Play-solicitation gestures in chimpanzees in the wild: flexible adjustment to social circumstances and individual matrices

Social play is a frequent behaviour in great apes and involves sophisticated forms of communicative exchange. While it is well established that great apes test and practise the majority of their gestural signals during play interactions, the influence of demographic factors and kin relationships between the interactants on the form and variability of gestures are relatively little understood. We thus carried out the first systematic study on the exchange of play-soliciting gestures in two chimpanzee (Pan troglodytes) communities of different subspecies. We examined the influence of age, sex and kin relationships of the play partners on gestural play solicitations, including object-associated and self-handicapping gestures. Our results demonstrated that the usage of (i) audible and visual gestures increased significantly with infant age, (ii) tactile gestures differed between the sexes, and (iii) audible and visual gestures were higher in interactions with conspecifics than with mothers. Object-associated and self-handicapping gestures were frequently used to initiate play with same-aged and younger play partners, respectively. Our study thus strengthens the view that gestures are mutually constructed communicative means, which are flexibly adjusted to social circumstances and individual matrices of interactants.


Introduction
Humans' unique creativity and innovation skills have been suggested to be longer-term outcomes and benefits of playfulness [1].
2016 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited.

Study sites and subjects
The study was conducted at two different chimpanzee communities in the wild: Kanyawara in Kibale National Park, Uganda and Taï South in Taï National Park, Côte d'Ivoire. Detailed descriptions of the study areas can be found in Wrangham et al. [56] and Boesch & Boesch-Achermann [46], respectively. During the two study periods, the size of the Kanyawara group varied between 53 and 56 individuals, and 26 and 33 individuals in Taï South, respectively. In addition, we had access to long-term data concerning the chimpanzees' demography and relatedness. We observed play interactions of a total of 16 infants (10 from Kanyawara, 6 from Taï South) with their mothers and conspecifics, with ages ranging from 9 to 74 months (electronic supplementary material, table S1). Conspecifics that the infants interacted with during social play included 10 juveniles/sub-adults (three females, seven males) and 11 adult individuals (eight females, three males) at Kanyawara, and 8 juveniles/sub-adults (one female, seven males) and four adult individuals (four females, no males) at Taï South. In terms of age classes, individuals were categorized as juvenile/sub-adults if aged between 6 and 13/15 years (females/males) and adults if aged 14/16 years or older (females/males) [46].

Data collection
Observations were made on chimpanzees during four periods between October 2012 and June 2014 (Kanyawara: Mar-May 2013; Mar-June 2014; Taï South: Oct-Dec 2012; Oct-Dec 2013). We used a focal next to a behaviour sampling approach, that is one individual was observed for a set period of time, while play behaviour of non-focal individuals in the immediate proximity was also recorded [57]. All play interactions of infants (i.e. mother-infant interactions and infant-conspecific interactions) were recorded using a digital high-definition camera (Canon HF M41) with an external unidirectional microphone (Sennheiser K6). This method resulted in a total of 81.9 hours (Kanyawara: 44.3, mean ± s.d. per infant = 6.6 ± 4.6; Taï South: 37.6, mean ± s.d. = 10.7 ± 5.4, see also electronic supplementary material, table S1) of video footage of play interactions recorded during approximately 1154 hours (Kanyawara: 557, Taï South: 597) of focal observations.

Coding procedure
To establish the signal repertoires of chimpanzees used to solicit play and enable subsequent analyses, a total of 643 high-quality video files of play interactions were coded using the program Adobe Premiere Pro CS4 (v. 4.2.1.). Behavioural definitions were based on established ethograms of two long-term studies of Eastern Chimpanzees at Gombe [22] and Mahale [58] and several gesture studies [21,59,60]. A specific coding scheme was developed based on parameters used in previous work on great ape gesturing [28,36,61]. 'Play solicitations' comprised dyadic play initiation from the start of a play bout, and also dyadic play reinitiation after social play was paused or interrupted by a third individual for at least 10 s. While coding all play interactions, we differentiated play-soliciting gestures from physical actions. An action was defined as any behaviour that resulted in play through direct manipulation of another's body via physical force (e.g. throw on) or one's own locomotion (e.g. move backwards). Contrarily, a gesture was defined as directed, mechanically ineffective movement of the extremities, the body or body postures that elicited (requested) a voluntary response by the recipient [62]. For our analyses, we only included play-initiating gestures that were accompanied by key characteristics of intentional communication: sensitivity to the recipient's attentional state, response waiting, apparent satisfaction of the signaller and goal persistence (for definitions, see [36,63,64]). Signals were clustered into three categories: audible (signals generate a sound while being performed, e.g. SLAP GROUND), tactile (signals include physical contact with the recipient, e.g. TOUCH) and visual (signals generate a mainly graphic component, e.g. RAISE ARM) gestures [28]. To identify play solicitations, the behaviour of both the signaller and the recipient throughout the interaction, from first initiating action/gesture until the start of play, was taken into account to assess the success of communicative attempts [65]. Finally, for each signal or action case, we coded the following parameters: signaller's and recipient's age (range = 0-57 years), if signaller = infant: infant age (range = 9-69 months), age class relative to recipient (three levels: same age class, older age class, younger age class), sex of signaller and recipient (two levels: male, female), kin relationship between play partners (three levels: infantmother, infant-maternal kin, infant-non-kin); play intensity (three levels: low-touching and tickling, intermediate-wrestling and biting, high-rough and tumble). Fifteen per cent of all coded interactions were coded for accuracy by a second observer and tested using the Cohen's κ coefficient to ensure inter-observer reliability [57]. The following results were found: a 'very good' level of agreement for play intensity (κ = 0.868), gesture type (κ = 0.827), signal category (κ = 0.884), object-associated gesture (κ = 0.916) and self-handicapping gesture (κ = 0.886) and a 'good' level of agreement for intentional usage of signal (κ = 0.703).

Analyses
To assess the influence on sampling size, we plotted the cumulative numbers of observed gesture types over time for each study site. If an asymptote was reached (i.e. no further gesture types were observed), we concluded that we had observed the individual's full repertoire for the communicative function of play solicitation [36].

Model specification
To test to which extent sex, age and the kin relationship between the play partners influenced (1) intensity of solicited play, (2) frequency of gestures overall (response variables: audible, tactile, visual gesture), (3) frequency of object-associated gestures, and (4) [66]) with a Poisson (1) or binomial error structure (2)(3)(4) and log (1) or logit link function (2)(3)(4). Since infant age varied considerably between infants, we used the method of within-subject centring [67] to tease apart whether the effect of infant age was particularly relevant within and/or between infants. Hence, we included into the model the average age of each infant (being constant across all data points of the respective individual; between-age) and also the difference between the infant's actual age and its average age (within-age). To rule out that age effects do not simply result from higher rates of conspecific-directed signalling with age, we initially included the interactions between relation and both age variables into the model (analyses 1 to 3, removed if nonsignificant). To control for confounding effects, we also included site as further fixed effect into the model. As random effects (intercepts) we included the identities (ID) of signaller, recipient and the play dyad. To keep type 1 error rates at the nominal level of 5%, we also included the random slopes components of within-age, age difference, sex of signaller/recipient and kinship within signaller ID and/or recipient ID [68,69]. For the other fixed effects, we did not include random slopes because they were usually constant within signaller and recipient ID. We also did not include correlations between random slopes and random intercepts to keep model complexity at an acceptable level and because neglected random slopes do not compromise type 1 error rates [69].

Model implementation
The models were implemented in R [70] using the function glmer of the package lme4 [71]. To test the overall significance of our key test predictors [72], we compared the full models with the respective null models comprising only the control predictor (study site) and all random effects using a likelihood ratio test [73]. Prior to running the models, we z-transformed between-age, within-age and age difference [74]. To control for collinearity, we determined variance inflation factors (VIF; [75]) from a model including only the fixed main effects using the function vif of the R package car [76]. This revealed collinearity to not be an issue (maximum VIF = 1.49). Confidence intervals were derived using the function sim of the R package arm [77]. Tests of the individual fixed effects were derived using likelihood ratio tests (R function drop1 with argument 'test' set to 'Chisq'). All statistical analyses were performed using the R statistical package, version R.3.1.1 [70], with a level of significance set at 0.05.

Datasets
For each of our four analyses, we used different datasets: for analyses 1 and 2, we included play signals of infants towards all possible play partners, including mothers, and used age and sex of the signaller as well as relation to the recipient (mother, maternal kin, non-kin) as fixed effects (for details see dataset in electronic supplementary material, S2). Analysis 3 used the same dataset, but here we excluded the few gestures towards mothers, since object-associated gestures did not seem to play a role in infant-mother play solicitation. Here, we additionally included age difference and sex of the recipient as additional test predictors. Finally, for analysis 4, we included gesture cases by or towards infants to examine the role of self-handicapping (i.e. signallers could be of all age classes, for details see dataset in electronic supplementary material, S3). This analysis used the same test predictors as utilized in analysis 3.  figure S1). Hence, we assumed that the approximate repertoire of play-soliciting gestures has been captured at both study sites.

Overview on gesture types in play solicitation
To examine which gestures chimpanzees used to initiate play, we compiled gesture repertoires for different age classes (infants, juveniles/adolescents and adults; for classification see Material and methods) of the two sites separately. Across study sites, we identified a total of 48 gesture types. Out of the 48 gesture types, three were audible, 24 were tactile and 21 were visual (for definitions of gesture types, see electronic supplementary material, table S2). We identified five group-specific gesture types: HEAD BUTT, HIDE SELF, PRESENT BACK, PRESENT LEG, SHAKE HEAD (Kanyawara-unique: N = 4, Taï Southunique: N = 1, see electronic supplementary material, table S3). With regard to age classes, across sites infants produced all 48 identified gestures (Kanyawara: N = 47, Taï South: N = 49). In the juvenile/subadult class the number decreased, with a total of 37 gesture types used across sites (Kanyawara: N = 32, Taï South: N = 27). Finally, the adult age class, including mothers, showed the smallest repertoire with a total of 25 gesture types across sites (Kanyawara: N = 24, Taï South: N = 16). A detailed overview on gesture types used in play interactions with regard to age class and community/study site can be found as electronic supplementary material, table S3.

Influence of demographic parameters and kin relationship on intensity of initiated play
Before examining the effect of sex on play-solicitation gestures, we analysed whether male (N = 10) and female (N = 6) chimpanzee infants differed with regard to the intensities of play solicited (N = 1157). Males produced on average 52.5 ± 34.4% (mean ± s.d.) gestures resulting in soft play bouts (low intensity), 28.0 ± 23.5% resulting in wrestle bouts (intermediate intensity) and 19.5 ± 24.5% resulting in rough and tumble bouts (high intensity). Females produced on average 68.1 ± 21.0% gestures resulting in soft play bouts, 30.0 ± 19.7% resulting in wrestle bouts, but only 1.8 ± 3.5% resulting in rough and tumble bouts (see also electronic supplementary material, figure S2). We used GLMMs to test the effects of signaller's sex, age (i.e. both between-and within-age) and a prevailing kin relationship to the recipient (i.e. mother, maternal kin and non-kin) as well as the interactions between kin relationship and age (comprising two interaction terms) on play intensity in infant chimpanzees. Overall, the test predictors had a clear impact (likelihood ratio tests comparing null and the full model (LRT): χ 2 = 21.515, d.f. = 7, p = 0.003). After removal of the (non-significant) two interactions, our results showed that intensity of play was significantly higher in older signallers (between-age: estimate ± s.e. = 0.152 ± 0.048, χ 2 = 8.210, d.f. = 1, p = 0.004) and significantly lower in interactions with mothers (relation/mother: −0.281 ± 0.089, χ 2 = 9.892, d.f. = 1, p = 0.002). There was no effect of signaller's sex (sex/male: 0.004 ± 0.090, χ 2 = 0.002, d.f.

Influence of demographic parameters and kin relationship on play solicitation
We used GLMMs to test whether the effects of sex, age (i.e. both between-and within-age) and kin relationship to the recipient (i.e. mother, maternal kin and non-kin) as well as the interactions between kin relationship and age (comprising two interaction terms) affected the production of visual, tactile and audible gestures in infant chimpanzees. Overall, the test predictors had a clear impact in all three models (LRT for audible gesturing: χ 2 = 28.676, d.f. = 7, p < 0.001; tactile gesturing: χ 2 = 14.976, d.f. = 7, p = 0.036; visual gesturing: χ 2 = 21.906, d.f. = 7, p = 0.003).
The two interactions between kin relationship and (between-/within-) age were non-significant in the other two models and were removed before testing the individual effects. With respect to tactile gestures, we found a clear sex difference: males produced significantly more tactile gestures to initiate play than females (sex/male: 0.387 ± 0.189, χ 2 = 4.216, d.f. = 1, p = 0.04; figure 1). In addition, younger individuals were more likely to use tactile gestures than older ones (between-age: −0.237 ± 0.088, χ 2 = 6.732, d.f. = 1, p = 0.009; figure 1), and tactile gestures were more likely to be produced towards mothers than other Visual gestures, on the contrary, were significantly more produced by older individuals (betweenage: 0.241 ± 0.115, χ 2 = 4.294, d.f. = 1, p = 0.038), and were significantly less often produced towards mothers than towards other individuals (relation/mother: −2.008 ± 0.559, χ 2 = 14.123, d.f. = 1, p < 0.001; figure 2). Sex, maternal kinship or site affiliation had no significant effect on the production of visual gestures (sex/male: −0.369 ± 0.251,   Table 1. Factors influencing (a) audible, (b) tactile and (c) visual gesture production to solicit play. Generalized linear mixed models (GLMMs) were used with sex, age, maternal kinship and site as fixed effects, while identities of signallers, recipients and play dyads were included as random effects. Significant effects are highlighted in italics (number of observations: N = 1157 across 16 subjects). (

Object-associated gestures in play solicitations
To test whether signallers' age (within-/between-age) and sex, recipients' sex, age difference and maternal kinship affected the production of gestures accompanied by objects outside the mother-infant bond, we again used GLMMs with a binomial error function. While controlling for site, we found that the test predictors had a clear impact in our model (LRT: χ 2 = 21.651, d.f. = 6, p = 0.001). Results showed that object-associated gestures were more frequently used with increasing age of the same infant (within-age: 0.530 ± 0.143, χ 2 = 8.002, d.f. = 1, p = 0.005) and used more by older compared with younger infants (between-age: 0.572 ± 0.230, χ 2 = 5.396, d.f. = 1, p = 0.020). In addition, the production of object-accompanied gestures was more likely for smaller absolute age differences between the partners (−0.090 ± 0.042, χ 2 = 4.925, p = 0.026; figure 3). No other effects in the model (signaller's sex, recipient's sex, maternal kinship or site) reached significance (table 2a).

Self-handicapping in play solicitations
Finally, we used a sixth GLMM with a binomial error structure to investigate the influence of signallers' age (within-/between-age) and sex, recipients' sex, age difference (older versus younger partners) and maternal kinship on the production of self-handicapping gestures, while controlling   Table 2. Factors influencing (a) object-associated gesturing and (b) self-handicapping to solicit play. Generalized linear mixed models (GLMMs) were used with signaller's sex and age, recipient's sex, age difference relative to play partner, maternal kinship and site as fixed effects, while identities of signallers, recipients and play dyads were included as random effects.   for study site. The test predictors had a significant impact on the occurrence of self-handicapping gestures (LRT: χ 2 = 18.369, d.f. = 7, p = 0.010). Chimpanzees used self-handicapping signals significantly more often towards recipients that were younger in age (age difference (+): 4.751 ± 1.065, χ 2 = 9.341, d.f. = 1, p = 0.002; figure 4). Additionally, they were more likely to be produced in younger individuals (between-age: −0.477 ± 0.230, χ 2 = 3.933, d.f. = 1, p = 0.047). Taï South chimpanzees produced more selfhandicapping gestures than Kanyawara chimpanzees (site/Taï South: 2.041 ± 0.849, χ 2 = 5.400, d.f. = 1, p = 0.020). None of the other effects in the model (signaller's sex, recipient's sex and maternal kinship) reached significance (table 2b)

Discussion
This is the first study that systematically investigated gestures employed to solicit play in chimpanzees of two different subspecies and communities in two chimpanzee communities (Kanyawara, Kibale National Park, Uganda, and Taï South, Taï National Park, Côte d'Ivoire) living in their natural environments. Our aim was to examine the influence of demographic factors and kinship on gestural play solicitations via 'pure', object-associated and self-handicapping gestures. To target these aims, we analysed play interactions to address the following three questions: first, do age and sex of the play partners influence the production of audible, tactile and visual gestures? Second, to which extent does the kin relationship with the play partner influence the modality of gestures used to initiate play? Third, do chimpanzees take into consideration recipients' attributes when using object-associated and self-handicapping gestures? Overall, our results showed that both signallers' age and sex as well as a prevailing kin relationship with the play partner significantly influenced the category of gestures being employed. Moreover, we found that object-associated and self-handicapping gestures had a crucial importance for initiating play with same-aged and younger play partners, respectively. These findings imply that chimpanzees are able to adjust their use of gesture flexibly to distinct attributes of conspecifics. We thus expand recent findings on the degree of gestural flexibility and underlying cognitive tool-kits in chimpanzees and great apes by showing that chimpanzees not only take into consideration attentional states of recipients [21,27,28] and social contexts [21] but also distinct characteristics of their conspecifics. In the following paragraphs, we will discuss each of our research questions and the related findings in detail.
In a first step, we compiled repertoires of play-solicitation gestures in relation to age class and study site resulting in around 50 gesture types used by chimpanzees. This result is in line with previous studies showing that the play context comprises a particularly large number of gesture types [21]. We also strengthen observations on the behaviour of free-ranging chimpanzees, showing that play-soliciting gestures in chimpanzees consist of a remarkable variety of behavioural patterns [4]. This included audio-visual attention-getters [78], self-handicapping signals [55] and elements of both solitary and social play [4]. Intriguingly, many of the reported attention-getting gestures (e.g. POKE, SLAP GROUND and HIT WITH OBJECT) do not involve specific information about the context of play, which might be rather expressed through the context-typical 'play face' and/or the play postures. Tomasello et al. [78] highlighted this signal combination as vital element in chimpanzee communication. Since attentiongetting gestures might be linked to an understanding of the recipient's intention and attention, Tomasello [79] argued that this gesture class is particularly novel and complex in the animal kingdom (however, note that great apes' abilities to understand others' intentions is still subject to an ongoing debate; [80,81]). Moreover, our results showed that age classes varied considerably in signal repertoires, with decreasing repertoire sizes used by individuals of older age classes. This is corroborated by earlier studies showing that gestural repertoires in several ape species increase sharply as infants develop, and then decrease again in adult individuals [26,28,29]. Since these studies demonstrated that many gestures are used in the play context, this highlights the major role of play for immature individuals as opposed to adults [82]. It is noteworthy that gesture counts, regardless of category, were always (except for self-handicapping gestures, see below) slightly larger in individuals of the Kanyawara community despite comparable observation times across sites. However, this difference should not be overstated since the amount of recorded play interactions was considerably larger for Kanyawara compared with Taï South (see Material and methods). We controlled for site effects in all our analyses and could thus show that study site had no significant influence on our results. Self-handicapping gestures comprised the only exception, because they were significantly more often produced by members of the Taï South community. We found evidence for five group-specific gesture types, yet each of these has been observed in (at least) one other community of Eastern chimpanzees [58,60] and can, therefore, not be considered as novel gestures unique to the respective community. However, since these communities do not exchange members, they could provide evidence for parallel invention. For instance a distinct grooming solicitation gesture used by chimpanzees, the DIRECTED SCRATCH, has so far only been reported from the Ngogo and Budongo communities in Uganda [61,83]. It is thus likely that the underlying developmental process (social negotiation, see below) from a non-communicative, undirected self-scratch into a communicative behaviour is the same [84]. The only other gesture type with geographical variation in wild chimpanzees, the grooming HAND-CLASP, has been suggested to demonstrate cultural transmission within a group [85,86]. It is, however, not clear whether the HAND-CLASP qualifies as gesture, since its communicative function is still debated [82] and it might simply act as a stabilizing tool. Our findings provide further support for the social negotiation hypothesis proposed by Fröhlich et al. [36] which postulates that gestures originate from the exchange of social behaviours resulting in a mutual recognition that (i) certain behaviours can be used communicatively, (ii) carry distinct meaning linked to particular social contexts, and (iii) are produced to achieve distinct goals. However, further fine-grained analyses of great ape gestural production are necessary to reveal the form of gestures in relation to age, context and social partner. In addition, we cannot completely rule out that the group-specific gesture types found resulted from our study design focusing only on interactions of infant chimpanzees with conspecifics.
With regard to the first question addressing the effect of age and sex on the production of distinct gesture categories, we found that visual and audible gestures were employed more frequently with increasing infant age. The production of tactile gestures was however higher in younger individuals. These findings are in line with results of a recent study at the same sites [36], at Gombe, Tanzania [41,87], and another study conducted in captivity [40]: chimpanzee infants become intentional agents throughout development, capable of manipulating the attentional and maybe also the mental states of their conspecifics [28,41,61]. To investigate the influence of sex differences, we firstly analysed whether males and females differ concerning the intensity of play that they are initiating. This enabled us to rule out that sex differences in signalling merely result from differences in play intensity for males and females. We found that the two sexes did not differ significantly with regards to the intensities of the initiated play bouts. In terms of gestural solicitations, however, males used significantly more tactile gestures than females. Tactile gestures differ from visual gestures in both physical effectiveness and potential demonstration of physical strength. We thus interpret these results as evidence for a sex difference in signal directness in terms of the level of physical contact involved. In addition, a recent study by Lonsdorf et al. [15] showed pronounced sex differences in social play interactions of chimpanzees, with males showing higher playing rates at earlier ages, a more diverse set of social partners and a larger frequency of interaction with adult males. They thus argued for the higher importance of socialization for young males given the importance of social bonds and apprenticeship during development and social dominance in adulthood [88]. In addition, this learning period might have crucial implications for the reproductive success in a given community: it could result in moving around with one of the highest ranking males and being his 'apprentice', having access to valuable food resources, resulting in stable and linear growth and higher reproductive success in the longer term [89,90]. In light of the social matrices of chimpanzees, it can be reasonably assumed that investment in play during development can take priority over physical development for skill acquisition [2]. A similar phenomenon has been reported for human infants concerning general play behaviour, with male infants exhibiting more independent exploratory and vigorous play behaviour than female infants [91,92]. However, virtually nothing is known about whether humans and primates differ between sexes in terms of their usage of communicative signals to initiate or solicit social play. Studies have shown that play signalling in humans is reduced to the face, taking the form of laughter and smiling [93]. Nonetheless, more research is needed to elucidate the role of body postures and non-manual gestures in this context. So far, most human gesture research has focused on visual signals in space [62,94,95], which might be one crucial factor responsible for the diverging gesture definitions in human and primate gesture research. Nevertheless, our results indicate that biological factors, that is, the selection pressures based on the sex-specific behavioural roles [96], rather than environmental factors (i.e. socialization processes), may influence sex differences in human play behaviour [14].
Regarding the influence of relationship (mother, maternal kin or non-kin) to the play partner, we found that tactile gestures were more likely to be used in interactions with mothers than with other individuals (i.e. maternal kin and non-kin partners). Visual gestures on the other hand, which were more commonly used with increasing infant age, were produced more frequently to solicit play with conspecifics than with mothers. We ruled out that this increase is just an effect of infants gesturing more frequently towards conspecifics with increasing age, by showing that there is no interaction between kin relationship and age. Thus, we demonstrated that both results, visual gestures being more employed with increasing age on the one hand and towards conspecifics on the other, are independent from each other. The same independence was confirmed concerning the use of tactile gesturing. For audible gesturing, however, we found a more frequent usage towards mothers with increasing age of an individual (within-age), while the effect of between-age (older versus younger infants) was independent from the one of kin relationship.
Hence, our results are partly in line with our predictions regarding higher frequencies of tactile signals within mother-infant interactions, while a maternal kin relationship between the play partners, as in the case of siblings, did not have an influence. Considering the difference between signalling within and outside the mother-infant bond, it appears that the mother-infant attachment is a stronger predictor than familiarity based on established predictable outcomes. Nonetheless, some unrelated individuals could be very 'familiar' to each other due to the frequent association of their mothers, which may significantly impact upon their communicative exchanges. However, because we did not collect data on affiliative relationships, we cannot test for this potential effect on play solicitation. Plooij [41] described a developmental shift from physical acts without social-communicatory intention to intentional tactile acts in interactions between chimpanzee infants and their mothers at Gombe, Tanzania. During this shift, the infant learns to use tactile behaviours, whose meanings have been established in earlier sessions, not only to maintain an interaction but also to initiate it. Our findings support the notion that the communicative development of great apes is supported by the infant leaving the 'security range' provided by the mother and entering its complex social environment [36,41]. Play interactions with peers and other 'nonmother' individuals may serve as essential grounds for experimentation, where great apes can test for the effectiveness of and practise intentional gestures that might gain vital importance in their adult life [87].
Intriguingly, about one-fifth of all gestures used to initiate or reinitiate play were accompanied by mobile or immobile objects. This behaviour was more common with advancing development and increasing infant age, probably due to growing proficiency in object handling. A recent study of chimpanzees and bonobos in the wild showed that chimpanzee infants start to manipulate objects as early as from nine months of age [97]. These object manipulations became more diverse and 'tool-like' (i.e. sticks) with increasing infant age, which was interpreted as preparation for tool use in adulthood [97]. However, one striking finding of our study is that this subset of gestures seems to play a larger role in interactions with same-aged peers, and a lesser role in interactions with considerably younger or older individuals (i.e. play partners with larger absolute age difference). This result strongly supports the view of play as training ground for social interactions in adult life. Young chimpanzees are possibly much more confident to use additional 'tools' of their communicative tool-kit if their play partners are of similar size and developmental stage. These interactions provide a safe environment and playground to train them for later, possibly more dangerous situations. Since the Taï South communities are renowned for their sophisticated tool-using technique of nut-cracking [13], it is surprising that gesture-object combinations do not play a larger role in Taï [98] found that sticks are commonly used and prepared at all three sites, while Taï South chimpanzees modified the material more before using it. Moreover, only Taï South chimpanzees pounded objects with tools and combined two different tool uses to get access to one food item [98]. However, the tool-using profiles of the two communities are distinctively different, and to our knowledge there is thus far no evidence that Kanyawara chimpanzees differ substantially from those in Taï South in terms of frequency of overall tool usage [99]. Nevertheless, the results obtained in this study might corroborate to some extent the complementation theory of language and tool use discussed in relation to human evolution [100].
Turning to self-handicapping gestures, we found that these were predominantly used to solicit play with younger, as opposed to older or same-aged, social partners. In addition, this behaviour was more common in young chimpanzees, indicating that this communicative strategy plays a larger role for juveniles and/or young sub-adults. These results thus expand findings by Flack et al. [23] on a captive group of chimpanzees, which showed that with larger age difference between play partners, the older play partner was more likely to play at lower intensity. They also reported a significant relation between the frequency with which an older play partner emitted play signals (e.g. play face, laughing) and the proximity of the younger play partner's mother, especially when play intensity was higher. If a young chimpanzee observes an older individual engaging in a compromising posture or performing a selfhandicapping movement, it may understand that this potential playmate poses no threat. That is, an individual lying in a supine position cannot attack promptly, and an individual shaking its head cannot leap precisely, because 'its judgement of distance and direction is momentarily blurred' [19, p. 146]. It has been suggested that self-handicapping strategies crucially limit the risk of harming younger play partners, serve to maintain the playful mood and prevent mother's intervention [35]. These observations led Flack et al. [23] to conclude that chimpanzees are able to perceive customary social rules about play. The causes underlying the inter-site difference in self-handicapping signals, which were more frequently produced in Taï South chimpanzees, remains elusive. One explanation may be that, due to the smaller community size of Taï South, the same individuals might interact more frequently and form better-established play dyads, resulting in a more common usage of self-handicapping strategies.
The means-end dissociation between signal and context is a key hallmark of flexible signal use [29]. Our study thus demonstrated signal flexibility in chimpanzees with regard to multiple gesture types used for a single communicative function (play solicitation). Since many of the play-soliciting gesture types described in our study have been shown to be employed in entirely different social contexts in previous studies [29,60,101], we could also show that single gesture types can be used for different communicative purposes. Hobaiter & Byrne [83] recently found that play-related meanings of gestures occurred with high generality, that is they were not dependably communicated by the outcome in other, 'serious' contexts. Flexible communication, as demonstrated by gestural play solicitations of chimpanzees in this study on an individual and a group level, is deficient of a one-to-one correspondence between single signals and responses, and assumed to involve higher cognitive complexity than the reflexive signals tightly linked to single contexts commonly found in the animal kingdom [82,102]. Intriguingly, a number of studies on great apes showed that social play is the context in which they test and practise the majority of gestures [21]. Hobaiter & Byrne [83] thus argued that play gestures are often not related to their normal meaning. Moreover, play signals mediating the play bout not only include gestures and body postures, as investigated in our study, but also facial expressions, spatial cues and relaxedness of movement. An exciting avenue for future studies thus would be to examine how different types of signals in play and other social contexts are integrated to achieve communicative goals and resolve ambiguity of signal meanings. Importantly, the overwhelming body of work on great ape gestures still stems from captive environments [103], which necessarily suffered from human intervention with the natural communication system [104]. However, we can only understand the communicative complexity of our closest living relatives if we also consider between-site variability [104]. Additional studies including more communities, thereby taking into account potential withinspecies variability, are needed to gain in-depth understanding of the cognitive mechanisms underlying ape communication.

Conclusion
In sum, by investigating the impact of demographic factors and kin relationships on play interactions in two different chimpanzee communities, our study has provided hitherto undocumented findings on how age, sex and a prevailing kin relationship to the recipient can influence gestural usage in natural environments. We showed that chimpanzee infants take into account distinct attributes of the play partner, using a higher frequency of visual and audible gestures towards individuals other than their mothers and a lower frequency of tactile (risky) gestures. Our study demonstrated that both object-associated and self-handicapping gestures are essential parts of the communicative tool-kit of chimpanzees that serve different functions depending on the age difference to the respective play partner. These findings further highlight the role of the play context as training ground for tool use and maintaining social relationships in later life. Moreover, this study provides evidence for a mutual construction of gestures by interactants and the cognitive ability to flexibly adjust these communicative means to social circumstances and individual matrices. We thus demonstrated gestural flexibility in great apes at an unprecedented scale. Although we will never be able to entirely grasp chimpanzee behavioural plasticity in its full complexity, systematic comparisons of the behaviour between communities and between subspecies are a vital step into this intriguing endeavour.
Ethics. Our study was purely non-invasive, with audio and video recordings taken from a minimum distance of 7 m, in an effort to avoid influencing the natural behaviour of the individuals, parties and communities observed.