same as QoL. Both include the state of the individual's coping systems, including those responding to pathology, various behavioural and physiological responses, and cognitive processes associated with suffering or pleasure. Hence, both welfare and QoL include health and the extent of positive and negative feelings. Many papers referring to animal welfare include objective quantification whilst few papers referring to QoL do so. Some human studies assess QoL by the less objective method of questions asked of subjects. Neither QoL nor welfare should be assessed using solely subjective measures. Assessment of welfare must take account of the wide variety of coping systems and coping strategies used. A range of measures including those of behaviour, physiology, brain function, immune system function, and damage is needed. The ease or difficulty of coping should be interpreted within the framework of the abilities of the animal. Animals with more sophisticated cognitive functioning may have the best abilities to cope with problems. The scheme presented here for assessing welfare over time facilitates ethical decisions regarding whether welfare is good or whether it is unacceptably poor.

The vocalisations of animals are results of particular emotional states. For example, the stress screams of pigs may be indicators of disturbed welfare. Our objective was to develop a system to monitor and record levels of stress calls in pigs, which could be employed in environments of breeding, transportation and slaughter. Using a combination of sound analysis by linear prediction coding and artificial neural networks, it was possible to detect the stress vocalisations of pigs in noisy pig units with few recognition errors (<5%). The system (STREMODO: stress monitor and documentation unit) running on PCs is insensitive to environmental noise, human speech and pig vocalisations other than screams. As a stand-alone device it can be routinely used for the objective, non-invasive measurement of acute stress in various farming environments. The system delivers reliable, reproducible registrations of stress vocalisations. Its detection quality in commercial systems was found to correlate well with that of human experts. STREMODO is particularly well-suited for comparisons of housing and management regimes. Since the system can be trained to recognise various animal vocalisations, its use with other species is also well within its scope.

The scientific study of animal welfare has generated a welter of complex, equivocal and often contradictory results. Consequently, there is little agreement about how impairment of welfare should be measured. While some solutions to this have been suggested, these have usually relied on more sophisticated versions of, or more control over, existing measures. However, we argue that the difficulties arise because of questionable assumptions in the definition and measurement of welfare, in particular the measurement of suffering and the assumed importance of individual well-being. We contend that welfare can be interpreted only in terms of what natural selection has designed an organism to do and how circumstances impinge on its functional design. Organisms are designed for self-expenditure and the relative importance of self-preservation and survival, and the concomitant investment of time and resources in different activities, varies with life history strategy. The traditional notions of coping and stress are anthropomorphisms based on homeostatic mechanisms of self-preservation in a long-lived species. Suffering-like states are viewed as generalized subjective states that are geared to avoiding deleterious circumstances with which the organism does not have specific adaptive mechanisms to deal. Attempts to measure suffering-like states directly are likely to remain inconclusive, at least for the foreseeable future, because such states are private and subjective, may take many forms fundamentally different from our own and are likely to depend on the operation of phenotype-limited priorities and decision rules. However, measuring the impact of circumstances on functional design via the organism ‘s decision rules provides a practicable means of giving benefit of the doubt by indicating when suffering, or an analogous subjective state, is likely.

The welfare of transported pigs can be compromised both by physical and psychological stresses. The animals’ responses can be assessed using records of mortality and trauma, physiological and behavioural observations and, to some degree, by measurements of meat quality since this can reflect the animals' physiological state at death. These assessments may, therefore, be used as measures of animal welfare. During transport pigs show weight loss, increased circulating concentrations of catecholamines, Cortisol and creatine phosphokinase (CPK), and an increase in heart rate and packed cell volume; sometimes there is evidence of dehydration. Increased levels of dark, firm, dry (DFD)meat after long transport reflect muscle glycogen depletion and possibly indicate some element of fatigue. There is experimental evidence that transport is aversive to pigs, which may be partially due to the fact that they become travel sick. Mortality in transport has ranged from < 0.1 to > 1.0 per cent in different European countries. Mortality is higher in more stress-susceptible breeds and at higher ambient temperatures. It is increased in pigs fed within 4h of transport, at higher stocking densities and after longer journeys at ambient temperatures greater than 10°C. Pigs may be fasted long enough before slaughter to prejudice their welfare through hunger. Long fasts may also reduce muscle glycogen levels and cause fatigue. Fighting between unfamiliar animals which have been mixed during the marketing procedure is also stressful, however, longer transport may actually reduce this problem by allowing animals to get used to one another under conditions in which it is difficult to fight.

Possible welfare benefits of qualitative rather than quantitative food restriction were investigated with growing female broiler breeder chickens (Ross 1). In Experiment 1, body-weight gains from 2 to 6 weeks of age were compared among different diet dilution, appetite suppression and low protein treatments, with free access to food at all times, to identify qualitative treatments causing weight gains similar to that recommended in the Ross 1 Parent Stock Management Manual. Based on these results, four diet dilution (400g kg−1 unmolassed sugar-beet pulp, 300 and 600g kg−1 oat hulls, 500g kg−1 softwood sawdust) and one appetite suppression (50g kg−1 calcium propionate) treatments were compared with two quantitative restriction (the recommended daily ration and twice that amount) and one ad libitum control treatments, from 2 to 10 weeks of age, in Experiment 2. As well as growth, food intake, excreta production and digestibility, measurements were also made of behaviour and blood indices of stress. Several conclusions were drawn. Different methods of qualitative food restriction can be used to control growth rate within desired limits. Problems with these methods include reduced uniformity in weight gain, increased excreta production and/or increased cost. Although they appear to suppress abnormal oral behaviours, they do not alter the increased general activity which is correlated with suppression of growth rate, and which may more accurately reflect associated hunger. Suppression of abnormal oral behaviours may only rarely correspond with reduction in blood indices of stress, and so cannot be taken to indicate improved welfare. Some of these methods can add to physiological stress. Finally, there was insufficient evidence of improved welfare, based on both behavioural and physiological criteria, to justify advocating the suitability of any of these methods for commercial use.

Individual differences in behaviour may be examined at two levels. First, individuals may differ in terms of frequencies, durations and/or patterning of particular measures of their behaviour. Second, individuals may differ in their temperament, ie in the way they react to environmental change and challenge. Individual differences in temperament are particularly relevant to animal welfare studies, for the welfare of an individual largely depends on whether it can cope with environmental challenge. Whereas the study of individual differences in behaviour at the first level may be achieved by using standard behavioural methods, the study of individual differences in temperament requires the use of more unusual methods, namely observers’ ratings and behavioural tests. Observers’ ratings provide information on subtle aspects of an individual's behaviour that could otherwise be overlooked. Behavioural tests facilitate comparisons between individuals in a more standardized way. It is suggested that both systems should be used together.

Taking individual differences into account when designing experiments may help reduce variability in studies on welfare issues and understanding the causes of individual differences in temperament may allow us to reduce the incidence of some welfare problems.

Refinement techniques are being increasingly employed in all fields of animal research to try to ensure that the highest standards of welfare are upheld. This review concerns one of the main emerging techniques for the assessment of welfare itself, namely the non-invasive measurement of glucocorticoids (GCs) as indicators of stress. The paper is divided into three sections. The first discusses the relationship between GCs and stress. The second section considers whether factors other than stress are linked to rises in GCs, eg exercise, oestrus cycle and diet. The final part examines the reliability of the non-invasive techniques that measure GCs from samples of saliva and faeces. Although it is important to take into account some caveats associated with the methodologies employed, it is concluded, nevertheless, that these techniques can give accurate and reliable information regarding the welfare status of an individual or group of animals without the procedures themselves causing any kind of distress to the subjects.

The prevalence and risks for abnormal gait in finishing pigs (Sus scrofa), gilts and pregnant sows from a representative cross-section of indoor and outdoor herds in the United Kingdom were investigated. The prevalence of abnormal gait in finishing pigs, maiden gilts, pregnant gilts and pregnant sows from 88 herds was 19.7, 11.8, 14.4 and 16.9%, respectively. In a multivariable analysis of 98 herds, there was an increased risk of abnormal gait in pregnant sows housed on slatted floors compared with pregnant sows housed on solid concrete floors with straw bedding or sows housed outdoors on soil. The lowest prevalence of abnormal gait in finishing pigs occurred in pigs housed outdoors (3.4 vs 19.7% in indoor-housed finishing pigs) however, the difference was not significant because only three farms in the study housed finishing pigs outdoors. In indoor-housed finishing pigs, there was an increased risk of abnormal gait in pigs housed on solid concrete floors with sparse bedding, partly-slatted floors or fully-slatted floors compared with those housed on solid concrete floors with deep bedding in all areas. However, there were no significant associations between floor type and abnormal gait in gilts. There was an increased risk of abnormal gait associated with increasing callus, bursitis and capped hock score on the limbs of finishing pigs. This might have occurred because limb lesions cause discomfort or because lame pigs spend more time lying and this increases the risk of limb lesions developing.

The concept of quality of life in animals is closely associated with the concepts of animal sentience and animal welfare. It reflects a positive approach that inquires what animals like or prefer doing. The assessment of farm animal welfare requires a good understanding of the animals' affective experience, including their emotions. However, affective experience in animals is difficult to measure because of the absence of verbal communication. Recent studies in the field of cognitive psychology have shown that affective experience can be investigated without using verbal communication by examination of the interactions between emotions and cognition. On the one hand, appraisal theories provide a conceptual framework which suggests that emotions in humans are triggered by a cognitive process whereby the situation is evaluated on a limited number of elementary criteria such as familiarity and predictability. We have applied these appraisal theories to develop an experimental approach for studying the elementary criteria used by farm animals to evaluate their environment and the combinations of those criteria that trigger emotions. On the other hand, an increasing body of research, first in humans and then in other animals, suggests that emotions also influence cognitive processes by modifying attention, memory and judgement in a short- or long-term manner. Cognitive processes could therefore be manipulated and measured to provide new insights into how not only emotions but also more persistent affective states can be assessed in animals. Further work based on these cognitive approaches will offer new paradigms for improving our understanding of animal welfare, thus contributing to ‘a life of high quality’ in animals.