Playful defensive responses in adult male rats depend on the status of the unfamiliar opponent

Play partner preferences among groups of unfamiliar juvenile male rats

Like many mammals, as juveniles, rats engage in play fighting, which in the laboratory is typically studied in dyads, and consequently, it is the researcher who determines a rat's play partner. In real-life conditions, a rat would have many partners with whom to play. In a previous study, we found that rats do prefer to play with some individuals more than others, and surprisingly, when given the choice, unfamiliar partners are preferred to familiar ones. In this study, we assessed partner choice when all the available partners are strangers. Eight groups of six unfamiliar juvenile male rats were observed for 10 min play trials. One of the six in each group was selected as the 'focal' rat and his play towards, and received by, the others were scored. Social networks revealed that five of the eight groups formed preferences, with preferred partners also engaging in more play with the focal rat. The mechanism by which these preferences were formed remains to be determined, but it seems that there are individual differences, potentially in the amount and style of play, that allow an individual to select the most suitable partner from a group of strangers.

Quality not quantity: Deficient juvenile play experiences lead to altered medial prefrontal cortex neurons and sociocognitive skill deficits

Reduced play experience over the juvenile period leads to adults with impoverished social skills and to anatomical and physiological aberrations of the neurons found in the medial prefrontal cortex (mPFC). Even rearing rats from high-playing strains with low-playing strains show these developmental consequences. In the present study, we evaluated whether low-playing rats benefit from being reared with higher playing peers. To test this, we reared male Fischer 344 rats (F344), typically thought to be a low-playing strain, with a Long-Evans (LE) peer, a relatively high-playing strain. As juveniles, F344 rats reared with LE rats experienced less play and lower quality play compared to those reared with another F344. As adults, the F344 rats reared with LE partners exhibited poorer social skills and the pyramidal neurons of their mPFC had larger dendritic arbors than F344 rats reared with same-strain peers. These findings show that being reared with a more playful partner does not improve developmental outcomes of F344 rats, rather the discordance in the play styles of F344 and LE rats leads to poorer outcomes.

Atypical play experiences in the juvenile period has an impact on the development of the medial prefrontal cortex in both male and female rats

In rats reared without play, or with limited access to play during the juvenile period, the dendrites of pyramidal neurons of the medial prefrontal cortex (mPFC) exhibit more branching than rats reared with more typical levels of play. This suggests that play is critical for pruning the dendritic arbor of these neurons. However, the rearing paradigms typically used to limit play involve physical separation from a peer or sharing a cage with an adult, causing stress that may disrupt pruning. To limit this potentially confounding source of stress, we used an alternative approach in this study: pairing playful Long Evans rats (LE) with low playing Fischer 344 (F344) rats throughout the juvenile period. We then examined the morphology of medial prefrontal cortex (mPFC) neurons, predicting that pruning should be reduced. LE rats reared with another LE rat had significantly greater pruning of mPFC pyramidal neurons compared to LE rats reared with a F344 partner. Furthermore, in previous studies, only one sex or the other was used, whereas in the present rearing paradigm, both sexes were tested, showing that play influences neuronal pruning in both. The neurons of the play deficient LE rats not only occupied more space, as determined by convex hull analyses, but the dendrites were also longer than in rats with more typical play experiences. Unlike studies using more stressful rearing paradigms, the present effects were limited to the apical dendritic projections, suggesting that the previously reported effects on the basilar dendrites may have resulted from developmental disruptions caused by stress. If correct, the present findings indicate that play experienced over the juvenile period affects how mPFC neurons develop and function.

Play fighting and the development of the social brain: The rat's tale

The benefits gained by young animals engaging in play fighting have been a subject of conjecture for over a hundred years. Progress in understanding the behavioral development of play fighting and the underlying neurobiology of laboratory rats has produced a coherent model that sheds light on this matter. Depriving rats of typical peer-peer play experience during the juvenile period leads to adults with socio-cognitive deficiencies and these are correlated with physiological and anatomical changes to the neurons of the prefrontal cortex, especially the medial prefrontal cortex. Detailed analysis of juvenile peer play has shown that using the abilities needed to ensure that play fighting is reciprocal is critical for attaining these benefits. Therefore, unlike that which was posited by many earlier hypotheses, play fighting does not train specific motor actions, but rather, improves a skill set that can be applied in many different social and non-social contexts. There are still gaps in the rat model that need to be understood, but the model is well-enough developed to provide a framework for broader comparative studies of mammals from diverse lineages that engage in play fighting.

High-Frequency Local Field Potential Oscillations May Modulate Aggressive Behaviors in Mice

Aggressive behavior is one of congenital social behaviors in many species, which could be promoted by social neglect or isolation in the early stages of life. Many brain regions including the medial prefrontal cortex (mPFC), medial amygdala (MeA) and ventromedial hypothalamus (VMH) are demonstrated to relate to aggressive behavior; however, the dynamic patterns of neural activities during the occurrence of this behavior remain unclear. In this study, 21-day-old male CD-1 mice were reared in social isolation conditions and cohousing conditions for two weeks. Aggressive behaviors of each subject were estimated by the resident-intruder test. Simultaneously, the local field potentials of mPFC, MeA and VMH were recorded for exploring differences in the relative power spectra of different oscillations when aggressive behaviors occurred. The results showed that the following: (1) Compared with the cohousing mice, the socially isolated mice exhibited more aggression. (2) Regardless of "time condition" (pre-, during- and post- attack), the relative power spectra of beta band in the cohousing mice were significantly greater than those in the socially isolated mice, and inversely, the relative power spectra of gamma band in the cohousing mice were significantly smaller than those in the socially isolated mice. (3) The bilateral mPFC exhibited significantly smaller beta power spectra but greater gamma power spectra compared with other brain areas regardless of rearing patterns. (4) For the right VMH of the socially isolated mice, the relative power spectra of the gamma band during attacks were significantly greater than those before attack. These results suggest that aggressive behaviors in mice could be shaped by rearing patterns and that high-frequency oscillations (beta and gamma bands) may engage in mediating aggressive behaviors in mice.

Social preference in Roman rats: age and sex variations relevance for modeling negative schizophrenia-like features

Social withdrawal is one of the most relevant negative symptoms of schizophrenia. Animal models that mimic schizophrenia's symptoms, in general, and negative symptoms, in particular, are difficult to develop because of the high complexity of symptoms and neurochemical disturbances that schizophrenia patients display throughout their lives. In recent years we have shown that Roman High- Avoidance (RHA) rats exhibit some phenotypes that are thought to represent positive symptoms, cognitive/attentional symptoms, as well as some negative symptoms of the disease. In the present study, we aimed at elucidating whether the social interaction (SI) deficits exhibited by adult male RHA rats, compared to their Roman Low-Avoidance (RLA) counterparts, are also present during adolescence, as well as whether there are between-strain differences in adolescent and adult female rats. The results of the present study show that adult male RHA rats exhibited a deficit in social preference compared to their RLA counterparts. Such a deficit was not observed in adolescent RHA rats or female rats of any age. The results also show that the adult male rats of both strains had significant decreases in social preference compared to the adolescent male rats. Additionally, we also show that female adult RHA rats have greater social preference than their male counterparts. These results seem to be in line with previous rodent and human studies and add face validity to the RHA rats as a model of schizophrenia.

Measuring Play Fighting in Rats: A Multilayered Approach

Rough-and-tumble play or play fighting is an important experience in the juvenile period of many species of mammals, as it facilitates the development of social skills, and for some species, play fighting is retained into adulthood as a tool for assessing and managing social relationships. Laboratory rats have been a model species for studying the neurobiology of play fighting and its key developmental and social functions. However, play fighting interactions are complex, involving competition and cooperation; therefore, no single measure to quantify this behavior is able to capture all its facets. Therefore, in this paper, we present a multilayered framework for scoring all the relevant facets of play that can be affected by experimental manipulations and the logic of how to match what is measured with the question being asked. © 2022 Wiley Periodicals LLC.

Using the 'stranger test' to assess social competency in adult female Long Evans rats reared with a Fischer 344 partner

Rats reared with limited access to a play partner during the juvenile period develop into adults with impairments in various cognitive, emotional, and social skills. The present study assesses the consequences of play deprivation on adult social skills in female Long Evans (LE) rats that were reared with a low-playing Fischer 344 rat over the juvenile period. As adults, their social skills were assessed using the stranger paradigm, by pairing the deprived LE rats with a novel LE partner in a neutral arena. While the deprived rat engages its partner in play there were alterations in key aspects of play, such as reduced pinning and a longer latency to begin playing, that suggest there are impairments in the social ability of the deprived rat. Most notable were the changes in the behaviour of the typically reared partner, a reduction in the amount of play it initiated and fewer actions that produced reciprocal and prolonged interactions. The changes in the behaviour of the normally reared partner suggest that it detected subtle changes in the play deprived LE rats. These findings support the hypothesis that peer-peer play experiences during the juvenile period are important for the development of socio-cognitive skills.

Let's unite in play! Play modality and group membership in wild geladas

Two of the main hypotheses put forth to explain the function of immature social play are the Social Skill Hypothesis and the Motor Training Hypothesis focussing on whether play can improve social competence to develop cooperative social networks or physical abilities to outcompete others, respectively. Here, we tested these hypotheses on a monkey species, the wild gelada (Theropithecus gelada) from the Kundi plateau, Ethiopia. This species is organized in bands divided in One-Male Units (OMUs), united only via social play. Immatures form 'play units' in which individuals from the same and different OMUs interact. We analysed the potential differences between inter- and intra-OMU play to verify which of the two hypotheses (Social Skill or Motor Training Hypothesis) best explains the function of play in geladas. We analysed 527 video-recorded social play sessions and found mixed support for both hypotheses. In agreement with the Social Skill Hypothesis, we found that play in geladas shows scarce social canalization being similarly distributed across age, sex and group membership. In line with the Motor Training Hypothesis, we detected higher levels of competition (shorter and more unbalanced sessions) in inter-OMU compared to intra-OMU play. Hence, in geladas play can be a tool for both the development of social relationships and the improvement of the physical skills necessary to cope with either future mates or competitors. In conclusion, neither hypothesis can be discarded and both hypotheses concur in explaining why immature geladas peculiarly form 'play units' embracing both ingroup and outgroup members.

The influence of breed and environmental factors on social and solitary play in dogs (Canis lupus familiaris)

The domestic dog is an ideal model species in which to study the genetic and environmental factors that influence play behavior. Dogs exist in a wide variety of breeds and frequently engage in multiple forms of play. In the present study, we investigated whether the levels of solitary and social play differed between dogs of three breed types with distinct predatory motor pattern sequences (herding dogs, retrievers, and livestock guarding dogs [LGDs]). Furthermore, we investigated how environmental factors (social and nonsocial contexts) influenced play in dogs of these breed types. Groups of breed-matched dyads with working experience and of equivalent age, sex, and neuter status ratios were exposed to four experimental test conditions and two control conditions in randomized orders. With respect to solitary play, environmental context did have a significant effect, with toys reliably producing the highest levels of solitary play across all breed types. Retrievers engaged in significantly higher levels of solitary play overall than LGDs, and there was a trend in comparison to herding dogs. In contrast, neither environmental context nor breed had a significant effect on social play levels; however, neuter status of the dyads did have a significant effect on social play, with mixed-status dyads engaging in significantly higher levels of social play than same-status dyads. Our findings provide experimental evidence for identifying proximate, environmental stimuli that reliably facilitate social and solitary play and discuss possible genetic (i.e., breed type) and lifetime influences on the form of play in domestic dogs.

In wolves, play behaviour reflects the partners' affiliative and dominance relationship

Puppy packs (consisting of only puppies) and mixed-age packs (composed of puppies and adults) were observed to test whether social play can be used for assessing and establishing social relations in wolves, Canis lupus. Differently from previous studies, we looked at play behaviours in detail, allowing us to categorize play interactions as either competitive or relaxed, and predicted that different types of play would be associated with different relationships between individuals. We found that the more time dyads spent in relaxed play, the more affiliative interactions they exchanged outside of play. In the mixed-age packs, dyads that spent more time in both relaxed and competitive play showed fewer exchanges of aggressive behaviours. Conversely, in puppy packs, the more time dyads spent in competitive play, the more aggressive interactions were exchanged outside of play. Since clear dominance relationships emerged in the mixed-age packs, but not in puppy packs, we suggest that play can help to reduce the frequency of aggressive interactions only when a clear hierarchy exists between pack members. Furthermore, we found that in both puppy and mixed-age packs, dominance relationships were reflected and rarely reversed during play. Finally, dyads with a less clear dominance relationship spent more time playing in a competitive way. Overall, our results support the social assessment hypothesis suggesting that social relationships outside of play are reflected during playful interactions. Moreover, we revealed how different types of play, that is, playing in a competitive or relaxed way, may be related to different social relationships. This distinction between play types has not been acknowledged before but could help researchers better understand the functions of play in different species.

•

Wolves may use play to establish and assess social relationships.

•

Play behaviour is correlated with both affiliative and dominance relationships.

•

Play may reduce the rate of aggressions when a clear hierarchy exists in the pack.

•

Different types of play may be related to different social relationships.

Avoiding escalation from play to aggression in adult male rats: The role of ultrasonic calls

Play fighting is most commonly associated with juvenile animals, but in some species, including rats, it can continue into adulthood. Post-pubertal engagement in play fighting is often rougher and has an increased chance of escalation to aggression, making the use of play signals to regulate the encounter more critical. During play, both juvenile and adult rats emit many 50-kHz calls and some of these may function as play facilitating signals. In the present study, unfamiliar adult male rats were introduced in a neutral enclosure and their social interactions were recorded. While all pairs escalated their playful encounters to become rougher, only the pairs in which one member was devocalized escalated to serious biting. A Monte Carlo shuffling technique was used for the analysis of the correlations between the overt playful and aggressive actions performed and the types and frequencies of various 50-kHz calls that were emitted. The analysis revealed that lower frequency (20-30kHz) calls with a flat component maybe particularly critical for de-escalating encounters and so allowing play to continue. Moreover, coordinating calls reciprocally, with either the same call mimicked in close, temporal association or with complementary calls emitted by participants as they engage in complementary actions (e.g., attacking the nape, being attacked on the nape), appeared to be ways with which calls could be potentially used to avoid escalation to aggression and so sustain playful interactions.

Are 50-kHz calls used as play signals in the playful interactions of rats? II. Evidence from the effects of devocalization

During playful interactions, juvenile rats emit many 50-kHz ultrasonic vocalizations, which are associated with a positive affective state. In addition, these calls may also serve a communicative role - as play signals that promote playful contact. Consistent with this hypothesis, a previous study found that vocalizations are more frequent prior to playful contact than after contact is terminated. The present study uses devocalized rats to test three predictions arising from the play signals hypothesis. First, if vocalizations are used to facilitate contact, then in pairs of rats in which one is devocalized, the higher frequency of pre-contact calling should only be present when the intact rat is initiating the approach. Second, when both partners in a playing pair are devocalized, the frequency of play should be reduced and the typical pattern of playful wrestling disrupted. Finally, when given a choice to play with a vocal and a non-vocal partner, rats should prefer to play with the one able to vocalize. The second prediction was supported in that the frequency of playful interactions as well as some typical patterns of play was disrupted. Even though the data for the other two predictions did not produce the expected findings, they support the conclusion that, in rats, 50-kHz calls are likely to function to maintain a playful mood and for them to signal to one another during play fighting.

Maternal care associates with play dominance rank among adult female rats

Variations in maternal care influence important life history traits that determine reproductive fitness. The adult female offspring of mothers that show reduced levels of pup licking/grooming (LG; i.e., low-LG mothers) show increased defensive responses to stress, accelerated pubertal development, and greater sexual receptivity than the female offspring of high-LG mothers. Amongst several species an accelerated pattern of reproductive development is associated with increased dominance-related behaviors and higher social rank. We hypothesize that rats from low-LG dams may thus also secure higher social rank as a means to compete for limited resources with conspecifics. In this study, social interactions were observed in triads of adult female rats aged p90 that received low, mid, and high levels of pup LG over the first week of life. Low- and mid-LG females had the highest pinning scores and high-LG rats the lowest, showing that low- and mid-LG adult females engage in greater play dominance-related behavior. Likewise, low- and mid-LG rats spent significantly more time drinking following 24 hr of water deprivation in a water competition test thus allowing them to secure a limited resource more easily than high-LG rats. Interestingly, pinning by play dominant females was increased when subordinates were sexually receptive (proestrus/estrus), suggestive of a process of reproductive suppression. Some evidence suggests that low-LG and mid-LG rats also show greater fecundity than high-LG rats. Variations in maternal care may thus have a long-term influence on the development of play dominance and possibly social rank in the female rat, which might contribute to reproductive success within a competitive environment.

Dysfunctional play and dopamine physiology in the Fischer 344 rat

Juvenile Fischer 344 rats are known to be less playful than other inbred strains, although the neurobiological substrate(s) responsible for this phenotype is uncertain. In the present study, Fischer 344 rats were compared to the commonly used outbred Sprague-Dawley strain on several behavioral and physiological parameters in order to ascertain whether the lack of play may be related to compromised activity of brain dopamine (DA) systems. As expected, Fischer 344 rats were far less playful than Sprague-Dawley rats, with Fischer 344 rats less likely to initiate playful contacts with a playful partner and less likely to respond playfully to these contacts. We also found that Fischer 344 rats showed less of a startle response and greater pre-pulse inhibition (PPI), especially at higher pre-pulse intensities. The increase in PPI seen in the Fischer 344 rat could be due to reduced DA modulation of sensorimotor gating and neurochemical measures were consistent with Fischer 344 rats releasing less DA than Sprague-Dawley rats. Fast scan cyclic voltammetry (FSCV) revealed Fischer 344 rats had less evoked DA release in dorsal and ventral striatal brain slices and high-performance liquid chromatography revealed Fischer 344 rats to have less DA turnover in the striatum and prefrontal cortex. We also found DA-dependent forms of cortical plasticity were deficient in the striatum and prefrontal cortex of the Fischer 344 rat. Taken together, these data indicate that deficits in play and enhanced PPI of Fischer 344 rats may be due to reduced DA modulation of corticostriatal and mesolimbic/mesocortical circuits critical to the execution of these behaviors.

Coevolutionary relationship between striatum size and social play in nonhuman primates

The striatum is a region of the brain specifically tied to the experience and anticipation of pleasure, reward, appropriate behavioral sequencing, cognition, learning, and social modulation. Furthermore, the striatum is connected neurologically and functionally to other brain regions associated with the exhibition of social play, such as the neocortex, cerebellum, and limbic system. For these reasons, the striatum is especially interesting to researchers of play behavior. Moreover, the caudate-putamen area of the striatum has been specifically implicated in laboratory studies of social play behavior. This study uses the phylogenetic comparative method of independent contrasts to test for an evolutionary relationship between striatum volume and a measure of social play in nonhuman primates. Relative volume of the primate striatum correlates with rate of social, but not nonsocial, play behavior across species, suggesting a coevolution of traits. The pleasurable and procedural aspects of social play behavior may be mediated in part by the striatum and further to its connection to dopaminergic pathways in the primate brain.

The effects of orbital frontal cortex damage on the modulation of defensive responses by rats in playful and nonplayful social contexts

In a series of 3 experiments on rats, 2 hypotheses were tested: (a) that damage to the orbital frontal cortex (OFC) would alter the socially relevant context for executing defensive responses but not their performance and (b) that damage done to the OFC in early infancy would produce more deficits in social behavior than similar damage occurring in adulthood. Bilateral or unilateral OFC damage in adult males did not impair their ability to defend themselves during play fighting and when protecting their food but did impair their ability to modify the pattern of defense in response to different partners. Rats that sustained bilateral damage at 3 days of age not only had deficits in partner-related modulation of defense but also exhibited hyperactivity in their play. The findings thus supported the proposed hypotheses.

Acute effects of cage cleaning at different frequencies on laboratory rat behaviour and welfare

In rodents, cage cleaning increases cardiovascular and behavioural activity for several hours, which are commonly interpreted as stress responses. In mice, post-cleaning activity also includes aggression, which can cause serious injuries. This study was part of a long-term investigation into the effects of cage cleaning frequency on rat behaviour and welfare. Here we aimed to ascertain whether post-cleaning activity is stress- or aggression-related, thereby leading to recurrent acute reductions in welfare, or simply a result of non-aversive stimulation. Male Wistar (n = 160) and Sprague–Dawley (n = 160) rats, kept in four animal units, had their cages cleaned twice per week, once per week or once per fortnight, and were kept on one of two types of bedding. Behaviours were recorded in detail before and after cage cleaning for 20 weeks, as was the aversion-related Harderian gland secretion, chromodacryorrhoea (‘red tears‘). Cage cleaning caused decreased resting and increased feeding, walking, bedding manipulation and sheltering for at least 30 min after the disturbance. Skirmishing also increased markedly for 15 min after cleaning, but decreased thereafter to below baseline levels. Unlike in mice, all skirmishing was non-injurious and play-like. The frequency of cage cleaning did not affect the magnitude of this skirmishing peak, but rats that had their cages cleaned more frequently settled more quickly after cleaning. Surprisingly, chromodacryorrhoea decreased after cage cleaning; this could mean that rats find soiled cages stressful or alternatively, like many disturbances, cage cleaning might provoke frequent, curtailed bouts of grooming, thereby removing the secretion. Rats also manipulated aspen bedding more than paper bedding. Overall, we found no evidence that cage cleaning caused rats any acute decrease in welfare — a finding consistent with additional data we have obtained on the lack of preference by rats for soiled over clean cages, and a lack of long-term, behavioural and physiological responses to being cleaned frequently or infrequently.

Play fighting between kindling-prone (fast) and kindling-resistant (slow) rats

Differences in the play behavior of 2 strains of rats suggest that different components of play fighting can be modified independently. The development of play fighting in cross-strain pairs of familiar and unfamiliar rats was examined to determine whether interacting with a non-congruent pair-mate would alter the pattern of play typical for each strain. In both strains, changes in play fighting were observed throughout development, but partner identity appeared to influence play fighting in different ways depending on age. These data suggest that some components of play may be more impervious to changes in social environment than other components.

Brain system size and adult-adult play in primates: a comparative analysis of the roles of the non-visual neocortex and the amygdala

Recent studies have shown that contrary to expectation, larger-brained species within mammalian orders are not more likely to engage in play. This is true for juvenile rodents, juvenile marsupials and adult primates. Neither does the relative size of the neocortex predict the prevalence of play in species of marsupials and primates. Two methodological limitations may account for the lack of such relationships. Firstly, play may only vary systematically with specific brain areas, not overall size increases in brain tissue. Secondly, the play indices used to measure the variation in play across species may be insufficiently sensitive to the effects of changes in brain size. In this study, we attempt to deal with the first methodological problem. The adult-adult play fighting among species of primates was correlated with the relative size of the non-visual cortex and the amygdala. The statistical analyses used took into account the problems of scaling and corrected for degree of phylogenetic relatedness among the species. The size of the non-visual cortex failed to predict the prevalence of play fighting occurring in either sexual or non-sexual contexts. In contrast, the size of the amygdala significantly predicted the prevalence of sexual play, but not non-sexual play. That is, species with larger sized amygdala are more likely to engage in sexual play. These findings provide new insights into the role of different brain systems in the regulation of play behavior.

Accelerated nervous system development contributes to behavioral efficiency in the laboratory mouse: a behavioral review and theoretical proposal

The emergence of the laboratory mouse as a favored species for genetic research has posed a number of problems for scientists interested in the reflection of genetic influences in mouse behavior. It is commonly thought that rat behavior, which has been studied more extensively than mouse behavior, could be easily generalized to mice. In this article, a number of categories of behavior displayed by the mouse (motor, spatial, defensive, social) are reviewed and contrasted with the same categories of behavior displayed by the rat. The comparison suggests that mouse behavior is simpler and more dependent upon elementary actions than the behavior of the rat. We suggest that the behavioral simplification in the mouse adapts it for a different ecological niche than that occupied by the rat. We propose that this simplification may be mediated by accelerated brain maturation during development. We further propose that this developmental acceleration in the mouse renders it less dependent upon complex social behavior and plastic nervous system changes associated with learning than the rat. This difference poses problems for the development of relevant methods of behavioral analysis and interpretation. Since the mouse's biological adaptations will be reflected in laboratory behavior, suggestions are made for behavioral approaches to the study and interpretation of mouse behavior.

The roles of phylogeny and sociality in the evolution of social play in muroid rodents

A composite index incorporating the frequency and structure (target, type of defence, etc.) of play fighting was used to compare the complexity of such play in 13 species of muroid rodents whose behaviour has been previously described. A phylogenetic comparison of the distribution of the complexity of play fighting revealed that relatedness did not predict complexity. The most likely pattern for the ancestral rodent was moderate levels of complexity, from which increases or decreases in complexity then appeared to have evolved independently, at the level of subfamily and genus. Given that phylogeny did not predict the distribution of the pattern of play fighting, an alternative hypothesis was tested. That is, that instead, the distribution was produced by species differences in sociality, as reflected by the degrees of male-female association amongst adults. The analysis revealed that play complexity was unrelated to species differences in sociality, with both highly social and relatively asocial species being equally likely to have high or low levels of play complexity. The implications of these results for the evolution of mammalian play are considered. Copyright 1999 The Association for the Study of Animal Behaviour.