Rearing density influences pig productivity and welfare, but its behavioral and physio-logical effects remain unclear. This study evaluated how increasing space allowance from 0.57 to 0.97 m2/pig affects growth, agonistic behavior, and stress in growing pigs. Seventy-six 12-week-old pigs were allocated to high or low rearing density (HRD: 12 pigs/pen, n = 4 pens; LRD: 7 pigs/pen, n = 4 pens) for 28 days by varying pig numbers within identical pens. Growth performance was recorded weekly, while agonistic behavior was continuously monitored using RGB cameras and detected with a YOLOv8-based model (overall mAP50 = 0.953; aggression = 0.960, ear biting = 0.927, tail biting = 0.972). Ear base temperature was measured at baseline and twice weekly, lesion scores were assessed at trial completion, and blood biochemical parameters were also assessed. Pigs under LRD exhibited higher (p < 0.01) body weight, daily gain, and feed intake, with a lower feed conversion ratio than HRD pigs. Increased space allowance reduced (p < 0.05) agonistic behavior, lesion scores, plasma glucose, free fatty acids, cortisol, and ear base temperature. These findings indicate that increased space allowance improves growth and welfare and demonstrate the value of AI-based behavioral monitoring in pig production systems.

Behaviour recognition of tail and ear biting in pigs using AI-based computer vision

Slaughterhouse monitoring provides a cost-effective and suitable tool for large-scale surveillance of tail-biting, which is a major welfare issue in pig production. The European Union Council Directive 2008/120/EC prohibits routine tail-docking as a preventive measure against tail-biting. Nevertheless, compliance remains inconsistent, and tail-docking is still widely practiced in Europe. This study aimed to assess the occurrence of tail-biting and tail-docking in slaughtered pigs (n = 15,000) from Italy, Netherlands and Spain using digital images. Results indicate that most pigs were tail-docked (88.1%), with substantial variation among countries: tail-docking was most common in Spain (99.4%), followed by Netherlands (86.5%), and least common in Italy (78.5%). Overall, tail-biting lesions were observed in 5.4% of pigs, with the highest prevalence in Italy (11.6%), followed by Netherlands (3.4%), and Spain (1%). The differences among the three countries were significant (p < 0.0001), tail lesions being more frequent in pigs with undocked tails than docked tails (p < 0.0001). The risk of having a lesion was substantially higher in pigs with undocked/intermediate tails (relative risk = 4.6). The severity of lesions was scored using two different methods, which showed an almost perfect agreement (weighted Cohen's kappa coefficient 0.826; p < 0.0001). Lesions were most frequently detectable in the two lateral views, whereas the central view alone was inconclusive in most of pigs (99%).

Weaning is a major stressor for pigs, often increasing agonistic behaviors such as aggression, ear biting, and tail biting, which can impair growth and welfare. This study evaluated the combined effect of rubber sticks and Italian ryegrass hay as environmental enrichment (EE) on growth performance, agonistic behavior, ear and tail biting lesion development, fecal consistency, and blood biochemical parameters. A total of 64 pigs (8 pigs × 4 pens × 2 groups) at 7 weeks of age were assigned to control (without EE) and treatment (with EE) groups for four weeks. Pens were the experimental unit for growth, fecal scores, lesion scoring, and behavioral outcomes. Growth and fecal consistency were measured weekly, while ear and tail lesions were scored at the end. Agonistic behavior was quantified using overhead RGB cameras and a YOLOv8-based AI system with high accuracy, mAP50 = 0.953, validated against manual observations, with behavioral outputs aggregated at the pen level from a single representative pen per group. Combined EE reduced lesion severity, lowered free fatty acids, improved fecal consistency, and decreased agonistic behavior, without affecting growth. AI-based monitoring offers a promising tool for quantifying social stress, although further studies with greater pen-level replication are warranted.

Evaluating loose materials and space allowance to improve welfare in commercial weaner and finisher pigs.

This study investigated the effect of group size on tail damage and growth performance in growing-finishing pigs with intact tails. A total of 432 pigs were housed indoors on fully-slatted floors and assigned to either small (nine pigs per pen) or large (18 pigs per pen) groups, with equal space and resource allocation per pig. No environmental enrichment was provided. From nine to 23 weeks of age, pigs were monitored weekly for tail injuries using a 5-point scale (0 = no injury, 4 = partial or total loss). The most severe score observed during each four-week period was used for analysis, and outbreaks were defined as the occurrence of one or more pigs per pen with a tail score ≥ 2. Group size did not influence average daily gain, feed intake, or feed to gain ratio. However, pigs housed in small groups experienced more frequent and severe tail injuries, including a higher proportion of removals due to tail wounds. In contrast, pigs in large groups were more likely to receive healed tail scores (score 1) or mild injuries (score 2), and experienced fewer removals. While these results suggest that tail damage may be less severe in larger groups, the total number of pigs affected by tail biting was similar across treatments. These findings highlight the importance of managing tail-damage severity and suggest that group size can influence welfare outcomes in systems where pigs are raised with intact tails.

Activity and synchrony patterns obtained by a tracking-by-detection algorithm as potential predictors of tail biting at pen and individual level in pigs.

Leveraging microbiota-gut-brain axis as a novel nutritional strategy to promote stress resilience and welfare in swine production: A review.

7. Characterization of vocalisations associated with behaviours related to tail biting

Unravelling bio-climatic thermal stress driven behavioral pattern shifts in crossbred pigs.

91. Metabolic biomarkers associated with lysine, methionine, or threonine deficiencies, growth and tail biting in intact-tail pigs

89. Measuring temperatures in pig sheds to predict tail biting: do we need experimental weather stations for accurate data?

85. A systematic review of risk factors for tail biting in pigs

87. Animal, time and place: non-genetic risk factors for tail biting in Australian pig production

Inter-alpha-trypsin inhibitor protein heavy chain 4 (ITIH4), also named porcine major acute phase protein (Pig-MAP), is a positive acute phase protein (APP) in pigs and can be measured in plasma and also in saliva to assess the animal health. The objectives of this report were (1) to evaluate if different commercially available porcine ELISA kits can measure ITIH4 concentrations in saliva samples of pigs, and (2) to study the distribution of ITIH4 components in saliva and compare it to the distribution in plasma. The results showed that two of the ELISA kits used in this report could measure salivary ITIH4 with precision and accuracy, but only one showed significant differences between pigs with tail biting and control animals without this condition. Western blotting analysis revealed the presence of a different distribution of ITIH4 bands in saliva samples compared to plasma samples. In conclusion, in saliva of pigs ITIH4 can be quantified using a commercial ELISA kit increasing its concentration in cases of tail biting. In addition, ITIH4 shows bands at western blot in saliva that are different from serum but are compatible with different ITIH4 forms. These data confirm that ITIH4 can be detected in saliva and be potentially used as a biomarker of inflammation.

Abstract Understanding the dynamics of tail biting in pigs, such as which pigs are involved and what roles these pigs play in this damaging behavior, can help us develop management strategies to minimize tail biting outbreaks. We evaluated social structures and positions of pigs involved in tail biting outbreaks (TBO) using social network analysis. Pigs (n=252, initial weight=28.9±3.4 kg) with intact tails were grouped (7 pigs/pen, 36 pens) based on litter origins (littermates, half-group littermates, non-littermates). Pigs were housed in barns with slatted floors (0.98m2/pig) for 12 weeks until market weight (123.1±11.9 kg). Behavior of pigs was recorded continuously using the NUtrack Livestock Monitoring System. Tail injury was evaluated weekly to monitor TBO. Thirteen pens (littermates=3 pens, half-littermates=5 pens, and non-littermates=5 pens) experienced TBO in which at least one pig in a pen had blood on the tail caused by tail biting. Videos for these pens, recorded the day before and during the first TBO, were manually reviewed to document tail biting events and identify the roles of the pigs involved. The top 25% of pigs responsible for the most tail-biting incidents were classified as tail-biters and others as non-biters. Network metrics were calculated using RStudio. Data were analyzed using Glimmix and NPAR1WAY procedures. No difference in network metrics at the pen-level (all P ≥ 0.128; Table 1) was detected among litter origins. Average density ranged from 0.27 to 0.33, indicating that 27 to 33% of possible pairs (dyads) of pigs engaged in tail biting. Among these pairs, 17 to 37% (reciprocity) were biting each other. Out-, in- and all-degree centralizations were closer to 0 than 1, indicating that tail biting was not performed mainly or received by few pigs. At the pig-level, no difference was detected in unweighted or weighted centralities among litter origins (all P ≥ 0.32). Compared to non-biters, tail-biters had greater unweighted out-degree, closeness, and betweenness, and greater weighted out-edge strength and betweenness centralities (all P ≤ 0.04, Table 2), indicating that tail-biters bit more pigs, were more connected to other tail-biters, and committed more tail biting events. No difference was detected in unweighted in-degree, weighted in-edge strength, out- and in-closeness centralities between tail-biters and non-biters. Both tail-biters and non-biters were bitten about 3 times (in-edge strength) by 1.76 pigs (unweighted in-degree centrality) on average, suggesting that tail-biters were bitten as often as non-biters. Gilts tended to have greater (all P ≤ 0.08) unweighted closeness and betweenness, weighted out-edge strength and betweenness centralities than barrows, suggesting that gilts were more closely connected with other pigs that were involved in tail biting and performed more tail biting than barrows. These results demonstrate that social network analysis can provide insight into the dynamics of tail biting among pigs.

Abstract The concept of animal welfare has fluctuated over time, following changes in values and beliefs, together with advances in the scientific understanding of animals. A relevant driver in the new definitions of animal welfare has been the higher interest in including affective states, especially positive ones, under a new approach known as Positive Animal Welfare (PAW). Under PAW approach promoting animal welfare goes beyond minimizing suffering or the avoidance of negative states, because an absence of suffering does not mean that all the proclivities of an animal are being fulfilled; particularly, it can overlook the potential benefits of providing environments that enable animals to express their whole behavioural repertoire. The sustainability of animal production systems is often defined by their economic and/or environmental impact. However, the Sustainability Assessment of Food and Agriculture Systems outlines two additional pillars of sustainability: governance and social. The social pillar covers the quality of life for both humans and animals involved in the system, and the promotion of positive animal welfare (PAW) is, therefore, also inherent to making a system more sustainable. Under the framework of the European Partnership of Animal Health and Welfare (EUPAHW), the trade-offs and synergies between systems promoting animal welfare with the other sustainability pillars have been explored. Three examples of best practices identified in the EUPAHW that promote PAW or help prevent negative welfare outcomes in pig production will be presented, along with their links to sustainability: non tail-docking, free-farrowing and provision of outdoor access. Tail docking is a common practice in European intensive production systems, although routine tail docking is not allowed by legislation; being considered a painful procedure not tackling the underlying causes of tail biting. At present, European farmers are conducting trials to develop strategies to prevent tail biting in undocked pigs, which have direct consequences on the sustainability of the production system. Besides, the European Citizens’ Initiative “End the cage age” was brought to the EU commission, who initiated plans to prohibit cages and improve animal welfare in multiple farmed animal species. In farrowing and lactating sows, crating has been recognized as severely restricting postural movements, nest building and maternal behaviours and, thus, inducing stress and frustration. The new housing designs under zero or semi confinement for sows will be presented, with regards to its possibilities to promote PAW and its potential impacts in sustainability. Finally, the conventional keeping of growing pigs in an indoor, climate-controlled building with fully or partly slatted floors and minimum space allowance has been said to limit the fulfilment of ethological and physiological needs, such as foraging, and will be presented as an example of best practice promoting PAW, together with its sustainability trade-offs.

Tail biting in pigs remains a persistent welfare and economic challenge in commercial pig production, despite decades of research. This harmful behavior is driven by a combination of genetic, environmental, and management factors, resulting in injuries, infections, and reduced growth performance. Tail-bitten pigs showed 10-21 g/day lower average daily gain compared to non-bitten pigs. Additionally, tail-bitten pigs exhibited poorer feed efficiency, resulting in increased production costs. Farmers have traditionally relied on tail docking to prevent tail biting. However, this practice causes pain and fails to fully address the issue, as studies show that 96% of docked pigs still exhibit minor tail damage. To combat tail biting effectively, a holistic approach is necessary. Providing manipulable substrates, such as straw, ropes, or wood, reduces tail biting by 65-88%. Ensuring adequate feeder space and improving ventilation are also key environmental interventions. Enhanced human-animal interaction during rearing has also been shown to reduce tail biting in piglets. Beyond housing adjustments, Precision Livestock Farming (PLF) technologies, including automated monitoring systems, enable early detection of stress and behavioral changes by monitoring tail posture (e.g., 3D cameras, RFID sensors), allowing timely intervention. Nutritional strategies, such as balanced protein/amino acid levels, further support pig health and reduce aggression. Genetic selection offers long-term potential by breeding for resilience traits (h² = 25-36%) and pigs less prone to tail biting, while improved husbandry practices strengthen overall herd resilience. When outbreaks occur, rapid response protocols such as isolating affected pigs and increasing enrichment help minimize damage. Farmer education is critical to ensure these strategies are properly implemented. By integrating environmental, technological, nutritional, and genetic solutions, the pig industry can move away from tail docking while improving welfare and productivity.

Promoting sow-piglet information transfer and playful foraging by changes in pen design and feeding strategy to ease weaning: effects on pig performance and behaviour.

Using sensors, the health and welfare of growing-finishing pigs can be continuously monitored by detecting deviations from pigs' normal behaviour, but the validity of such algorithms requires improvement. As changes in the environment influence pig behaviour, monitoring temporal changes in environmental factors may help identify periods with a higher risk of welfare issues. The real-time relationships between pig welfare and many environmental factors are, however, not well-understood. This study examined the short-term associations of ambient ammonia with indicators of pig production and health. Ambient ammonia concentrations were monitored with sensors during the growing-finishing period of three rounds at a German (farm G, n = 110 pigs/round) and one round at a Danish farm (farm D, n = 144 pigs). Body weight was estimated daily using 3D cameras (both farms), feed intake was recorded using electronic feeding stations (only farm G), and health indicators were recorded during twice- (farm G) or thrice-weekly (farm D) farm visits. Using splines (generalised additive models), ammonia concentrations were linked to indicators of pig production and health in real time and, for body weight, at a lag of 1, 2, 3 and 7 d. We found a range of relationships between ambient ammonia (5 - 50 ppm) and production or health indicators (i.e. body weight (real-time and lagged), feed intake, coughing, sneezing, pen fouling, diarrhoea, and tear staining), but they were highly inconsistent between farms and pig rounds. Part of this inconsistency may be due to differences in manure management and sensor locations, or could be explained by age or seasonal effects (e.g. heat stress). More robust relationships were identified for clinical measures related to pig behaviour, where tail damage and skin lesions linearly increased with ammonia from low concentrations (5 - 10 ppm) onwards, hence suggesting more tail biting and aggression at higher ammonia concentrations. In conclusion, ambient ammonia did not clearly associate with pig performance and health in the short term, while higher ammonia concentrations were related to higher occurrences of clinical signs reflecting undesirable behaviours. Therefore, daily ammonia measurements using sensors may be of limited value in identifying health issues in pigs, but they may aid in detecting periods with high risk of aggressive or tail biting behaviours that require interventions. As ambient ammonia was confounded with other environmental measurements, such as ambient temperature or carbon dioxide concentrations, identified associations should be interpreted cautiously or with ammonia as general indicator of air quality.