Behaviour as a tool in the assessment of animal welfare

A central issue in animal welfare research is how to assess the welfare state of animals objectively and scientifically. I argue that this issue can be approached by asking two key questions: 1) is the animal physically healthy and 2) does the animal have what it wants? Behaviour is used to answer both of these questions. In the assessment of physical health, it can be used for clinical and pre-clinical diagnosis. In the assessment of what animals want, it has a major role through choice and preference testing. It is particularly important that applied ethologists develop methods for assessing welfare in situ--in the places where concern for animal welfare is greatest such as on farms and in zoos.     

Use and non-use values as motivational construct dimensions for farm animal welfare: impacts on the economic outcome for the farm

This study explored how farmers’ motivation in terms of use values and/or non-use values to work with farm animal welfare are associated with the economic outcome for the farm. Use values in farm animal welfare refer to economic value derived from productivity and profitability considerations. Non-use values in farm animal welfare refer to economic value derived from good animal welfare, irrespective of the use the farmer derives from the animal, currently or in the future. The analysis was based on detailed information about the income statements of a sample of Swedish dairy farmers, obtained from the Swedish Farm Economic Survey, complemented with survey information about their perceived use and non-use values in farm animal welfare. The findings suggest that farm economic outcome is significantly associated with motivation in terms of use values, but not so much with motivation in terms of non-use values. This is interesting from a policy point of view, because it indicates that farmers with different approaches to farm animal welfare may experience different economic outcomes for their farms. Findings can, for instance, be used to strengthen farmers’ engagement in various private quality assurance standards, which generally focus on values of non-use type, by pointing to that realisation of such values will not impair the economic outcome of the farms. Moreover, findings also suggest that farmers’ economic incentives for engagement in such standards may need to be further strengthened in order to become more attractive, as findings point to that a focus on non-use values generally is not associated with more favourable economic outcomes.     

Road transport of farm animals: effects of journey duration on animal welfare

Transport of farm animals gives rise to concern about their welfare. Specific attention has been given to the duration of animal transport, and maximum journey durations are used in legislation that seek to minimise any negative impact of transport on animal welfare. This paper reviews the relatively few scientific investigations into effects of transport duration on animal welfare in cattle, sheep, horses, pigs and poultry. From the available literature, we attempt to distinguish between aspects, which will impair welfare on journeys of any duration, such as those associated with loading, and those aspects that may be exacerbated by journey time. We identify four aspects of animal transport, which have increasing impact on welfare as transport duration increases. These relate to (i) the physiological and clinical state of the animal before transport; and - during transport - to (ii) feeding and watering; (iii) rest and (iv) thermal environment. It is thus not journey duration per se but these associated negative aspects that are the cause of compromised welfare. We suggest that with a few exceptions, transport of long duration is possible in terms of animal welfare provided that these four issues can be dealt with for the species and the age group of the animals that are transported.     

Register-based predictors of violations of animal welfare legislation in dairy herds

The assessment of animal welfare can include resource-based or animal-based measures. Official animal welfare inspections in Denmark primarily control compliance with animal welfare legislation based on resource measures (e.g. housing system) and usually do not regard animal response parameters (e.g. clinical and behavioural observations). Herds selected for welfare inspections are sampled by a risk-based strategy based on existing register data. The aim of the present study was to evaluate register data variables as predictors of dairy herds with violations of the animal welfare legislation (VoAWL) defined as occurrence of at least one of the two most frequently violated measures found at recent inspections in Denmark, namely (a) presence of injured animals not separated from the rest of the group and/or (b) animals in a condition warranting euthanasia still being present in the herd. A total of 25 variables were extracted from the Danish Cattle Database and assessed as predictors using a multivariable logistic analysis of a data set including 73 Danish dairy herds, which all had more than 100 cows and cubicle loose-housing systems. Univariable screening was used to identify variables associated with VoAWL at a P-value<0.2 for the inclusion in a multivariable logistic regression analysis. Backward selection procedures identified the following variables for the final model predictive of VoAWL: increasing standard deviation of milk yield for first lactation cows, high bulk tank somatic cell count (⩾250 000 cells/ml) and suspiciously low number of recorded veterinary treatments (⩽25 treatments/100 cow years). The identified predictors may be explained by underlying management factors leading to impaired animal welfare in the herd, such as poor hygiene, feeding and management of dry or calving cows and sick animals. However, further investigations are required for causal inferences to be established.     

Assessment of welfare indicators in dairy farms offering pasture at differing levels

In terms of animal welfare, farming systems of dairy cows are perceived positively by consumers when compared to pigs or poultry. A main reason is that the majority of consumers associate dairy farming with pasture, which in turn they relate with benefits for animal health and welfare. However, holistic scientific assessments of the effects of pasturing on animal welfare are rare. Hence, it was the aim to study the animal welfare level in 61 German loose housing dairy farms by using the measures of the Welfare Quality® protocol for dairy cattle (WQP). Data were collected twice per farm at the end of the pasture season (July to October) and approximately 6 months later at the end of the barn season (December to April). Farms were classified based on the duration cows had access to pasture per day during the pasture season: group 1 (G1)>10 h; group 2 (G2) 6 to 10 h; group 3 (G3)<6 h and group 4 (G4) without pasture access. The average herd size was 129 Holstein-Friesian or Red-Holstein cows (range 58 to 527). In addition to WQP data, performance data were gathered from routine herd data recordings. The indicators were aggregated to criteria applying the scoring system of the WQP. G4 received lower scores at the first than at the second visit for the criterion absence of hunger, while there were no differences between visits in the other groups (P=0.58 – group×farm visit effect). All pasturing groups were scored better at the end of the pasture season than G4 for the criterion comfort around resting (P<0.01). Compared with G1 for both farm visits and G2 for the end of the barn season, G4 reached inferior scores for the criterion absence of injuries, including indicators such as hairless patches, lesions, and swellings and lameness. At both assessments G2 was scored higher than the other groups for the criterion absence of diseases (P=0.04). In conclusion, pasture access had positive effects only on selected welfare indicators, however, these effects were not maintained throughout the barn season.     

Effect of milking frequency and nutritional level on aspects of the health and welfare of dairy cows

The objective of this experiment was to investigate the effect of milking frequency and nutritional level on some aspects of animal health. Holstein-Friesian cows (n = 60) were assigned post calving to a factorial arrangement of treatments; twice a day (TAD) milking on a high or low nutritional level; once a day (OAD) milking on a high or low nutritional level. Milking characteristics were recorded daily. Blood samples to evaluate changes in the composition of the blood cells, milk leakage, udder tension and locomotory ability were measured on four occasions. Teat-ends were classified for hyperkeratosis (HK) monthly post partum. TAD had longer daily milking times (P < 0.001) compared with OAD cows. There was no effect of milking frequency or nutritional level on morning milking time, time to milk letdown or peak milk flow rate (P>0.05). High nutritional level cows had higher average flow-rates (P < 0.05) than low nutritional level cows. Neither milking frequency nor nutritional level affected HK (P>0.05). However, HK values were positively correlated with daily milking time for OAD cows for 6 months of lactation (P < 0.05). This correlation was significant (P < 0.01) for cows milked OAD on high nutrition during the peak lactation period. OAD cows had higher levels of milk leakage compared with TAD cows during the month of May (P < 0.01). Cows on high nutrition milked OAD showed higher udder firmness scores than cows milked once or twice daily at the low nutrition level in June and July (P < 0.05). OAD cows had higher locomotion scores compared with TAD cows (P < 0.001). Locomotion and udder firmness scores were significantly correlated for OAD in June (P < 0.05). OAD cows had lower blood lymphocyte counts, numerically higher counts of neutrophil and a higher monocyte count at peak lactation compared with TAD cows suggesting that OAD cows had altered immune responses. The increase in milk leakage, higher udder firmness and locomotion scores in conjunction with changes in blood cells, suggests that OAD milking may have caused some discomfort to the cows during peak lactation. A reduction in the concentrate input to OAD cows during this period could ameliorate this problem.     

Animal-based measures on fattening heavy pigs at the slaughterhouse and the association with animal welfare at the farm level: a preliminary study

Monitoring animal welfare (AW) in pig farms requires both proper indicators and a feasible approach. Animal-based measures (ABMs) are well-established AW indicators. Furthermore, AW screening at the slaughterhouses could be useful for identifying problems on farm. The aim of this study was to evaluate ABMs at the slaughterhouse and, when possible, to compare these ABMs with those collected on the farm. The study was carried out in northern Italy in a commercial abattoir and in a sample of farms. Animal-based measures were recorded on pigs from 62 batches of 54 farms, during ante-mortem (n=10 085 pigs) and post-mortem (n=7952 pigs) inspections. Sixteen of 54 farms were selected to compare ABMs collected at the slaughterhouse with ABMs collected on the farm. Overall, 2295 pigs (mean pigs examined per farm 119±45) were inspected at the slaughterhouse (group S) and 420 pigs (mean pigs per farm 26±5) on the farm (group F). Non-animal-based measures were also collected at the 16 farms. Differences between groups S and F, at the animal level, were assessed by a two-tailed paired Wilcoxon–Mann–Whitney test. Differences at the site of observation level (farm and slaughterhouse) were assessed by Fisher’s exact test using a hierarchical log-linear modelling for contingency tables. The most frequent ABMs at the slaughterhouse were manure on the body (47.7%), followed by dermatitis (28.0%), white spot (25.4%) and bursitis (24.7%). Recording ABMs at the slaughterhouse and on the farm usually yielded similar results; however, there were some exceptions. In particular, significant differences were found for non-uniformity of size (P<0.05) and dermatitis (P<0.001), which were higher at the slaughterhouse than on the farm. Results of log-linear modelling underlined the effect of the farm of origin on the percentage of pigs with bursitis, manure on the body and ear injuries that were observed at the slaughterhouse. In group S, significant associations between manure on the body and insufficient presence of clean and dry areas in the corresponding farm were found (P<0.05). Although these results should be interpreted with care due to the limited sample of farms, the slaughterhouse could be a feasible site of observation of ABMs, which could then be integrated in monitoring of AW on farm. Considering the number of slaughtered batches per farm, it would be possible to repeat assessments several times throughout the year for each farm, which could help provide an index for the continuous monitoring of AW.     

Aggregation of measures to produce an overall assessment of animal welfare. Part 1: a review of existing methods

Several systems have been proposed for the overall assessment of animal welfare at the farm level for the purpose of advising farmers or assisting public decision-making. They are generally based on several measures compounded into a single evaluation, using different rules to assemble the information. Here we discuss the different methods used to aggregate welfare measures and their applicability to certification schemes involving welfare. Data obtained on a farm can be (i) analysed by an expert who draws an overall conclusion; (ii) compared with minimal requirements set for each measure; (iii) converted into ranks, which are then summed; or (iv) converted into values or scores compounded in a weighted sum (e.g. TGI35L) or using ad hoc rules. Existing methods used at present (at least when used exclusively) may be insufficiently sensitive or not routinely applicable, or may not reflect the multidimensional nature of welfare and the relative importance of various welfare measures. It is concluded that different methods may be used at different stages of the construction of an overall assessment of animal welfare, depending on the constraints imposed on the aggregation process.     

Ethical considerations and animal welfare in ecological field studies

This paper addresses some of the ethical and welfare considerations implicit in the application of general techniques in common use during the course of collecting data in ecological field work. Even if they are not explicitly constructed as manipulative experiments, many field studies involve some degree of intervention during routine monitoring programmes: through disturbance caused merely by the presence of an observer or where specific sampling techniques themselves involve capture, handling and marking. Such interventive techniques may cause discomfort, distress or loss of fitness, even in the extreme may result in incidental mortality — and the ethical scientist should critically evaluate the implications of each methodology before adopting any procedure. The paper reviews by way of example the types of objective information now available for both small and larger mammals in relation to: (i) distress and mortality during capture operations; (ii) mortality or distress caused at the time by marking; (iii) longer-term consequences of handling and marking in terms of subsequent [delayed] mortality or loss of fitness, before considering a formal framework for assessment of costs and benefits of any given field programme.     

Aggregation of measures to produce an overall assessment of animal welfare. Part 2: analysis of constraints

The overall assessment of animal welfare is a multicriterion evaluation problem that needs a constructive strategy to compound information produced by many measures. The construction depends on specific features such as the concept of welfare, the measures used and the way data are collected. Welfare is multidimensional and one dimension probably cannot fully compensate for another one (e.g. good health cannot fully compensate for behavioural deprivation). Welfare measures may vary in precision, relevance and their relative contribution to an overall welfare assessment. The data collected are often expressed on ordinal scales, which limits the use of weighted sums to aggregate them. A sequential aggregation is proposed in the Welfare Quality® project, first from measures to welfare criteria (corresponding to dimensions with pre-set objectives) and then to an overall welfare assessment, using rules determined at each level depending on the nature and number of variables to be considered and the level of compensation to be permitted. Scientific evidence and expert opinion are used to refine the model, and stakeholders’ approval of general principles is sought. This approach could potentially be extended to other problems in agriculture such as the overall assessment of the sustainability of production systems.     

Rapid animal welfare assessment: an archaeological approach

The welfare of an individual depends on its capacity to overcome suboptimal conditions in its environment; otherwise, its physical and psychological health becomes compromised. A situation that clearly indicates lack of control of the environment is the expression of abnormal behaviours, such as stereotypies. This study aimed to verify the well-being of police horses using a new rapid form of welfare assessment: an archaeological approach. To this end, we sampled and quantified marks found on the stables, deposited as a result of abnormal behaviour. We cross-referenced these physical marks with veterinary records of diseases, such as colic, known to be associated with stress. A total of 46 horses were sampled and the results showed a significant medium-strength, positive correlation between bite mark frequency on stable doors and the incidence of colic. A weak significant positive correlation was found between length of scratch marks (from pawing) and the incidence of lameness. We conclude that these marks reflect the accumulated expression of abnormal behaviour and can provide rapid insight into the welfare of individual animals.     

Assessing animal welfare in sow herds using data on meat inspection,medication and mortality

This paper aims to contribute to the development of a cost-effective alternative to expensive on-farm animal-based welfare assessment systems. The objective of the study was to design an animal welfare index based on central database information (DBWI), and to validate it against an animal welfare index based on-farm animal-based measurements (AWI). Data on 63 Danish sow herds with herd-sizes of 80 to 2500 sows and an average herd size of 501 were collected from three central databases containing: Meat inspection data collected at animal level in the abattoir, mortality data at herd level from the rendering plants of DAKA, and medicine records at both herd and animal group level (sow with piglets, weaners or finishers) from the central database Vetstat. Selected measurements taken from these central databases were used to construct the DBWI. The relative welfare impacts of both individual database measurements and the databases overall were assigned in consultation with a panel consisting of 12 experts. The experts were drawn from production advisory activities, animal science and in one case an animal welfare organization. The expert panel weighted each measurement on a scale from 1 (not-important) to 5 (very important). The experts also gave opinions on the relative weightings of measurements for each of the three databases by stating a relative weight of each database in the DBWI. On the basis of this, the aggregated DBWI was normalized. The aggregation of AWI was based on weighted summary of herd prevalence’s of 20 clinical and behavioural measurements originating from a 1 day data collection. AWI did not show linear dependency of DBWI. This suggests that DBWI is not suited to replace an animal welfare index using on-farm animal-based measurements.     

Aligning health and welfare principles and practice in organic dairy systems: a review

This review provides an assessment of research findings into the current practices and standards and the principles and aspirations for organic dairy production, with respect to the health and welfare of the dairy cow. The relationships between the four main factors: management, environment, genetics and nutrition and their impact on the health and welfare status of organic dairy cows are considered. The concept that good animal health and welfare is more than merely the absence of disease, with behavioural aspects of health and welfare such as physiological and psychological needs, is also discussed. These factors are inter-related and important in all dairy systems, irrespective of whether the system is organic, low-input or intensive. Incidences of individual clinical and sub-clinical diseases that are recorded in conventional dairy systems also occur in organic dairy systems, with infertility, lameness and mastitis being the major problems. However, the magnitude of the incidence of many of these diseases may be either lower or higher in organic systems due to different management practices and the standards defined for organic milk production that, for example, prohibit the routine use of conventional medicines and require the feeding of high-forage diets. In relation to different systems, it is important to note that contrary to a common assumption, good welfare does not necessarily occur with more extensive systems. The type of organic system (self-sufficient, purchased-feed) also has the potential to have a major influence on the incidence of health problems and the reproductive status of organic dairy herds. Health status is also influenced by environmental and welfare factors, including the method of rearing replacement animals, type of housing and the geographical and climatic conditions of individual farms. Overall, this review identifies where conflicts arise between current practice and the organic principles and standards, and aims to provide suggestions to bring about further improvement in organic dairy health and welfare.     

The behaviour and welfare of buffaloes (Bubalus bubalis) in modern dairy enterprises

This review deals with the behaviour of river buffaloes (Bubalus bubalis), in confinement and in extensive conditions, also focusing on the effects of different housing and rearing conditions on their welfare. The behavioural repertoire expressed by buffaloes in extensive and intensive conditions is similar to those displayed by other domestic ruminants. However, through natural selection, buffaloes have also acquired several morphological, physiological and behavioural (i.e. wallowing) adaptations to hot climatic conditions. Buffaloes kept in intensive conditions and having no access to pasture and water for wallowing extend their periods of idling and are less often involved in investigative activities. Confinement is also associated with a reduction of space; however, no specific studies have been carried out to determine the specific requirements of this species. Space restriction can adversely affect various aspects of buffalo welfare, such as health (increased levels of lesions and injuries), social behaviour (increased number of agonistic interactions) and heat dissipation. The buffaloes, originating from tropical areas, are well adapted to large variations in food availability and quality, and to dietetic unbalances. As to human animal relationship, it has been observed that the incidence of stepping and kicking behaviour of buffaloes in the milking parlour is positively correlated with the frequency of oxytocin injections, whereas the frequency of positive stockperson interactions with the animals such as talking quietly, petting and gentle touching are negatively correlated with the number of kicks during milking. Data from farms where both dairy cattle and buffaloes are present show that avoidance distance measured in the pen is lower in buffaloes than in cattle. This may be attributed to the fact that buffaloes are generally recognised to be curious animals. Finally, the effects of different farming practices on animal-related indicators are described. However, these measures should be integrated into a monitoring protocol, such as the Welfare Quality® scheme, to reliably assess buffalo welfare in the current intensive farming conditions.     

Effects of cubicle characteristics on animal welfare indicators in dairy cattle

Cubicle characteristics such as cubicle dimensions or management factors such as cow-to-cubicle ratio could affect health and behaviour of dairy cows. The objective of this study was to estimate effects of cubicle characteristics on animal welfare indicators in dairy cattle. A total of 64 loose housing farms in Germany were assessed once during the winter housing period by one experienced assessor. Nearly 15% of the dairy cows had access to pasture during summer months for <6 h/day, whereas 85% were zero-grazing farms. Selected animal welfare indicators (duration of the lying down process, collisions of cows with cubicles, cows lying outside cubicles, cow cleanliness, integument alterations, lameness and subclinical mastitis incidence) of the Welfare Quality® protocol and cubicle characteristics such as cow-to-cubicle ratio and cubicle dimensions were recorded. Data were statistically analysed using a multiple linear regression approach. Pasture access and cubicle type were considered as potential influencing factors. Wider cubicles positively affected the proportion of dairy cows with dirty flanks (−18.5% per 10 cm increase) but increased the number of cows with severe integument alterations (+8.9% per 10 cm increase). Larger lying areas reduced the percentage of cows with dirty udders (−2.9% per 10 cm² increase). Longer distances from neck rail to curb were associated with higher prevalence of cows with dirty flanks (+3.1% per 10 cm increase) and subclinical mastitis incidence (+1.2% per 10 cm increase). With increasing neck rail height, the duration of the lying down process (−0.1 s per 10% increase), the percentages of cows with dirty legs (−8.4% per 10 cm increase), dirty udders (−7.0% per 10 cm increase) and severe lameness (−3.0% per 10 cm increase) decreased. Compared with rubber mat-equipped cubicles, deep-bedded cubicles showed a reduction in the lying down duration (−0.8 s), percentages of cows with dirty legs (−34.2%), dirty udders (−30.5%) and lesions and swellings (−13.1%). Compared with farms that did not provide any summer grazing, pasture access was associated with an increase of cows with severe lameness (+5.6%). Contrastingly, the number of cows with subclinical mastitis incidence was lower when cows had access to pasture in summer (−5.4%). Findings of the present study indicate several associations between cubicle characteristics and animal welfare in dairy cattle. Bedding type was found as the most influencing factor in terms of health and behaviour. Results of this study are valuable for farmers to identify the optimal cubicle design and improve the animal welfare level.     

Friends with benefits: Social support and its relevance for farm animal welfare

Despite growing interest in promoting positive welfare, rather than just alleviating poor welfare, potential measures of good welfare, and means to provide it, have remained elusive. In humans social support improves stress-coping abilities, health, and promotes positive psychological welfare. Therefore, social support may be a key approach to promote positive physical and psychological welfare in farm animals. However, the roles of positive social behaviors and social support have been overlooked in comparison to negative social behaviors such as agonistic interactions. The benefits of social partners on an animal's stress coping abilities and welfare are yet poorly understood.The purpose of this paper is to review the protective or buffering effects of social support against stressful challenges and its potential implications for farm animal welfare. The biology of social support is first presented with its behavior, endocrine, autonomic and immune components. The major factors modulating the social support process are then synthesized. Research and implications for animal welfare in different farm species are discussed. Lastly, this review identifies research areas that especially deserve further attention in our effort to understand and implement social support in farm animal welfare.Social support could constitute one of the foundations for welfare researchers to leap from the absence of negative welfare to the provision of positive welfare and emotional experiences.     

Current available strategies to mitigate greenhouse gas emissions in livestock systems: an animal welfare perspective

Livestock production is a major contributor to greenhouse gas (GHG) emissions, so will play a significant role in the mitigation effort. Recent literature highlights different strategies to mitigate GHG emissions in the livestock sector. Animal welfare is a criterion of sustainability and any strategy designed to reduce the carbon footprint of livestock production should consider animal welfare amongst other sustainability metrics. We discuss and tabulate the likely relationships and trade-offs between the GHG mitigation potential of mitigation strategies and their welfare consequences, focusing on ruminant species and on cattle in particular. The major livestock GHG mitigation strategies were classified according to their mitigation approach as reducing total emissions (inhibiting methane production in the rumen), or reducing emissions intensity (Ei; reducing CH₄ per output unit without directly targeting methanogenesis). Strategies classified as antimethanogenic included chemical inhibitors, electron acceptors (i.e. nitrates), ionophores (i.e. Monensin) and dietary lipids. Increasing diet digestibility, intensive housing, improving health and welfare, increasing reproductive efficiency and breeding for higher productivity were categorized as strategies that reduce Ei. Strategies that increase productivity are very promising ways to reduce the livestock carbon footprint, though in intensive systems this is likely to be achieved at the cost of welfare. Other strategies can effectively reduce GHG emissions whilst simultaneously improving animal welfare (e.g. feed supplementation or improving health). These win–win strategies should be strongly supported as they address both environmental and ethical sustainability. In order to identify the most cost-effective measures for improving environmental sustainability of livestock production, the consequences of current and future strategies for animal welfare must be scrutinized and contrasted against their effectiveness in mitigating climate change.     

国内外高校动物福利教育发展历史与前景展望

自20世纪80年代动物福利作为一个严谨的科学分支被人们接受以来,动物福利学科得到了迅猛发展,相关概念在科学研究结果的支持下被更精确地定义,评价动物所处福利水平的体系也逐步被建立起来.伴随着动物福利科学的发展,动物福利教育也逐步被纳入高等教育的课程体系.综述了国内外高等院校特别是高校的动物医学学院近30年间的动物福利教育进程,并以英国、美国、澳大利亚等国的动物福利教育为例说明国际动物福利教育的发展现状与趋势,结合中国现状,思考中国与国际动物福利教育之间存在的差异.随着动物福利相关科学、伦理学以及法学研究的深入,社会对动物福利的关注将会日益加深,相信至少对准动物医生普及动物福利教育必将成为发展的趋势.     

Inter-observer reliability of animal-based welfare indicators included in the Animal Welfare Indicators welfare assessment protocol for dairy goats

This study was conducted within the context of the Animal Welfare Indicators (AWIN) project and the underlying scientific motivation for the development of the study was the scarcity of data regarding inter-observer reliability (IOR) of welfare indicators, particularly given the importance of reliability as a further step for developing on-farm welfare assessment protocols. The objective of this study is therefore to evaluate IOR of animal-based indicators (at group and individual-level) of the AWIN welfare assessment protocol (prototype) for dairy goats. In the design of the study, two pairs of observers, one in Portugal and another in Italy, visited 10 farms each and applied the AWIN prototype protocol. Farms in both countries were visited between January and March 2014, and all the observers received the same training before the farm visits were initiated. Data collected during farm visits, and analysed in this study, include group-level and individual-level observations. The results of our study allow us to conclude that most of the group-level indicators presented the highest IOR level (‘substantial’, 0.85 to 0.99) in both field studies, pointing to a usable set of animal-based welfare indicators that were therefore included in the first level of the final AWIN welfare assessment protocol for dairy goats. Inter-observer reliability of individual-level indicators was lower, but the majority of them still reached ‘fair to good’ (0.41 to 0.75) and ‘excellent’ (0.76 to 1) levels. In the paper we explore reasons for the differences found in IOR between the group and individual-level indicators, including how the number of individual-level indicators to be assessed on each animal and the restraining method may have affected the results. Furthermore, we discuss the differences found in the IOR of individual-level indicators in both countries: the Portuguese pair of observers reached a higher level of IOR, when compared with the Italian observers. We argue how the reasons behind these differences may stem from the restraining method applied, or the different background and experience of the observers. Finally, the discussion of the results emphasizes the importance of considering that reliability is not an absolute attribute of an indicator, but derives from an interaction between the indicators, the observers and the situation in which the assessment is taking place. This highlights the importance of further considering the indicators’ reliability while developing welfare assessment protocols.