Le bâillement, du réflexe à la pathologie Le bâillement : de l'éthologie à la médecine clinique Le bâillement : phylogenèse, éthologie, nosogénie  Le bâillement : un comportement universel La parakinésie brachiale oscitante Yawning: its cycle, its role Warum gähnen wir ?   Fetal yawning assessed by 3D and 4D sonography Le bâillement foetal Le bâillement, du réflexe à la pathologie Le bâillement : de l'éthologie à la médecine clinique Le bâillement : phylogenèse, éthologie, nosogénie  Le bâillement : un comportement universel La parakinésie brachiale oscitante Yawning: its cycle, its role Warum gähnen wir ?   Fetal yawning assessed by 3D and 4D sonography Le bâillement foetal http://www.baillement.com

mystery of yawning 

 

 

mise à jour du
18 août 2025
Curr Zool
2024;71(2):137-151
Yawning in sync:
implications for social cohesion in horses
Alice Galotti, Martina Romano,
Paolo Baragli, Elisabetta Palagi

Chat-logomini
Abstract
The increasing interest in the study of spontaneous (SY) and contagious yawning (CY) was so far focused on several taxa, especially primates. Here, we focused on SY and CY in horses, a suitable species due to their complex social dynamics that has been largely overlooked in research on these phenomena. By analyzing videos of 48 horses on pasture, we identified 2 yawning morphologies: Covered (Y __) and Uncovered Teeth (Y UcT). Using EquiFACS, we quantitatively demonstrated that Y __ and Y uct differ in terms of muscle recruitment. Moreover, we provide the first evidence for the presence of CY by comparing 2 different conditions: chewing-yawn-chewing versus chewingchewing-chewing. Supporting the Social Modulation hypothesis, in our mares, CY was more prominent among subjects sharing good relationships. Moreover, subjects responded more rapidly to kin compared with non-kin and kin frequently grooming each other responded even more rapidly to each other yawns. The high familiar yawn sensitivity can provide selective advantages increasing behavioral synchronization and group cohesion.
 
Résumé
L'intérêt croissant pour l'étude des bâillements spontanés (SY) et contagieux (CY) s'est jusqu'à présent concentré sur plusieurs taxons, en particulier les primates. Ici, les auteurs se sont concentrés sur le SY et le CY chez les chevaux, une espèce appropriée en raison de sa dynamique sociale complexe qui a été largement négligée dans la recherche sur ces phénomènes. En analysant les vidéos de 48 chevaux au pâturage, nous avons identifié 2 morphologies de bâillements : Les dents couvertes (Y __) et les dents non couvertes (Y UcT). En utilisant EquiFACS, nous avons démontré quantitativement que Y __ et Y uct diffèrent en termes de recrutement musculaire. De plus, ils fournissent la première preuve de la présence de CY en comparant 2 conditions différentes : mâcher-yawn-mâcher versus mâcher-mâcher-mâcher. A l'appui de l'hypothèse de la modulation sociale, chez ces juments, le CY était plus important chez les sujets partageant de bonnes relations. En outre, les sujets réagissaient plus rapidement aux proches qu'aux non-apparentés et les proches qui se toilettent fréquemment répondaient encore plus rapidement aux bâillements de l'autre. La forte sensibilité aux bâillements familiers peut fournir des avantages sélectifs augmentant la synchronisation comportementale et la cohésion du groupe.
 
 
Yawning is present in all vertebrates, from the fetal stage to the adult age (Blanton 1917; Schiller 2002; Provine 2005; Matikainen and Elo 2008; Massen et al. 2021; Gallup 2022). It is considered a fixed action pattern being unstoppable, uncontrollable, and morphologically similar across species (Deputte 1994; Walusinski and Deputte 2004; Provine 2010). From a mechanistic perspective, yawning involves basically 3 different phases: 1) a slow and wide opening of the mouth, with a deep inhalation, 2) a quick closure of the mouth accompanied, and 3) by a brief exhalation (Barbizet 1958).
 
Although challenging to distinguish, the functions attributed to yawning can be categorized based on the distinct physiological conditions of the yawner (Physiological domain) and the social situations in which yawning occurs (Social domain) (Guggisberg et al. 2010). Obviously, all these hypotheses, both between and within domains, are not necessarily mutually exclusive, probably being the 2 faces of the same coin (e.g., proximate vs ultimate explanations). Yawning can be implicated in regulating physiological functions, including blood oxygen levels (Guggisberg et al. 2010), thermoregulation (Miller et al. 2010; Gallup and Eldakar 2013), brain cooling (Gallup and Eldakar 2013), states of drowsiness and arousal (Deputte 1994; Zilli et al. 2008; Guggisberg et al. 2010; Gallup 2022). The Thermoregulation hypothesis suggests that yawning helps reduce body and brain temperature, responding to environmental temperature changes (Massen et al. 2014; Eldakar et al. 2015; Massen and Gallup 2016). The Drowsiness hypothesis links yawning to changes in alertness during periods of inactivity, and it is largely explored in both humans (Greco et al. 1993; Provine 2005; Giganti et al. 2010; Gallup and Meyers, 2021) and in different non-human mammals (Otaria flavescens, Palagi et al. 2019; Loxodonta africana, Rossman et al. 2017; Crocuta crocuta, Casetta et al. 2022; Pan troglodytes, Vick and Paukner 2010; and Theropithecus gelada,Leone et al. 2014). The Brain Cooling Hypothesis, a form of thermoregulation, suggests that yawning dissipates heat by increasing blood flow, acting as a radiator (Gallup and Gallup 2008; Gallup et al. 2016, 2020; Massen et al. 2021). Lowering brain temperature improves mental performance and alertness. Yawning promotes cooling through inhaling cool air (Gallup and Gallup 2008; Gallup et al. 2016; Massen et al. 2021). Baenninger (1997) highlights that spontaneous yawning can be also influenced by social context. The Social Distress hypothesis (Altmann 1967; Deputte 1994) states that yawning as well as self-scratching, a self-directed behavior predicting anxiety in mammals (Palagi et al. 2019), rises after agonistic events, suggesting a connection of yawning to anxiety (Melopsittacus undulatus: Miller et al. 2010; Rattus norvegicus:Moyaho and Valencia 2002; Otaria flavescens:Palagi et al. 2019; Lemur catta and Propithecus verreauxi: Zannella et al. 2015; and Theropithecus gelada: Leone et al. 2014).
 
Although yawning has been extensively studied in primate and carnivore species, literature on ungulates is deficient (Baenninger 1997; Gallup 2011). Domestic horses (Equus ferus caballus) are a rare case where yawning has been described (Fureix et al. 2011). The authors found that horses exhibit more yawning and stereotypical behaviors (e.g., Lip play, head shaking and nodding, weaving) in pre-feeding than during/after feeding, suggesting a possible linkage between yawning and anxiety also in horses. In Przewalski horses (Equus ferus przewalskii), a study revealed an association between yawning and aggressive behavior especially in bachelor males that frequently engage in agonistic encounters. Additionally, stallions yawned more than adult females and immature males (Górecka-Bruzda et al. 2016). The authors suggest that stallions' yawning can be triggered by higher testosterone levels and social stress deriving from male-male interactions.
 
Despite the fixed nature of yawning, data exist underlining the presence of a certain degree of variability (Walusinski and Deputte 2004; Guggisberg et al. 2010; Vick and Paukner 2010; Provine 2012; Leone et al. 2014; Gallup et al. 2016; Massen et al. 2021). In primates, larger yawns (also called "threat yawns," Altmann 1967) are expressed more by males than females and occur under high-tension situations (Leone et al. 2014). Moreover, especially in species with high sexual dimorphism, males exhibit larger yawns during territorial defence or competition for females (Zannella et al. 2015, 2017). In primate species, the diverse yawn morphologies, defined based on teeth visibility, are distributed differently as a function of the social context (Pan troglodytes:Vick and Paukner 2010; Theropithecus gelada:Palagi et al. 2009; Macaca sp.: Zannella et al. 2021).
 
Despite the widespread presence of spontaneous yawning across vertebrates (Heusner 1946; Baenninger 1997; Massen et al. 2021), contagious yawning (CY) (the response to others' yawns with a yawn) seems to be present in highly cohesive and social species (Guggisberg et al.2010). CY is described as a response to an innate releasing mechanism (Provine and Hamernik 1986; Bartholomew and Cirulli 2014) and it has been observed between individuals of the same species (e.g., ungulates: Yonezawa et al. 2017; Norscia et al. 2021; primates: Palagi et al. 2009; Demuru and Palagi 2012; van Berlo et al. 2020; carnivores: Romero et al.2014; Wojczulanis-Jakubas et al. 2019; Casetta et al. 2021, 2022; Ake and Kutsukake 2023; birds: Miller et al. 2012) or different species (proboscidates: Rossman et al. 2020; primates: Cambpell and de Waal 2014; Gallup and Wozny 2022; Pedruzzi et al. 2022; carnivores: Romero et al.2013). One study on horses (Equus caballus) provided no evidence for CY in this species (Malavasi 2014).
 
Although the issue is highly debated and far to be completely demonstrated (Massen et al. 2012; Massen and Gallup 2017; Neilands et al. 2020), different authors suggest that the phenomenon of CY could be influenced by social factors (Social Modulation hypothesis): the better the relationship quality of the subjects, the higher their reciprocal susceptibility to CY and shorter response latency (Romero et al. 2014). Most of the data about the Social Modulation hypothesis come from primate species such as redcapped mangabeys (Pedruzzi et al. 2022), geladas (Palagi et al. 2009; Pedruzzi et al. 2022), chimpanzees (Campbell and de Waal 2011, 2014), bonobos (Demuru and Palagi 2012), and humans (Norscia and Palagi, 2011; Palagi et al. 2014).
 
Recent findings suggest that CY helps synchronize group activities. In lions (Panthera leo), yawning is a reliable indicator of the imminent behavioral state changes. CY between the 2 agents promotes their subsequent behavioral alignment (Casetta et al. 2021). Behavioural synchronization between group members is crucial to social life and plays a central role in maintaining inter-individual cohesion (Engel and Lamprecht 1997; Gautrais et al. 2007; King and Cowlishaw 2009). A study based on a social network involving more than 100 feral horses living in a multi-level society showed that subjects belonging to the same unit synchronize their activities (e.g., resting or moving) more than subjects belonging to different unit (Maeda et al. 2021).
 
Synchronisation also contributes to increase efficiency in group vigilance (Pays et al. 2007; Beauchamp 2015) thus reducing predation risk, again especially in ungulates (Equus caballus: Souris et al. 2007; Bos taurus: Sárová et al. 2013).
Here we focus on spontaneous and CY in horses. As a first step, by analyzing spontaneous yawns in female groups of horses on pasture, we aim at investigating the possible presence of different morphs of yawning and which factors can influence their expression (e.g., context, age, dominance rank). As a second step, due to the horse social cohesiveness and propensity to synchronize activities and movements, we expect to find CY in our groups (Hypothesis 1). Moreover, in line with the Social Modulation hypothesis, we expect that CY is more frequent and rapid (i.e., shorter response latency), among subjects acting in a cohesive way thus sharing preferential relationships (Hypothesis 2).
 
Discussion
Despite the highly stereotyped nature of yawning, universally considered as an excellent example of fixed action pattern, there are some differences in the way this behavior can be displayed (Provine 2005). It has already been demonstrated in several primate species that different morphs of yawn exist (Vick and Paukner 2010; Leone et al. 2014; Zannella et al. 2021; Galotti et al. 2024). Yawning in herbivores is still neglected with no studies on yawning in ungulates investigating its variability. Via EquiFACS we tested whether a priori distinction of 2 different yawn morphologies based on teeth visibility (Yawn Uncovered Teeth, LucT; Yawn Covered Teeth, -cr, Supplementary Figure S1) was reliable. By analyzing a subset of well-visible yawns, our results confirm that horses show 2 distinct yawn morphologies (Figure 2A) as it has been found in social primates (Vick and Paukner 2010; Zannella et al. 2021; Galotti et al. 2024).
 
Through a network analysis (NetFACS) we assessed the likelihood of each AU and AD being present in each yawning event showing that Yuct was associated with 9 AUs/ADs, while Yct with 4 AUs/ADs (Figure 2B). The total absence of AU overlapping and the higher number of AUs recruited during Yuct suggest that this form of yawn is more complex than Ycr. In addition, our results can have a direct application. When it is difficult to obtain videos of good quality due to challenging observational settings, teeth visibility per se can be used to discriminate different morphs of yawns, thus providing a useful tool in yawning investigation.
 
Yawning may have communicative functions (Guggisberg et al. 2010) especially in those species that live in large social groups and with complex social dynamics, as horses (Maeda et al. 2021). In this perspective, in primates, different yawn morphologies have been found to be associated with different social contexts (Theropithecus gelada, Leone et al. 2014; Pan troglodytes, Vick and Paukner 2010; Macaca tonkeana, Zannella et al. 2017). Contrary to other studies, we did not find any influence of the imminent individual context (social or solitary) on the emission of the different types of yawns. This could be related to the limitation of our study groups including only adult females with their offspring. Contrary to the agonistic encounters involving stallions (Olczak and Klocek 2014), female agonistic interactions mostly involve threatening actions probably provoking less anxiety variations in animals. Indeed, a study on Przewalski horses showed that after agonistic interactions, stallions yawned more than females (Górecka-Bruzda et al. 2016).
 
If the context did not apparently influence the different yawn morphologies, the breed seems to have a role, with ponies of Monterufoli performing more Yuct compared to Italian Heavy Draft horses (Figure 3). Although, the interpretation of this outcome is anything but straightforward, a possible explanation can be found in the Brain Cooling Hypothesis (Gallup et al. 2016; Massen et al. 2021). We could hypothesize that to cool a relatively larger brain, the animal can display larger yawns (Galotti et al. 2024). Shetland ponies have been observed to have a number of distinctive cranial features, including a broad skull with a rounded ribcage, a concave face, and large orbits, all characteristics predictive of a larger brain compared to other breeds of horses (Hanot et al. 2021). These traits, that are typical of the juvenile phases in horses, are retained in pony adults probably due to a pedomorphosis phenomenon (Goodwin et al. 2008). In short, ponies manifest larger skulls compared with the body size even as adults (Hanot et al. 2021). The higher frequency of Yuct displayed by the pony of Monterufoli could be an indirect cue supporting the Brain Cooling Hypothesis (Gallup et al. 2016; Massen et al. 2021), although it remains highly speculative.
 
Some authors state that the social value of yawning is a derived feature, while the primitive function of yawning is mainly physiological (Gallup 2011; Palagi et al. 2020). One of the most relevant results of our study is the demonstration of CY in horses (Figure 4). Since CY can have a role in social and motor coordination (Casetta et al. 2021; Gallup 2022), our finding fits with the high level of group cohesion of this species (Maeda et al. 2021). The possible role of CY in favoring inter-individual cohesiveness is supported by the data showing that our mares were more likely to respond to kin compared to non-kin (Figure 5A) and to groupmates with whom they frequently engaged in mutual grooming (Figure 5B). On the other side, CY was not affected by the mares' dominance rank. Females remain in their natal groups throughout their entire lives allowing them to form long-term bonds resulting in coordinated movements during their daily activities and active participation in affiliative exchanges, such as mutual grooming (Arnold and Grassia 1982; Waring 1983; Keiper and Sambraus 1986). Among mares, dominance ranks do not shape mutual grooming thus making the mares' society highly fluid from a social viewpoint (Kimura 1998). Hence, the social modulation of CY reflects the sociobiology of the species with a possible role to increase synchronization of social activities that do not seem to be mediated by dominance relationships.
 
Although some empirical data indicate that CY helps synchronize activities and not viceversa (lions, Casetta et al. 2021; humans, Gallup and Meyers 2021), the potential impact of circadian rhythms on the synchronization of yawns cannot be discounted (Massen et al. 2017). The tendency of social animals to engage in shared activities, such as waking, feeding, and resting, leads to the alignment of their circadian rhythms. It is a reasonable hypothesis that spontaneous yawns cluster temporally within social groups, due to the alignment of their circadian rhythms. This could explain why yawns occur in rapid succession and, unlike chewing events, may be attributed to both contagion and common neurophysiological changes, such as fluctuations in brain or body temperature (Landolt et al. 1995). Additionally, kinship and grooming can facilitate the synchronization among socially bonded animals (Massen and Gallup 2017; Gallup 2022).
 
To promptly synchronize their activities, humans and non-human animals need to rapidly change their behavior according to others' (Nagasaka et al. 2013; Hattori 2021). CY was faster in mares that were kin compared to non-kin (Figure 6A) and reached the lowest latency response between kin mares who mostly groomed each other (Figure 6B). disen Data on humans (Gallup and Meyers 2021) and social carnivores (Casetta et al. 2021) indicate that perceiving a yawn from a conspecific increases the alertness state in the receiver that will be more prone to respond in a congruent and rapid way. From an evolutionary viewpoint, the ability to respond more and faster to familiar yawns can provide selective advantages due to an increased behavioral coordination. This hypothesis needs to be addressed in future studies where the temporal relation between the CY event and the subsequent behavioral alignment of the 2 interacting agents should be evaluated. We are confident that expanding the exploration of yawning to other ungulate social species living in complex societies can open new scenarios on the functions of spontaneous yawning, CY, and their predictable immediate consequences for the social partners.