Social play is a common behavior in group-living animals. While there is an ongoing debate regarding its specific function, it is generally considered to have a significant role in the development of social relationships, social learning, and the development of motor skills, particularly among juveniles (Fagen et al., 1993; Pellegrini et al., 2007). Like all types of play behavior, social play needs to meet the following five criteria as proposed by Burghardt (2005): 1) it is not fully functional in the expressed form; 2) it is a spontaneous behavior done for its own sake; 3) play is to some extent an incomplete behavior, exaggerated or awkward, or involves modified behaviors; 4) it is a repetitive but not stereotyped behavior; and 5) it is initiated when an animal is in a relaxed state due to low competitive demands, or low predation pressure (Burghardt, 2005).

Although social play might be expressed in various ways across different taxa, there are commonalities in behavioral patterns (Pellis et al., 2010). The most common version of social play, rough-and-tumble play or play fighting, is characterized by behaviors that resemble agonistic interactions (Aldis, 1975; Burghardt, 2005; Pellis and Pellis, 1998) and it involves some degree of turn-taking (Pellis and Pellis, 2017). In primates, this type of play behavior typically includes actions like chasing, fleeing and jumping, but is not intended as being aggressive or harming the partner (Pereira et al., 1988; Palagi, 2009; Pellis et al., 2010). Yet, participation in such activities that involve playfighting carries certain risks, such as potential injuries or social strife, that might result from miscommunication or misinterpretation of the behavior. These behaviors are therefore often accompanied by gestural and visual communication to prevent these risks (Fagen et al., 1993; Burghardt, 2005). Thus, the choice of play partner and effective communication between play partners during social play are crucial elements to maintain the play status (Bekoff, 1975; Burghardt, 2005; Palagi et al., 2016; Brooks and Burghardt, 2023; Adriaense et al., 2024).

In primates, one of the shared visual signals during social play is relaxed open-mouth display, also referred to as primate play face or full play face (lemurs, spider monkeys and patas monkeys: Pellis and Pellis, 1997; bonobos and chimpanzees: Palagi, 2009; marmosets: de Boer et al., 2013; macaques: Iki and Kutsukake, 2023; orangutans: Kunz et al., 2024). The play face is described as an opened mouth with either the lower teeth or both the lower and the upper teeth visible (Palagi, 2006). It is suggested that an exaggerated display of teeth serves as a signal to the partner, indicating that the behavior is playful rather than indicative of genuine aggression (Palagi et al., 2014). Signaling in social play can also include other behaviors that include either body postures (Adriaense et al., 2024; Godard et al., 2025) or the use of the tail (Palagi, 2009). Another signal associated with social play is self-handicapping, which is often displayed when there is an age gap and thus potentially also a size and strength difference between playing partners (Burghardt, 2005). This gestural signal provides an opportunity for the partner to take a turn in attacking, especially when there is an inequality in power dynamics (Pellis and Pellis, 2017).

Although social play mostly occurs among conspecifics due to the shared communication and shared behavioral repertoire, interspecific play behavior, wherein individuals from different species engage in play together, has been observed between various species (reviewed in Brooks and Burghardt, 2023). Examples of interspecific social play, defined by its willing participation in the play by all participants (Brooks and Burghardt, 2023), are mostly from human-nonhuman interactions. However, by analyzing nonhuman interspecific play through social media videos and reports, Brooks and Burghardt (2023) found that most literature reports from instances of nonhuman interspecific play come from marine mammals or nonhuman primates. Interestingly, most reported cases of interspecific play among nonhuman primates have been documented in haplorrhines, including both monkeys and apes. From an evolutionary perspective, interspecific social play is predicted to occur more often among animals in human care compared to their wild counterparts (Brooks and Burghardt, 2023). Brooks and Burghardt (2023) propose two non-mutually exclusive explanations for this pattern. First, animals in human care often share a common space and are raised together, resulting in levels of familiarity that are uncommon in the wild. Secondly, social play, including interspecific social play, is associated with a time and cognitive investment (Smaldino et al., 2019), which is expected to favor the evolution of interspecific social play in species with greater cognitive capacity and behavioral flexibility.

Though accounting for approximately a quarter of all known extant nonhuman primates, across all reported cases of interspecific play, only a single article has documented interspecific play in strepsirrhines, namely in lemurs (Freed, 2007). In that report, juvenile individuals of crowned lemurs (Eulemur coronatus) and Sanford’s lemurs (Eulemur sanfordi; previously Eulemur fulvus sanfordi) regularly engaged in playfighting. That this remains the only documented case of interspecific play in lemurs raises the question of whether the underrepresentation originates from a research bias or from lemurs participating less in instances of interspecific play. Overall, little is known about what facilitates interspecific play or what the advantages are for individuals to choose a heterospecific partner rather than a conspecific.

Here, we describe multiple sequences of interspecific social play behavior between black-and-white ruffed lemurs (Varecia variegata) and ring-tailed lemurs (Lemur catta) in a semi-free-ranging environment with the presence of conspecifics of both species. Social play in black-and-white ruffed lemurs is described as playfighting on the ground and in trees, including wrestling behavior initiated by jumping toward, or grabbing a play partner, chasing and fleeing with exaggerated movements (Pereira et al., 1988) whereas playfighting in ring-tailed lemurs is characterized by jumping on, grabbing, slapping, biting, rolling over, tail pulling and running (Palagi, 2009). The relaxed open-mouth display is observed in both ring-tailed lemurs (Pellis and Pellis, 1997) and black-and-white ruffed lemurs (van Hooff, 1967).

The animals were housed at the Affenwald Wildlife Park, Sondershausen, Germany, living in a semi-free-ranging environment. The park housed two larger groups of ring-tailed lemurs, which were free to mingle at all times. In addition, the park housed multiple groups of black-and-white ruffed lemurs, some of which were indoors with access to outdoor enclosures and some of which were mixed with the ring-tailed lemurs in the park. At the time of the observation, the park held 87 ring-tailed lemurs (age range 1–20 years; mean = 7 years; 38 females) and 35 ruffed-lemurs (age range 1–27 years; mean = 10 years; 24 females). In addition, nine Bennett’s wallabies (Notamacropus rufogriseus) were living and roaming with the ring-tailed lemurs and the ruffed lemurs in the park. The lemurs in this study roamed freely in a natural forest covering approximately two hectares, which was primarily composed of natural vegetation. The park also included huts that were used for sleeping year-round and for containment during the winter months The animals were fed two times per day with fresh vegetables and fruits and had ad libitum access to water. Additionally, the animals fed on leaves, grasses and weeds inside the park. Both lemur species could breed in the facility and offspring were regularly born into the groups between the beginning of March and the end of April. In the last ten years, the ring-tailed lemurs had, on average, seven immatures per breeding season, whereas ruffed lemurs had on average less than two immatures per season. Note that both species experienced the typical shift of their breeding season when held captive in the northern hemisphere (Rasmussen, 1985).

This population is regularly studied through both cognitive experiments and behavioral observations. Also, the park is open to the public and, from early April to late October, is frequently visited by tourists. Hence, the animals are used to the presence of humans and handling by researchers. All individuals in the park are chipped and individually identifiable via a Trovan^® transponder reader. Individuals were identified after a play session had ended to avoid interrupting the ongoing play behavior. Sequences of social play behavior were per chance observed during our regular observation sessions between experimental sessions. None of the individuals involved in the interspecific play interactions participated in (cognitive) experiments at the time of observation. All instances of interspecific play behavior described were partially filmed by either handheld camcorders, GoPros or Smartphones. We analyzed the sequences using Mangold Interact (see Supplementary Table S1 for the ethogram). We conducted continuous sampling from the recorded videos and calculated, for each behaviour, its proportional occurrence based on the total number of observed events.

A typical play sequence unfolded as follows: a ring-tailed lemur would jump toward or onto a black-and-white ruffed lemur (Figure 1a) and gently slap or grab her. In response the ruffed lemur either dropped into a supine position, showing self-handicapping behavior (Figures 1a, b), or tumbled and wrestled with the approaching lemur. Notably, the ruffed lemurs were the only ones to display self-handicapping behavior and to show a relaxed open-mouth display toward the ring-tailed lemurs. Among all the interspecific plays we observed, the most common behavior in ring-tailed lemurs were jumping toward or onto a ruffed lemur (~34%), wrestling (~14%), and gentle grabbing (~13%). As is typical for their play repertoire, they also engaged in tail-play by rubbing their wrists against their tails (~8%). However, they did not wave their tails at the ruffed lemurs with the intensity generally observed during sexual behavior toward conspecifics. Ruffed lemurs’ behavior was characterized primarily by taking a supine position (~21%) and waiting for the lemurs to jump at her, followed by wrestling with the approaching individual (~19%). They also spent considerable time in the branches engaging in acrobatic play (~16%) and displayed relaxed open-mouth display (~7%) (Figure 1c), which occurred twice during acrobatic play, once after being gently grabbed, and once when a ring-tailed lemur reengaged in play.

The first instance of playfighting between a black-and-white ruffed lemur and a ring-tailed lemur was documented on September 9^th, 2025, at approximately 2 p.m. during a period of free-ranging prior to the scheduled feeding time (Supplementary Video S1). At this time, other conspecifics from both species were present and in a resting state. The exact identities of both individuals could not be verified at the time of observation. However, the ruffed-lemur was an adult, and the ring-tailed lemur was a subadult of less than 2.5 years of age (Jolly et al., 2006), based on general appearance. We did not observe the beginning of the sequence and hence have no information about who initiated the play behavior. From all the instances we were able to record, this was the longest play bout (2:40 min of play bout recorded). Playing was characterized by a sequence of rough and tumble play in which the ruffed lemur was the individual that showed self-handicapping behavior and a relaxed open-mouth display.

The second and third instances were consecutive play sessions between first Emmy and Mowgli (Table 1), followed by a second play session between Emmy and Otis (Table 1). Both instances were recorded on September 24^th, 2025, at around 10 a.m. after the regular feeding. Unfortunately, we could not capture who started the play sessions each time, but only recorded the last 12 seconds of the first play bout and only the last 14 seconds of the following play bout. Since we were present in the area when these two play sessions occurred, we suspect that they were relatively short compared to the previously observed session. However, we cannot determine the exact duration of the play sessions. Nevertheless, we could capture some repetitive patterns from both species in the play sessions in which they jump on and towards each other in a non-aggressive way and multiple occurrences of open-mouth display and self-handicapping to signal play behavior.

The last observed instance occurred on October 24th at approximately 9 a.m. (Supplementary Video S2) and the observed time lasted for one minute and twenty seconds. In this case, Estella (an adult female black-and-white ruffed lemur) was playing with three ring-tailed lemurs, who were also playing among themselves. The observed sequence began with Estella engaging with one individual and then switching mid-play to another, while the first lemur continued playing with a conspecific. Unfortunately, we could not observe who started the session in the first place. Overall, the observed sequence was shorter than the first one. Once Estella moved away, the young ring-tailed lemur reengaged with its conspecifics rather than with her. As in previous observations, the play sequence was characterized by frequent play jumps from the ring-tailed lemurs and a more submissive play style from Estella.

Here, we observed multiple instances of interspecific social play, involving play signals and self-handicapping behaviors, in co-housed but ecologically allopatric species. The function of social play remains a topic of discussion. A plausible explanation for this potentially risky behavior, is that it plays a significant role in an individual’s physical and social development (Fagen et al., 1993; Pellegrini et al., 2007). It enables individuals to improve their body control and rehearse their movements in a less risky situation where there is no actual target source to compete over. However, playing can escalate into a real fight if one does not remain in the play or the play signals are misinterpreted by the partner (Pellis and Pellis, 2017). Social play thus requires efficient communication to avoid conflicts, particularly during playfight. Therefore, individuals often exhibit continuous behavioral and visual signaling during social play (Palagi et al., 2016). Similarly, partners should be able to interpret those signals and intentions, such as assessing whether the social partner is in the state of an actual fight or playfight. In this context, mutual communication in social play, particularly in playfight, has a crucial role in interpreting the intentions of others (Bekoff and Allen, 1998).

Playfighting exhibits actions similar to those found in agonistic behavior (Palagi, 2009). This illustrates that the same behavior can be utilized and understood differently based on the context in which it occurs. Therefore, understanding whether the social partner is engaging in playfight behavior or a genuine fight depends on several factors. Firstly, the context of the situation, such as the presence of a resource to compete for or whether individuals are in the state of rest, can provide valuable cues to make this decision. In other words, recipients need to evaluate the contextual cues to be able to attribute a meaning to the initiated behavior (Wheeler and Fischer, 2012). Secondly, maintaining the play behavior is as important as initiating or successfully interpreting the initiation of a play. Therefore, individuals use specific play signals, like self-handicapping or open-mouth display, to remain in the play. Self-handicapping, in particular, allows the partner to gain a temporary advantage during play, which is especially useful in cases of a power mismatch between the play partners (Pellis and Pellis, 2017). Consequently, it is essential for heterospecific partners to establish a shared basis for this kind of communication, which might be challenging, as play signals may appear differently in each species’ typical interactions with conspecific partners. In the present observations, the ruffed lemurs showed relaxed open-mouth display and self-handicapping behavior, whereas the ring-tailed lemurs did not reciprocate these signals. Rather, they performed more jumping-on and grabbing behaviors. This asymmetry may be explained by a power mismatch between the adult ruffed lemurs and the juvenile ring-tailed lemurs (Pellis and Pellis, 2017). Nevertheless, both species did engage in wrestling behavior, indicating some level of shared play signaling and a mutual understanding of play context.

It remains unclear what facilitates interspecific play or why individuals prefer a heterospecific social partner when conspecifics are available. One potential influencing factor could be age. Research indicates that sometimes juveniles or infants are less hesitant to approach an adult of a different species compared to an adult of their own species for social play and are more flexible in the cues they use to engage in a play sequence (Rose, 1977). The underlying rationale for this preference may be attributed to potential conflicts arising from the hierarchical structures within conspecific groups. In contrast, such dynamics are less likely to occur in interactions with heterospecific partners, where the absence of hierarchical competition may facilitate more peaceful engagement. Similarly, we observed that interspecific social play occurred between juvenile male ring-tailed lemurs and an adult female ruffed lemur. Ring-tailed lemurs have female-dominant social structure, which might lead juvenile male ring-tailed lemurs to engage in play with a heterospecific adult partner instead of an adult female conspecific.

Another factor that might enhance interspecific play is the presence of a shared territory and the habituation of one species to another. Studies on free-ranging vervet monkeys and guerezas in overlapping territories demonstrate that proximity and regular exposure to one another can enhance social interactions and play behaviors between different species (Rose, 1977). Likewise, the subjects of this study cohabit in a semi-free-ranging environment, where they receive food together and interact with each other in a naturalistic setting. Although interactions are most often limited to communal feeding, occupying the same shelters and occasional agonistic behaviors, this co-feeding arrangement allows for a richer social dynamic to develop between the two species than is usual. It is important to note that black-and-white ruffed lemurs and ring-tailed lemurs do not share overlapping territories in the wild. The former inhabits the eastern regions of Madagascar, where they thrive in the rainforests, while the latter are found in the southern part of Madagascar, living in dry forests, rocky terrains, or savannahs.

While it is a rare behavior, interspecific play can help us to explore the evolutionary function of social play. The function of play is often associated with the development of social relationships, social learning, and the development of motor skills (Fagen et al., 1993; Pellegrini et al., 2007). However, Spinka et al. (2001) hypothesized that play, especially play through self-handicapping, also helps develop flexible responses to events where the animal unexpectedly loses control, for instance when falling over. Practicing this might also help the animal emotionally recover faster from stressful situations (Spinka et al., 2001). This skill is not only useful for juveniles, adult animals may also benefit from this, which may have been the case in the play interactions described here. Nevertheless, for both adults and juveniles, interspecific play could also be the result of a surplus of energy due to reduced foraging time and predation pressure under human care (Sharpe et al., 2002). Obviously, the probability of interspecific play is higher in mixed-species enclosures compared to the wild since animals that would rarely or never (as in this case) meet in the wild spend constant time around each other. Hence, interspecific play could just be a result of opportunity and additional time and energy. Further research is needed to examine these possibilities and clarify whether the interspecific social play is ultimately the same as intraspecific social play or whether there are additional aspects of interspecific play interactions at the proximate level that are not found in interactions with conspecifics (Oliveira and Bshary, 2021).

In conclusion, research on social play can provide valuable insights into communication dynamics. Within this context, observations of interspecific play highlight the underlying motivations or advantages of social play. Here, we describe several occurrences of interspecific play between ring-tailed lemurs (Lemur catta) and black-and-white ruffed lemurs (Varecia variegata), a behavior rarely described in primates. Consequently, much remains unclear about the function of interspecific social play, and further investigation is necessary to understand this behavior.