Rodent management presents a major but under-recognized animal welfare challenge in Nigeria, particularly within small- and medium-scale poultry and farming systems. In many settings, rodent infestations are addressed primarily through indiscriminate use of toxic rodenticides, including highly concentrated anticoagulants and, alarmingly, repurposed human pharmaceuticals such as indomethacin. Evidence from field assessments in Kwara State demonstrates that these practices result in severe and prolonged suffering in target rodents through internal hemorrhaging, while also causing substantial non-target harm to poultry, domestic animals, wildlife predators, and humans through secondary poisoning and occupational exposure. Drawing on two grant-supported interventions implemented by RodentPlus Conservation Initiative, this paper highlights both the scale of the welfare problem and emerging humane alternatives. The first intervention involved training 70 poultry farmers across three production clusters in Kwara State on Ecological-Based Rodent Management (EBRM). This approach prioritizes prevention, exclusion, sanitation, and rapid-kill mechanical trapping over poison use. Through pre- and post-training evaluations, we seek to reduce reliance on inhumane rodenticides, improve poultry health and biosecurity, and increase farmer awareness of animal suffering and ethical pest control. Importantly, EBRM directly addresses welfare by avoiding slow-acting poisons and minimizing prolonged distress in rodents, while simultaneously protecting non-target species. The second intervention explores fertility control as a non-lethal alternative for managing populations of *Mastomys** **natalensis*, a major agricultural pest and Lassa fever reservoir. Through baseline ecological surveys, controlled field trials of oral contraceptive baits, and community engagement in Lassa fever-endemic areas of Ondo and Kwara States, this project seeks to reduce rodent population growth without causing pain, distress, or mortality. Rigorous assessment of non-target risks and ethical safeguards underpins the work, positioning fertility control as a potentially transformative welfare-positive strategy for rodent management in low-income settings. Together, these projects illustrate how animal welfare science can inform context-appropriate, humane pest management in the Global South. Addressing rodent welfare alongside poultry, wildlife, and human health offers a more ethical and sustainable pathway than conventional lethal control, with implications for policy, farmer education, and future welfare-focused innovation in Nigeria and similar regions.

Surgical castration of beef calves is a routine management practice in Argentina, yet when performed without analgesia, it represents a major welfare concern due to the acute nociceptive and affective responses it elicits. The procedure activates the autonomic nervous system and the hypothalamic-pituitary-adrenal axis, disrupting homeostasis and altering behavioral and emotional states. This study aimed to evaluate the effects of three practical analgesic protocols—lidocaine ring block (L), systemic flunixin meglumine (FM), and their combination (L+FM)—on behavioral indicators of pain intensity, emotional state, temperament, and short-term productive performance in Angus calves undergoing surgical castration. Fifty-one Angus calves from the National University of La Plata, aged 238±40 days, were enrolled in a blocked 2x2 factorial design. Ethical approval was granted by the Animal Use Committee (reference 001-25). Animals were randomly assigned to four treatments: no drug (ND), 2% lidocaine ring block (L), flunixin meglumine (1mg/kg IM) (FM), or the multimodal combination (L+FM), alongside a control group of intact bulls. Lidocaine onset was verified via sensitivity testing, ensuring a 1-5 minute latency before performing an open scrotal procedure. Behavioral and affective assessments were conducted at six time points (0, 0.5, 1, 2, 24 h, and 15 days post-castration) using Temperament Scores (TS), pain-related body language (BL-Pain), and an emotional-state scale (ESS). Live weight was recorded using electronic scale to calculate immediate weight change and short-term gain. Analgesic treatment had a significant effect on BL-Pain (p=0.0011), with all protocols markedly reducing pain scores compared to the ND group (54.95±8.42). The L+FM protocol produced the lowest scores (7.18±2.08), though differences between active drug treatments were not statistically significant (p>0.90). While TS did not differ during the procedure, a significant difference emerged at 24 hours (p=0.048), suggesting residual arousal or persistent inflammatory processes. Immediate weight loss (~10-12 kg) at 24 hours was uniform across all castrated treatments regardless of analgesic intervention (p=0.967). Only intact bulls showed a significantly steeper growth trajectory compared to castrated groups (p=0.023). Strong correlations were observed between the "tense" ESS rating and BL-Pain (r=0.64). In conclusion, analgesic protocols are essential for mitigating acute behavioral indicators of pain during surgical castration. While short-term weight loss appears governed by systemic stress responses rather than momentary nociception, the clear behavioral benefits and improved animal comfort justify the routine implementation of analgesia as a welfare and ethical imperative in beef production systems.

Accurate assessment of cattle welfare is essential for ensuring ethical, sustainable, and productive livestock systems. Among the range of behavioural and physiological indicators available, vocalization has gained increasing attention as a sensitive, non-invasive tool for evaluating welfare in cattle. This review explores the relevance of cattle vocalizations in welfare assessment, examining their biological significance, associated affective states, and potential applications in both research and on-farm settings.Cattle vocalizations play an important role in social communication and are closely linked to emotional and physiological states. Variations in vocal characteristics such as call frequency, duration, intensity, and acoustic structure have been associated with key welfare-related conditions. Increased vocalization rates and higher-pitched calls have been observed in situations of stress, frustration, hunger, pain, and social separation, including during weaning, transport, restraint, and milking. Conversely, reduced vocal activity or low-frequency contact calls may be associated with calmer states and social cohesion within the herd. These vocal responses provide insight into cattle’s subjective experiences and complement traditional welfare indicators.This review synthesizes current research linking specific vocal parameters in cattle to negative and positive welfare states, with consideration of age, sex, production system, and social context. Calf vocalizations, particularly during maternal separation and weaning, have been extensively studied and are highlighted as a robust model for understanding affective communication in cattle. Advances in bio-acoustic analysis and precision livestock farming technologies have further strengthened the applicability of vocal measures, enabling automated detection and classification of cattle vocalizations in real time. Such technologies offer promising opportunities for continuous welfare monitoring, early detection of welfare compromise, and improved management interventions. However, several challenges remain in the interpretation and application of cattle vocalizations for welfare assessment. Environmental noise, group housing, individual variation, and context-dependent meaning of calls can complicate analysis. Moreover, vocalizations alone may not fully capture welfare status and should be integrated with other behavioural, physiological, and health indicators for a holistic assessment. Standardization of recording methods and validation across different farming systems are also needed. In conclusion, cattle vocalizations represent a valuable and under-exploited indicator of welfare. When combined with modern analytical tools and multidisciplinary approaches, vocalization-based assessment can significantly enhance our ability to understand and improve cattle welfare, supporting evidence-based practices.

Stocking density is a critical factor of animal welfare in small ruminant production systems. Inadequate space allowance can increase competition, restrict expression of natural behaviour, and increase risks of injury and stress. This study aimed to quantify effective space allowance and resource availability in commercial sheep and goat farms in the region of Eastern Macedonia and Thrace, Greece. Data was collected from 18 farms. Housing conditions were evaluated at pen level (n=91). For each pen, animal category and number of animals were recorded, pen dimensions were measured, and the area occupied by feeders and water troughs subtracted to calculate effective floor area per animal. Feeding and watering space per animal were estimated based on respective dimensions. Effective floor space per animal averaged 2.4 ± 1.30 m2 for males, 1.8 ± 0.86 m2 for adult females and 1.5 ± 1.00 m2 for primigravid animals. Those values were compared with recommendations in the available literature considering that EU Directive 98/58/EC does not provide species specific minimum values. The evidence in the literature suggests typical benchmarks ≥2.0–2.5 m² per adult animal depending on category, body weight, and production system. Housing units that did not meet these benchmarks accounted for 66.6% of male pens, 68.7% of adult females pens, and 38.0% of primigravid pens with average space deficit of 0.59 m2, 0.63 m2, and 0.23 m2 per animal, respectively.  Access to water varied significantly compared to recommended allowance of ≥30-45 cm per animal; mean space in water trough was 12.2 ± 6.35 cm per male, 3.8 ± 4.05 cm per female and 5.0 ± 3.80 cm per primigravid animal. Insufficient feeding space availability for concentrates and roughages was observed in 68.1% and 39.5% of pens, respectively. The study, although regionally focused, reveals variable deficiencies in stocking density and resource provision in studied farms. The results highlight stocking density and resources availability as critical welfare challenges in commercial small ruminant farms in this area of Greece. This study was supported by external research funding (ELKE-AUTH project code: 76290).

While UFAW is celebrating its centenary in 2026, Wild Animal Initiative is more modestly celebrating its seventh anniversary. As Wild Animal Initiative aims to extend the scope originally covered by animal welfare science to include wild animals, this age difference between both organizations likely reflects the many decades that were necessary to explicitly recognize wild animal welfare for its own sake. Moreover, while the impact of human activities on wild animal welfare has long been considered (e.g., humane methods of trapping rabbits and rats advocated by UFAW as early as the 1930s), wild animal welfare science goes beyond these initial considerations by also covering the natural experiences of free-living wild animals, which can be both negative (e.g., suffering because of diseases, starvation, or harsh weather conditions) or positive (e.g., enjoying life when being healthy, satiated, or exposed to mild conditions). Paradoxically, even though many animals undoubtedly have subjective experiences in their natural environment, very little research has been conducted so far to better assess and understand these experiences. During this talk, we will discuss the recent growth of wild animal welfare science, and show how this new field can constructively build on the foundations already laid by animal welfare science (e.g., using pre-existing theoretical frameworks to assess affective states). We will also illustrate the specific challenges faced by wild animal welfare science (e.g., heterogeneity of experiences between species, variable welfare assessments between species) and show that if these can be overcome, wild animal welfare science is likely to be extremely impactful (given its current overlook contrasting with the countless number of animals living in the wild). Importantly, the field of wild animal welfare science has already started to take off, in part thanks to the support and efforts provided by Wild Animal Initiative since its inception (e.g., funding of > 80 external wild animal welfare research projects, publication of ca. 20 wild animal welfare articles by our grantees and ourselves). Hopefully, these promising early signs will accumulate and consolidate in the future, which should ultimately provide us with a clearer understanding of wild animal welfare before Wild Animal Initiative celebrates its own centenary.

In the European Union’s Common Agricultural Policy, farmers can receive financial support if they implement animal welfare friendly management procedures. This support has gained importance in the programming period 2014-2022 with public expenditures rising from 55 million Euro in 2014 to a total of 3.7 billion Euro in 2022*. Although support for animal welfare is becoming increasingly important, analyses of its effectiveness are still scarce. Our aim is to contribute to this area of research by evaluating the animal welfare measure 'Grazing' in the German federal state of North Rhine-Westphalia. We analysed the effect of the measure on animal health using milk recording data. To establish causality, we applied a panel event study design using two-way fixed effects regressions including year as well as farm-level fixed effects. The indicators included in the analysis were: - mortality, - longevity, - proportion of cows with < 100,000 somatic cells as an indicator for cows without mastitis and - proportion of cows with a milk to fat ratio > 1.5 as an indicator for metabolic disorders. We found no effect of participation on the animal welfare indicators investigated (see Fig. 1). One reason for this result is that supported farms often do not change their management practices in order to comply with the requirements. Although we found no causal effects of participation, the values of the supported farms indicator differ from those of the control group, suggesting structural differences between the groups (see Fig. 2). Overall, the indicators show positive developments, with the exception of mortality. In line with the meta-analyses of Arnott et al. (2017) and Mee and Boyle (2020), supported farms have lower mortality but a higher proportion of cows with metabolic disorders. However, contrary to the cited literature, the supported farms have a lower proportion of cows without mastitis. The animal welfare measure “Grazing” does not induce changes in the observed indicators. It supports farms with better indicator results for mortality and longevity but worse results for mastitis and metabolic health. We conclude that: - To induce changes, the payment of a premium is not sufficient but other instruments such as the tightening of legislation or the addition of health parameters to labelling schemes are needed. - While access to pasture has many positive effects on animal welfare, some of which we were unable to investigate due to a lack of data, it is not a solution to all animal welfare issues.

Ensuring that all animals in a country experience the legally mandated minimum level of welfare requires a well-functioning animal welfare control system. In this context, consistency in inspectors’ assessments, decisions, and actions is decisive. However, achieving such consistency is highly challenging, as the legislation requires complex assessments. Furthermore, inspections are carried out by inspectors with diverse professional backgrounds and experiences, and take place across a wide range of settings with substantial contextual variation. In this mixed-methods study, we examined how Swedish animal welfare inspectors perceive the current level of consistency within the official animal welfare control. In addition, challenges related to achieving a reasonable level of consistency were identified, as well as strategies that might improve consistency. Two datasets were collected: (1) a quantitative component consisting of a digital questionnaire completed by 119 animal welfare inspectors, and (2) a qualitative component comprising twelve semi-structured interviews with animal welfare inspectors. The data were analysed using a convergent mixed-methods design, and the qualitative and quantitative data were integrated using joint displays. Our analysis shows that consistency was not understood by inspectors as ‘always doing the same’ or as algorithmic case handling. Rather, consistency was perceived as a shared professional framework within which inspectors make contextual and situation-based assessments and decisions. Inspectors generally perceived the level of consistency to be sufficient at the local level (within offices or regions), but insufficient at the national level (between offices or regions). Challenges to achieving consistency included a lack of structured inter-office discussions, differences in inspectors’ experience and training, limited resources, insufficient national guidance and support from the Central Competent Authority (CCA), and differences in delegation of authority (delegation schemes). Several strategies perceived by inspectors to enhance consistency were identified. Of these, regular peer discussions, paired inspections, and clear guidance from the CCA were highlighted as the most effective. The findings highlight the complexity of animal welfare inspectors’ work and underscore the crucial role of context and professional judgement in animal welfare inspection assessments and decisions.

Impairments in reward learning have been observed in human mood disorders and may represent a trans-species ‘behavioural biomarker’ of affective state relevant for animal welfare science. Mice can discriminate between rewards varying in either quantity, quality, or probability, but the comparative strength and rate of discrimination between rewards of varying magnitudes has not been directly compared. The effect of spatial separation of simultaneously presented rewards was also uncertain. This study therefore investigated reward learning rates in mice using different reward contingencies varying in magnitude and/or by spatial location. 24 CD1 mice were trained in modified home cages (1hr/d for up to 30d) using automated reward dispensing devices (FED3, n=2/cage) operated by nose-poke sensors (n=2 sensors/device). Mice were assigned to 1 of 4 variations in reward magnitude/spatial separation, differing either in quality (chow pellets vs. reward pellets; SQUAL), probability (0.2 vs. 0.8; SPROB), quantity across 2 devices (1 vs. 3 reward pellets; SQUAN), or quantity differing across sensors on the same device (1 vs. 3 reward pellets; LQUAN). Reward contingency significantly predicted bias towards the more highly rewarded option (% total nose pokes directed at highly reward option; Figure 1) and the rate at which subjects learnt to successfully discriminate between high and low reward options (number of sessions required for subjects to exceed ‘bias threshold’ where highly rewarded option received >50% total nose pokes; Figure 2). These results indicate that the type of variation in reward magnitude, as well as the spatial separation of simultaneously presented rewards, may influence discriminability. Results also suggest that reward magnitude differences varying in either quality, probability, or quantity when in close spatial proximity, may be used in studies designed to compare reward learning abilities, though the rate of discrimination learning between rewards differing in quality (SQUAL) may be too great for use in research where learning rate is to be compared. Though fewer mice that received the more spatially separated quantity variation (SQUAN) learnt to favour the more highly rewarded option, it is possible that the number of test sessions was not sufficient to detect a potentially slower learning rate rather than reflecting an inability to learn this contingency. Further, though only compared across one reward magnitude variation (quantity), the difference in learning rate at different levels of spatial separation suggests that the location of simultaneously presented rewards may be an important consideration for future research. Figure 1. Figure 2.

Animal welfare legislation presumes clear standards, assignable responsibilities, and enforceable compliance. In practice, however, animal welfare offences, maybe particularly in animal transport, unfold in situations marked by time pressure, conflicting interests, and profound practical constraints. This paper examines how animal welfare is negotiated and enforced in such contexts through a socio-legal analysis informed by the One Welfare framework. The concept of One Welfare emphasises the interconnectedness of animal welfare, human welfare, and societal well-being. While endorsed at a conceptual level, the framework is less often applied to the practical realities of animal welfare enforcement. The paper draws on courtroom observations of Danish animal welfare cases concerning animal transport, conducted as part of the interdisciplinary research project *ANIMALAW – An interdisciplinary study of Danish animal welfare offences in animal transport*, and presents examples from court cases. The paper addresses the question: How is animal welfare enforcement practically enacted in situations characterised by conflict, time pressure, and competing obligations, and what do these structural conditions imply for how responsibility is produced and assessed in animal welfare law? Across cases, the analysis identifies a recurring figure: an individual actor positioned in what participants themselves describe as “the art of the possible in an impossible situation.” These actors are required to make immediate decisions balancing animal welfare considerations with legal obligations, evidentiary requirements, economic pressures, and interpersonal conflict. While animal welfare law formally centres on the welfare and protection of animals, responsibility is in practice concentrated on these frontline actors, whose actions can become the focal point of legal evaluation. Applying a One Welfare perspective, the paper demonstrates how animal welfare outcomes cannot be addressed in isolation, as human welfare (including stress, professional risk, and moral responsibility) and societal welfare (such as legal legitimacy, regulatory transparency, and trust in enforcement) shape welfare outcomes for animals. The analysis shows how legal frameworks both rely on and obscure the human dimensions of animal welfare enforcement. The paper argues that understanding animal welfare enforcement as a socio-legal practice is essential for developing more robust, legitimate, and humane systems of animal welfare governance. Animal welfare enforcement cannot be adequately understood or improved without acknowledging the structural conditions under which responsibility is enacted.

The effect of increasing automation of the dairy sector on cow welfare is hotly debated. Some expect positive effects (as time freed up by automation can be devoted to improving welfare). Conversely, others fear that automation reduces opportunities to detect welfare problems by reducing stockman-herd interactions. To study this in commercial practice, we focused on one type of automation: automated milking using robots placed in the cow house, meant to be voluntarily visited by cows without human intervention. We evaluated the welfare of seven robot herds and thirteen parlour herds during the housing season, using the Welfare Quality® protocol. We report on nine out of the eleven criteria included in the protocol. We excluded the criteria ‘absence of disease’ (because data is currently incomplete) and ‘ease of movement’ (because herds were loose-housed on all farms, thus obtaining the maximum score). Statistical analysis was performed using linear models with farm type (robot vs. parlour) as the independent variable and the criterion score (ranging from 0 (worst) to 100 (best)) as the dependent variable. The only two criteria that differed significantly between robot farms and parlour farms were ‘absence of thirst’ (better on robot farms; means±SEM parlour: 37±7, robot: 67±9, p=0.01) and ‘expression of other behaviour’ (better on parlour farms; 60±10 vs. 11±12, p=0.01). These two criteria respectively reflect the level of access to / cleanliness of drinking water, and the number of days per year with pasture access. None of the other criteria were significantly affected by farm type (p>0.10). Numerically, differences between means were small and not uniformly in favour of one system over the other (‘absence of hunger’: 81±5 vs. 80±7 , ‘comfort around resting’: 39±4 vs. 33±5, ‘absence of injuries’: 61±4 vs. 57±5, ‘absence of pain’: 73±3 vs. 75±5, ‘expression of social behaviours’: 77±3 vs. 84±5, ‘good human-animal relationship’: 51±4 vs. 48±5, and ‘positive emotional state’: 72±6 vs. 84±8). Our results suggest that robotic milking does not result in an overall improvement or deterioration of cow welfare during the housing season. However, robotic milking was associated with low pasture access. Pasture access improves several aspects of welfare during the grazing season (as shown in our other abstract submitted for this conference). Thus, robotic milking may indirectly reduce cow welfare, unless more attractive ways to combine grazing and robotic milking are designed.

The early-life environment has a profound influence on offspring development. More complex environments, for example, may enhance cognition and growth. However, there is limited research on how early life conditions influence individuals across socio-cognitive domains. The aim of this study was to investigate socio-cognitive differences between pigs housed in barren and enriched conditions. Twenty-two male pigs were studied from birth until 12 weeks of age in tests related to spatial cognition (maze, spatial holeboard), intra- and inter specific social cognition (social foraging paradigm and human-help-seeking), cognitive flexibility (reversal phase in spatial holeboard), and behavioural strategies. Pigs were housed in either barren conditions with a concrete floor (B, n = 13) or in enriched conditions (E, n = 9) with straw bedding and twice weekly a novel item. At 12 weeks of age, pigs were sacrificed and five major brain regions were analysed for DNA methylation using the GBS-MeDIP method to assess epigenetic profiles. From weaning, E pigs were growing faster (p=0.03), resulting in a 4.4 kg difference at 12 wk age. In a 3-min sow-piglet foraging test, orientation to the sow  (B13, E21 times) or feed items did not significantly differ between pigs from different housing conditions. However, in a 1-min foraging paradigm with a human, E pigs oriented less to the human than B pigs (10 vs 16 sec; p = 0.04). In the spatial holeboard task, housing conditions did not influence the number of trials to reach a reference memory of 0.8 (15 trials) or the duration to complete the task (24 sec). In the reversal phase, housing conditions did not significantly influence cognitive flexibility. In the spatial holeboard, B pigs had a more consistent behavioural strategy in obtaining rewards, with more consistent spatial patterns (B 0.64 vs. E 0.47% similarity in patterns). Pigs from enriched environments had a higher epigenetic variability in the nucleus accumbens, hippocampus, frontal cortex and hypothalamus, suggesting increased epigenetic plasticity. The opposite was found for the amygdala, with low variation in E pigs and high variation in B pigs. These preliminary results suggest that housing conditions result in socio-cognitive differences which are not necessarily reflected in a single test. Hence, using measures across multiple domains assists in gaining a better understanding of the widespread implications of how we house animals.

Animal-based welfare indicators (ABWIs) are foundational to assessing animal welfare and supporting the core concept of refinement (3Rs: replacement, reduction, and refinement) when working with animals in research and teaching. The selection and application for ABWIs are core components in the monitoring of animals. Given their importance, the current consensus in ABWIs for routine monitoring for mice, rats, pigs, poultry, sheep, horses, cattle, fish, and rabbits are limited, inconsistent, or unavailable. This potentially creates confusion and concerns around variability, confounding effects, and inconsistencies when monitoring the experimental and welfare outcomes of animals. The absence of consensus in ABWIs highlights a gap across the animal care and use industry with a potential lack of consistency in training and competency, regulatory enforcement, policy and legislation, guidelines and standards, and auditing and welfare assessments. This is particularly relevant to the care of animals and their use in teaching and research. Establishing consensus to harmonise the use of timely, practical, contextually, and species appropriate ABWIs has the potential to mitigate some of these issues. As part of this research, a novel approach was developed by blending aspects of a modified Delphi method and co-design to identify and harmonise ABWIs to support better animal welfare and research outcomes. The method was trialled across animal research and teaching industry in Australia and New Zealand. Participants consisted of researchers (including postgraduate researchers), academics, teachers, animal care staff, veterinarians, and animal ethics staff. Sessions consisted of a 90-minute in-person facilitated workshop and 3 rounds of consensus building followed by expert thematic analysis. There was a total of 8 workshops with 24 groups and 232 participants. Consensus in use of ABWIs for macaque, fish, poultry, pig, equine, rabbits, cattle, mice, sheep, and rats and nine overarching harmonised ABWIs were identified. A further suite of ABWIs were generated for potential use which included general and species-specific indicators. The use of a novel approach to finding consensus in ABWIs has the potential to offer well-defined, consistent, practical, evidence-based recommendations for policy makers, animal managers and carers, accreditation standard owners, auditors, regulators, scientists, educators and others in the animal industry to support better animal welfare.

Automatic milk feeders provide the opportunity to wean calves with autonomy over meal size and frequency, which may positively affect weight gain and milk feeding behaviors, such as milk intake, drinking speed, and rewarded and unrewarded visits to the milk feeder. To explore this, female Holstein calves were housed in 3 group pens (n=50-55 calves/pen; 4m2/calf) with access to automatic milk feeders (25-28 calves/teat) and ad libitum concentrate. Calves were weighed at d0 and weekly until complete weaning at d63 of age. On d35, calves were weighed and assigned in a randomized block design balanced by weight to one of three weaning treatments that differed in daily meal size and frequency (Small: 5 meals/d, n=44; Large: 2 meals/d, n=41; and Unlimited: any number of meals until daily allowance reached, n=41). During pre-weaning (d0-d34) calves had ad libitum milk allowance; during weaning, for all treatments, milk was gradually reduced from ad libitum to 10L/d (d35-d42), held at 10L/d for 13d, then gradually reduced until complete weaning on d63. Milk feeding behaviors and weight gain were analyzed using mixed linear regression models (covariates: birthweight, pre-weaning illness; random effect: calf nested within pen). Log-transformed variables are reported as back-transformed means and 95% CI. During pre-weaning, calves averaged 8.52±0.17L/d of milk and averaged 0.96±0.03kg/d. During weaning, treatment did not affect milk intake (7.9±0.12L/d; *P=*0.11) nor average daily gain (1.0±0.02kg/d, *P=*0.14). However, during weaning, drinking speed was greater (or tended to be greater) for Unlimited (477±8mL/min) compared to Large (432±8mL/min; *P**<*0.001) and Small (457±8mL/min; *P=*0.08), and was greater in Small compared to Large (*P=*0.02). Rewarded visits were greater for Unlimited (6.4 visits/d [6.2,6.6]) compared to Small (5.8 visits/d [5.6,6.0]) (*P*<0.001) and Large (4.8 visits/d [4.7,5.0]) (*P**<*0.001), and greater for Small compared to Large (*P*<0.001). Unrewarded visits were greater for Small (3.8 visits/d [3.5,4.2]) and Large (3.7 visits/d [3.7,4.0]) compared to Unlimited (2.7 visits/d [2.5,3.0]) (*P*<0.001) but did not differ between Small and Large (*P*=0.84). These results suggest that when given autonomy over meal patterns, calves consume smaller, more frequent meals during weaning and perform fewer unrewarded visits. While these results apply to calves managed at high milk allowances and stocking rate/teat, it provides practical insight for designing milk-feeding plans in automatic feeder systems to manage hunger-related behaviors without compromising growth.

In domestics pigs, tail postures is an early indicator of tail biting risk used in both scientific research and farms. Most recently, tail postures and motions have also been explored as indicators of the emotional state of the animal making them promising measures to study positive welfare states. To validate their use, research on the reliability of these indicators is still needed. The aim of this study was to evaluate the inter-rater reliability (IRR) in the observation of pig tail postures and motions. Forty-eight partially docked Duroc pigs (castrated males and females, ≈160 days of age), reared on concrete slatted floor in groups of six, were assessed. Two observers conducted independently but simultaneously three scan-sampling sessions on three consecutive workdays; each sampling session consisted of seven scans, every 10 minutes, carried out between 11 AM and 12 AM. For each scan, pigs were assessed sequentially. Tail postures and motions were scanned instantaneously for each pig. For the posture, an ethogram was used to assess tails held upwards (“up”), sideways (“sideways”), fully or mostly aligned with the spine (“erect horizontally”), pointing downward (“hanging”) or kept near the body (“tucked”). The ethogram also assessed stationary (“motionless”), wagging (“loosely” and “intense wagging”), briefly wagging (“flicking”), rotating (“circling”) or contracting (“jamming”) tail motions.  Scans per category were then summed up. For the IRR, intraclass correlation coefficients (ICC) were calculated for each assessed posture and motion. We found that the IRR was good for the “up” (ICC=0,79) and “erect horizontally” (ICC=0,69) postures. Reliability was acceptable for the “hanging” posture (ICC= 0,50) and weak for the “sideway” posture (ICC= 0,38). Due to the low frequency of the “tucked” posture, IRR could not be determined. Regarding tail motion, the IRR was excellent (ICC = 0,93) for “motionless” tails. The reliability of “loosely wagging” and “intense wagging” tails was weak (ICC = 0,396 and ICC = 0,319, respectively). The other categories of tail motions (i.e. “jamming”, “circling” and “flicking”) were too rare to calculate their IRR. The good reliability of the “up” and “erect horizontally” postures, as well as the “motionless” tails likely reflects the fact that they were easily identifiable and frequent. In contrast, low reliability for less frequent or more ambiguous categories suggests that future assessments can be improved by increasing the number of scans or the sample of animals assessed, and consider continuous recording to capture brief or rare tail motions more consistently.

To enhance animal welfare, it is essential to understand how animals perceive their environment, particularly through their sensory systems. Olfaction plays a central role in social interactions and is involved in social recognition, mother–young bonding and transmission of information related to individual identity, social status and emotional state. Contrary to the numerous studies on human emotional odors, very few studies focused on farm animals’ emotional odors and their possible effects on behavior and welfare. Sheep remain unstudied. Therefore, the present study aimed to assess whether sheep are able to discriminate between neutral and stress-related odors emitted by conspecifics. All the experiments were conducted on French commercial sheep farms. Emotional odors were collected from ewes through their sweat, secreted by the apocrine glands, using cotton pads rubbed for 10 seconds on the inguinal area during two events: shearing (sitting position; a well-known stressful event; n=33) for stress odor (SO) and gathering (standing position; n=30) for neutral odor (NO). These odors were subsequently presented to unfamiliar ewes in two tests: a small-group test (lasted 2 minutes) and a large-group test (lasted 8 hours). The small-group test involved presenting a single odor (SO or NO), placed in a bucket in a test pen and presented to groups of three ewes (n = 17 groups) to assess their exploratory, locomotor, and social behaviors. The large-group test involved one group of 18 ewes tested in their home pen, which was divided into three odor zones (SO, NO, no odor), using to perforated boxes containing odor-impregnated cotton pads, allowing continuous diffusion. We assessed ewes’ spatial distribution, avoidance or preference behaviors. In the small-group test, exposure to SO resulted in a significant reduction of exploratory and locomotor behaviors, accompanied by an increase in social interactions, suggesting a shift in behavioral priorities toward social proximity and vigilance. In the large-group test, ewes showed a significant preference for the no odor zone and avoided the SO zone, indicating sensitivity to conspecific olfactory cues and safety-seeking behavior modulated by flock social dynamics. Overall, these results demonstrate that stress-related and neutral odors modulate both individual behavior and collective dynamics in ewes, suggesting that ewes are able to discriminate between emotional odors emitted by conspecifics. This study emphasizes the role of olfactory cues in social regulation and highlights new perspectives for animal welfare.

Background. Management strategies that enhance agency and provide rewarding experiences, such as outdoor access and environmental enrichment, are increasingly recognised as key components of positive animal welfare. Blood-based physiological markers and gut microbiota composition may serve as complementary indicators responsive to housing conditions and environmental complexity. Aim. Evaluating whether outdoor access and/or environmental enrichment are associated with changes in faecal microbiota diversity and blood-based inflammatory and stress-related biomarkers in sheep. Methods. Twenty-four adult sheep (*Ovis **aries*) from the same flock were allocated to four groups (n = 6/group) for 14 weeks: CTRL (indoor control), ENI (indoor + enrichment), OD (outdoor access), and ENO (outdoor access + enrichment). Enrichment increased environmental complexity and was rotated every two weeks. Faecal and blood samples were collected at T0 (start), T3 (midpoint), and T6 (end). Faecal DNA was extracted and 16S rRNA gene amplicon sequencing performed (Illumina NextSeq 1000, 2×300 bp). Reads were processed using DADA2 and taxonomy assigned against the SILVA 138 database. Alpha diversity was analysed using Mann–Whitney tests. Serum TNF-α, IL-6, haptoglobin, cortisol and IFN-γ were quantified, and analysed using mixed-effects models (REML) with time as a repeated factor. Results. Sequencing generated ~16.5 million raw reads (mean ~686k/sample), yielding 10,841 ASVs after processing. From T0 to T6, CTRL remained relatively stable, whereas ENI and OD showed significant increases in alpha diversity. Observed Features increased in ENI (p = 0.039) and OD (p = 0.039), and Faith’s PD increased in ENI (p = 0.039) and OD (p = 0.012). ENO showed a similar trend without statistical significance. For blood biomarkers, TNF-α showed a significant effect of time (p < 0.0001), with no effect of housing condition or time x condition interaction. IL-6 and haptoglobin remained low with substantial inter-individual variability and no consistent group differences. Cortisol concentrations remained within physiological ranges. Variability in inflammatory markers tended to decrease over time, suggesting physiological adaptation. Conclusions. Outdoor access and enrichment were associated with increased faecal microbiota alpha diversity compared with indoor controls, supporting microbiota profiling as a complementary welfare indicator. In contrast, systemic inflammatory markers were influenced primarily by temporal factors rather than housing system, suggesting adaptive physiological modulation rather than chronic inflammatory stress. Integrating longitudinal biomarker and microbiota data may enhance welfare assessment sensitivity in sheep. Ethical authorisation: All procedures complied with National Implementation of EU Directive 2010/63 authorisation no. 159/2024-PR. Project was co-funded by the European Union’s Horizon Europe Project 101136346 EUPAHW

Laying hens are motivated to seek out dark, secluded places to lay, such as nest boxes. Aviaries provide hens with nest box access, but accessibility is an important factor that determines their use. Genetics may also play an important role in mitigating nest box use, with studies suggesting white strains lay more eggs in nest boxes than brown strains, potentially due to differences in body size and mobility. To investigate the effects of aviary design and genetic strain on daytime usage of nest boxes, we randomly assigned 320 Hy-Line Brown and 320 Hy-Line White hens, by strain, to standard (3-tier, metal round perches) or enhanced (3-tier, plastic mushroom shaped perches, ramps) aviaries (4 pen replicates per housing and strain combination; 40 hens per pen). We recorded the number of hens on each aviary tier using instantaneous scan sampling with 15 min intervals conducted during a 2 h morning and 2 h afternoon period on two consecutive days per week at 23, 28, 33, 37, 43, and 48 weeks of age. We noted the locations (nest box, aviary tiers, floor, other) of all eggs on one day per observation week. Data were analyzed via generalized linear mixed models with binomial distributions using R software. Regardless of age and strain, more hens (P < 0.001) used the nest box tier (tier 2) in the morning (proportion of hens/observation ± SE: 0.057 ± 0.003) than the afternoon (0.014 ± 0.001), and if housed in enhanced vs. standard aviaries (*P* < 0.001; brown/enhanced: 0.03 ± 0.003, brown/standard: 0.007 ± 0.001; white/enhanced: 0.072 ± 0.008, white/standard: 0.042 ± 0.005). The nest box tier was used by more white vs. brown hens at all ages in both aviary types. Nest box tier use increased over time for white but not brown hens. Brown hens laid more nest box eggs when housed in the enhanced vs standard aviary (enhanced: *P* < 0.001; 0.894 ± 0.052) and there was a tendency for this pattern in white hens (*P = *0.057; 0.839 ± 0.072). Under standard housing, white hens laid more nest box eggs than brown hens (*P *< 0.001; white: 0.554 ± 0.131; brown: 0.057 ± 0.031). Overall, the provision of ramps promoted use of the nest box tier and substantially improved nest box eggs laying by brown hens. These results highlight the importance of considering strain-specific needs in aviary design.

Stereotypies are repetitive behaviours commonly manifested in restrictive environments, in several species. In horses, behaviours like wood-chewing and cribbing often represent lifelong challenges and become established as quickly as within one month following artificial weaning, characterised by early and/or sudden rupture of mare-foal bonds. Detection of foals predisposed to performing such behaviours is valuable for investigating risk factors and establishing preventive measures, fostering resilience and promoting better welfare. Towards this goal, in this study 62 Brasileiro de Hipismo foals, 34 females and 28 males, were observed at artificial weaning for assessments of individual variability in expression of repetitive object-directed oral behaviours. Foals were weaned in groups (7 groups of 6 to 15 foals) at ages 186 to 304 days by confinement in a 167.5 m² pen for 28 to 52 hours, under routine husbandry conditions, and filmed continuously. A preliminary observation phase was conducted to develop a behaviour analysis protocol before applying to the full cohort. Ten foals were observed continuously for 24 hours and all occurrences of object-oriented oral behaviours were recorded in start-end timestamps, frequency, duration and targets: wooden fence, wooden trough, strings. These observations registered 773 occurrences, between 10 and 329 per foal; bout durations varied between 1 and 120 seconds; 723 occurrences targeted wooden structures; and individual between-bouts intervals averages varied from 00:03:14 to 2:23:48 (hours:minutes:seconds). Following protocol refinement, all groups were observed during 2 hours before and after each feeding, plus 1 hour in the evening, totaling 11 to 16 hours/foal. Observations maintained behavioural sampling, with addition of 4 action classifications: biting, chewing, grasping and cribbing-like movements. These observations yielded 5109 occurrences, between 0.005 and 0.382 bouts per minute per foal (median 0.076; IQR 0.050 – 0.127). Foals spent 00:00:06 to 00:38:29 performing these behaviours, up to 4% observed time (median 0.985%; IQR 0.570% - 1.815%). 4644 occurrences were classified by action, comprising 2793 chewing (expressed by 100% foals; 1.59% to 87.83% of individual occurrences), 1169 biting (98.27%; 0% to 61.54%), 447 grasping (87.93%; 0% to 59.09%) and 235 cribbing-like events (65.52%; 0% to 96.83%). Further individual variability was observed in bout durations, between-bouts intervals, daily expression patterns, and preferences for behaviour types and targets. These findings provide novel insights on the expression of repetitive behaviours by foals immediately after artificial weaning, demonstrating great individual variability which contributes to detecting repetitive-behaviour-prone horses in early life and promoting interventions to shape their long-term trajectories.

Episodic memory has various features in humans. In a famous what-where-when-experiment it is tested whether scrub jays have some of those features (Clayton & Dickinson, 1998; Clayton et al., 2001). Scrub jays are able to cache two types of food at different times. Later they get the chance to recover the food. Sometimes this will be when the worms are still good. So, for a successful completion scrub jays would need to recover the worms. However, sometimes they get to recover the food only after the worms have already degraded. In this case, the scrub jays should search for the peanuts. These experiments serve to show that scrub jays have some features of human episodic memory: they can remember what, where and when something happnened in an integrated experience that can be used flexibly (Clayton & Dickinson, 1998; Clayton et al., 2001). They have episodic-like memory. In this paper, I explore whether and why episodic-like memory is ethically important for some moral theories. I two different strands in animal welfare: hedonic approaches and capability-focused approaches (see for example Bentham, 1780; Duncan & Dawkins 1983; Duncan 2004; Fraser 2008; Nussbaum 2023; Singer 2024). I argue that for both approaches having episodic memory is morally significant. This implies that a creature with this capacity deserves careful, moral consideration relating to their memory capacities. For episodic memory to be morally significant both demand certain features of episodic memory. However, episodic-like memory might not include those features, like the conscious re-experience which is not tested for. Nevertheless, it would be too quick to conclude that creatures to which researchers only attribute episodic-like memory, should not be deserving of the same treatment as an animal with episodic memory. Instead, a more precautionary conlusion is justified when considering animals that show various features of human episodic memory and are taken to be conscious. This is because while episodic memory is not necessary for developing PTSD or to experience distress, for example, after a traumatic event, one’s ability to re-experience one’s harmful past increases the likelihood for future severe distress. While the above theories will evaluate the effects differently, I argue that these risks justify adopting a precautionary conclusion given the already pervasive evidence for both approaches. Thus, welfare decisions should be made *as if* the animal has episodic memory although experiments might only show that it has episodic-like memory.

Research using animals is guided by the3Rs: Replacement, Reduction, and Refinement. On the premise that the best welfare is achieved by not using animals at all, both UK and EU law require researchers to use non‑animal methods whenever they are scientifically suitable. Despite this, replacement is often not considered thoroughly, and the current system of checks and balances frequently fails. Existing guidance on how to search for and identify replacement approaches is often confusing, unclear, or misleading, which contributes to poor awareness, low confidence, and structural barriers that prevent researchers from fully addressing replacement. In the [Replacement Checklist paper](https://doi.org/10.1177/02611929251319265), we outline simple improvements to current processes and introduce a practical tool that helps researchers find and evaluate scientifically valid replacement options. It has since been used and endorsed by establishments, named veterinary surgeons and home office inspectors. We will present the checklist and show how researchers, AWERBs and other reviewers can use it to ensure that all non-animal options have been thoroughly investigated before any decision to make animals is made. This helps ensure full implementation of Replacement meets both legal and ethical responsibilities and protects animal welfare through the proactive identification and use of alternatives. Our checklist is a practical and adaptable tool already used in various institutions to strengthen challenge, improve assurance, and enhance decision-making around animal research. **Dukes JP**, Beale A, Camp C. Reviewing Current Guidance for the ‘R’ of Replacement and Rethinking it with the ‘Replacement Checklist’. *Alternatives to Laboratory Animals*. 2025;53(2):72-83. doi:[10.1177/02611929251319265](https://doi.org/10.1177/02611929251319265)

Mice can use their prior knowledge of spatio-temporal patterns to predict a reward. Previous studies have mainly focused on evaluating behavioural responses in anticipation of a reward by measuring their activity levels. However, increased activity can indicate either positive excitement or frustration. It has been demonstrated that vocalisations are a reliable indicator of animals' emotions in many species, including rats, which vocalise in anticipation of a reward. As few researches have yet been conducted on ultrasonic vocalisations (USVs) in mice in non-social contexts, our study aimed at determining whether and how mice vocalise in anticipation of a sweet caloric reward. Nineteen adult female CD-1 mice (6 non-gestating and 13 in the last 3rd of gestation) were first trained from days 1 to 6 in an anticipation procedure involving a "neutral" stimulus (bottle of tap water). The anticipation procedure consisted of depriving the mice of water for 5h at the end of their inactive diurnal phase, presenting them with the "neutral" stimulus during 5min, without allowing them to drink (bottle placed on the cage grid, nipple out of reach), and then allowing them to drink. From days 7 to 14, the mice underwent the same anticipation procedure, but with a sweet caloric reward (bottle of water containing 1% sucrose). USV recordings were performed on days 7 and 14 to compare anticipatory responses before and after any positive associative learning (day 7 = naïve; day 14 = experienced). Mice emitted their first USV earlier when experienced compared to naïve (p=0.04, 34±9 *vs.* 99±27 sec respectively). They vocalised more and with longer USVs' duration when experienced compared to naïve (p<0.01, 5.7±1.2 *vs.* 1.6±0.5; p<0.01, 28±5 *vs.* 8±2 sec). The minimum, mean, and maximum frequency of the USVs were higher for experienced than naïve mice (p<0.05 for the 3 variables; min: 50±3 *vs*. 40±2 kHz; mean: 54±3 *vs.* 43±2 kHz; max: 58±3 *vs.* 45±3 kHz). Our results demonstrated that mice vocalised in a non-social context, *i.e.* when they anticipated a sweet caloric reward. They vocalised differently when they had learned the rewarding value of the sweet caloric water compared to when they were naïve. The frequencies of the USVs could suggest a negative emotional context for the mice. Further behavioural studies are needed to assess whether the gestational state of the mice may have influenced their behaviour and assessment of the situation.

We developed a new set of measurements for a comprehensive assessment of animal welfare in dairy herds at farm level. The measurements cover all animal groups on a dairy cattle farm (dairy cows, youngstock, calves) and reflect five dimensions of welfare: nutrition, health, environment, behavior, and emotional state. The aim was to create an assessment method that would be scientifically sound but practical and time efficient, thus enhancing wide scale implementation. The resulting assessment supports improvement of animal welfare on farms, and it can be integrated into quality systems. Potential measurements for on-farm use were collected from scientific literature and previously published assessment systems. During development, a holistic understanding of welfare as the animal’s own subjective experience was emphasized. Therefore, animal‑based measurements were chosen as the primary focus. Resource‑ and management‑based measurements were included when they had a valid connection to animal welfare and when satisfactory animal‑based measurement was not available based on reliability or feasibility. The on-farm measurements were tested on dairy farms in three stages. Measurements that had low inter-observer reliability or were not deemed practical were modified or removed. The required sample size for animal‑based measurements was determined using data collected on farms and from the literature. The assessment consists of 44 measurements in total; 18 are questions for the farmer, 9 variables are extracted from Finnish cattle databases, and 17 are measured by a trained assessor on farm. Unlike in many earlier welfare assessment protocols for dairy cattle, we have included assessments for pre-weaned calves and other youngstock. Both measurements for adult animals that were suitable for calves and youngstock, and new measurements specifically developed for young animals are used. The number of measurements used varies between animal groups. Pre-weaned calves and dairy cows are observed individually (e.g. body condition score, skin lesions, and symptoms of disease ). For safety and reliability reasons, measurements for youngstock are conducted at group level from outside the pens. Questions to the farmer are used to gather information about management practices that cannot be assessed during a farm visit (e.g. calf feeding practices, outdoor access, and pain mitigation). The assessment system is designed to be dynamic as animal welfare research and technology continuously develop. The assessment protocol can be modified in the future, and new measurements can be added to better include e.g. aspects of positive welfare. An aggregation method to form welfare indices is under development.

While biosecurity and animal welfare are widely understood to be essential for sustainable livestock production, tensions can arise when measures to reduce disease risks limit welfare opportunities, or when welfare-enhancing practices increase the likelihood of pathogen introduction. As part of the BIOSECURE project, we have explored farm decision-makers’ perspectives on the relationship between biosecurity and animal welfare, with a particular focus on situations where the two may come into conflict and eliciting potential solutions for those conflicts. The methodology used was semi-structured interviews with farmers or veterinarians where we presented short vignettes designed to prompt discussion. The vignettes were developed based on previous input from livestock experts from different sectors and one was common between all sectors (access to outdoor rearing) while the others were sector specific (e.g. barn systems for layers, slow growing breeds for broilers, enrichment material for pigs, and group pens for dairy cattle). Participants were selected to reflect a range of herd sizes and management systems across each sector. Interviews were recorded and afterwards analysed using thematic analysis to identify key themes. This work was done between May and December 2025 in three different countries: 9  poultry farmers (4 layer and 5 broiler) were interviewed in Ireland; 9 dairy farmers (small to large farms) were interviewed in Sweden; while 10 pig veterinarians (5 intensive and 5 extensive) were interviewed in Spain. Although farmers are the main decision-maker for poultry in Ireland and dairy cattle in Sweden, pig veterinarians are considered the main farm-level decision maker on Spanish pig farms. From the analysis there were common themes between sectors: farmers consistently described animal welfare and biosecurity as dependent on and essential to each other, participants repeatedly described biosecurity as a pre-requisite for animal welfare. Outdoor access was considered “essential” for welfare but poses challenges for biosecurity, which are difficult to manage albeit fencing was a suggested way to minimise exposure to wildlife. Other welfare factors such as material for enrichment and contact with people, were identified as potential sources of pathogen introduction in all sectors. While solutions were found for many of these, some are not easy to implement on farm. This study has identified potential conflicts between biosecurity and animal welfare by decision-makers at farm level and local context appeared differently relevant for each sector. For some conflicts ready to use solutions could be identified whereas others may require a trade-off between welfare and biosecurity.

Pig production’s sustainability is the result of three different pillars: Animal production efficiency, Animal welfare, and Environmental sustainability. A systematic review was conducted to identify “iceberg” indicators that enable monitoring of the three pillars of pig production’s sustainability with a few measurements. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, a literature search was conducted for each of the three production pillars. Peer-reviewed scientific literature related to pig production published from 2000 to 2025 was collected from three online search engines: ScienceDirect, Scopus, and PubMed. Manuscripts were reviewed if they had at least an open-access abstract, and articles in which the abstract was the only open-access section were counted as valid if information about animal-based indicators and measurement techniques was explicitly mentioned. Studies performed on species other than pigs, in vitro studies, and manuscripts in languages other than English were omitted. The included studies were all performed under research or production conditions. The search terms to find relevant literature were selected based on the primary research question: *“What are the indicators to monitor pigs’ sustainability in production conditions regarding animal performance, animal welfare, and environmental **sustainability?”*The systematic literature search yielded the following results: animal production pillar, 28 indicators from 31 manuscripts, with body weight, average daily gain, and feed conversion rate as the most used indicators; animal welfare pillar, 23 indicators from 119 articles displaying fecal soiling, feeding behavior and body temperature as the indicators more repeated in the revised articles; and environmental sustainability pillar, 17 indicators from 31 papers reporting gases emissions, soil acidification, and water eutrophication as the measurements more recorded. The identified indicators were assessed through direct associations among the indicators, looking for those with more direct associations with indicators from other pillars. The indicators that showed the highest associations with indicators from other pillars were body condition, feeding behavior, body lesions, and body temperature, potentially working as iceberg indicators. The findings of this literature review encourage researchers and stakeholders to utilize the proposed iceberg indicators to efficiently gather relevant information regarding the three pillars of pig production sustainability under practical production conditions.

Aviaries promise freedom of movement. Laying hens are ground‑dwelling, high‑wing‑loading birds that perform flight in brief, explosive bursts. When resources are placed on upper tiers, hens descend using flight, and their ability to adjust trajectory or decelerate appears biomechanically constrained. To improve understanding of these constraints and how to manage them, we synthesized findings from three biomechanical studies (Leon et al., 2021; vanStaaveren et al., 2023; Hong et al., 2024) of hens spanning ages, strains, housing, and health states. We measured flight kinematics and ground reaction forces. We used feather clipping to reduce wing area and challenge the capacity of 34-week-old LohmannLSLLite hens (n=18) to produce aerodynamic force. Hens descending from a 1.5m tower with 50% of wing area showed rapid vertical velocity (3.9m/s), high horizontal acceleration, and did not modulate wingbeat frequency, amplitude, and angular velocity. These findings indicate hens operate at the limits of their muscle power and cannot adjust wing control when lift is reduced. We tested whether painful conditions, foot pad dermatitis (FPD), keel bone fracture (KBF), or healthy, combined with meloxicam or placebo affected landing forces. Seventy‑week‑old LohmannLSLlite hens (n=37) descended from 30 or 170 cm onto a force plate. Landing velocity and force increased with height. At 30 cm, hens landed at forces of ~3.3× body weight (BW); hens with KBF showed higher landing velocities (+17–21%) and tended to generate higher peak forces (+24%) than those with FPD. At 170 cm, hens landed at ~6.6–7.1× BW. Meloxicam did not reduce loads at 30 cm and tended to increase velocity and landing force at 170 cm in hens with FPD (~+16% BW vs. placebo). We also tested for effects of ramp training using a 16-week wing assisted incline running (WAIR) regime. We measured kinematics in 37-week-old hens (n=82; DekalbWhite, Hy-LineBrown) descending from a 1.55m tower. WAIR training increased initial flight velocity 0.2m/sec (6.5%) and tended to decrease descent acceleration, while descent velocity and trajectory angle (33–37°) were not affected. Although observational research on hen movement is common, biomechanical studies in laying hens remain limited. Our research begins to address this gap by identifying physical limits that shape descent and landing. Trained hens require long landing distances, and impact forces increase with height.  Factors that increase descent speed, feather loss, keel bone/foot injuries, further elevate risk. Aviary design must match hens’ biomechanical capacities.

Dairy cows have been traditionally housed on pasture, but this has been largely replaced by indoor barn housing with the rise of intensified systems in the developed world. However, pasture access provides benefits to cows’ health and welfare, such as decreased lameness and mastitis. Furthermore, cows typically prefer pasture over indoor housing and indoor-housed cows remain highly motivated to access pasture. The public also value pasture-based systems; they frequently place emphasis on naturalness in dairy farming and are not supportive of practices that restrict the animal’s freedom of movement. As such, pasture-based housing is often a key component of high-welfare dairy product labels. Despite the abundance of research conducted to understand stakeholders’ (defined here as people or groups having a vested interest in how dairy cows are housed, including the general public, consumers, producers, and industry professionals) attitudes towards pasture access on dairy farms, a synthesis of this research has not been conducted. In this systematic review, we aimed to 1) summarize and describe the scientific literature exploring stakeholder perceptions of pasture access for dairy cows, 2) identify similarities and differences in views between stakeholder groups, and 3) highlight arguments used to justify these views. Forty-three studies were included for analysis of quantitative (e.g. Likert scale responses) and qualitative data. Quantitative and qualitative data were thematically analysed and 7 themes were identified: Animal welfare, management and the farmer, environment, consumer preferences and product characteristics, socio-cultural or tradition, economics, and trade-offs. Overall, stakeholders had neutral to positive attitudes towards pasture access, with the public showing the strongest support and producers and industry professionals carrying more variable views. Pasture access was widely perceived to improve cow welfare, but fully outdoor systems were viewed negatively in some countries due to adverse environmental conditions. Farmers’ views were strongly influenced by farm type, region, climate, land availability, market and experience, and they generally supported the benefits of their own systems while acknowledging the drawbacks of alternative systems. The findings from this review demonstrate stakeholder support for pasture access in dairy farming, particularly when incorporated into mixed-housing systems. Greater inclusion of industry professionals such as veterinarians, and research targeting multiple stakeholder groups are recommended. Improved intra- and inter-country studies are needed to recognise regional barriers to pasture access and inform evidence-based industry recommendations.

Accurately and objectively quantifying affective state is essential for improving the welfare of laboratory rodents. In rats, ultrasonic vocalisations (USVs) are widely recognised as affective communication, 50 kHz calls are associated with positive experiences, whereas 22 kHz calls are linked aversive or threatening events. The affective bias test (ABT) is a validated behavioural paradigm sensitive to acute psychosocial and pharmacological manipulations, producing dose-dependent effects. Importantly, it provides an objective “ground truth” measure of how animals recall positive or negative experiences. Our previous work demonstrated that tickling induces a positive affective bias at the population level and that the number of 50 kHz calls correlates with the magnitude of positive bias across individuals. Animals that failed to emit 50 kHz calls also showed no evidence of a positive affective state. The present study aimed to further validate whether acute induction of either positive or negative affective states elicits corresponding USV patterns across both sexes and two commonly used laboratory rat strains (Lister Hooded and Sprague Dawley), and whether the number of calls reflects individual differences in their affective state. To test our hypothesis, animals underwent positive manipulations (tickling, exposure to a playpen, opposite-sex odour) and negative manipulations (restraint stress and administration of the anxiogenic drug FG7142). Using the ABT as a benchmark, we found that positive manipulations reliably induced a positive affective bias at the group level. Importantly, at the individual level, the number of 50kHz calls emitted by animals significantly positively correlated with the magnitude of the % choice bias, indicating a graded relationship between vocalisation rate and affective experience. In contrast, although negative manipulations produced a negative affective bias, animals did not emit 22 kHz calls as hypothesised. In the FG7142 study specifically, treatment was associated with a reduction in 50 kHz vocalisations. Taken together, our work provides robust empirical evidence that the number of 50kHz ultrasonic vocalisations is directly related to the animal’s positive affective experience, and may serve as simple standalone indicators of rats’ affective experience. All research work was conducted in accordance with the UK Animals (Scientific Procedures) Act 1986 and were approved by the University of Bristol Animal Welfare and Ethical Review Body and UK Home Office (PPL number PP4130557).

Objective As part of the EU Horizon project “aWISH” (Animal Welfare Indicators at the Slaughterhouse), “Best Practice Guides” are being developed to improve animal welfare in pig farming. To ensure both scientific accuracy as well as suitability of the identified “Best Practices”, animal welfare experts and pig farmers were consulted in two surveys and asked to evaluate the measures. Material and Methods First, a literature review was carried out to identify Best Practices to reduce the occurrence of tail/ear lesions, skin lesions and bursitis in fattening pigs. The practices were then integrated into two surveys: one was sent out to animal welfare experts (aWISH partners, the project’s expert panel and external scientific experts from 21 countries), the other was aimed at pig farmers from the DACH region. Participants were asked to rate the proposed Best Practices on a scale from 0% (unsuitable) to 100% (very suitable). Data analysis, including a mixed linear regression model, was performed with the software R. Results Overall, 47 completed questionnaires were collected from experts and 80 from farmers. Almost all proposed Best Practices received a higher score from experts than from farmers. However, both groups showed the same tendencies in their ratings. The top Best Practices for tail/ear lesions included “Good health status” (experts: 81.5% vs. farmers: 75.0%; p = 1.000) and “Good barn climate/air quality” (84.0% vs. 72.0%; p = 1.000). Measures such as “Enough feeding space per pig (preferably 1:1)” (experts: 80.0% vs. farmers: 46.0%; p<0.001) and “Lowering stocking density” (71.5% vs. 40.0%; p<0.001) received very different scores in the two groups. Regarding skin lesions, the Best Practice “Little regrouping/mixing of animals” received the highest score for both groups (88.0% vs. 71.0%; p = 0.006). For bursitis, soft floor alternatives were identified as top measures: “Bedding (e.g. straw)” (87.0% vs. 54.0%; p<0.001) and “Soft or soil floors, no slats” (80.0% vs. 55.5%; p = 0.008). Conclusion In this study, a clear difference in the evaluation of Best Practices was detected between animal welfare experts and pig farmers, even though the top Best Practices were often the same in both groups. It is therefore important to consider practical constraints like costs, feasibility and workload when developing guides to improve animal welfare.

Across species, play vocalisations (e.g. human laughter, primate play calls, rat 50 Hz chirps) can indicate positive social contexts. In dogs, the ‘play-pant’), a pant-like exhalation is observed during play with humans but not in other contexts such as rest or training. This study aimed to investigate how play context and play partner impacted the occurrence of the play-pant. We hypothesised that dogs would play-pant more in contexts that were more playful (positive, and intrinsically motivated states, characterised by low stress and spontaneity). More playful contexts were assumed to involve owners, unstructured play, and outdoor settings. Additional aims were to describe other vocalisations made during play and the acoustic characteristics of the play-pant. Twenty-two family dogs participated in structured (2 fetches & 60s tugging) and unstructured play sessions (90s free play) with their owner indoors and outdoors and eighteen family dogs participated in structured and unstructured play sessions indoors only with their owner and a stranger. Vocalisations were recorded using AudioMoth devices attached to the dog using a harness. Vocalisations were extracted using Praat, and acoustic features and statistics were completed in R, using poisson models of play-pant or play-pant bout count with a random effect of dog. In the indoor/outdoor structured/ unstructured comparison, dogs panted between 0-131 times and exhibited the most play-panting during outdoor play (mean play-pant count ± SE, unstructured indoor: 12.8± 3.54, unstructured outdoor: 14.6± 4.25, structured indoor: 14.9± 3.76, structured outdoor 20.6± 10.2, poison glm, z =3.77, P=0.0002). In the indoor owner/stranger structured, unstructured comparison, bouts of play-panting were between 0 -12, with the most play-pants performed most during unstructured play with their owner (mean play-pant bout count ± SE, unstructured with owner: 3.56± 0.80, unstructured stranger: 2.50± 0.64, structured owner: 2.44± 0.58, structured stranger 0.61± 0.24, poison glm, z =2.71, P=0.007). The play-pant could be distinguished from panting using the amplitude of the sound. A pig-like vocalisation was recorded mainly in brachycephalic dogs and had some acoustic similarities to the play-pant. Growls and barks were also noted during play sessions. The dog play-pant appears to be influenced by a combination of environmental and individual factors, supporting its role as an emotionally expressive signal and a positive welfare indicator in play contexts. The findings emphasise the acoustic and behavioural richness of dog–human play and raise new questions about how anatomy and context influence vocal communication in dogs.

Providing farmed animals with opportunities for agency, such as choice and control over social and physical resources in their environment, is considered a key component of good animal welfare. Automatic milking systems (AMS) are promoted as allowing cows to voluntarily present themselves for milking; thereby, increasing cow agency compared to conventional parlour-based systems in which cows are fetched for milking by farmers at specific times. However, the behaviours of cows using AMS are also constrained. Most notably, cows do not have the choice to opt out of being milked, and milking is incentivised through rewards (e.g., concentrate feed) otherwise not accessible. Our review explores how cow agency is affected by management decisions (e.g., fetching cows that do not visit the AMS voluntarily), system-based factors associated with AMS designs (e.g., free-flow versus forced-flow AMS), and individual characteristics (e.g., social rank, parity, personality). In the absence of direct measures of agency, our evaluation of factors influencing choice and control relies on behavioural measures reported in AMS studies (e.g., waiting times, visit frequencies, fetching events). We argue that AMS provides only limited agency over the milking process. Within different AMS designs, free-flow systems, where valued resources (feed bunk, lying stalls) can be accessed independently of milking visits, impose fewer restrictions on cows compared to the one-way gate systems used in forced-flow AMS. However, feed reward predictability tends to be lower in free- than forced-flow systems, as some AMS entries are unrewarded. At the individual level, primiparous or lower-ranking individuals may be more susceptible to barriers (e.g., social competition) limiting their ability to control AMS access. By identifying these restrictions, we highlight where further research and refinement are required to accommodate individual differences and enhance cow agency in AMS.

Open research practices ensure transparency, reproducibility, and long-term data reuse in the field of animal behaviour science. We conducted a pilot study, examining repositories used by animal behaviour researchers, to evaluate data sharing practices in the field. The study was part of a larger project (“ABOR”, [https://osf.io/kw7h9](https://osf.io/kw7h9)) that is investigating the current state of data sharing practices in behavioural research and the influence of journal data sharing policies on researchers’ sharing practices. We randomly selected 25 studies (approximately 15% of a larger sample) that reported sharing their underlying raw data. We found the use of three repository types: 15 studies utilised general-purpose repositories (Zenodo, Figshare, Dryad), five studies favoured code-oriented open research platforms (GitHub, OSF), and five employed local institutional repositories. Each study was evaluated for multiple indicators, including working link status, repository type, accessibility, presence of raw behavioural data, types of data formats, completeness of metadata, licensing, and the presence of analysis code and/or documentation. Overall, only 80% of the links we tested were functional. The primary causes of broken links were malformed DOIs (digital object identifiers), missing secondary links, or inaccessible repositories. Of the functional and partially functional links, 21 datasets were openly accessible, four were closed, unavailable, or unworkable due to non-functional links. Raw behavioural data were shared in 19 studies, while two studies did not share raw data. Using a scoring system for data format and metadata quality, we found that seven studies provided thorough metadata (score 3), six provided complete but less comprehensive documentation (score 2), six offered only partial metadata (limited data description or code; score 1), and six had insufficient or inaccessible metadata (score 0). The highest scoring datasets were those that included both behavioural data and accompanying analysis code and documentation, particularly found in GitHub hosted projects. CC BY 4.0 was the most common public copyright license. Dryad and Figshare were frequently used together to host original datasets and supplementary materials, and institutional repositories, often had inconsistent or inaccessible metadata. One dataset exceeded 300 GB, presenting additional challenges for access and reuse. This pilot provides insights into how repositories and data sharing practices are implemented in animal behaviour research. It also highlights barriers, such as inaccessibility and metadata incompleteness, that must be addressed for more consistent and effective open research practices. The findings are now being used to guide an important large-scale investigation.

The upcoming revision of the EU animal welfare legislation represents a once-in-a-generation opportunity to modernise an outdated, fragmented, and scientifically obsolete framework that directly affects over 8.4 billion farm animals raised and slaughtered in the EU each year. Despite strong and stable public support for higher welfare standards, as reflected in the 2023 Eurobarometer and in major European Citizens’ Initiatives such as End the Cage Age (1.4 million signatures) and Fur Free Europe (1.5 million), current legislation has remained largely unchanged for more than two decades. At present, more than 430 national animal welfare laws coexist across the EU, creating regulatory fragmentation, legal uncertainty for farmers, and uneven competition. This presentation highlights why an ambitious legislative revision is essential for improving animal welfare, strengthening the competitiveness and resilience of EU agriculture, and enabling a predictable regulatory environment for farmers. The current framework fails to incorporate at least 42 EFSA scientific opinions, and still relies primarily on input-based standards and the outdated “Five Freedoms” model. Modernising EU rules with science-based, animal-based indicators aligned with the Five Domains would ensure more coherent, preventive, and outcome-oriented welfare standards across species. A harmonised EU framework would also reduce administrative burden, simplify compliance, and offer farmers greater clarity for long-term investments. Legal certainty over a 10–15 year horizon is essential for operators planning housing upgrades, transitioning away from cages, or adopting precision livestock technologies. Evidence from higher-welfare poultry systems shows the potential: breeds aligned with the European Chicken Commitment use 80–85% fewer antibiotics than conventional broilers, demonstrating how better welfare supports animal health, public health, and the fight against antimicrobial resistance. Modernisation is equally important for the EU’s competitive position. Clear and measurable EU-level welfare standards would ensure a level playing field with imports, improve coherence with CAP instruments, and reduce duplication between legislation, certification schemes, and support programmes. Better-aligned standards would also diminish the need for Member States to legislate independently, helping to streamline the regulatory environment and reduce costs linked to disease outbreaks and mortality. Finally, an ambitious revision is necessary to close existing gaps, including the absence of specific rules for aquatic animals, and to fulfil the commitment to phase out the use of cages, which we are expecting by the end of 2026. Updating EU legislation in line with scientific consensus and societal expectations is not a regulatory burden but a strategic investment in sustainability, resilience, and market fairness.

Practical and valid indicators are necessary to ensure that dairy cattle welfare can be adequately monitored on-farm. We explored the validity of three categories of non-invasive physiological indicators of welfare – hair cortisol, heart rate variability indicators (including SD1 for short-term variability, LF for the relative power of the low frequency band and HF for the relative power of the high frequency band) and milk-derived indicators (including fat and lactose content, as well as daily milk yield fluctuation) – while accounting for three personality traits: activity, fearfulness, and sociability. Primiparous cows (n = 48) were first housed under reference conditions (9 wk) and subsequently exposed to either worsening or improving living conditions (6 wk). Living conditions were adjusted weekly, based on three facets: stocking density, social (in)stability, and level of enrichment. We hypothesised that physiological indicators would vary across living conditions in a personality-dependent manner, given that personality influences how individuals perceive their environment. Against expectations, no significant differences in hair cortisol or lactose content were found between cows under negative and positive living conditions. In contrast, worsening living conditions led to an increase in heart rate and a decrease in milk yield – regardless of personality. Other milk-derived indicators and heart rate variability indicators showed personality-dependent responses to living conditions. Daily milk fluctuations increased in more fearful cows under worsening living conditions. For heart rate variability, SD1 and HF increased while LF decreased in less active cows under improving conditions, and HF increased in less fearful cows and less social cows under improving conditions – as expected. Unexpectedly, however, more active cows also showed increased HF under worsening living conditions. Finally, the relative milk fat content increased both in more fearful cows and more social cows under worsening conditions, as well as in less fearful cows and less social cows under improving conditions. Taken together, these results indicate that heart rate variability indicators and daily milk yield fluctuation hold promise as non-invasive indicators of dairy cow welfare. Daily milk yield fluctuation is of particular interest, as it can be routinely recorded on many dairy farms without additional animal handling, and therefore shows strong potential for monitoring of dairy cattle welfare. This study also illustrates that similar housing conditions can affect individuals differently depending on their personality, thereby further questioning the existence of “one-size-fits-all” housing conditions.

The ways that people view animals and their needs can impact their interactions with animals under their care. Farmers interact frequently with the animals under their care and their decisions can affect the animal’s welfare. However, little research has investigated how farmers understand the needs of the animals under their care. Our study aimed to describe how dairy farmers understand the needs of calves under their care and the language that farmers use when conceptualizing these needs. We conducted a secondary analysis of interview data with 17 dairy farmers across 15 farms that aimed to understand the process that dairy farmers in British Columbia, Canada undergo, when transitioning between individual and social housing for milk-fed calves. During these interviews, the interviewer noticed that interviewees, unprompted by the interviewer, would describe their understanding of their calves’ needs, prompting this secondary analysis. Interview transcripts were thematically analyzed in NVivo. Our findings indicate that some farmers understood their calves’ needs in anthropomorphic terms (i.e., ascribing human characteristics to their calves), while other farmers did so in theriomorphic terms (i.e., using language that referenced the calves’ perspective). In both cases, farmers’ viewpoints impacted their calf care and housing decisions. For example, some farmers who described social housing negatively and decided to individually house calves referenced anthropomorphic concerns about calf health and co-mingling young animals, suggesting that their understanding of their calves’ needs may have impacted their on-farm management practices. This research provides context to existing research on how to motivate change on farms by calling attention to how individuals’ language and perspectives can influence animal care. Efforts to change farmer behaviour and management practices must therefore consider how farmers understand the animals under their care, and communications with farmers may require using language that is more aligned with farmers and their perspectives.

Animal welfare is increasingly recognised as a core component of veterinary and animal health professional responsibilities. In Latin America, however, veterinary education has traditionally prioritised clinical and zootechnical training, and updated information on the status of animal welfare education in the region has been lacking for more than a decade. This study, conducted between 2023 and 2024, aimed to assess how animal welfare is taught in tertiary animal health programmes in Latin America, identify key challenges and drivers influencing its integration, and inform future curriculum development. An exploratory mixed-methods approach was used, combining semi-structured interviews with eight animal welfare experts from six Latin American countries and an online survey of veterinary and animal science teaching institutions. Survey responses were obtained from 57 participants representing 39 institutions in 12 countries. Qualitative data from the interviews were analysed thematically using an inductive approach, and quantitative survey data were analysed descriptively. Interviewees reported a positive trend in the integration of animal welfare into veterinary curricula, although they noted considerable variation between institutions in both the extent of integration and the teaching modalities employed. Similarly, survey results indicated that animal welfare was included in the majority of participating institutions’ curricula (72%), and in most of these (85%) it was delivered as a mandatory subject. Nevertheless, substantial variation was observed in curriculum structure, including whether animal welfare was taught as a stand‑alone, transversal, or hybrid subject, the stage of the programme at which it was taught, and the total time allocated to the subject. In most institutions, animal welfare was taught in only one year. Teaching focused predominantly on farm and companion animals, while fish, wildlife and laboratory animals received comparatively limited attention. Key challenges to effective animal welfare education included limited curriculum time, insufficient numbers of lecturers with formal training, low institutional prioritisation, budget constraints, and bureaucratic barriers to curriculum reform. Drivers supporting progress included capacity building, academic and professional networks, market and trade requirements, integration within One Health and One Welfare frameworks, and the availability of teaching resources. The findings indicate that animal welfare education in Latin America has expanded in the last decade but remains uneven and insufficiently standardised. Strengthening teacher training, fostering regional collaboration, and embedding animal welfare as both a mandatory and cross-cutting component of veterinary curricula are essential to improve educational quality and ultimately promote better animal welfare outcomes in the region. **Keywords**: Animal welfare education, Latin America

Reliable assessment of animal welfare depends on the use of valid and repeatable indicators that can be applied consistently across contexts. In pigs, tail length is an Results-oriented welfare indicator closely linked to tail biting and compliance with the European ban on routine tail docking. However, systematic monitoring of tail length at slaughterhouses remains challenging due to the subjectivity, time requirements and limited scalability of manual assessment. This study aimed to develop and validate a standardised approach for assessing pig tail length under commercial slaughterhouse conditions with a particular focus on repeatability, validity and feasibility for large-scale welfare monitoring. Two categorical tail length scoring systems (three- and six-category schemes) were designed based on a representative population of undocked pigs in Austria. Inter- and intra-observer reliability were evaluated using weighted Cohen’s kappa statistics and categorical scores were validated against direct tail length measurements in centimetres collected at the slaughter line. Both scoring systems demonstrated substantial to almost perfect observer agreement, with the three-category system showing the highest repeatability and the six-category system providing increased resolution. To explore scalability and objectivity, a supervised machine-learning model was trained on more than 4,500 labelled images and validated on an independent dataset. The automated system achieved high agreement with human assessments, indicating that image-based classification can reliably support tail length assessment in practice. These findings demonstrate that pig tail length can be assessed in a robust and standardised manner at slaughterhouses and that automated tools can support, rather than replace, welfare-relevant human judgement. By addressing issues of repeatability, validation and practical implementation, this work contributes to improving the robustness and applicability of animal welfare assessment methods and supports evidence-based monitoring of welfare legislation in pig production.

There are large numbers of non-human primates (NHPs) housed in captivity globally, kept for public display, education, to support conservation, as pets, and for biomedical research. To improve their welfare, primates are increasingly required to be pair- or group-housed, rather than kept in solitary cages. However, group housing, particularly when combined with complex enclosures, can make welfare monitoring challenging, and it is difficult to know which measures will be effective. There is a growing need to identify, consolidate and evaluate the validated welfare measures available for use with group-housed NHPs, to aid decision making for caretakers, as well as to guide future research directions. Information about the ease or difficulty of applying a given welfare measure, i.e. its practicality, would also be valuable to caretakers and researchers. This study therefore aimed to 1) identify the existing measures available for measuring welfare in group-housed primates, and 2) assess their validity and practicality. Using systematic review methods we identified 337 sources, across 65 journals, involving 100 different species. MetaBeeAI, a pipeline using Large Language Models and Vision AI, was used to extract, analyse, and benchmark information from sources. It was found that the number of studies of NHP welfare published yearly has increased steadily since the 1980s. The most studied species were Chimpanzees (n=79) and Rhesus Macaques (n=65). This is unsurprising given chimpanzees are the NHP species most closely related to humans, raising concerns about their welfare in captivity, and rhesus macaques are the most predominant NHP in laboratory research. The most used measures were behaviour (129) and cortisol/corticosterone (n=81). Within behaviour, the use of abnormal and/or stereotypic behaviours was most common (n=71), however these were typically used in combination with additional behaviours, such as aggression, feeding, or social behaviour. There was relatively little work on positive welfare indictors or affective state assessments, such as cognitive or judgement bias. Future work should aim to validate markers of positive welfare, such as play, for group-housed NHPs. Despite its importance in understanding the capability of a measure for assessing welfare, validity was mentioned in only 37% of sources. Similarly, comments on practicality were made in just 34% of articles. In the interest of improving reproducibility in welfare science, moving forwards authors should be encouraged to report measure validity in all publications. In this presentation we will explore our outcomes in more detail, and discuss their implications for NHP welfare and future research.

CO₂ gas stunning is widely recognized as inherently aversive: exposure to high CO₂ concentrations leads to the formation of carbonic acid on moist tissues, causing chemical burning, as well as respiratory distress and an overwhelming urge to breath (air hunger) through rising blood CO2 levels - compounded by fear and prolonged time to loss of consciousness. Despite extensive research, no alternative gas method has yet proven both humane and practical at commercial scale. Electric stunning, when correctly applied, induces immediate and painless unconsciousness. However, conventional automatic electric stunning systems in large slaughterhouses have historically been associated with substantial pre-stunning stress. This stress is primarily caused by excessive line speed, poor facility design, movement of large pig groups, bottlenecks at single-file entrances, and frequent use of electric prods to force pigs forward. Consequently, electric stunning has often failed to reach its full welfare potential in practice. This study describes the complete redesign and successful commercial implementation of an improved automatic electric stunning system in a large pig slaughterhouse processing 660 pigs per hour. The objective was to eliminate pre-stunning stressors while maintaining throughput and meat quality, thereby enabling electric stunning to function as a genuinely humane alternative to CO₂ gas. The redesign was guided by pig behavioural principles and applied ethology and involved fundamental changes to facility layout, stunning capacity, crowd-pen and raceway design, restraint configuration, visual and acoustic environments, timing of pig movement, and staff training. Key improvements included increasing stunning capacity through a four-stunner configuration to reduce time pressure, limiting pig movement to small groups (≤6 animals), allowing pigs to enter parallel single-file raceways side by side with visual contact, eliminating aversive handling tools, and minimizing visual distractions and noise. Electric prodding was completely eliminated when moving pigs from group into the single files. Limited and strictly regulated low-voltage electrical stimulation is now used only at the point where pigs step onto the conveyor-belt restrainer immediately before stunning. Artificial intelligence–based monitoring was implemented to ensure compliance with group size limits and stimulator use. Since 2023, the system has operated successfully without bottlenecks or forceful handling, while maintaining production efficiency and improving animal welfare. This case study demonstrates that, with appropriate design and management, automatic electric stunning can be applied humanely at high throughput and represents a viable alternative to CO₂ gas stunning.

Reliable welfare assessment is essential for ethical and scientifically reproducible animal research, particularly in long-term disease models with high cumulative severity. Assessing welfare at the level of individual laboratory animals remains challenging, as no single indicator reliably captures stress and wellbeing across different contexts and disease stages. Case-based analyses of trained behaviours therefore offer a complementary approach, enabling closer examination of behavioural sensitivity and validity. This pilot study evaluated trained behavioural tasks as sensitive, non-invasive indicators of stress and wellbeing in rats with diethylnitrosamine-induced hepatocellular carcinoma. Six male Wistar rats were trained prior to tumour induction to perform two tasks – lockbox solving and scale climbing – allowing repeated behavioural assessment throughout disease progression. Hepatocellular carcinoma was induced via diethylnitrosamine supplementation in drinking water for up to 122 days. Selected findings are presented as case reports to examine welfare assessment at the level of the individual animal. Task performance was recorded daily as time to completion, with a maximum duration of 60 s per task; trials exceeding this limit were recorded as ‘time over’. Behavioural performance was interpreted alongside established physiological and clinical parameters, including hair corticosterone and testosterone, salivary corticosterone and immunoglobulin A (IgA), body weight, and clinical observations. One case (Rat 5) showed a lack of concordance between behavioural, physiological, and clinical indicators during the terminal study phase. From day 124 onwards, task performance deteriorated markedly in this animal, including repeated ‘time overs’ in both tasks, coinciding with clinical signs of respiratory disease, elevated Grimace Scale scores, and declining general condition. In contrast, salivary corticosterone and IgA showed no consistent elevation, and body weight changes remained modest, highlighting the added value of behavioural assessment during acute welfare decline. Two further case reports (Rats 2 and 6), characterised by early and pronounced tumour development, revealed a complementary pattern. These animals exhibited the highest mean hair corticosterone and testosterone concentrations, IgA values in extreme (Rat 2 very high vs Rat 6 undetectable) and consistently poorer lockbox-solving performance, while scale-climbing performance remained comparatively unaffected, indicating task-specific sensitivity to tumour-associated distress. These case reports illustrate that trained behavioural tasks can reveal individual welfare trajectories not captured by physiological or clinical measures alone, supporting their use for refined, individual-level welfare assessment.

Stereotypic behaviours (SBs) are widely-used indicators of poor welfare. However, they can persist after attempts to improve well-being (e.g. via environmental 'enrichment'). Furthermore, Tilly et al. (2010) found that in ageing, conventionally-housed (CH) laboratory mice provided with large well-resourced enclosures, being little-motivated to use these enclosures predicted persistent, treatment-resistant SBs. This raises the question: are treatment-resistant SBs merely ‘scars of the past’ that fail to reflect welfare improvements (as often assumed), or do they instead indicate that attempts to improve welfare have failed? To investigate, two complementary studies assessed behavioural correlates of SB reduction following ‘enrichment’. Study 1 longitudinally investigated predictors of SB treatment-resistance (inferred from absolute and proportional SB reductions, and residuals from regressing post-treatment SB time budgets against pre-treatment SB) following transfer of DBA-strain CH mice to large, well-resourced enclosures. This reduced SB with great individual variation. Mice with the most treatment-resistant SBs engaged less with ‘enrichments’, according to two of the three metrics used in regression analyses (proportional reduction: *t*34=4.084, *p*=0.00025; residuals measure: *t*34=-3.38, *p*=0.0018). Furthermore, SB reductions significantly exceeded time spent using enrichments (*t*35=2.50, *p*=0.009), indicating effects beyond mere behavioural competition. Mice with treatment-resistant SBs may thus have benefitted less from the new resources. Study 2 used judgement bias (JB) tests to investigate whether mice with treatment-resistant SBs benefit less in terms of improved mood. CH mice (Balbs and DBAs) were again transferred to large well-resourced cages. Before and after transfer, affective states were assessed using a validated test for mouse JB (responses to ambiguous olfactory cues in a digging task). Again, this transfer typically reduced SBs, and by extents exceeding the time spent using the new resources (*t*22=5.18, *p*<0.0001; although this engagement now only predicted the residuals measure of SB reduction: *t*5.71=-3.13, *p*=0.022). Furthermore, mice with smaller SB reductions exhibited smaller improvements in JB (absolute reduction: *t*16=2.64, *p* = 0.012; residuals measure: *t*15.94=1.82, *p*=0.087; proportional reduction: latency x strain interaction [*t*6.19=-3.67, *p*=0.0099], with effects present in BALB mice only [β=0.23 ± 0.08, *p*=0.023]), again suggesting that mice with treatment-resistant SBs benefitted less from the new resources. More research is needed, perhaps refining how SB, 'enrichment'-use, and affective states are all assessed, and investigating whether anhedonia helps explain such patterns. Nevertheless, together such studies indicate that treatment-resistant SBs may reflect treatment-resistant negative affective states, and thence worrying failures of attempts to rectify animal welfare.

Beef-on-Dairy (BoD) calves are present on dairy farms worldwide; however, social housing is not yet a common practice for these animals. As a result, little is known about their behavior or about how social housing with dairy calves may influence the behavioral patterns. This study evaluated the effects of breed composition (Holstein vs. BoD [Angus × Holstein]) and pair composition on daily behavior. Sixty-two calves were assigned (19 ± 2 d) to one of three possible pair compositions: Holstein calves in homogeneous pairs (Holstein + Holstein; n = 13 pairs), mixed pairs (Holstein + BoD; n = 11 pairs), and BoD calves in homogeneous pairs (BoD + BoD; n = 7 pairs). Behavioral assessments were conducted weekly using scan sampling, twice daily, for 1 h around milk meals (30 min before and 30 min after), based on a predefined ethogram. Data were analyzed using mixed-effects logistic regression and Poisson models, with week included as a random slope, pair and calf ID as random effects, and age and pair composition fitted as fixed effects. Regarding breed composition, Holstein calves showed more odds of exploration (OR = 2.84; 95% CI: 1.98–4.08; P < 0.01), whereas BoD calves showed more relative risk of idle behavior (RR = 1.33; 95% CI: 1.06–1.67; P = 0.02). In addition, feeding and non-nutritive oral behaviors showed an age-by-breed interaction, with the between-breed differences in these behaviors increasing over time relative to Holstein calves. Pair composition affected daily calf behavior, with more pronounced effects in Holstein calves when paired with BoD calves. Compared with Holstein calves in homogeneous pairs, Holstein calves in mixed pairs spent less time idle (RR = 0.75; 95% CI: 0.58–0.98; P < 0.01). In addition, feeding behavior and abnormal oral behaviors showed an age-by-pair composition interaction, indicating that over time Holstein calves housed in mixed pairs had higher odds of feeding and non-nutritive oral behaviors. To our knowledge, this is among the first studies comparing BoD and Holstein calves under paired housing. From a welfare perspective, mixed pairing improved Holstein calves’ time budgets (more feeding, less idle), but the increase in non-nutritive oral behaviors suggests a potential welfare concern, possibly linked to milk-feeding management or social transmission. Thus, mixed housing may offer welfare benefits while requiring optimized liquid-feeding strategies to limit abnormal oral behaviors.

Anti-methanogenic feed additives (AMFAs) are proven to reduce methane emissions from dairy cows, but their impacts on animal welfare are not well documented. This study assessed whether two commercial AMFAs, 3-nitrooxypropanol (3-NOP – commercial name Bovaer) and dietary nitrate (NO3 – commercial name SilvAir), impacted the welfare of indoor-housed lactating Holstein cows. Eighty-seven recently calved Holstein cows were allocated to three equal groups, balanced for parity and milk yield. A Control group received a total mixed ration (TMR) of grass silage plus a concentrate; a 3-NOP group received the same TMR supplemented with Bovaer fed at 7.5 g/kg dry matter/day; and a Nitrate group received the same TMR supplemented with SilvAir fed at12 g/kg dry matter/day.  Seven cows from each group were housed together in 2-week blocks, during which their welfare was assessed. Posture and movement were monitored using hind-leg accelerometers (IceQubes). Feeding behaviour was recorded automatically using Biocontrol feed stations. Pro-social and agonistic behaviours were assessed using twice-weekly video recordings of the feeding area with assessment made of displacements (by contact or threat) and grooming behaviours (allo- and autogrooming). Potential discomfort was quantified from direct observation using the Gleerup cow pain scale, analysed as the probability of a score reaching 3 or above, a threshold indicative of possible pain-related behaviour. The mean ± SD time spent feeding for the Nitrate group (4.59±0.367h/day) was significantly longer than for the control (3.56±0.309h/day, p<0.0001), and the Nitrate group also had significantly more feeding bouts (53.0±15.21 vs 21.6±6.49, p<0.0001) than the control group. However, total feed intake was not significantly different between the treatments.  The mean ± SD number of daily lying bouts for the Nitrate group (6.40±0.981) was significantly less than the control group (7.06±1.032, p=0.0075), however, there was no significant difference in total lying time between the treatments.  Probabilities of agonistic and pro-social behaviours were not significantly different between groups, nor was the probability of an animal displaying an elevated pain score. Nitrate supplementation was associated with the cows taking longer to eat the same amount in more frequent, shorter bouts, and this may indicate some palatability issues.  The reduced number of separate lying bouts associated with this group may simply be an inverse relationship with the increased time spent at the feeder.  Overall, these results did not identify behavioural evidence consistent with a negative welfare impact from either AMFA under the conditions studied.

Many dogs experience fear during veterinary visits, which compromises dog welfare and increases staff safety risks. Despite growing interest in stress-reducing practices, limited research has examined how veterinarians make day-to-day handling decisions when assessing and responding to dog fear and aggression during routine examinations. This qualitative study explored how veterinarians conduct routine dog examinations, the factors that shape their handling decisions, and their perceived benefits and challenges of stress-reducing practices. Semi-structured virtual interviews were conducted with 17 veterinarians practicing in Canada and the United States between November 2024 and February 2025. Transcripts were analyzed using inductive content analysis within an interpretivist framework to examine veterinarians’ experiences and decision-making processes. Four major categories emerged: (1) approaches to dog handling, (2) factors affecting handling, (3) veterinarian professional well-being (i.e., compassion fatigue and satisfaction), and (4) perceptions of stress-reducing practices. All participants described assessing dogs for signs of fear and reported using stress-reducing practices (e.g., treat provision, using minimal restraint initially), citing benefits such as reduced dog fear, improved cooperation, and decreased injury risk. However, some participants perceived these practices as impractical or ineffective in certain contexts or for specific patients. Handling choices were influenced by prior experiences and training (e.g., mentorship, conferences, and workshops), clinic-level factors (e.g., management practices, appointment structure), patient characteristics (e.g., dog behavior), and owner presence and expectations. Many participants reported facing constraints to implementing stress-reducing practices, including high caseloads, time pressure, and limited support staff; under such conditions, some described omitting stress-reducing practices on fearful dogs to save time. Veterinarians’ responses to dog aggression varied, including the use of restraint equipment (e.g., muzzles), having technicians handle the dog in the treatment area, increasing physical restraint, using chemical restraint, or rescheduling appointments with pre-visit medications (e.g., trazadone). Nearly all participants experienced compassion fatigue at some point in their careers, particularly during early years in practice and/or working in emergency roles. While some felt their patient interactions were unaffected from compassion fatigue due to effective coping strategies (e.g., compartmentalization), others described less patience and decreased use of certain stress-reducing practices. Positive interactions with clients and patients, and successful clinical outcomes contributed to compassion satisfaction, resiliency, and professional purpose. These findings suggest that clinic-level support and attention to practitioner well-being are important for sustaining stress-reducing practices within a One Welfare framework that recognizes the interconnected well-being of animals and veterinary professionals.

Farm animals have received relatively limited attention in animal cognition research, despite their numbers and biomass surpassing all other mammalian species, including humans. Nevertheless, studying farm animal cognition can benefit applied research, as understanding how farm animals perceive and interact with their environment helps clarify how cognitive processes develop under different husbandry conditions and, in turn, informs management practices that better align with their abilities and welfare needs. Diverse and openly available datasets are essential for accounting for genetic, developmental, and environmental factors that influence cognitive performance, thereby promoting generalisability, comparability, and transparency. To assess the diversity and openness of the farm animal cognition literature, we conducted a scoping review of the available cognition literature on cattle, pigs, sheep, and goats, assessing study and sample characteristics and whether data were openly available. In total, 252 papers, from years 2000-2024, fulfilled our inclusion criteria. Most studies were conducted on single breeds at single sites (which were most often research sites), with an overrepresentation of females and young animals (i.e., under one year), limiting the generalizability of findings and interpretation of fully developed cognitive capacities in these species. Geographically, most studies originated from North America or Europe, restricting international representation and limiting insights into local breeds across the globe. None of the papers published between 2000 and 2014 contained a data availability statement or shared data openly. This changed between 2016 and 2019, when a small number of studies (8 of 62 papers in total) made data openly available via electronic supplementary materials (ESM) or repositories. The trend continued from 2020 to 2024 (61 papers in total), with data openly shared via repositories (10) or ESM (14). Nevertheless, most recent papers (2020-2024) still did not provide open data, either lacking a data-availability statement altogether (21) or indicating that data were available only upon request (16). These sample biases and lack of data sharing reduce diversity, compromise generalizability, limit reproducibility, hinder independent validation, and constrain opportunities for meta-analyses and cumulative knowledge building. In the long run, it weakens the potential for a broader understanding of farm animal cognition, which is crucial for understanding the impact of domestication on the brain and for improving husbandry practices to meet farm animals’ perceptual and cognitive needs. Strategies to achieve more representative study populations are needed, as well as open and well-documented data sets in farm animal cognition.

Animal welfare remains a critical yet under-examined element of smallholder dairy production in Kenya, with implications for herd productivity and health, particularly for disease conditions like subclinical mastitis which is a significant, but underdiagnosed constraint affecting dairy cattle in smallholder systems. This study investigated knowledge, attitudes, and practices related to animal welfare, determined prevalence and risk factors for subclinical mastitis among smallholder dairy farms in Njoro sub-County, Kenya. A total of 179 dairy cattle farmers were interviewed using structured questionnaires, and milk samples collected from 248 lactating cows for California Mastitis Test. The data were analyzed using R software for descriptive statistics and logistic regression to determine risk factors associated with sub-clinical mastitis in herds. Results indicated that awareness of animal welfare by the dairy farmers was low, with only 30% of farmers having heard of the concept, and 62% rating their knowledge of the concept as poor. Despite these gaps, most farmers demonstrated positive attitudes and were willing to adopt improved dairy farming practices. The overall prevalence of subclinical mastitis in herds was 58 %, with higher rates in Kihingo ward (72 %) and in younger cows aged below six years (64 %). Multivariate analysis revealed that poor hind limb hygiene, prolonged lactation of more than 7 months, and co-rearing of sheep in farms were significant predisposing factors with increased likelihood of infection by 4, 1.8, and 2.3 times, respectively. This study highlights critical gaps in knowledge and management practices that contribute to high incidences of mastitis and suboptimal animal welfare. Strengthening farmer training, improving herd management, and enhancing access to veterinary and extension services are essential for improved dairy cattle productivity, animal welfare, and livelihood support for farming households in areas with similar production settings.

Animal shelters expose dogs to stressors which may lead to excessive arousal, negative welfare, reduced adoptability, increased risk of euthanasia, and increased human safety risks. Current recommendations for reducing arousal include pharmaceutical and enrichment interventions; however, little evidence exists to support these recommendations. Thus, we conducted a randomized control trial to evaluate the influence of these interventions on excessive arousal behavior in adult shelter dogs. Inclusion criteria included adult dogs of any breed housed at a municipal shelter in Texas (USA), aged 1-6 years, that displayed excessive arousal behavior defined as at least three instances of jumping, vocalizing, mouthing, pacing, and/or circling during one-minute observations conducted in their kennel and on walk. Dogs (N=50) were randomly assigned to one of five groups: 1) control (5-min walk, 2x/day), 2) enrichment-only (15-min walks 2x/day and 15-min human-animal interaction), 3) play enrichment (15-min play 2x/day and 15-min human-animal interaction), 4) pharmaceutical-only (20 mg/kg gabapentin and 6mg/kg trazodone and 5-min walk 2x/day), 5) pharmaceutical and enrichment (20 mg/kg gabapentin and 6mg/kg trazodone, 15-min walk 2x/day, and 15-min human-animal interaction). Testing included one control week (days 1-7, 5-min walk, 2x/day) followed by one week of assigned treatment (days 8-14), with behavioral assessments conducted on days 1, 3, 7, 10, and 14. Behaviors (jumping, mouthing, vocalizing, locomotion) were assessed and overall arousal and transition (kennel exit) scores were assigned using Likert-scale ratings via 1-min videos collected in-kennel during pre-enrichment and post-enrichment interventions and during a 1-min walk. Statistical analyses using mixed-effects regression and negative binomial models indicated lower overall arousal scores during treatment week compared to Day 1 of the control week regardless of treatment group (day 10, p<0.001; day 14, p=0.003). Arousal scores were lower during walks than pre-enrichment for both the pharmaceutical-only (p<0.001) and pharmaceutical and enrichment groups (p<0.001). Locomotion was lower during the post-enrichment versus pre-enrichment phases (p=0.012). Dogs in the enrichment-only group had higher odds of vocalizing than those in the pharmaceutical and enrichment group (95% CI:1.93, 27.69; OR: 7.3; p=0.003), while a lower odds of vocalization during walks (95% CI:0.001, 0.007; OR: 0.002; p<0.001) and post-enrichment (95% CI:0.03, 0.09; OR:0.05; p<0.001) compared to the pre-enrichment phase. Results suggest that enrichment and/or pharmaceutical interventions provide the greatest improvements to arousal-related behaviors and overall welfare for shelter dogs displaying excessive arousal. Future research should identify predictors of excessive arousal to inform targeted prevention and intervention strategies in shelter dogs.

Bio-acoustic monitoring is a non-invasive, low-impact, and cost-effective method for the longitudinal assessment of animal health and welfare. Acoustics are of growing interest within Precision Livestock Farming (PLF), particularly within poultry production, where large flock sizes and diverse housing systems makes the use of alternative monitoring techniques (e.g. cameras or wearables) challenging. Chicken calls provide a meaningful welfare indicator: They possess a highly conserved, information-rich vocal repertoire, comprising referential calls that reflect both environmental stimuli, and internal physiological or affective states. Despite this promise, a bottleneck exists when translating raw audio into robust, real-time welfare monitoring, namely data labelling to annotate calls. Contemporary machine learning methods, such as those used to automatically detect vocalisation type, require a large volume of accurately labelled “ground-truth” data. Adult chickens are estimated to produce around 30 distinct vocalisations making annotation of call types a specialised skill. Annotators typically manually scan spectrograms using specialist software to identify a vocalisation, noting the onset and offset of the call and call type. The process is time-consuming, and researchers frequently opt to work with small flocks or controlled laboratory recordings to simplify the task. While such recordings are acoustically cleaner, they cannot easily be generalised to commercial farm environments, where overlapping calls from thousands of birds and high ambient noise are the norm. To address these technical barriers, we developed Squawk-Spot, an RShiny-based application designed to simplify and democratise the annotation process. Squawk-Spot integrates a machine learning backend that automatically pre-isolates candidate vocal events from continuous audio recordings. These candidate calls are then presented to the user through an intuitive graphical and audio interface, enabling rapid, single-click classification of call types. The system requires only a modest amount of initial manual annotation to train a Random Forest classifier, which is subsequently embedded within the workflow to prioritise calls of interest. As projects evolve, iterative refinement of the training dataset and model parameters maintains and progressively increases the proportion of high-interest calls, enhancing overall performance. By reducing annotation time and lowering technical barriers, Squawk-Spot enables efficient data labelling by both experienced bio-acousticians and non-expert users. This approach creates a platform for more open research while addressing the data bottleneck limiting progress within acoustic monitoring. By harnessing the intuitive expertise of diverse users, Squawk-Spot facilitates the generation of large, annotated datasets necessary to develop scalable, automated welfare monitoring solutions for the global poultry industry, and expediate poultry welfare research.

In mammals, 2-arachidonoylglycerol (2-AG) is an endogenous endocannabinoid associated with social bonding and positive affect when elevated, and with stress and negative affect when reduced. We examined 2-AG as a novel physiological indicator of affective state in dairy cattle. Cow-calf separation and reunion was used as a model to induce positive and negative affective states in Holstein dairy cows. Cow-calf pairs (primiparous=8; multiparous=11) were housed in individual pens for 36–60 h postpartum before 3 test phases (pre-separation, separation, and reunion) of 90 min each between 1330-1830h. During separation, the calf was removed to a separate pen with auditory but not visual contact; during reunion, the calf was returned to the cow in the home pen. Cow behaviors (duration of calf touching, allogrooming, exploring/searching, and self-grooming; bouts of calf-checking) were continuously recorded during the first 60 min of each phase with a subsequent blood sample collected. Behaviors were analyzed using Kruskal-Wallis tests. Serum was analyzed for 2-AG concentrations using mass spectrometry, and linear regression models analyzed percent change in 2-AG concentrations across phases and between parity (covariate: time in lactation). Before separation, 2-AG was similar between primiparous (1.71±0.23 ng/mL) and multiparous cows (1.72±0.19 ng/mL) (*P*=0.43). Contrary to expectation, 2-AG did not decrease following separation for either parity (primiparous: 7.4±15.9% change; multiparous: 27.7±13.6% change; *P*=0.35). Upon reunion, 2-AG increased more in primiparous (28.9±10.5% change) than multiparous cows (8.2±9.0% change) relative to separation (*P*=0.02), and tended to increase relative to pre-separation levels (primiparous: 37.7±13.4% change; multiparous: 9.3±11.4% change; *P*=0.09). Observed behaviors supported the predicted emotional states: cows increased exploration/searching, self-grooming, and rubbing behaviors during separation (*P*=0.001), suggestive of stress and/or frustration; in contrast, cows increased affiliative behaviors (touching, allogrooming, and calf-checking) during reunion (*P*≤0.01), suggestive of a positive state. Overall, 2-AG concentrations did not change during separation despite behavioral indicators of a negatively valenced state, suggesting limited sensitivity to acute stress in this social context. In contrast, reunion elicited both positively valenced affiliative behaviors and increased 2-AG concentrations, consistent with mammalian evidence linking endocannabinoids to social bonding and positive affect. Greater changes in 2-AG in primiparous cows may reflect heightened sensitivity to calf reunion after first calving, whereas multiparous cows showed more stable responses consistent with homeostatic regulation. Together, these findings suggest that endocannabinoids may be informative as physiological indicators of positive affect when combined with behavioral measures.

Free-roaming cat caretakers can substantially influence cat welfare and the effectiveness of community management programmes, yet caretakers are often treated as a homogeneous group. This study examined heterogeneity among cat caretakers, focusing on behavioural engagement, motivations and barriers relevant to cooperation and compliance with management initiatives. A cross-sectional online questionnaire was completed by a convenience sample of 509 respondents recruited via a national survey panel in Israel (February–March 2025). Respondents were categorised by feeding intensity based on the number of cats fed per outing (“low intensity”: 1–7 cats, n=398, total=1,519 cats; “high intensity”: 8–35 cats, n=111, total=1,287 cats). Higher-intensity feeders reported significantly greater engagement in welfare- and management-related activities beyond feeding, including reporting cats for neutering, caring for sick/injured cats, and involvement in rehoming. They also reported higher feeding frequency and higher monthly expenditure on cat food. These findings indicate that a subset of feeders may play a disproportionate role in the welfare of free-roaming cats. Differences were also observed in attitudes and perceived responsibility, with higher-intensity feeders more likely to agree that without their feeding, cats would suffer greatly or die. Knowledge gaps were common across the sample, including those shared by both groups, such as limited awareness of recommended feeding practices and sanitation challenges related to the removal of leftover food. To complement the survey, semi-structured interviews were conducted with 12 very high-intensity feeders (feeding 30–200 cats each, total >850 cats), providing insight into the extreme end of feeding involvement that was only sparsely represented in the survey. Feeding was frequently framed as a moral responsibility and personal identity, accompanied by positive emotions such as satisfaction and relief, as well as financial strain, lifestyle constraints, social conflict and emotional burden. Participants described adapting feeding practices to avoid conflict, expressed mixed levels of trust in municipal and veterinary services, and reported limited willingness to cease feeding even under restrictive policies. Together, the findings highlight meaningful heterogeneity among free-roaming cat caretakers. Feeding practices were associated with different motivations, perceived responsibilities and knowledge gaps. Interventions may need to be tailored to distinct groups: lower-intensity feeders may represent an important audience for awareness-raising and behaviour-change initiatives related to cat health, welfare and environmental sanitation, while higher-intensity feeders may represent key stakeholders for welfare-supportive activities such as reporting, neutering and rehoming. Recognising these distinct roles may help inform more realistic and welfare-focused free-roaming cat management strategies.

Despite two decades since the introduction of the Animal Welfare Act 2006 (AWA), literature suggests that companion animal owners have limited knowledge of the legislation. This study explored awareness and knowledge of the AWA amongst rabbit owners in the United Kingdom and associations with husbandry practices. Between December 2023 and February 2024, companion rabbit owners in the UK were surveyed to explore their awareness and knowledge of the legislation around companion animal care. The seven-part questionnaire explored demographic variables, pet ownership, acquisition practices and awareness of the AWA. Binary logistic regression modelling was used to identify which variables are associated with the husbandry practices of companion rabbit owners. 519 rabbit owners responded to the survey. 66.3% reported that they were aware of the legal responsibilities of companion animal owners in the UK. However, only 29.5% of the sample could name the AWA and 39.1% were able to recognise all five welfare needs identified within the legislation. Variables including the different species owned by participants and rabbit acquisition factors were significantly associated with multiple rabbit husbandry practices. Awareness and knowledge of the AWA and welfare needs were only significantly associated with the provision of a rabbit companion. Participants who were able to name the AWA were 1.8 times more likely (OR= 1.834, 95% CI= 1.096-3.069, *p*=0.021) to house their rabbit with another rabbit compared to those who did not name the Act. Those who acquired a rabbit at least in part due to the wants of their children were significantly less likely to provide their rabbit with: opportunities to display natural behaviours daily; access to fresh hay at all times; a rabbit companion; and annual vaccinations. Undertaking pre-acquisition research and taking longer when planning to get a rabbit were associated with better welfare practices. These findings suggest that awareness of the AWA amongst companion rabbit owners in the UK is limited but raise an important question about the importance of such knowledge amongst owners. Factors relating to rabbit acquisition, including the motivation behind acquisition and the research carried out before acquisition, appear to be more strongly associated with husbandry factors which could influence welfare. Pre-acquisition research may be protective against some husbandry practices that can be detrimental to welfare. Therefore, opportunities for messaging to raise awareness of appropriate husbandry practices during the acquisition journey should be optimised to improve rabbit welfare.

Many animal welfare scientists and applied ethologists use research to improve the lives of animals, yet their studies at times cause harm to animals. However, this dilemma has not been well-explored. Our aim was to describe the perspectives of these scientists on research methods or approaches in their work that may compromise the welfare of animals, and the extent to which this may provoke internal conflict. We conducted a workshop at the 2025 International Society for Applied Ethology meeting in Utrecht, Netherlands. A total of 24 participants attended the workshop and were divided into 5 focus groups (n = 4-5 per group). Participants included university professors, graduate students, and industry researchers working in the field of applied ethology and/or animal welfare. One of the goals of the workshop was to provide a chance for researchers to discuss in a supportive setting the internal struggles they may have encountered in their research using animals. All participants provided informed consent, and focus group conversations were audio-recorded and transcribed. We performed an inductive reflexive thematic analysis of the transcripts and developed three preliminary themes. The first theme describes participants’ challenges with causing harm to animals as part of their research, where they specifically identified euthanasia, pain, fear, isolation of social animals, and housing animals in barren environments. The second theme focused on the (mis-)alignment between participants’ moral values and their actions as researchers, such as the internal conflict that occurred when their actions or methods opposed their beliefs, and how they struggled to balance the trade-offs between the costs to individual animals in their studies and the potential benefits of the research results to society. The third theme encompassed participants’ perceived ability or inability to control situations where research animals were harmed, such as being a student unable to choose or change their research topic or methodology. These results provide important foundational knowledge to drive future work aimed at supporting researchers in coping with and achieving greater agency to improve the welfare of animals in their research.

Mortality can be defined as the incidence of death within a specific population, group, or area during a distinct time period. The estimation may include animals that died on farm without human assistance (unassisted death), animals that were euthanised on farm to avoid prolonged suffering (emergency killed), or an otherwise healthy animal that was slaughtered on farm due to an accident (emergency slaughter) (see Regulation (EC) No 1099/2009 for complete definition of the terms). Mortality in dairy cattle is usually calculated including both unassisted death and euthanasia, and, in some cases, emergency slaughter is also included, effectively putting everything in the same bag. Be noted that well-known and widely used protocols for the assessment of animal welfare, e.g. Welfare Quality and Classyfarm, include all three categories in the estimation of mortality. This may be an issue and of particular interest when mortality is used as an animal welfare indicator. The proportion of animals dying unassisted, euthanised and emergency slaughtered may vary within a farm over time as well as between farms. The cause of death leading to unassisted death, euthanasia and emergency slaughter can be different between the three categories, as well as the management decisions made immediately before the death. This may result in different lengths of suffering before death occurs. This study will compare the duration of suffering before death in cases of euthanasia, emergency slaughter and unassisted death. Cases of death among dairy cows recently reported to national Competent Authorities will be further investigated by applying a survey to the farmer to find out about the events and duration of the events contributing to the death of each cow. Additionally, farmer perceptions about animal suffering and factors affecting farmers’ decision-making process will also be investigated. This study will be carried out in the dairy cattle industry in Denmark and Ireland. Data collection will start in March 2026, and preliminary results will be presented at the conference.

Transhumant pastoralism in Jammu & Kashmir (J&K), India, supports an estimated 0.5–0.6 million tribal people, primarily from Gujjar, Bakarwal, and Gaddi communities, who migrate seasonally with sheep, goats, and cattle across distances of 300–600 km, typically over 30–45 days each migration cycle. This system contributes significantly to regional food security and conservation of indigenous livestock genetic resources, yet remains one of the most underserved livestock production systems in terms of animal welfare support. The transhumant population of Jammu & Kashmir is more than individual population of as many as 55 countries (Maldives, Iceland, Malta, Suriname, Luxemburg, Brunei, Iceland) in the world. Drawing on data from the First Survey of Transhumance (2021) and complementary field observations, this study synthesizes the major constraints affecting animal welfare and identifies priority training needs of transhumant livestock owners. Survey evidence indicates that livestock rearing is the primary livelihood for over 70% of transhumant households, with an average holding of 26–30 small ruminants per family. Despite this dependence, only 14–15% of livestock-owning households reported access to regular veterinary services, while over 85% lacked any veterinary support during migration. Preventive vaccination coverage was reported to be below 25%, and fewer than 20% of households had ever received formal training in animal health or welfare practices. Key constraints reported by pastoralists include prolonged walking without adequate rest or watering points, fodder scarcity along migration routes, limited disease surveillance, and unsafe transport conditions. These constraints are associated with high prevalence of lameness, hoof injuries, body condition loss, disease outbreaks, and mortality, particularly during peak migration months. Importantly, more than 60% of respondents expressed difficulty in recognising early signs of stress and disease, yet over 75% showed strong willingness to participate in welfare-focused training when linked to improved animal survival and productivity. The findings underscore an urgent need for context-specific, welfare-oriented training programmes for transhumant livestock owners. Priority areas include humane herding and handling, migration planning, welfare-friendly transport, early disease detection, preventive healthcare, and nutritional stress management. Strengthening pastoralist capacity through multilingual, mobile, and culturally appropriate training interventions, combined with improved veterinary outreach, is critical for enhancing animal welfare, protecting livelihoods, and integrating transhumant systems into national and international animal welfare frameworks.

Animal transport continues to pose welfare challenges in Europe, where e.g. the assessment of animals’ fitness for transport relies heavily on veterinary science. Evaluating welfare during transport involves interactions between clinical indicators, behavioural signs and legal requirements. Despite the centrality of veterinary expertise to these assessments, little is known about how veterinary science is applied and interpreted when welfare concerns lead to legal action. In Denmark, the Danish Veterinary Health Council (VHC) has, for nearly a century, advised criminal justice authorities on the severity of animal welfare offence cases, including those related to transport. This long‑standing practice has never been systematically examined, leaving an important gap in understanding how veterinary assessments influence legal decisions in regard to animal welfare during transportation. This project investigates what types of veterinary science and evidence are used in Danish legal cases concerning animal transport – and how it is used. The study forms part of an interdisciplinary research project integrating veterinary, legal, and anthropological perspectives to explore how knowledge about animal welfare during transport is produced, interpreted, and applied across institutional settings. Using data from legal cases, we analyse the types of veterinary evidence submitted—such as clinical findings, post‑mortem evaluations, behavioural observations, and assessments of fitness for transport—and examine how these are weighed in determining legal violations. A literature review evaluates the scientific evidence related to common clinical conditions observed during transport, highlighting potential discrepancies between veterinary science and its practical use in legal case evaluations. The aim is to clarify how veterinary science is used in legal decision-making and how it reflects contemporary veterinary knowledge. Preliminary findings from the case review, including animal species, animal category and the types of violations involved, will be presented at the conference.

Large-scale operational datasets from animal welfare organisations offer substantial opportunities to understand current animal welfare challenges at scale. Working in partnership with the Scottish Society for the Prevention of Cruelty to Animals (Scottish SPCA), this project draws on records of over twelve thousand dog admissions over seven years to generate insights that support proactive and targeted approaches to improving dog welfare in Scotland. The first stage of the project built a detailed picture of the dogs entering and moving through the Scottish SPCA’s Animal Rescue and Rehabilitation Centres (ARRCs). The data recorded at intake and throughout their stay captures key characteristics of the dogs and their subsequent journeys. Preliminary analysis indicates that some of these characteristics—specifically body size, age and sex—are associated with differences in fostering, with smaller, younger, female dogs being more likely to be fostered than other groups. Breed and breed-related factors such as body size, coat type and elements of conformation (including cephalic index) are also explored. This stage establishes a population-level profile of the dogs entering the system and an understanding of how breed-related conformational characteristics impact their outcomes. The second stage will examine how admissions and dog profiles vary across Scotland over time. Patterns in arrivals and movement through the ARRC system will be explored by season, year and geographic area, with the aim of identifying temporal trends and potential “hotspots” for particular dog types and cases. This work aims to support the Scottish SPCA by enabling the anticipation of “busy seasons” and emerging or intensifying welfare challenges, allowing staffing, foster carers, outreach activities, and kennel capacity to be planned for more effectively - all for the purpose of improving dog welfare. It also provides evidence of the charity’s impact to support fundraising efforts, as well as offer compelling data for policymakers considering future dog-related legislation. In the longer term, this research will be extended to other species arriving at Scottish SPCA centres and enriched by incorporating relevant contextual factors such as social, economic and policy shifts into the analysis. Ultimately, the goal of this project is to translate this rich dataset into positive impact that improves the lives of animals across Scotland.

Use of environmental enrichment to support natural hiding behavior, and its potential effects on welfare, have not been thoroughly investigated in artificially-reared dairy goat kids. This study evaluated whether the presence versus absence of a cardboard hide in the home pen of pair-housed goat kids impacted early weight gain and behavior during separation from their pen-mate. A total of 50 Saanen, Alpine, and Lamancha breed and crossbred pair-housed kids were allocated to either a pen with a solid-sided cardboard hide (H) occupying half of the pen (n=12) or a pen without a hide (NH; n=13) using a matched-pairs design followed by random allocation. Eight pens were arranged in a row balanced by group, with hide placement (left or right) counterbalanced. Kids were enrolled on a rolling basis as they were born and remained in the pens for 10 days, with weights collected at the start and end of the study period. On day 8, one kid per pen was randomly selected and removed for 2-5 min for another test, leaving the focal kid without visual access to its pen-mate. On day 9 the other kid was removed such that each kid had a turn with removal/separation. The focal kid was video recorded and the first 1.5 min of this separation period was later scored to compare how kids respond to this common early life stressor. Statistical analyses included descriptive statistics and regression modelling. During separation, focal kids in group H spent an average proportion of 93.2% (SD: ±12.4%) of time on the open side of the pen, whereas NH kids numerically spent more time on the side of pen-mate removal, averaging 60.0% (SD: ±32.6%). Only two kids (one per group) showed lying behavior during separation, while average rates ± SD of locomotion and standing vigilant were H: 20.6%±11.5%, NH: 19.8%±13.7%; H: 42.9%±23.2%, NH: 45.2±24.1%, respectively. A mixed logistic regression indicated no difference in rearing behavior (OR=4.1, 95% CI [0.9, 18.5], p=0.06) nor urinations (OR=0.7, 95% CI [0.1, 4.9], p=0.68) between groups. A linear mixed model indicated no difference in weight gain between groups (F(1, 19)=0.09, p=0.77). Overall, the presence of a cardboard hide did not impact early weight gain or behavioral responses to pen-mate separation. Further research will assess kid behavior in the home pen to elucidate the implications of hide provision for supporting the natural behavior and welfare of dairy goat kids.

**Introduction** Feline overweight and obesity are major welfare concerns and are typically attributed to imbalances between energy intake and expenditure. While dietary and activity-related risk factors are well documented, the role of the human–animal relationship and broader environmental context in shaping feline body condition remains less well understood. This study examined how social, behavioural, and environmental factors jointly relate to Body Condition Score (BCS) in domestic cats. **Methodology** Data were collected from 111 cats and 77 owners using a mixed-methods design combining an online questionnaire, photographic assessment, home visits, and a photo-interview. During home visits, BCS of 59 cats was assessed using the Royal Canin 9-point scale. Multinomial logistic regression analysis showed associations between BCS and owners’ age, play frequency, gender, neutering state, and indicators of the human–animal relationship. Owners’ ability to assess BCS was evaluated by comparing their ratings with those of the researchers using six photographs of unfamiliar cats. Inter-rater agreement within both groups was quantified using the intraclass correlation coefficient (ICC). **Results** Healthy BCS was associated with lower owners’ age (p<0.001); higher BCS was associated with higher play frequency (p<0.001), and with neutering (p<0.001). A stronger human–animal relationship was associated with an optimal or higher BCS (p<0.001). Overall, owners were generally able to assess their cats’ BCS accurately; however, when misjudgements occurred, they predominantly underestimated body condition. **Conclusions and implications** These findings indicate that feline body condition is shaped by an interplay of biological, behavioural, and relational factors. The association between a stronger human–animal relationship and BCS suggests that social context may play a protective role against under- and overweight and merits further investigation. Targeted owner education focusing on nutrition, activity, and body condition monitoring is essential for effective obesity prevention and improved feline welfare.

During the last decade, research on tear staining (TS) in pigs has provided evidence that it may be a potential new welfare indicator, with TS associated with certain stressors and other welfare issues. However, questions remain on how reliably tear stains can be scored. Existing scoring scales are discrete and compare the size of the stain relative to the eye. While in reality, TS size varies along a continuum. Further, our previous work with naïve assessors scoring with a continuous scale resulted in weak to moderate reliability. Therefore, this study assessed intra- and inter-observer reliability after training using a 0 to 100 mm continuous tagged Visual Analog Scale (tVAS) based on the discrete DeBoer-Marchant-Forde 6-point scale. The test was performed using videos of both left (L) and right (R) side stains on pigs taken before the scalding tank at a Dutch slaughterhouse. A trained person preselected pigs across the available scoring range, which were in turn scored by two other trained people. Their three scores were averaged as a silver standard. After selection, an intra- and inter-observer reliability testing video was produced with 44 pigs (28 unique pigs + 16 repeats), with a score range of 1 to 65 mm for  L  and 1 to 63 mm for R. Ten assessors with varying experience in scoring pig welfare were trained, via one in-person session with live pigs (familiarisation with scale and pig morphology) and five online exercises (step-wise scoring familiarisation and calibration). The intra- and inter-observer reliability of the assessor scores of the final test was analysed using intra-class correlation coefficients (ICC [95% CI]) for L and R stains separately and combined. Inter-observer reliability was deemed moderate to good across the L (0.706 [0.585, 0.822]), R (0.689 [0.565, 0.810]), and the overall dataset (0.694 [0.607, 0.780]). The general intra-observer reliability across raters (Eliasziw et al.’s 1994 model; no 95% CI) was good for L stains (ICC: 0.807), and moderate for R stains (ICC: 0.640) and overall (ICC: 0.724). This indicates that assessor-related variability remains important when drawing conclusions from TS results scored with a tVAS, even after training. Compared with naïve assessors, training and calibration do seem crucial for improving reliability. Improving training and the tVAS could potentially even further advance this. In conclusion, trained assessors can score TS using a continuous scale with moderate to good reliability.

Holistic concepts of animal welfare include affective states as animal-based outcome measures that reflect how animals themselves experience their environments, including the balance of positive and negative emotional states. The difficulty remains in how to measure affective states in most animals. Traditional methods, such as detailed behavioural coding or cognitive bias testing, are often suggested but can be resource-intensive and challenging for everyday use. Animal staff ratings have been proposed to measure animal welfare, but may be difficult to score reliably and consistently across taxa. Additionally, the implementation of positive welfare-related actions may be difficult based on arbitrary scoring systems. We introduce ZOOMOOD, a simple app designed to collect subjective ratings of animals using adjectives derived from Qualitative Behavioural Assessment (QBA). The app translates staff familiarity and observational expertise into structured affective profiles without requiring complex statistical processing, making it suitable for routine welfare monitoring in daily care. By focusing on the balance between positive and negative affective descriptors, ZOOMOOD offers a pragmatic early-warning welfare indicator that can inform more targeted follow-up assessments or interventions when needed. We present preliminary results of the first trials of the app's usage in real-life contexts and discuss how staff-driven affective monitoring can be used as a welfare-relevant practice that is practical for busy animal professionals and can contribute to more animal-centered decision-making in everyday management.

Heat stress is a major cause of transport-related distress and mortality in farm animals. Rising temperatures across the EU have raised concerns about whether current regulations adequately protect animals during transport. The EU Commission recently set a proposal for new regulatory measures (currently under discussion). These include, for example, limiting daytime journeys to a maximum of 9 hours when temperatures range between 25°C and 30°C (down from the current threshold of 30°C ±5°C), restrict transport at temperatures above 30°C to nighttime hours (21:00–10:00) whilst also increasing space allowance by 20%. To assess whether proposed regulations are future-proof, we conducted a systematic review including 98 peer-reviewed studies covering seven farm animal species. The literature provides insight into thermal thresholds, mitigation strategies such as shorter journeys, increased space allowance and nighttime transport. Two-thirds of studies reporting heat stress observed adverse effects at temperatures below 30°C. Heat stress was observed in both short and long journeys, although journey duration is a known predictor of heat stress severity. These findings support that long journeys should be limited to lower temperatures than those permitted under current regulations. However, thresholds varied widely (16.6–42°C), reflecting differences in species, methods, indicators, and transport conditions such as humidity, ventilation, stocking density, and journey duration. Internal trailer temperatures consistently exceeded ambient conditions (by 1-18°C), challenging the notion that regulatory proposals should rely solely on forecasted outdoor temperatures. Rest stops often increased internal temperatures. In contrast, night-time transport and increased space allowance, both considered in proposed regulatory changes, generally reduced heat stress, although evidence remains limited and some species-specific differences may apply. Major data gaps persist for rabbits, horses, turkeys, goats, and cattle. Our findings suggest that lowering the temperature threshold to 25°C would better protect animals, but proposed regulations may still fall short, particularly for vulnerable animals (e.g. young, sick, lactating). This review provides a consolidated evidence base to inform ongoing policy discussions and the development of science-based regulations that more effectively account for heat stress and animal welfare. Beyond the European context, these results are relevant to animal transport systems globally and can support regulatory evaluation and refinement, particularly in regions facing increasing temperatures associated with climate change. It also highlights key knowledge gaps and priorities for future research to improve animal welfare and the management of heat stress during transport.

The objective of this study was to evaluate whether early-life characteristics are associated with recovery from diarrhea in neonatal dairy calves. Eighty-six Holstein calves were group housed (15 calves/pen), fed ad libitum via automated milk feeders, and monitored daily for disease from 3 to 28 d of age. Activity was continuously recorded using accelerometers placed at 3 d of age, and body weights were collected weekly. Diarrhea was diagnosed using fecal scoring (loose feces for ≥ 2 d or watery feces for ≥ 1 d) and classified as mild or severe based on clinical signs of dehydration or depression. Early-life characteristics included colostrum drinking ability (suckling strength), drinking speed, days to learn the feeder, milk intake, ADG, and activity during the first 7 d of life. Recovery was defined by the number of days to resolution of clinical diarrhea for > 2 d and days required to return to baseline milk intake and activity. Associations between early-life characteristics and recovery outcomes were evaluated using mixed linear regression models with pen as a random effect while accounting for age at diarrhea onset, diarrhea severity, birth weight, order of entry into the pen, and serum Brix%. Mildly diarrheic calves were diagnosed at 12.5 ±3.8 d of age and had a diarrhea duration of 3.0 ±2.0 d, whereas severely diarrheic calves were diagnosed earlier at 9.4 ±1.5 d of age and had a diarrhea duration of 5.0 ±1.5 d. Later age at diarrhea onset (β = −0.24 d, *P* = 0.05 ), earlier entry into the group pen (β = −0.11 d, *P* = 0.04), and faster feeder learning (β = −0.41 d, *P* = 0.05) were associated with fewer days to diarrhea resolution, whereas greater milk intake during the first 7 d of life was associated with longer time to diarrhea resolution (β = 0.52 d, *P* = 0.05). There were trends toward faster return to baseline milk intake among calves with greater milk intake, faster drinking speed, and quicker feeder learning (β = -0.10 d for each; *P* = 0.09). No associations were observed with return to baseline activity. Overall, these findings suggest that early-life behaviors related to learning ability and feeding may provide insight into a calf’s resilience and capacity to recover from disease, potentially allowing for earlier identification of calves that require additional support during illness.

Farm animal welfare has been, and is currently, impacted by selective breeding. At the same time, there can be immense opportunities to improve farm animal welfare using careful and sustainable selective “welfare-breeding” strategies. However, progress in this area is limited by challenges in collaboration such as different definitions of animal welfare or discipline-specific constraints (i.e., data volume, feasibility). These differences in definition have been highlighted by a systematic review of the scientific literature across welfare, production and genetics domains. Therefore, a series of six workshops were designed to facilitate a dialogue between livestock industry stakeholders on the topic of animal welfare in animal breeding. Workshop 1, in 2 parts, attempted to produce separate definitions and visions for animal welfare in sustainable breeding from animal welfare scientists and animal geneticists at Danish institutions, respectively. Workshop 2 brought these parties together for a facilitated discussion of the similarities and differences between the two positions, with the goal of producing a combined roadmap for collaboration that respects the visions and constraints of each scientific discipline. Workshops 3-6 expand on this combined position on animal welfare in sustainable breeding through discussions with other industry stakeholders; an international industry (incl. breeding companies) perspective (W3), social scientists and ethicists (W4), and in the future with animal science and veterinary students (W5). Finally, workshop 6 includes policy makers and influencers (i.e., European Forum of Farm Animal Breeders). The overarching goal of this initiative is to improve communication between livestock industry stakeholders on the subject of animal welfare in animal breeding, and to open a dialogue about how to practically and feasibly conduct research and make improvements in this space using methods agreeable to all parties. We here present results of the systematic literature review and perspectives from the workshops 1-3, giving a snapshot of stakeholder views on welfare breeding in the Danish farm animal and genetics fields.

The welfare of animals is one of the most challenging debates affecting animal industries today including agricultural production, pet care, equine activities and fish farming. As animal welfare becomes increasingly important within modern society with current public concerns, it requires organisations to provide methods of objectively monitoring and addressing welfare. The Animal Welfare Assessment Grid (AWAG) ([www.awag.org.uk](http://www.awag.org.uk) ) is a validated user-friendly cloud-based tool which encompasses the five domains of welfare and draws attention to the temporal component that is often overlooked. It assesses the four parameters of physical health, psychological wellbeing, environmental quality, and veterinary/management procedural events.  It is used in a wide range of species and types of animal use including zoos, pet dogs, racing greyhounds, horses, and research laboratories providing objective evidence of improving or deteriorating welfare. This knowledge and the visual representation of the animal’s welfare state can be used to ensure they have the best possible quality of life, and to support and maintain public trust about health and welfare. Outreach to promote its use will increase knowledge and also provide evidence-based education to promote and maintain public trust around welfare such that a clear transparency of welfare standards could be established in the industries to meet public concerns. The integration of precision livestock technology will facilitate input of data routinely collected. This could provide a beneficial impact to UK trade and profitability through improved public confidence and increased sales of British food with high welfare standards. Different sectors have different approaches to welfare. Some focus standards primarily on productive and physically healthy animals. The general public is often distracted by the cuteness of the animal which actually can be detrimental to its welfare. Animal welfare is occasionally considered in parliament, but it has not been a major political concern, although the DEFRA Animal Welfare Strategy was published recently. This includes action points about regulations but there is there is an urgent need to build greater trust in the regulators by improving enforcement. The AWAG has been cited as a potential tool when investigating animal welfare offences and is also relevant to food labelling for which the outputs from the AWAG could be applied. An effective welfare strategy needs to include - welfare standards, - effective communication and education to respond to evolving social and political expectations, - robust data and evidence, required to drive improvements, tackle areas of risk, and provide the proof of performance on welfare.

Gestation housing for sows remains one of the most contested animal welfare issues in U.S. livestock production. Legislative bans on gestation stalls have expanded rapidly, with eleven states currently prohibiting their use and additional states restricting the sale of pork produced using stalls. Scientific evidence has increasingly documented welfare concerns associated with prolonged stall confinement, including chronic stress, behavioral restriction, and musculoskeletal problems. While group housing systems can alleviate some confinement-related harms, they may also introduce distinct welfare risks such as injury and social stress from agonistic interactions, particularly in suboptimally designed or managed facilities. To date, however, no study has quantified how these diverse welfare impacts compare in magnitude, or how their impact accrues across the lifespan of sows. Accordingly, the net welfare difference between housing systems, and the degree to which it depends on implementation and management, remains unknown, precluding quantification of the welfare gains that housing reforms deliver and identification of key practices to realizing them. To address this gap, this study applies the Welfare Footprint Framework (WFF) to quantify the cumulative welfare impacts of gestation housing systems. The WFF translates diverse welfare outcomes into a common currency - time spent in negative and positive affective states - by integrating the intensity, duration, and prevalence of each experience encountered in commercial production. This approach enables explicit, transparent comparison of welfare trade-offs across systems that differ in risk profiles. The analysis covers large-enterprise operations across thirteen U.S. states representing 97% of the national swine population. Over 40 negative affective experiences — including pain from lameness, gastric, integumentary, reproductive, respiratory, and circulatory conditions, thermal stress, hunger, social and handling stress, and frustration from thwarted foraging, locomotion, social, resting, and feeding motivations — and 10 positive experiences (e.g., fulfillment of motivated behaviors, positive social interactions) are quantified for each scenario. One gestation stall scenario is compared with three group housing scenarios designed to bracket realistic U.S. practice: (1) poor-practice group housing, reflecting features associated with high welfare risk; (2) best-practice scenario, incorporating commercially achievable strategies known to mitigate major risks; and (3) an intermediate scenario defined by moderate probabilities of key welfare challenges. In addition to quantifying the cumulative welfare impact of each housing scenario, per animal and per Kg of pork meat produced, the analysis also quantifies the cost-effectiveness of housing reforms.

Global amphibian declines have declined sharply, leading to increased reliance on ex situ conservation programmes. As amphibians are increasingly maintained under human care, understanding how housing conditions influence behaviour is important for informing welfare-oriented husbandry decisions. Despite this, the relevance of housing conditions for amphibian welfare remains underexplored. This study investigated the behavioural responses of two fire salamander subspecies, *Salamandra **salamandra** **bernardezi* and *S. s. terrestris*, to different pair-housing conditions in captivity, with the aim of exploring behavioural patterns with potential welfare relevance. Thirty-four individuals (20 *S. s. **bernardezi*, 14 *S. s. terrestris*) each participated in five trials: one single-housing, two same-subspecies pairings, and two mixed-subspecies pairings. An ethogram was developed to categorise observed behaviours, which were interpreted along behavioural axes commonly used in animal personality research, including exploration versus avoidance, sociability, and aggression. Salamanders exhibited higher levels of exploratory behaviour under pair-housing conditions compared to single housing, a behavioural pattern commonly associated with positive welfare states in captive animals across taxa, though not validated as a direct welfare indicator in amphibians. Interactive behaviours varied by subspecies combination, and a previously undocumented foot movement was recorded in some pairings. These findings suggest that pair housing may influence behavioural expression in captive fire salamanders and may have relevance as a form of enrichment under appropriate conditions. However, interpretation of these behaviours in welfare terms requires caution, and further research is needed to clarify how social composition and individual variation shape behavioural outcomes, supporting future welfare-oriented amphibian husbandry.