The inclusion of Russell and Burch's Three Rs (replacement, reduction and refinement) in guidelines, codes of practice and law reflects their current position as the guiding principles of ethical assessment of research involving animals. This article explores some activities within the contemporary livestock industry that constitute the experimental use of animals on a local and global scale. The elucidation of correlated responses during trait selection in genetic improvement programs provides one example of experiments occurring within the commercial livestock industry. This experimentation is largely conducted without scrutiny of its conformity to the Three Rs. Experimentation to improve the management of the livestock industry is consistent with the principle of refinement, and experimentation to increase productivity per unit of livestock is consistent with the principle of reduction; however, experimentation to increase total livestock production conflicts with the principle of replacement. Some approaches regarding the appraisal of the ethics of research involving animals, which could avoid arbitrary boundaries associated with the location or purpose of experimentation, are considered together with the relationship between experimentation and other anthropogenic impacts on animals.

Various animal welfare assurance programs are being used to encourage or require the adoption of animal welfare standards in food production, and to assure the public that such standards are followed. The programs involve five main formats. Non-mandatory codes/guidelines are relatively easy to institute and appear well-supported by the industry, but provide only minimal assurance to the public unless measures are taken to ensure compliance. Programs based on government regulations and inter-governmental agreements are more challenging to institute; they are likely to generate less industry acceptance, but may provide more public confidence if enforcement is adequate. Product differentiation programs, and retailer policies requiring products to meet certain standards, serve a range of functions; these may generate public confidence but only for products covered. The various programs include several types of requirements. Requirements that are designed to maintain animal health and functioning have a widely accepted scientific basis, are often easy to incorporate into existing production systems, and often provide economic benefits, but do not fully address public concerns over animal welfare in some cultures. Requirements that address pain, distress and other affective states, and those that accommodate certain natural behaviour, have a growing but less traditional scientific rationale and appear likely to generate public confidence; however, they sometimes require significant changes to existing practices. Requirements for more natural surroundings (outdoor, free-range) seem to generate public confidence, but appear most likely to increase costs, least likely to be supported by the existing industry, and may involve trade-offs with productivity and with other aspects of animal welfare. The various formats and requirements provide a range of policy options for addressing animal welfare concerns in different cultural, industry and market contexts.

The welfare of non-human primates used in scientific research must be safeguarded to promote scientific validity and for ethical reasons. Welfare can be improved by the refinement of practice, particularly if these refinements are applied to every aspect of the life of an animal used in the laboratory, from birth to death with the aim of both minimising harm and maximising well-being. Many refinement methods have been described in nationally and internationally accepted guidelines on laboratory practice, but awareness of these guidelines is not universal. In Part I of this review, we examine the influence of humans on non-human primates and summarise and evaluate methods of refinement that are or could be used to reduce suffering and improve welfare. In particular we focus on staff selection, education and training, human–animal bonds, staff communication, and training primates. In Parts II and III, refinements of housing, husbandry and experimental procedures are reviewed.

Some supporters of religious slaughter methods claim that efficiency of bleed-out is adversely affected by stunning. Our previous study carried out in sheep at an abattoir comparing the Muslim method of slaughter without stunning with pre-slaughter stunning using a captive bolt or by electrical methods concluded that bleed-out is not adversely affected by stunning, nor improved by a neck cut without stunning. In this paper, a similar study carried out in cattle is reported. In this study, captive bolt stunning followed by neck cutting was compared with the Muslim slaughter method without stunning. The total blood loss, percentage blood loss expressed as a percentage of live weight and percentage loss of estimated total blood were calculated and compared between each group. In addition, the time periods taken to reach 25, 50, 75 and 90% of total blood loss were also calculated. There was no significant difference between the two stunning groups for any of these blood loss variables. The results, subjected to statistical analyses, also showed no apparent difference in the PCV levels and meat quality parameters between treatments. These results confirm the findings with sheep and show that the bleed-out is not adversely affected by captive bolt stunning, nor improved by a neck cut without stunning in cattle. It is anticipated that these findings may help promote the use of stunning methods during Halal slaughter.

The effectiveness of electrical water bath stunning of 172 individual broilers for 1 s with a constant root mean square (RMS) current of 100, 150 or 200 mA, delivered using a variable voltage/constant current stunner with 200, 400, 600, 800, 1000, 1200 or 1400 Hz sine wave alternating current (AC), followed by slaughter using a unilateral or ventral neck-cutting procedure, was evaluated. A binary logistic regression of broilers showing epileptiform activity or not following stunning showed that both the electrical frequency (Hz) and RMS current (mA) had a significant effect on the probability of the electroencephalogram (EEG) manifestation. The univariate analysis of variance showed that the time to the onset of less than 10% of the pre-stun relative power contents was significantly affected only by the interaction between electrical frequency and slaughter method. A similar analysis of variance of the time to reach less than 10% of the pre-stun total power content showed slaughter method, RMS current, the slaughter method/frequency interaction and the RMS current/frequency interaction to be either significant or approaching significance. Based on these results it is recommended that effective water bath stunning of broilers with a minimum constant current of 100, 150 and 200 mA could be achieved with electrical frequencies of up to 200, 600 and 800 Hz, respectively. In addition, It is likely that electrical frequencies of above 800 Hz would have required a minimum current of greater than 200 mA to induce epileptiform activity in the EEGs of broilers.

In order to safeguard the welfare of laboratory-housed non-human primates, refinement techniques should be applied to every aspect of the life of animals used in the laboratory, from birth to death, with the aim of both minimising harm and maximising well-being. In this second part of a three-part review we summarise published information on housing and husbandry practices, and describe ways to minimise contingent inhumanity associated with the use of primates in laboratories and their breeding and supply (where inhumanity is defined as the infliction of distress). We also discuss methods by which the welfare of these animals can be maximised on a day-to-day basis. The principles of enrichment, aspects of the housing environment, social and physical enrichment and acquisition are discussed. Refinement of the influence of humans and experimental procedures are discussed in Parts I and III of this review, respectively.

Three housing systems for fattening rabbits were compared using a stocking density of 15 rabbits m–2. The rabbits were housed in large pens (3.67 m2, 50 rabbits), small pens (0.66 m2, 10 rabbits) or in conventional standard cages (0.39 m2, 6 rabbits) from 31 to 72 days of age. Rabbits housed in each small pen or in each cage belonged to the same litter, and the 50 rabbits housed in each large pen were from six or seven litters. At the end of fattening, when rabbits were 72 days old, there was no significant difference in the weight of rabbits from the three different housing systems, even though the rabbits from small pens were slightly heavier in weight compared with rabbits from large pens. The best feed conversion ratio was found in rabbits from cages, but was only significantly different from rabbits housed in small pens. No significant differences were found in the main activities: nutrition, social behaviour, resting, and standing; however, the frequency of runs, hops and consecutive hops was significantly higher in rabbits from large pens compared with rabbits from the two other housing systems. These results confirm that the total surface area available for animals is the most important factor for such locomotory activities. This study did not reveal any significant difference in aggressive behaviours between rabbits from different litters housed in large pens and between rabbits from the same litter housed in small pens or cages.

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 effectiveness of electrical water bath stunning of broilers (n = 96) for 1 s with a constant average current of 100, 150 or 200 mA delivered using a variable voltage/constant current stunner with 200, 800 or 1400 Hz pulsed direct current (DC), with a mark:space ratio of 1:1, followed by slaughter using a unilateral or ventral neck cutting procedure, was evaluated. The results of a binary logistic model showed that both the electrical frequency and average current had significant effects on the probability of inducing epileptiform electroencephalogram (EEG) and therefore, of effective stunning. The results of univariate analysis showed that only slaughter method had highly significant effects on the power contents in the 13–30 Hz and 2–30 Hz EEG frequency bands. Based on these results, it is recommended that a minimum of 200 mA average (400 mA peak) current per bird should be delivered using 200 Hz pulsed DC, with a mark:space ratio of 1:1, to achieve effective water bath stunning in 80% of broilers. Frequencies of above 200 Hz pulsed DC would presumably require average currents of greater than 200 mA. Electrical water bath stunning of broilers with 200 mA average current of 200 Hz resulted in cardiac arrest in six out of eight broilers that showed epileptiform activity. Two other broilers that had cardiac arrest showed no epileptiform EEGs indicative of effective stunning. Owing to the prevalence of cardiac arrest in conscious broilers, the use of pulsed DC for water bath stunning of broilers could be questioned on ethical and bird welfare grounds.

Standard techniques for assessing plumage damage to hens from feather pecking typically require capture and handling. Handling of individual birds for plumage assessment is relatively easy in experimental studies; however, close inspection of individual birds in commercial flocks is less feasible because catching birds is difficult, may compromise bird welfare and affect egg production. The aim of this study was to assess a non-intrusive method for scoring plumage damage in a commercial free-ranging flock of laying hens. Plumage damage was scored within a 2 m distance of the birds, without capture or handling, using a 5-point scale for 5 body regions. The feather scores, recorded at a distance, by two independent scorers were compared (distance scores), and were then compared with feather scores recorded by a scorer who caught and handled the birds to examine the plumage damage closely (capture scores). There was a significant and positive correlation between the distance scores and the capture scores, and the mean correlation coefficient for all plumage score traits was 0.89. There was also a significant and positive correlation between scorers, and the mean correlation coefficient for all plumage score traits was 0.84. The standard deviation of the residual mean difference between scorers and between methods was less than 1 point for individual body regions and less than 1.5 points for the total body score. Large variation in feather damage within a flock and small sample size increased the standard error of the mean total feather score. When feather damage variation within flocks is low (ie little observed feather damage), the current industry standard of scoring a sample of 100 birds is likely to provide a reliable estimate of flock feather damage; however, when there is large variation within birds of a flock (ie considerable observed feather damage) ≥200 birds should be inspected to accurately monitor changes in plumage condition. The non-intrusive method of feather scoring described in this paper may be useful for commercial-scale feather pecking studies or for farmers who need to assess the plumage damage of their flocks reliably, quickly and with minimal disturbance or stress to the birds.

There is an ethical and scientific need to minimise the harm experienced by animals used in scientific procedures and to maximise their well-being. Welfare can be improved by the refinement of practice, particularly if these refinements are applied to every aspect of the life of an animal used in the laboratory, from birth to death. Primates are considered likely to have a greater capacity for suffering than other sentient species and therefore refinement of their use is particularly important. The refinement of the human impact on laboratory-housed primates and of housing and husbandry practices are dealt with in parts I and II of this three-part review. In part III, methods of refinement that can be applied specifically to the use of primates in procedures, are summarised and discussed, together with a description of some current practices, and the scientific evidence that suggests that they should no longer be used. Methods of refinement of identification, capture and restraint, sampling, administration of substances, humane endpoints, and euthanasia are included. If these methods are used, taking into account species-specific differences and needs, it is concluded that harm can be minimised and primate welfare improved.

The effect of the pulse width of a direct current (DC) on the effectiveness of electrical water bath stunning, and slaughter, was evaluated in broilers (n = 29). Broilers were individually stunned in a water bath for 1 s with a constant peak current of 400 mA of 200 Hz DC delivered using a variable voltage/constant current stunner. The pulse width of the 200 Hz DC was set at 0.5, 1.5 or 2.5 ms (10, 30 or 50% of 5 ms current cycle). The results showed that pulse width had a significant effect on the incidence of epileptiform activity in the electroencephalograms (EEGs). A pulse width of 10% of the current cycle was less effective than pulse widths of 30 and 50% of the current cycle; there was no significant difference between a pulse width of 30 and 50%. The results of a univariate analysis showed that ventral neck-cutting resulted in a significantly shorter time to the onset of less than 10% of the pre-stun power contents in the 13–30 Hz and 2–30 Hz EEG frequency bands when compared with unilateral neck-cutting. It is concluded that a pulse width of 30 or 50% of the current cycle of 200 Hz DC, delivering 400 mA peak current, was better than using a pulse width of 10% of the current cycle.

It has been argued that the welfare of gestating sows is higher in groups than singly in stalls, in part because group housing offers them more space and social contact. This study set out to ascertain how important access to a group pen was to dominant sows housed in stalls, using a measure of motivation. Subjects were trained to perform a panel-pressing task, then housed in a stall and permitted each day to work for a day's access to a fully slatted group pen containing two familiar, subordinate sows at a stocking density of 2.7m2 per pig. Social ranking was determined by observations at mixing and from feed competition tests. The fixed-ratio schedule was increased daily and the highest schedule reached (the reservation price) was used as a measure of motivational strength. To interpret this measure, it was compared with the highest schedule that subjects reached when working for access to the last 1/16th of their estimated ad libitum daily food intake after having consumed the first 15/16ths free. Sixteen subjects were tested, eight working for access to the group pen first and eight for access to the food first. Seven subjects yielded useable data: four reached a higher schedule working for food and three reached a higher schedule working for the group pen. Overall, subjects attached no more importance to a day's access to the group pen than to the last 1/16th of their estimated ad libitum food intake. It is likely that the subjects were close to satiation when working for food because consumption frequently fell substantially short of the ‘ad libitum‘ allowance. These results suggest that dominant, stall-housed sows are only weakly motivated to gain access to a fully slatted group pen, although motivation might be higher when deprived of access to the group pen for longer than one day, if tested at a different time of day or if the quality of the group space was improved; these three possibilities still need to be tested.

In recent years there has been an increased interest in studies related to the welfare of avian species commonly kept as companion animals, specifically those in the order Psittaciformes, commonly referred to as ‘parrots’. During this time the biology and behaviour of wild parrots has also become better understood, aiding the assessment of welfare in captive environments. The impact of the pet trade on wild parrot populations has also become clear. This order now has more globally threatened species than almost any other major group of birds. Many significant aspects of parrot behaviour in the wild, such as flocking, social interaction with conspecifics, foraging on a variety of foods and flight, are denied to varying degrees to parrots kept as companion animals. Captive parrots show high levels of stereotypy, suggesting poor welfare. Welfare may be improved by appropriate environmental enrichment and changes in the social environment of captive parrots kept as companion animals; however, such changes require that caretakers have sufficient knowledge, resources and motivation to accommodate such conditions. The concept of companion animal suitability is an important consideration when developing regulations or policy aimed at improving the welfare of animals kept as companions. Although individual exceptions exist and the level of suitability may vary depending on species, in general, their presence in the pet trade has resulted in serious animal welfare and conservation challenges for parrots, indicating that these animals may be unsuitable as human companions.

Isolating an animal refers to the situation where the animal is physically fully demarcated from conspecifics without physical, visual, olfactory and auditory contact. Animals housed in separate cages in the same room are, although deprived of physical and visual contact, still in olfactory and auditory contact, and thus not totally isolated. During the fifties and sixties several studies claimed to show physiological and behavioural differences between individually and group housed rats and mice. The so-called ‘Isolation Syndrome’ characterised by changes in corticosterone levels, metabolism, growth, and behaviour was introduced, rather as a model for psychoneurosis than through any concern for animal welfare. Today, it is often stated as common knowledge in laboratory animal science textbooks that individual housing as well as isolation of rats and mice has an effect on physiology and behaviour. It is, however, unclear whether this effect actually impairs animal welfare.

The aim of this paper is to analyse studies on individual housing of mice and rats to evaluate whether there is documented proof that individual housing affects welfare, and, alternatively whether it is possible to house these animals individually without negative impact on welfare, eg by providing special housing improvements.

A range of studies have shown that individual housing or isolation has effects on corticosterone, the open field behaviour, barbiturate sleeping time and the metabolism of different pharmaceuticals in the animals. However, this review of 37 studies in rats and 17 studies in mice showed divergence in test results difficult to explain, as many studies lacked basal information about the study, eg information on genetic strains and housing conditions, such as bedding, enrichment and cage sizes. Furthermore, test and control groups most frequently differed in cage sizes and stocking densities, and behavioural tests differed in ways which may very well explain the differences in results. Overall, there seemed to be an effect of individual housing, although it may be small, and it seems reasonable to assume that, through making small changes in the procedures and housing environments, the effects can be minimised or even eliminated. More well-controlled and standardised studies are needed to give more specific answers to the questions this issue poses.

Male piglets are castrated primarily to prevent the unpleasant odours and flavours of entire male pig meat (boar-taint). Although castration can be legally performed without analgesia in the first seven days of life, available evidence shows that castration at any age is painful and may have a detrimental influence on health. Few anaesthetics or analgesics are licensed for use in piglets. The known methods for general and epidural anaesthesia cannot be run at the farm level for practical and/or legal reasons. Use of the local anaesthetic lidocaine is easy and allows the pain resulting from castration to be alleviated. Local destruction of testicular tissue by intra-testicular injection of chemical compounds (salts and acids) is an alternative to surgical castration but needs further investigation regarding welfare improvement and boar taint reduction. Immunocastration, by which castration is achieved using active immunisation (anti–GnRH immunisation) is an efficient alternative to surgical castration; however, there are no licensed vaccines in the EU and the consequences, in terms of pig welfare as well as its acceptability among EU consumers, need further evaluation.

Common marmosets, Callithrix jacchus, are widely used by research laboratories and are commonly provided with food in bowls. These centralised, unchallenging sources of food result in high foraging success for low foraging effort. Foraging devices, which require more skill and effort for foraging success, may broaden the behavioural profiles of marmosets by including more elements of their natural ethogram, reflecting improved welfare. The behaviour of eight female common marmosets was examined as a function of four different food distributions: food centrally located in a stationary bowl; food in a bowl that changed location each day; food centrally located in a stationary bowl, in addition to hidden food in a clustered food source (cluster feeder) or hidden food in dispersed food sources (dispersed feeders). Both the cluster and dispersed feeder distributions increased foraging, and there was a trend for reduced scratching and grooming in the presence of the feeders compared with the bowl-only treatments. The cluster feeder increased the amount of time a marmoset spent in a large room annexed to the home rooms more than the dispersed feeders, and this effect was sustained throughout the day after the feeders had been removed. Both feeders increased activity and movements within all areas of the annexed room compared with the bowl-only treatments; therefore, both feeder types improved the welfare of the captive marmosets more than the provision of food bowls.

Laboratory animals are frequently subjected to routine procedures, such as injections or the withdrawal of blood samples. Acute stress caused by such procedures is associated with physiological changes that can have a strong impact on experimental results. This study investigated the integrated effects of cage enrichment, social housing and handling on the acute stress response of animals subjected to routine experimental procedures. Female mice of two inbred strains (BALB/c and C57BL/6) were housed under either minimal husbandry conditions (MH: no cage enrichment, infrequent handling and a period of individual housing) or enriched husbandry conditions (EH: with cage enrichment, frequent handling and social housing at all times). One mouse in each cage was implanted with a radio-telemetry transmitter for measuring heart rate (HR) and body temperature (BT). The animals were subjected to intraperitoneal injections or short periods of restraint. In addition to telemetry measurements, thymus weight and tyrosine hydroxylase (TH) activity were assessed. It was found that individual housing under MH conditions, as compared with social housing under EH conditions, elevated both basal HR and BT, and significantly elevated the relative recovery time following routine experimental procedures. Thymus weight and TH activity suggested a long-term stress response under MH conditions following individual housing, although the influence of transmitter implantation and (repeated) acute stress remains to be investigated. The results emphasise that husbandry conditions should be taken into account when evaluating physiological measures after routine procedures.

Forty-eight budgerigars (Melopsittacus undulatus) were pairwise housed in small (80 × 40 × 50 cm) and large (160 × 40 × 50 cm) (length × breadth × height) boxes, as well as in 2.0 × 1.0 × 2.0 m aviaries. All boxes and aviaries had two or three (aviaries) perches and food was offered on the bottom. The body masses of female budgerigars increased significantly when they were housed in boxes instead of aviaries. The size of the box did not influence body mass. The frequency of flying was adjusted to the size of the box as both males and females flew approximately twice as often from perch to perch in a small box than in a large box of double length. Flying behaviour differed significantly between large and small boxes and aviaries. The ratio of invariantly flying from perch to perch to all flying events increased from aviaries < large boxes < small boxes. About 75% of all flying events in small boxes consisted of flying from perch to perch. Females in boxes spent more time on the bottom where the food was placed than females in aviaries. Both the significant body mass gain in female budgerigars, as well as the invariant, stereotypic flying movements, indicated reduced welfare in budgerigars housed in boxes, compared with those in aviaries. Providing a large box did not prevent body mass gain, but did increase the variation in flying patterns. Under long-term housing conditions, boxes could lead to at least two serious welfare problems; obesity and stereotypic behaviour, and should be avoided for budgerigars. Therefore, aviary housing should be chosen wherever possible.