Qualitative Behaviour Assessment of cattle expression using a fixed rating scale of 20 descriptors is one of the measures of the Welfare Quality® (WQ) assessment protocol for dairy cattle. As for other on-farm measures of welfare, reliability is an important issue especially if farms are to be certified. This study investigated the repeatability of QBA results across three different observation times during the day (early morning, late morning, early afternoon). For this purpose, 13 observers assessed a total of 30 video clips from ten commercial dairy farms using visual analogue scales to score the 20 QBA terms. QBA scores for ‘emotional state’ were computed according to the Welfare Quality® protocol (WQ_QBA) and, additionally, a Principal Component Analysis was carried out. The latter revealed two main dimensions which may be described as ‘mood’ and ‘activity’, the former thus corresponding to the ‘emotional state’ score of the WQ protocol. Both for scores derived from the WQ protocol and from PCA, mixed model analysis for repeated measures revealed a significant effect of observation time depending on the farm. Mixed model analysis for repeated measures revealed a significant effect of observation time for three farms out of ten on both the WQ_QBA score and the PCA ‘mood’ dimension; a similar effect was found for eight out of ten farms for the PCA ‘activity’ dimension. These results indicate that observation time potentially affects WQ (and other QBA) outcomes on a proportion of farms. However, given that outcomes for WQ_QBA and PCA ‘mood’ were consistent for the majority of farms, procedures suggested in the Welfare Quality® protocol may constitute a reasonable compromise between reliability and feasibility. If the QBA assessment should reflect the ‘mean mood’, multiple assessments throughout the day may be carried out.

This paper describes the development and pilot-testing of a horse welfare assessment protocol (HWAP). The HWAP consists of the collective measurement of numerous factors considered likely to affect a horse's welfare and is thereby designed to provide a holistic score of its welfare status and to identify potential risk factors. The draft protocol contains 47 measures: 15 animal-based, 24 resource-based and eight management-based. It was tested in the autumn at two Swedish riding schools using a total of 37 horses of varying breed, gender and age. Each assessment was repeated after 16-25 days. The results showed that 66% (31/47) of the measures had over 85% repeatability between assessments. Results indicated occurrence of behavioural issues, eg aggression and avoidance, and potential risk factors, such as inadequate management routines and feeding regimes. Using the HWAP, the assessment of up to 22 horses could be carried out in one day. Changes were proposed to the draft protocol which included incorporating an ethogram to assess the human-animal relationship and assessing bit-related injuries. We propose that the protocol might: i) provide a firm basis for the welfare monitoring of horses; ii) identify important potential risk factors; iii) guide welfare improvement and management practices for horse owners and stable managers; and iv) contribute to the development of certification schemes for horse facilities.

The integration of outcome-based measurements to animal welfare assessment programmes can provide a new perspective on the individual animal's experience. Identifying variability in individual experiences can facilitate understanding of animals at the periphery of the welfare spectrum, compared to those at the average. Welfare Quality® physical measurements and behavioural observations were taken from the same fifteen non-cage laying hens throughout their production cycle. The average amount of time performing each of nine behaviours and the amount of variation in each behaviour's performance was compared at four different ages: 19, 28, 48 and 66 weeks. The same analysis was performed for all Welfare Quality® physical measurements. To identify associations between a hen's behaviour and her physical condition, a cluster analysis was performed for all ages, as well as on data collated from all ages. No differences were observed among the four ages for the average amount of time performing many of the behaviours, but the amount of variability differed for most behaviours observed. Physical measurements taken at 19 weeks differed from those taken at later ages. Bodyweight consistently clustered with time spent preening, yet the patterns of clustering differed at each age. These results highlight the importance of age when conducting welfare assessments. Auditors also should report not only the average, but the variability of responses; while the average response of the flock may appear consistent across time or treatment, differences among the hens within the same flock may vary drastically.

The Welfare Quality® (WQ) protocols are increasingly used for assessing welfare of farm animals. These protocols are time consuming (about one day per farm) and, therefore, costly. Our aim was to assess the scope for reduction of on-farm assessment time of the WQ protocol for dairy cattle. Seven trained observers quantified animal-based indicators of the WQ protocol in 181 loose-housed and 13 tied Dutch dairy herds (herd size from 10 to 211 cows). Four assessment methods were used: avoidance distance at the feeding rack (ADF, 44 min); qualitative behaviour assessment (QBA, 25 min); behavioural observations (BO, 150 min); and clinical observations (CO, 132 min). To simulate reduction of on-farm assessment time, a set of WQ indicators belonging to one assessment method was omitted from the protocol. Observed values of omitted indicators were replaced by predictions based on WQ indicators of the remaining three assessment methods, resources checklist, and interview, thus mimicking the performance of the full WQ protocol. Agreement between predicted and observed values of WQ indicators, however, was low for ADF, moderate for QBA, slight to moderate for BO, and poor to moderate for CO. It was concluded that replacing animal-based WQ indicators by predictions based on remaining WQ indicators shows little scope for reduction of on-farm assessment time of the Welfare Quality® protocol for dairy cattle. Other ways to reduce on-farm assessment time of the WQ protocol for dairy cattle, such as the use of additional data or automated monitoring systems, should be investigated.

A central aim of animal welfare science is to be able to compare the effects of different ways of keeping, managing or treating animals based on welfare indicators. A system to aggregate the different indicators is therefore needed. However, developing such a system gives rise to serious challenges. Here, we focus specifically on the ethical aspects of this problem, taking as our starting point the ambitious efforts to set up an aggregation system within the project Welfare Quality® (WQ). We first consider the distinction between intra- and inter-individual aggregation. These are of a very different nature, with inter-individual aggregation potentially giving rise to much more serious ethical disagreement than intra-individual aggregation. Secondly, we look at the idea of aggregation with a focus on how to compare different levels and sorts of welfare problems. Here, we conclude that animal welfare should not be understood as a simple additive function of negative or positive states. We also conclude that there are significant differences in the perceived validity and importance of different kinds of welfare indicators. Based on this, we evaluate how aggregation is undertaken in WQ. The main conclusion of this discussion is that the WQ system lacks transparency, allows important problems to be covered up, and has severe shortcomings when it comes to the role assigned to experts. These shortcomings may have serious consequences for animal welfare when the WQ scheme at farm or group level is applied. We conclude by suggesting ways to overcome some of these shortcomings.

Regarding broilers, Brazil is the third largest producer and the leading chicken meat exporter in the world. Independent certifications are required by some importers in the European Union (EU) to guarantee compliance with minimum welfare requirements. Our objective was to compare broiler chicken welfare in GLOBALGAP® certified (C) and non-certified (N) intensive farms in the State of Paraná, Brazil, using the Welfare Quality® protocol. Ten farms in each group were evaluated and data were transformed using scores that ranged from 0 to 100, where 100 is the best condition. Results suggested that farms adhered to minimum welfare standards, regardless of certification and that the adoption of standards based on overseas rules may have limitations for the improvement of animal welfare. In order to further improve broiler welfare in Brazil, more rigorous standards should be developed.

This paper presents an account of a Welfare Quality® assessment of 92 dairy farms carried out by seven experienced assessors. The aim was to evaluate the potential of the Welfare Quality® assessment protocol with respect to its uptake by UK farm assurance schemes. Data collection, and measure aggregation were performed according to the Welfare Quality® protocol for dairy cows. This study examined the data itself, by the testing of how hypothetical interventions might be reflected in changes in the aggregated scores, and also investigated human-related aspects, through inter-assessor standardisation sessions to evaluate reliability, and an assessor focus group to collect feedback. Overall, three main ‘challenges’ were identified. The first challenge related to the large amount of missing data. Unexpectedly, this was such that it was only possible to calculate an overall classification for 7% of farms. The second challenge concerned the way in which aggregated scores did not always reflect hypothetical interventions. The final challenge was inter-assessor reliability, where not all assessors were found to achieve acceptable levels of agreement on a number of outcome measures by the third training session. Suggestions for managing these challenges included, follow-up to assessor training, the use of multiple imputation methods to fill in missing data, and, where applicable, not aggregating the scores. The conclusion of the study was that the protocol provided useful information from which to make an informed selection of measures, but that the challenges, combined with the lengthy assessment time, were too great for its use as a certification tool.

The Welfare Quality® (WQ) on-farm welfare assessment protocols for fattening pigs (n = 95 farms) and sows, as well as suckling piglets (n = 103 farms), were applied on Finnish farms. In order to identify distinct types of welfare problems (WPTs) Principal Component Analysis (PCA) was applied to the pooled animal-based items within both categories of animals. Measures describing suckling piglets did not contribute to the WPTs. The main WPTs (seemingly) reflected fighting in fattening pigs and lack of bedding in both fattening pigs and sows. The results imply that WQ includes biologically plausible shortlists of animal-based measures with decent to good internal consistency describing distinct types of welfare problems in growing pigs and in sows. The 20 descriptors of Qualitative Behavioural Assessment were analysed similarly to identify distinct mood types, which were named active positive, passive positive and passive positive behaviours. The different mood types had close to identical build-up in both fattening pigs and in sows and suckling piglets.

This study aimed to establish associations between the environment and animal-based measures of welfare collected on 158 Finnish farms according to the Welfare Quality® systems for pigs. The data consisted of 95 welfare assessments in fattening pigs and 103 in sows, including suckling piglets. Principal Component Analysis had previously been applied to animal-based welfare measures (ABWM) and to the 20 descriptors of QBA to identify distinct types of welfare problems (WPT) and mood (MT), respectively. Generalised linear modeling was used to investigate environmental (space allowance, group size, feeding arrangement, floor type and use of enrichment or bedding) effects on WPT and MT scores. Those ABWMs not contributing to the major WPTs, but occurring on more than 40% of the farms, were considered important and used as outcome variables as well. The most important environmental determinants of pig welfare were space allowance for fattening pigs, group size in gestation and in the use of bedding for both fattening pigs and gestating sows. Bedding decreased tail biting and signs of fighting when used as a fairly thick layer for fattening pigs. In sows, the benefits of bedding, including less frustration and bursitis, required a smaller amount of material than in fattening pigs. An increasing space allowance was advantageous for fattening pigs, although signs of fighting increased in very spacious bedded pens. The positive effects of space, including a decrease of tail lesions and a more positive mood continued at least up to 1.5 m2 per fattening pig. Signs of resource shortage in sows increased with a growing group size according to a steepening curve.

This study aimed to identify current weak points in animal welfare in Danish dairy production at herd level using the Welfare Quality® (WQ) protocol, and at national level using the Danish Animal Welfare Index (DAWIN) protocol. The DAWIN was developed as a monitoring tool for the welfare of the Danish dairy cow population, derived from the aggregation of DAWIN assessments at herd level. The DAWIN dairy cow protocol covers 29 measures (13 resource- and 16 animal-based measures) that were weighted and aggregated into a final overall population welfare score. A total of 3,591 cows from 60 dairy herds were assessed throughout 2015. Results from both the WQ and DAWIN were presented at six criteria levels in order to identify specific areas of concern relating to animal welfare at herd versus population level. Both protocols indicated a good general level of welfare across study herds, but also identified insufficient water supply as the main area of concern. In addition, resting comfort (ie time needed to lie down, collisions with barn equipment, cleanliness of rear body parts, animals lying outside of the designated lying area) and disease (in terms of the proportion of cows with chronically elevated somatic cell counts) were identified as problematic areas. The two assessment protocols both identified behavioural deficits, but in the WQ it was due to zero-grazing systems in contrast to the insufficient numbers of cow brushes in the DAWIN protocol. Despite differences in the aggregation, similar areas of concern were identified at criteria level.

Positive stockperson attitudes to his or her animals is associated with a positive behavioural response in the animals and in other indicators assumed to reflect a high level of welfare as well as increased productivity, eg in milk yield. Conversely, negative attitudes have been found to have the opposite effect. However, so far, no attempt has been made to link the attitudes of stockpersons to outcomes of comprehensive protocols assessing animal welfare at farm level. In the study reported here, attitudes of Danish dairy farmers are compared with the on-farm welfare of dairy cows as assessed by the Welfare Quality ® (WQ) protocol. In 35 dairy herds the welfare was assessed, and the farmers filled in a self-report questionnaire on their attitudes to dairy cows and how they should be handled. The farmers on the farms with the lowest total welfare score could, according to the self-report questionnaire, be characterised as those with a negative attitude to the handling of cows (eg less strongly disliked to kick a cow or use a stick, when necessary). As expected, farmer behavioural attitudes regarding different situations correlated: for example, farmers who liked more to work with and have positive interactions with the cows also agreed more on patient handling and the importance of regular positive contact. Furthermore, farms where farmers gave cows a positive characteristic were found to score highly on the WQ principle ‘Appropriate behaviour’.

The objective of the present study was to assess the sensitivity and feasibility of the Welfare Quality® (WQ) protocol for finishing pigs in 10 Spanish slaughterhouses. Sensitivity refers to the ability of the WQ assessment system to discriminate between slaughter conditions whilst feasibility denotes that the protocol is concise and easy to implement. On arrival at the plant, the incidence of dead, sick and panting animals was assessed in 1,002 (± 93) pigs per abattoir. During unloading, the percentage of pigs that slipped, fell, showed reluctance to move or turned back and were lame was also assessed. In the lairage pens, the stocking density and the percentage of pigs that were panting, shivering, and huddling was assessed in a total of 346 (± 81.0) pigs per abattoir. Stunning effectiveness, slaughter checks and skin lesions were also assessed in 60 animals per abattoir. For the majority of measures, any differences between slaughterhouses were found to be attributable to the installation itself and the management of the slaughterhouse, such as generalised fear, slipping and falling or stunning effectiveness, as opposed to measures taken to assess transport conditions or farm origin, such as lameness or sick and dead animals. The study protocol took 5.5 h for one observer to complete, in a slaughterhouse killing more than 550,000 pigs a year, although this time could increase dramatically in smaller abattoirs due to delays in the arrival of lorries. The protocol provides a general overview of the state of welfare of animals at the slaughterhouse and can readily identify specific problems in certain areas, such as stunning of animals.

The aim of this study was to assess the interobserver reliability of the measures forming the Welfare Quality® animal welfare assessment protocol for sows and piglets. The study was carried out at nine farms in Northern Germany. Two trained observers evaluated identical animals simultaneously but independently in 40 joint farm visits. Interobserver reliability was calculated at individual animal level using Cohen's kappa, weighted kappa and the prevalence-adjusted, bias-adjusted kappa (PABAK) and at farm level using Spearman's rank correlation coefficient (RS), the intraclass correlation coefficient (ICC), smallest detectable change (SDC) and limits of agreement (LoA). While a direct comparison of the adjectives of the qualitative behaviour assessment showed poor interobserver reliability, a Principal Component Analysis detected good interobserver reliability. The assessment of social and exploratory behaviours showed acceptable interobserver reliability, while the assessment of stereotypies displayed good interobserver reliability. The human-animal relationship test showed only poor interobserver reliability at individual animal and farm levels. In most cases, measures of health and physical state assessed in sows and piglets exhibited acceptable or good interobserver reliability. In conclusion, after some measures are revised, particularly those examining the human-animal relationship, the Welfare Quality® protocol for sows and piglets will represent a reliable approach in terms of interobserver reliability to assess the welfare of sows and piglets.

This study was carried out to compare the health of growing pigs in five different production systems in France and Spain using measures provided by the Welfare Quality® protocol. A total of 11,647 pigs housed on 91 commercial farms were evaluated over a two-year period (2007-2009). Farms considered as conventional were close to the European dominant production system, rearing ‘white’ pigs (eg Large White; Landrace * Pietrain) housed on concrete floors. Systems considered as differentiated had specifications to distinguish them from the conventional one. Farms that housed ‘white’ breeds of pigs on straw were then considered as a different production system. Mallorcan Black pigs managed extensively on family farms in the Balearic islands represented a third production system. The remaining two systems assessed were represented by the methods used for Iberian pig rearing extensively or intensively. Multiple Generalised Linear Mixed Models were performed for each animal-based measure of health. The straw-bedded and the conventional systems did not differ in the prevalence of any animal-based measures. Mallorcan Black pigs and Iberian pigs kept extensively had a lower prevalence of severe wounds than pigs in the conventional system and the lowest prevalence of tail biting. Focusing on pigs housed in the conventional system, several possible causal factors (such as the feeding system and the type of floor) were identified relating to severe wounds, tail biting and lameness. Therefore, the recording of simple environmental-based factors can be useful in detecting farms that are more likely to show these problems.

The Welfare Quality® (WQ) project developed protocols as international standards for farm animal welfare assessment. For dairy cattle, the WQ protocol may be performed any time during the indoor period when the herd has no access to pasture. However, timing of welfare assessments during the indoor period might influence the outcome, as pasture access influences many welfare aspects and such effects are likely to carry over to the beginning of the indoor period. In order to test this hypothesis, we assessed ten herds at both the beginning and end of the same indoor period. Assessment at the end of the indoor period resulted in a higher prevalence of the following welfare measures: severely lame cows, mild integument alterations, coughs per cow per hour, diarrhoea, and vulvar discharge. In addition, at the end of the indoor period, the mortality rate was higher and the Qualitative Behaviour Assessment (QBA) score was worse, but there were fewer cows with dirty udders and mean time to lie down was lower. This led to worse scores on the criterion and principle levels of WQ integration, but resulted in a lower welfare categorisation for one of the herds only. The better scores for several aspects of health and for the QBA at the beginning of the indoor period not only confirm the carryover of positive effects of access to pasture, but also imply a need for careful consideration of the timing and frequency of WQ assessments of herds that are housed outdoors for part of the year.

The Welfare Quality® (WQ) protocol for on-farm dairy cattle welfare assessment describes 33 measures and a step-wise method to integrate the outcomes into 12 criteria scores, grouped into four principle scores and into an overall welfare categorisation with four possible levels. The relative contribution of various welfare measures to the integrated scores has been contested. Using a European dataset (491 herds), we investigated: i) variation in sensitivity of integrated outcomes to extremely low and high values of measures, criteria and principles by replacing each actual value with minimum and maximum observed and theoretically possible values; and ii) the reasons for this variation in sensitivity. As intended by the WQ consortium, the sensitivity of integrated scores depends on: i) the observed value of the specific measures/criteria; ii) whether the change was positive/negative; and iii) the relative weight attributed to the measures. Additionally, two unintended factors of considerable influence appear to be side-effects of the complexity of the integration method. Namely: i) the number of measures integrated into criteria and principle scores; and ii) the aggregation method of the measures. Therefore, resource-based measures related to drinkers (which have been criticised with respect to their validity to assess absence of prolonged thirst), have a much larger influence on integrated scores than health-related measures such as ‘mortality rate’ and ‘lameness score’. Hence, the integration method of the WQ protocol for dairy cattle should be revised to ensure that the relative contribution of the various welfare measures to the integrated scores more accurately reflect their relevance for dairy cattle welfare.

Welfare assessment protocols using primarily animal-based measures are believed to give valid information about the welfare of animals. However, they tend to be time consuming and therefore costly, thus in practice there is often considerable reluctance to use them. In the present study, the relatively quick to use, simple but non-validated welfare assessment protocol for dairy cattle developed by the Danish Cattle Federation and the validated comprehensive Welfare Quality® protocol were compared in Danish conditions. In total, 44 Danish dairy herds were evaluated using the two protocols. The protocols were correlated on four sub-levels (corresponding to the ‘Principles’ in the Welfare Quality® protocol) and on the Overall welfare score. They correlated significantly with regard to the Principles ‘Good health’ and ‘Appropriate behaviour’. Significant correlations were not found for ‘Good feeding’, ‘Good housing’, and the Overall score. On the basis of this we changed the Danish Cattle Federation protocol by introducing six new measures, changing three measuring procedures and omitting two measures. This extended protocol was found to correlate significantly with the Welfare Quality® protocol in all four Principles and on the Overall score. The extended protocol still has the advantage of the original Danish Cattle Federation protocol whereby under Danish field conditions it will take only 2 h to apply as opposed to 7-8 h for the Welfare Quality® protocol. We believe that the extended protocol balances, in a good way, the demands of practicability against its value as a diagnostic test.

This paper focuses on the reliability of the multi-criteria evaluation model included in the Welfare Quality® protocol for growing pigs to aggregate the animal-based indicators, first to criteria, then to principle level and finally to an overall welfare score. This assessment was carried out in a practical application study on a sample of 24 farms in Germany. Altogether, 102 protocol assessments were carried out in repeated visits to these farms in order to evaluate the inter-observer and test-retest repeatability of the overall scores calculated by the multi-criteria evaluation system. Reliability is then assessed by the calculation of different reliability and agreement parameters: Spearman Rank Correlation Coefficients (RS), Intraclass Correlation Coefficients (ICC), Smallest Detectable Changes (SDC) and Limits of Agreement (LoA). Inter-observer repeatability was insufficient for the criteria comfort around resting, absence of injuries, expression of social behaviours, expression of other behaviours, good human-animal relationship and positive emotional state as well as for the principles good housing and appropriate behaviour. This is probably due in the main to insufficient repeatability of the underlying indicators that have been revealed in previous studies. Test-retest repeatability is predominantly insufficient. Overall, the present results highlight the importance of absolutely reliable indicators at the baseline level. Furthermore, it could be shown that the calculation procedure is partly incorrect and consequently needs correction. Therefore, this study is an important contribution to the future progression of the Welfare Quality® protocols and animal welfare assessment tools in general.

This paper describes the implementation and simultaneous promotion of an action plan designed to ensure animal welfare standards on-farm that exceed the requirements for acceptability in law. The approach is based on two action cycles, the producer and retailer cycles, The producer cycle, involving welfare audit and the implementation of an action plan for welfare has four stages: self-assessment; independent audit; creation of an action plan based on identification of principal hazards and critical control points; review; and revision of the action plan depending upon assessment of outcomes. The retailer cycle is designed to set quality standards for animal welfare, demonstrate compliance, promote proven high welfare products and reward producers. The paper reviews some incentives and constraints to action for both farmers and retailers and presents encouraging examples of the extent to which both producers and retailers have responded to increased public demand for high welfare products.

This study sought to investigate whether animal health and welfare state and changes were associated with technical efficiency in a sample of 34 Austrian dairy farms. Health and welfare were assessed twice using the Welfare Quality® (WQ) assessment protocol for dairy cattle. Following a baseline welfare assessment, farm-specific health and welfare planning was conducted on the farms. This included the identification, selection and implementation of measures in housing and management that aimed at improving health and welfare states. One year after implementation, farms were reassessed to evaluate changes in health and welfare states and consequences for farms’ technical efficiency were analysed using the Malmquist index. Our results indicated that farms with a higher health state (WQ principle score ‘Good health’) achieved higher technical efficiencies. However, we could not show that changes in the welfare state within a one-year period affected technical efficiency: across all farms, technical efficiency remained stable and Malmquist indices (indicating efficiency and technological change) could not be explained by the different welfare scores. Nevertheless, our study showed data envelopment analysis to be a valuable method for analysing the relationship between animal welfare and farm success and our results indicate substantial potential synergies between these two aspects.