ebook img

Playing it cool: Characterizing social play, bout termination, and candidate play signals of juvenile and infant Tibetan macaques (Macaca thibetana) PDF

2018·0.42 MB·
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Playing it cool: Characterizing social play, bout termination, and candidate play signals of juvenile and infant Tibetan macaques (Macaca thibetana)

ZOOLOGICAL RESEARCH Playing it cool: Characterizing social play, bout termination, and candidate play signals of juvenile and infant Tibetan macaques (Macaca thibetana) KaitlinR.Wright1,JessicaA.Mayhew1,2,LoriK.Sheeran1,2,JakeA.Funkhouser1,Ronald. S.Wagner1,3,Li-XingSun1,3, Jin-HuaLi4,* 1PrimateBehaviorandEcologyProgram,CentralWashingtonUniversity,EllensburgWashington98926,USA 2DepartmentofAnthropology,CentralWashingtonUniversity,EllensburgWashington98926,USA 3DepartmentofBiologicalSciences,CentralWashingtonUniversity,EllensburgWashington98926,USA 4SchoolofResourceandEnvironmentalEngineering,AnhuiUniversity,HefeiAnhui230601,China ABSTRACT juvenileandinfantmacaquesusedmultiplebodyand facial candidate play signals. Our data showed that Playbehaviorsandsignalsduringplayfulinteractions juvenileandinfantTibetanmacaquesuseaversatile with juvenile conspecifics are important for both the repertoireofplaybehaviorsandsignalstosustainplay. social and cognitive development of young animals. Keywords: Social play; Play signaling; Play face; Thesocialorganizationofaspeciescanalsoinfluence Macacathibetana juvenile social play. We examined the relationships INTRODUCTION among play behaviors, candidate play signals, and playboutterminationinTibetanmacaques(Macaca Playisoneofthemostconspicuousbehaviorsinwhichanimals thibetana) during juvenile and infant social play to engage. Social play combines elements of cooperation, communication, and reciprocity in participant actions. Play characterize the species play style. As Tibetan also incorporates behavioral modifications from other social macaques are despotic and live in groups with contexts, such as agonism, mating, and predation, and strict linear dominance hierarchies and infrequent thus the division between play and non-play is not always reconciliation,wepredictedthatplaywouldbeatrisk obvious (Burghardt, 2005). Although play behavior has been of misinterpretation by both the individuals engaged researchedformanydecades,itisvariouslydefineddepending in the play bout and by those watching, possibly on the field of study from which it is viewed. For example, Fagen (1981) defined play as behavior that “functions to leading to injury of the players. Animals living in develop,practice,ormaintainphysicalorcognitiveabilitiesand such societies might need to frequently and clearly social relationships,includingbothtactics and strategies, by signalplayfulintenttoplaypartnersandothergroup memberstoavoidaggressiveoutcomes. Wegathered video data on 21 individually-identified juvenile and Received: 31October2017;Accepted: 12April2018;Online: 04May infant macaques (one month to five years of age) 2018 fromtheValleyoftheWildMonkeys,Mt. Huangshan, Foundation items: This project was supported by grants from China. We used all-occurrence sampling to record Debra and Arlen Prentice, Central Washington University School play behaviors and candidate play signals based of Graduate Studies and Research, National Science Foundation on an ethogram. We predicted that play groups (NSF)InternationalResearchExperiencesforStudents(IRES)(OISE wouldusemultiplecandidateplaysignalsinavariety 1065589), and National Science Foundation of China (31372215, of contexts and in association with the number of 31672307) audiencemembersinproximitytotheplayersandplay *Correspondingauthor,E-mail:[email protected] boutlength. Inthe283playfulinteractionswescored, DOI:10.24272/j.issn.2095-8137.2018.048 272 SciencePress ZoologicalResearch39(4):272–283,2018 varying, repeating, and/or recombining already functional behaviorsandisincompleteorexaggerated; (4)isperformed subsequences of behavior outside their primary context” (p. repeatedly but not stereotypically; and (5) occurs when the 65).Fagen’sdefinitionfocusesonplayfunction;however,even organism is free of stress or social/physical pressures (i.e., todaytheultimatebenefitsofplayremaincontroversial. From the player is in a “relaxed field”). Burghardt (2005) further astructuralperspective,Burghardt(2005)proposedaworking posited that these criteria must be met to label a behavior as definition for identifying play behaviors using five key criteria: playfulinsolitaryorsocialcontexts. Fagen’sdefinitionofplay that is, play behavior (1) has limited immediate function; (2) behavior,coupledwithBurghardt’scriteriaforidentifyingthese hasanendogenouscomponentthatisvoluntary,rewarding,or behaviors, offer a method to recognize and characterize play autotelic;(3)isstructurallyortemporallydifferentfrom“serious” versusnon-playbehaviorinanimalsofallages(Table1). Table1Operationaldefinitions Term Definition Functionstodevelop,practice,ormaintainphysicalorcognitiveabilitiesandsocialrelationships, Playbehavior includingbothtacticsandstrategies,byvarying,repeating,and/orrecombiningalreadyfunctional sub-sequencesofbehavioroutsidetheirprimarycontext(Fagen,1981,p65). Communicatorybehaviorsthatfunctiontopromote,cultivate,andmanagesocialplayand Playsignal demonstrateplayfulintentions(Bekoff,1974;Fagen,1981;Yanagi&Berman,2014a,2014b). Successfulbout Startoftheboutmarkedbyexchangeofphysicalcontact,chasing,orotherplaytypeorplaysignal. Various solitary and social play behaviors are observed Thierry,2007).Despoticspecies,suchasJapanesemacaques across animal species (e.g., turtles, Trionyx triunguis: (Macaca fuscata), with high rates of aggression, low rates of Burghardt et al., 1996; elephants, Loxodonta Africana: Lee reconciliation, and high levels of nepotism are characterized & Moss, 2014; domestic dogs, Canis familiaris: Horowitz, by a competitive, defensive, and low-risk play style (Petit et 2009);however,playisparticularlyimportanttoprimates,with al., 2008; Reinhart et al., 2010). Additionally, mothers in relevancetosocialandcognitivedevelopment(Martin&Caro, despotic species may adopt a more restrictive rearing style, 1985; Palagi et al., 2007). Therefore, studying play can and mothers of low-ranking individuals may intervene when provide important insight on a species’ social relationships, theiroffspringattempttoplaywiththeoffspringofhigh-ranking how a species develops, and how they relate to cognitive females(Maestripieri,2004).Incontrast,speciesthataremore abilities. Playing with juvenile and infant (hereafter: juvenile) egalitarian and characterized by low rates of aggression and conspecifics is typically the first non-mother activity to occur high rates of affiliation and reconciliation, such as Tonkean in young animals (Bekoff, 1972; Poirier, 1970). Playing macaques(Macacatonkeana),haveamorecooperative,close with peers gradually increases in frequency, complexity, and contact play style and less restrictive mothers (Maestripieri, intensity as juveniles age and their social networks expand. 2004; Petit et al., 2008; Reinhart et al., 2010). These play The repetitive modification and practice of behaviors (e.g., patterndifferencesappeartovarywithsocialorganizationand mounting and biting) within the play context may yield both exist due to the potential devolvement of play from a friendly short- and long-term benefits in future hunting, mating, or interaction into an aggressive one that has likely negative social interactions (Bekoff & Allen, 1997). The frequency of repercussions for the players. Variability in play frequency playful behavior typically declines as juveniles transition into and form also extends to the individual level, in which an adulthood, although adults may maintain playful relationships individual’s“playfulness”islikelyinfluencedbymultiplefactors, with juveniles (e.g., geladas, Theropithecus gelada: Mancini including personality (Lampe et al., 2017; Pellis & McKenna, & Palagi, 2009; chimpanzees, Pan troglodytes: Palagi et al., 1992), prior experience (Cloutier et al., 2013), opportunity to 2004; Shimada & Sueur, 2014). Less frequently, adult-adult play(Panksepp&Beatty,1980),andneurochemistry(Siviyet play can occur in some primate species in both sexual and al.,2011). However,muchofthisresearchislimitedtohuman non-sexualcontexts(Pellis&Iwaniuk,1999;T.gelada:Mancini children and laboratory rats; therefore, examining individual &Palagi,2009;bonobos,P.paniscus: Palagi,2006;ring-tailed differences in playfulness across multiple species, including lemurs, Lemur catta: Palagi, 2009; Pellis & Iwaniuk, 2000; P. primates,isnecessary. troglodytes: Yamanashi et al., 2018) but may exhibit variable Social play fighting in juvenile animals may influence the formsandfunctionsinspecieswithdifferentsocialorganization. development of dominance relationships later in life (Pellis In primates, e.g., Macaca spp., social organization exerts & Pellis, 1996), and the style of play fighting may also be a pervasive influence on a variety of behaviors, including related to a species’ social structure and dominance style play (Ciani et al., 2012; Fagen, 1981; Maestripieri, 2004; (Fagen,1981;Palagi,2006;Petitetal.,2008).Althoughgentle ZoologicalResearch39(4):272–283,2018 273 playfightingmaybeusedtomaintainaffiliation, moreintense multifunctionalandmayalsoworktomodulatemood(Pelliset rough-and-tumble play may establish a dominance hierarchy al.,2011),inviteathirdpartyintotheplaybout,expressplayful inpost-pubertaljuveniles(especiallyinmale-maleplaybouts) intenttoathirdparty, oractasarewardforbothpartnersfor bytestingthestrengthofplayparticipants(Palagietal.,2007; playful engagement (Spijkerman et al., 1996). Play signals Pellis & Pellis, 1996). Thus, play style may be predictive of likely perform a critical role in prolonging play duration and socialdominancestyle,andthebalancebetweencooperation maintainingmultipleplayersinabout(polyadicplay).However, andcompetitionwithinaplayboutmaybedifferentdepending it is possible that certain factors also decrease a signal’s on the nature of a species’ social system (Palagi, 2006; Petit salience, including increase in bout length, bout intensity, et al., 2008; Reinhart et al., 2010). An individual may use misdirectionorimpairmentofthesignal,changesinaudience certain behaviors, such as slap or chase, to test, cultivate, or members, or the addition of more players (Bekoff, 1972). stabilize a competitive edge in a play bout (van Leeuwen et Short play bouts may be influenced by the misinterpretation al., 2011). By testing one’s competitive advantage against a of play signals and, similarly, long play bouts with aggressive peerduringplay,partnerscanpracticeaggressiveinteractions behaviors, such as wrestling, may need to include a higher thatmaybenecessarylaterinlifetodefend,maintain,orgain frequency of intention reinforcing play signals (Spijkerman et accesstoresources. Moreover, thefrequencyofplayfighting al., 1996). Research reflects the variation and degree of and use of play signals in certain social systems, such as frequency, flexibility, and functionality of play signals, which the egalitarian social structure characteristic of adult female also appear to be influenced by species social organization. bonobos(P.paniscus),mayindicatethenecessityofaflexible Therefore,possessingadiverserepertoireofsignals(suchas playstyletoassessandstrengthensocialrelationships(Palagi, vocalizationsandheadandbodymovements),inadditiontothe 2006). iconicplayfaceorplaybow,maybeadvantageousforjuveniles For many primates, social play is frequently coupled with tolearnandpractice;onemightexpectmultipleplaysignalsto playfulsignalsobservedthroughoutaplaybout(Fagen,1981; haveevolvedinanygiventaxon. Yanagi & Berman, 2014a). During play, partners transmit For the genus Macaca (23 species), visual signals during and receive signals, which include vocalizations (Biben & play, such as bared teeth or play face, are used differently Symmes, 1986; Kipper & Todt, 2002, Vettin & Todt, 2005), depending on the social organizational grade of the species body movements, gestures, and/or facial expressions (e.g., (Scopa & Palagi, 2016; Thierry et al., 2000). For example, relaxedopenmouthorplayface;Bekoff&Allen,1997;Pellis& bared teeth as a play signal might be held longer or Pellis,1996).Thesesignalsmayfunctiontoqualifysubsequent performedmorefrequentlyinonespeciescomparedtoanother. behaviors as playful, maintain a playful context, and/or help Macacaspeciessharesimilaritiesinsocialstructure,including to avoid an escalation to aggression, especially when the multi-maleandmulti-femalegroups,overlappinghomeranges, behaviors performed are risky or ambiguous (e.g., play bite, and female philopatry; however, interspecific variation can play slap, or play fight; Bekoff & Allen, 1997; Burghardt, be found in patterns of affiliation, reconciliation, dominance, 2005; Pellis & Pellis, 1996). Play signals can therefore be aggression,nepotism,andtemperament(Thierry,1985,1990; definedascommunicatorybehaviorsthatfunctiontopromote, Thierry et al., 2000). Macaques are categorized on a cultivate, and manage social play and demonstrate playful 4-gradetolerancescale,withGrade1beinghighlyhierarchical, intent (Bekoff, 1974; Fagen, 1981; Waller & Cherry, 2012; nepotistic, and despotic (e.g., M. mulatta or M. fuscata) and Yanagi&Berman,2014a;Table1). Thereisdebatethatsuch Grade 4 being more tolerant or egalitarian (e.g., M. tonkeana signals are only observed during the context of play and are or M. nigra) (Thierry et al., 2000). Tolerant macaque species distinct from the behaviors used within a play bout, and also use play signals interchangeably and redundantly to initiate predict the occurrence of play (Bekoff, 1974; Fagen, 1981; and/or terminate play and self-regulate mood (Pellis et al., Yanagi & Berman, 2014a). However, many potential signals 2011; Scopa & Palagi, 2016). They also engage in rapid discussed in the literature (e.g., play bow in canids or play facial mimicry with their play partner, where one partner face in primates) are unclear as indicators of the beginning mimics the facial expression of the other (Scopa & Palagi, of a play bout; instead, these signals often punctuate the 2016). Suchmimicryisanunconsciousmotormirrorresponse boutatdifferentpointsandarevariableintheirduration,form, and appears to contribute to increased play duration in more and intensity. For example, the play face, also known as the tolerant species (Mancini et al., 2013; Scopa & Palagi, 2016; relaxedopenmouthface,isafrequently-observedplaysignal see Palagi & Scopa, 2017 for discussion). This flexibility in in primates and is commonly associated with close-quarter signaling may be related to an unpredictable but cooperative contact, which occurs during play fights (Palagi et al., 2014; and less risky play style. Conversely, in despotic macaque Pellis & Pellis, 1996; van Hooff, 1967). The intensity, rate, species, ambiguous play behaviors and miscommunication and timing of the play face can change with the intensity or likely generate increased social repercussions and physical behavioral content of the play interaction, thus acting as a risks; therefore, play signals may be particularly inflexible, flexible message to indicate playful intent while the dynamics specific,andlessinterchangeableinsuchspeciesthatexhibit of a bout quickly change (Pellis & Pellis, 1996; Špinka et al., highlevelsofaggressionandcompetitiveplaystyle(Scopa& 2016;Symons,1978;Waller&Cherry,2012;Waller&Dunbar, Palagi, 2016; Thierry et al., 2000; Yanagi & Berman, 2014b). 2005). Thisflexibilitysuggeststhattheplayfaceismostlikely In such cases, play signals may alert both participants and 274 www.zoores.ac.cn third parties that the players are “only playing,” and may be as there are currently no published studies characterizing performed more often when a third party, such as a related the specific play behaviors and signals observed during their adult or juvenile conspecific, is present and participation or playful interactions. Insight on Tibetan macaque juvenile play interferenceislikely(Pellisetal.,2011). will provide an important foundation to examine their social In free-ranging despotic juvenile rhesus macaques (M. and cognitive development in relation to social structure and mulatta), play signals predict the imminent occurrence of organization. dyadic play (Yanagi & Berman, 2014a). Furthermore, most We characterized play behaviors and identified candidate signals are non-randomly associated with various initiations play signals in juvenile Tibetan macaques. Tibetan of play, indicating that signals are used selectively by the macaques show evidence of despotism, with a Grade 2 players depending upon the play content (Yanagi & Berman, social organization, linear dominance hierarchies, and low 2014b). For instance, chase play is often associated with conciliatory tendencies (Berman et al., 2004). It is expected a crouch-and-stare signal, and leg-peeks are typically used that this despotic social structure will impact juvenile play by the play recipient, not the initiator (Yanagi & Berman, behavior,playboutproximitytoadults(i.e.,potentialthird-party 2014b). For Grade 1 despotic macaques, the specific adult interference), frequency and distribution of play signals use of signaling might decrease miscommunication and help performed by juveniles, and composition of players within the reinforce,clarify,oremphasizetheplayfulintentofthesender, bout (i.e., audience members). Specifically, do juveniles use wherebyindicatingthatparticipantsare“onlyplaying”alleviates specificcandidateplaysignalsinvariousplaycontextsbased potentialrisingtensionwithinthegroup. Therefore,onemight on the length of the bout and number of audience members hypothesize that multiple play signals may be used in other in proximity? We therefore tested the following predictions: despoticmacaquespeciestoclarifysocialplaywhilemitigating 1) due to the despotic dominance style, third-party adult potential associated aggression (Yanagi & Berman, 2014b). interference will end play more often than other forms of play However,theextenttowhichthesesignalsarespecies-specific bouttermination(e.g., withdrawal)(Pellisetal.,2011; Thierry, isunknown. 2011; Yanagi & Berman, 2014b); 2) play behaviors and play Tibetan macaques (M. thibetana) are the largest macaque signals will show unequal distribution at the beginning of a species and the most derived of their particular lineage bout, with play signals occurring more frequently to indicate (Fooden, 1983; Thierry, 2011). Female macaques reach that the subsequent behaviors are playful (Yanagi & Berman, adulthoodatapproximatelyfourorfiveyearsofage,generally 2014a); 3) the number of play bout audience members will birththeirfirstoffspringbetweenfourandsixyearsofage,and be positively correlated with the number of observed play nurse infants for six to twelve months (Zhao & Deng, 1988). signalsduetoanincreasednecessitytoconveyplayfulintent In males, adulthood begins at approximately seven years of to more individuals (Bekoff, 1972; Spijkerman et al., 1996; age (Zhao & Deng, 1988). Tibetan macaques reproduce Palagi,2006);and4)playdurationwillbepositivelycorrelated seasonally, and the number of offspring sired by a male is with the number of observed play signals to reinforce, clarify, correlated with his dominance rank (Thierry, 2011; Xia et andemphasizeplayfulintent(Pellis&Pellis,1996;Spijkerman al., 2012). Tibetan macaque social organization consists of et al., 1996; Yanagi & Berman, 2014b). We also analyzed multi-male, multi-female groups of 15–50 individuals, with a the distribution of playful behaviors and signals across all female-skewed sex ratio (Berman et al., 2004; Li et al., 2007; individualsandinvestigatedthepossibleeffectsofsexandage Thierry et al., 2000; Thierry, 2011). This organization is on the frequency of dyadic engagement in playful behaviors centeredondominancehierarchiesandkin-bondedcoalitions andplaysignaling. (Thierry, 2011). The dominance rank of females is based on MATERIALSANDMETHODS matrilines, with a daughter generally attaining the dominance rankimmediatelybelowhermotherbutaboveheroldersiblings Subjectsandstudysite (Berman et al., 2004; Thierry, 2011; Zhao, 1997). This Video data were collected on 21 individually-recognized, hierarchy influences intergroup competition among females free-ranging infant and juvenile Tibetan macaques of the and preferential bonds between kin (Thierry, 2011). Most Yulingkeng A1 (YA1) group located at the Valley of the Wild males disperse at adulthood and transfer between groups Monkeys in the Huangshan Scenic District, Anhui Province, during their lifespan, regardless of dominance rank (Thierry, China(seeBerman&Li,2002formoreinformationaboutthe 2011; Zhao, 1997). Although group males occupy the top site). These individuals were between one month and five ranks,femalescanoccasionallyoutrankmales(Bermanetal., yearsoldbasedonMacacaagecategorization(Thierry,2011) 2004). Adult social relationships influence the socialization and current group structure maintained by the researchers of of immature individuals (Thierry, 2011), with the population Anhui University. At the time of the study (summer 2015), generally showing a kin bias and linear hierarchies (Berman 10 juvenile males were present, including (4) five year olds, et al., 2004). As such, it is expected that the Grade 2 social (3) three year olds, (1) two year old, and (2) individuals less organization of Tibetan macaques will affect juvenile play via than 12 months old; and 11 juvenile females were present, third-partyadultinterference, wherebyaggressiveoraffiliative including:(1)fiveyearold;(3)threeyearolds;(4)twoyearolds, behaviors are used to disrupt and terminate play. However, and (3) individuals less than 12 months old. Data collection little is known about the play style of Tibetan macaques, focusedonjuvenilesasadult-adultplayisrareandmaydisplay ZoologicalResearch39(4):272–283,2018 275 different frequencies of play signals than that which occurs video data. During preliminary data collection, inter-observer during immature play (Mancini & Palagi, 2009; Palagi et al., reliabilityinindividualidentification(κ=0.86)andintra-observer 2004; Shimada & Sueur, 2014). The members of YA1 have reliability using play behavior and signal ethograms were been habituated to human presence since 1986 for scientific established (play behavior: κ=0.93, play signals: κ=0.88, researchandsince1992fortourism(seeBermanetal.,2004). bout termination: κ=1.00, actor identification: κ=0.87, and The macaques at this site are provisioned with corn three to audience member identification: κ=0.86). All-occurrence fourtimesdailybytheparkstaff,andthefeedingsarevisibleto sampling was used to efficiently record high quantity playful tourists (see Berman & Li, 2002). The monkeys occasionally interactions(Altmann,1974).Allplayfulinteractionsdefinedby interact with people (McCarthy et al., 2009). We obtained Fagen(1981)andoutlinedbyBurghardt(2005)wererecorded, research approval from the Central Washington University including play behaviors (see Tables 1 and 2) and candidate Institutional Animal Care and Use Committee (#A021507), play signals (see Tables 1 and 3), using a Canon HD Vixia andourresearchprotocolsfollowedthelegalrequirementsof camcorder.Observationswereundertakenfromtourist-viewing the People’s Republic of China and the American Society of platforms and near feeding sites and other locations where Primatologists’PrinciplesfortheEthicalTreatmentofPrimates. juvenileandadultmacaqueswerevisible.Ifasecondplaybout occurredsimultaneouslywhileobservingaprecedingplaybout, Videodatacollection onlyplaybehaviorsandsignalsintheinitialboutwererecorded. Video data were collected at the study site from 3 August to After the players disengaged from the play bout, recording 19September2015fromapproximately0800hto1800heach continueduntil10shadelapsedwithoutboutre-initiation. day, resulting in 48 d in the field and approximately 400 h of Table2Definitionsandcomponentsofplaybehaviors Typeofplay Behaviorcomponent Definition Locomotiveactions,suchasrunning,climbing,and leapingtowardsorawayfromanotherindividual,in Chasing Leaping,running,walking whichanimalsalternaterolesofchaserandchasee, withouthavingbodycontactwitheachother. Slightlyresembleswrestling,butinanextremely Embracing,holding,hugging, mildform,i.e.,holdingeachotherwithveryslight Cuddling touching pushingofbody,butwithoutanybodydisplacement. Oftenresemblesembracing. Playinwhichanimalsgrappleandplacetheir Biting,dragging,embracing, mouthsoneachother’sbody.Typicallyinvolves grabbing,hitting,leaping,lying, Playbiting similarbehaviorpatternstowrestlingbutoccurswith pinning,pulling,pushing,rolling, biting.Bitingandavoidingbeingbittenwithbody running,tackling,touching,walking displacementarecentralactivities. Twoanimalshiteachotherwiththeirhandsfora Hittingwithhands,touching, Slapping periodwithoutproceedingtoaclearerformofplay, visualfixation norterminatingtheplayencounter. Dragging,embracing,grabbing, Alsoknownasrough-and-tumbleplay.Includes hitting,leaping,lying,pinning, playbehaviorpatternsinwhichtwomonkeys Wrestling pulling,pushing,rolling,running, engageinmutualgrasping,pushing,pulling, tackling,touching,walking androlling,withoutattemptstobiteeachother. Ethogramforrhesusmacaque(Macacamulatta)play:Yanagi&Berman(2014a;2014b). Videodataanalysis used to analyze the play bout video data, recording all play Fromthevideofootage,thetimestamp,actoridentity,audience behavior and candidate play signals observed throughout all member identity, and play signals and behaviors of the dyadicandpolyadicplaybouts. Behaviorspreviouslyunlisted participants were coded (QuickTime Player for Mac). A by Yanagi & Berman (2014b) were added to the play signal modified macaque ethogram (Yanagi & Berman, 2014b) was ethogram and included play threat and slap-and-play face 276 www.zoores.ac.cn (Table 3). We estimated each player’s proximity to group the audience member tally was categorized as zero. In this members as within arm’s reach (<50 cm) or beyond arm’s study,weconsideredplaybouts“successful”whenthestartof reach (>50 cm). If a group member (adult or juvenile) was theboutwasmarkedbytheexchangeofcontact, chasing, or in proximity to the play bout then he or she was considered otherplaybehaviors/signals(recordedas“startofplay”)(Table partoftheplaybout’saudience,regardlessoftheirorientation 1). Asuccessfulplayboutwasconsidered“terminated”when or activity. Each member of the play bout was counted to a player engaged in 1) non-play activities, e.g., grooming; 2) generateanaudiencemembertally. Ifanindividualperformed withdrewfromthebout;or3)adultinterferenceoccurred(Table aplaysignalorbehaviorwithoutagroupmemberinproximity, 4;recordedas“endofplay”). Table3Definitionsofplaysignals Playsignal Definition Animal’sventralsurfaceison/neargroundanditslimbsarefixed, Crouch-and-stare whilemaintainingvisualfixationonpartner(Symons,1978). Dangle-and-stare Animalstaresatpartnerwhilehangingfromanobjectbyhindlimbs(Levy,1979). Bobbing,highsteppinggaitinwhichforequartersandhindquartersarealternately Gamboling raised(Symons,1978).Oftenaccompaniedbyrotationofthehead(Sade,1973). Animalhidesbehindanobjectandthenpeeksatpartner, Hide-and-peek alternatingthetwobehaviorpatterns. Animalstaresatpartnerthroughitslegswiththetopofitsheadagainsttheground Leg-peek (Symons,1978).Animalmayholditsanklesorplaceforearmsonground. Animal’sbodyisorientedawayfrompartnerinafixedpositiononallfours,whilethe Look-back headisturnedtowardthepartnerovertheshoulder(Levy,1979;Symons,1978). Playface Relaxed,openmouthface,typicallyobservedduringplaybouts(Levy,1979). Roll-onto-back Animalrollsandliesonitsbackandstaresatpartner(Levy,1979). -and-stare Playthreat Animaldirectsalunge<2bodylengthstowardsanotherindividual, (candidatesignal) endingthemovementbyhittingtheground,withoutfacialexpression. Slap-and-playface Animalhitsanotherindividual’sbodywhilesimultaneously (candidatesignal) directinganopenmouthfacetowardstheindividual. AdaptedfromYanagi&Berman(2014b). Table4Definitionsofplaybouttermination Modeoftermination Definition Playersbegintoengageinanybehavior/ Non-playactivities activitynotconsideredunderthecategory orcriteriaofplay(Bermanetal.,2004). Playersmoveoutofproximityfromeach Withdraw other(outofarmsreach),andnosubsequent playbehaviorsorsignalsareobserved. Playboutisinterruptedbyanadultgroupmember Adultinterference performingaggressiveornon-aggressivebehaviors towardsanyplayer(Bermanetal.,2004). AdaptedfromBermanetal.(2004). ZoologicalResearch39(4):272–283,2018 277 Statisticalanalysis acrossindividuals. UsingIBMSPSSStatistics(Version23),VassarStatsWebsite Distributionofplayacrossageandsexclasses forStatisticalComputations(©RichardLowry,1998–2013),and We used MR-QAP regression analyses in UCINET to UCINET (Borgatti et al., 2002), we tested each prediction investigate possible correlations between the dyadic with α=0.05. We only analyzed successful bouts and the frequencies of play behaviors, play signals, age, and sex. signals that were displayed during such bouts. We used Being of the same age class had no effect on either the chi-squaregoodnessoffitteststoassesspredictions1,2,and dyadic frequency of playful behaviors (Y=–0.007X+0.86, 3. Spearman’s rank correlation tests were used to analyze P=0.41,r2=0.00)orplaysignaling(Y=–0.008X+4.53,P=0.43, the average number and rate (average number/min) of play r2<0.001). However, being of the same sex was significantly signalsacrossboutlength(prediction4).Additionally,weused related to both the dyadic frequencies of playful behavior chi-square goodness of fit tests to examine the distribution of (Y=1.22X+0.54, P=0.025, r2=0.023) and play signaling the total frequency of play signals across varying audience (Y=16.91X+2.73,P=0.031,r2=0.017). Althoughtheseresults numbers. We also used chi-squared goodness of fit tests to showedthatsexhadasignificanteffectonplaybehaviorand analyzethedistributionofplaybehaviorsandsignalsforeach signaling, the sizes of these effects were small (r2<0.023). individual. Spearman’scorrelationcoefficientwasusedtotest Giventheseresults,wedidnotdividethesubsequentanalyses the correlation between the number of observed play signals intodifferentdyad-specificsexorageclasses. acrossaudiencemembercategories. WealsousedMR-QAP matrix-based regression analyses in UCINET to analyze the Playbouttermination effectofsexandageclassesonthedyadicfrequencyofplayful We used chi-square goodness of fit tests to compare the behaviorsandsignals. threeplayterminationcategories(Table4)andtestprediction While these analyses might be less robust than others, 1 (third-party adult interference will end play more frequently chi-square goodness of fit tests are valuable as they can use thanotherformsoftermination). Resultsshowedastatistically entire datasets to determine overall patterns (Reinhart et al., significant difference from the expected values (χ2(2)=53.52, 2010). We used chi-squared analyses to investigate general P=0.001),indicatingthatterminationoccurredmorefrequently patternsinTibetanmacaqueplay. However, asotherauthors duetonon-playfulbehaviors(n=74)andwithdrawal(n=94)than deviatefrominvestigationsofgeneralpatterns,morecomplex tothird-partyadultinterference(n=16). Therefore,prediction1 analyseswouldbepreferabletoaccountforthewidevariation wasnotsupported. in assumptions that are violated by play-typical data (e.g., GLMM). Playsignals Intotal,415playsignalsinplayfulinteractionswerecoded.The RESULTS averagenumberofplaysignalsseenincompletedplaybouts Intotal,werecordedandanalyzed283dyadic(=2individuals) (n=94) was 1.6 play signals per bout. We used chi-square andpolyadic(≥2individuals)playbouts,with136observations goodness of fit tests to compare the total frequency of play of the start of play, 183 observations of the termination of bouts initiated by either a play behavior (Table 2) or a play play, and94observationsofacompleteplaybout(wherethe signal (Table 3). Results showed a statistically significant start and termination of play were clearly marked). In the 94 differencefromtheexpectedvalues,andthusprediction2was completed play bouts observed, average length was 64.7 s, not supported (χ2(1)=45.88, P=0.001): play bouts were most with a range of 1 to 585 s. The number of players present frequentlyinitiatedbyplaybehaviors(n=108)ratherthanplay in a play bout ranged from 1 to 5, and all 22 juveniles were signals (n=28) (an unequal distribution across bouts). In the observedtoengageinplayatleastonce. playboutsobserved(n=283),ninecandidateplaysignalswere recorded,includingcrouch-and-stare(n=41),dangle-and-stare Distributionofplayacrossindividuals (n=21),gamboling(n=3),hide-and-peek(n=1),look-back(n=1), Only positively identified individuals were included in the playface(n=263),roll-onto-back-and-stare(n=19),playthreat analyses (n=19). During play bouts, 16 juveniles exhibited (n=17), and slap-and-play face (n=49) (Table 5). Six of the both play behaviors and signals during bouts. We seven candidate body signals observed in rhesus macaques used chi-square goodness of fit tests to investigate the (Yanagi & Berman, 2014a) were also observed in Tibetan distribution of individual engagement in play behaviors and macaques (crouch-and-stare, dangle-and-stare, gamboling, signals. Play behaviors were differentially (as opposed to hide-and-peek, look-back, and roll-onto-back-and-stare). Two equally) and significantly distributed across individuals (n=19; additional candidate signals were also observed (play threat χ2(18)=4 016.06, P=0.001). Each juvenile engaged in playful andslap-and-playface). behaviors between 4 and 589 times (125.35±169.58). Play We compared the frequency of play signals observed for signaling was also significantly and differentially (as opposed zero (n=14), one (n=137), two (n=159), three (n=86), four toequally)distributedacrossindividuals(n=19;χ2(15)=561.52, (n=16),andfiveaudiencemembers(n=3)andpredictedthatas P=0.001). Eachjuvenileexhibitedplaysignalsbetween1and thenumberofaudiencemembersincreased,thefrequencyof 101times(25.56±30.93).Theseresultsindicatewidevariation playsignalswouldalsoincrease(prediction3;Table5).Results in the frequency and use of playful behaviors and signals showed a significant statistical difference from the expected 278 www.zoores.ac.cn values (χ2(5)=335.46, P=0.001); however, the Spearman’s weidentifiedwhenplaysignalswereusedbyacurrentplayer rankcorrelationbetweenfrequencyofplaysignalsandnumber as a new member joined or a member withdrew from a ofaudiencememberswasnotsignificant(r(4)=–0.371,P>0.05; bout.Resultsshowedasignificantdeviationfromtheexpected Figure 1). Thus, these results suggest the trend between values (χ2(1)=6.025, P=0.001), and most players did not play signals and audience members was not linear, but the use a play signal when a new member joined the bout (no frequencyofplaysignalsacrossaudiencemembercategories signal=230, signal=26)or when anexisting memberwithdrew wasdistributednon-randomly.Toexaminetheseresultsfurther, (nosignal=101,signal=24). Table5Playsignalsobservedforeachaudiencemembercategory Audience Play Dangle- Crouch- Hide- Look- Roll-onto-back- Play Slap-and- Total Gamboling number face and-stare and-stare and-peek back and-stare threat playface (n) Zero 0 1 8 0 0 0 5 0 0 14 One 84 11 18 1 1 1 7 7 7 137 Two 104 6 11 1 0 0 4 7 26 159 Three 61 3 3 1 0 0 2 1 15 86 Four 11 0 1 0 0 0 1 2 1 16 Five 3 0 0 0 0 0 0 0 0 3 Total(n) 263 21 41 3 1 1 19 17 49 415 We used Spearman’s rank correlation to test prediction 4, as aggressive interactions. In despotic species such as that play duration will be correlated with the rate of observed Tibetan macaques, observers of play may interpret a friendly play signals. Results showed a strongly significant positive exchange as an aggressive one if it does not contain clear correlationbetweentheobservedrateofplaysignalsandbout andfrequentplaysignals,andconsequentlymayterminateor duration (Spearman’s rank correlation: r(5)=0.991, P=0.001; disrupt the play bout. However, our results did not support Figure 2) and the average number of play signals and bout this prediction. In fact, play termination in juvenile Tibetan duration (Spearman’s rank correlation: r(5)=0.964, P=0.001; macaquesusuallyoccurredbecauseindividualswithdrewfrom Figure 3), thus indicating that as the length of a play bout the bout or engaged in non-play behaviors. We observed increased, the average number and rate of play signals also only 16 cases in which other monkeys terminated the playful increased. interaction:onecasebyalow-rankedadultmaleandonecase byalow-rankedadultfemale,twocasesbyamid-rankedmale DISCUSSION andfivecasesbyamid-rankedadultfemale,threecasesbya We examined the behaviors and candidate signals used in high-rankedadultmaleandtwocasesbyahigh-rankedadult juvenile Tibetan macaque play bouts to characterize the play female, and two cases by a young adult female. We have style of this species. Yanagi & Berman (2014b) found insufficient data to draw conclusions regarding the identities that juvenile rhesus macaques use play signals in selective of the individuals who terminated these play bouts, but our ways, as signals were non-randomly associated with various anecdotalevidencedoesnotfollowourpredictionbothinterms subsequent play behaviors, and further hypothesized that oftheoverallnumberofthird-partyterminationsortheidentities signalsmaybenecessarytoreinforceandclarifyplayfulintent. of the individuals who terminated the bouts. It is possible Despotic macaques may use play signals more frequently to that juvenile macaques avoid areas occupied by adults to emphasizetheaffiliative,ratherthanaggressive,natureoftheir directly manage the end of play themselves and avoid adult behaviors. OurstudyshowedthatjuvenileTibetanmacaques interference(Selfetal.,2013).Additionally,althoughprediction used similar play signals as rhesus macaques, but because 1 was not supported, due to their despotic dominance style these behaviors did not frequently precede the beginning of and based on the Burghardt (2005) criteria for play, young a play bout, their use may reinforce and clarify playful intent macaquesmayavoidadultstomaintainplayinarelaxedfield. ratherthansignaltheinitiationofabout. However, further comparative research is needed to test the Bermanetal.(2004)hypothesizedthatduetotheirdespotic hypothesis that social organization in all macaque species nature,Tibetanmacaqueadultswouldinterferewithplaybouts (including Grades 2 and 3) may influence play style (but see perceivedbythirdpartiesasaggressive.Manybehaviorsused Reinhartetal.,2010), andthereforeplayterminationbyadult in play, such as bite, wrestle, and chase, are also used in interference. aggressive contexts and could potentially be misinterpreted ZoologicalResearch39(4):272–283,2018 279 180 180 signalinginjuvenileTibetanmacaquesmaybeusedsimilarly 160 160 159 tosignalinginjuvenilerhesusmacaques,inwhichplaysignals 140 159 areusedtoclarify,reinforce,andprolongaplayboutbetween 140 137 n) 120 137 partners (Yanagi & Berman, 2014b). Furthermore, our data Frequency of play signals (n)Frequency of play signals(111468024680000000000 86 86 (smmdfaphrceoraoeemrwwedn(biecnsedtf=iirrgoos2annm6aa33npl;st;dohSseTonsuascuipebbcmpllueerblreere5smitnrr;uegeolFtnnfsidwgt.pauliratIbrhtyyeeatFsopw1iinpgg)e.eenueaarSneorlsesrtSvhgtte1hweer)anaotmneluiracnoamoudrtnedeibvcideetildhurninuascoanieaoflsnobmastuohuesucedmetari,enbppwnellbicaatreehsyy 2200 signals(n=152),regardlessofaudiencenumber.Thisindicates 00 Z1e4r1o4 One Two Three F1o6ur 16 Fi3ve 3 the salient nature of the play face signal when there is a Zero One Two Three Four Five receiverpresent(Pellis&Pellis,1996;vanHooff,1967)orwhen Number of audience members (n) Number of audience members (n) acombativeplaybehavior,suchasplayslap,mayescalateto FFiiggurue r1e Fr1eqFuernecqy uofe pnlacyysigonaflsp wlaityh dsififgernenat lasudwienitche mdeimffbeerrse ntaudience Figure 1 Frequency of playsignals with different audience members aggression(Palagietal.,2007).Thiswasfurthersupportedby members the lack of play face observed in the zero-audience member category (when no other group member was in proximity 12 to the initiator). Therefore, play face may be an important 12 (n)10 communicativetoolinpolyadicplaywhenaplayerisnearthe Average number ofplaysignalsAverage number of play signals (n)10246824680 1sotpt&zaooee9tllalhrPn6ocytoehd7elft-rauhee)lasl.rleiruis.ifirpndy,nHTlcitap1eeoaphyr9nwntlioeaes9ctxeryig6eosirvmo)enew.mrst,irihth,eueyaMeeslpmsttmnool(sabirnstuetyte=hehcgor4elefagvvac0neeebcad1rssoeb,ty)teu,nest(mgthaPatuorenmaheasrdyalyteilohcittsdcb(ehspncerei&lu=noatlhar1yfaraPa4ectfghnea)knveelcacwlgonieersefabei,enomsorotae1byufamlr9styasoes9uictfgirb6ecghvcnhe;dnaeoaandaalnvloilgsatnpwessengalixehagt(syroiHnsPniwtagfoeahatlniolnhnacltaifodegnesfl, 0 60 120 180 240 300 301+ Bout length (s) may be an artifact of this sampling. Additionally, the frequent 0 occurrence of play face may be an involuntary artifact of the 0 60 120 180 240 300 301+ sender’s enjoyment of the bout rather than a message for Bout length (s) 31 players(Spijkermanetal.,1996). Figure2Averagenumberofobservedplaysignalsperbout Onlythreeplaysignalswereobservedinthezero-audience length membercategory: (1)crouch-and-stare,(2)dangle-and-stare, Figure 2 Average number of observed playsignals per bout length 31a nd(3)roll-onto-back-and-stare.Thismayindicatetheneedto usecomplexbodyandfacialsignalstoattractplayerstobegin 0.2 a bout, rather than a face-only signal. For example, juvenile 0.18 chimpanzees often use attention-getting gestures when a play 0.16 partnerdoesnotseethesignalsender’splayface(Tomaselloet nals per 60 s00..1124 aothlf.e,tw1so9e8no9dr)e.mrI’snormteheisssigwsnaaagyl,esa.mpOalauyyrbrseeigsnnuealtclsethssasuatpripnyovtrootlvtaehmsispthslipefyeccoourmlraebtiiionnnfao,triocanes play sig00.0.81 jduivveenrsileebTeibheatvainoramlarceapqeurteosireustoedinmdiuclatitpelethpelairywsiilglinnganlsesfrsomtotphleaiyr Rate of 0.06 with conspecifics. However, a larger sample of zero-audience memberplayboutsisneededforfurtherinvestigation. 0.04 Additionally, across all audience member categories, the 0.02 leg-peek signal (see Table 3) was never observed. Yanagi 0 & Berman (2014b) found that in rhesus macaques, the 0 60 120 180 240 300 301+ Bout length (s) leg-peek was associated with play initiated by the receiver and may have served as a play invitation. However, this Figure3Rateofobservedplaysignalsper60sacrossbout signal was not present in Tibetan macaque play, indicating Figure 3 Rate of observed play signals per 60 s across bout length length that leg-peek may be a species-specific signal. Similarly, two candidateplaysignalswereobservedintheTibetanmacaque repertoire that were not observed in rhesus macaques: play We found that more play bouts began with play behaviors threat (Supplementary Figure S2) and slap-and-play face thanplaysignals(notsupportingprediction2). Therefore,play 280 www.zoores.ac.cn 32 (Supplementary Figure S3). Adult Tibetan macaques use and in various contexts, as they were observed before, during, threat behavior as part of an aggressive interaction, and andwhenwithdrawingfromplay(Pellisetal.,2011). it is characterized by an open mouth gesture and another Finally, we examined the length of play bouts in relation to body movement, such as a ground slap, raised eyebrow, or thenumberofplaysignalsused(prediction4;Figure2;Figure lunge(Bermanetal.,2004). Therefore,theplaythreatsignal, 3). Spearman’s rank correlation showed a strong statistically observedduringplaybouts, maypotentiallybeanincomplete significant correlation, thus supporting this prediction and adult threat behavior, with the absence of the open mouth indicating the likely importance of play signals in sustaining facial expression. Play threat may enable juvenile macaques aplaybout. Previousliteraturehasshownthatplayboutstend to practice components of adult behavior within the context tobelongerwhenplayersusetheplayface(Waller&Dunbar, of play (Martin & Caro, 1985). However, further research is 2005), playful facial expressions punctuate bouts rather than needed to confirm that the candidate play signals observed initiatethem(Palagietal.,2007),thirdpartiesjointheplaybout in Tibetan macaque juveniles do not occur under any other more often when the play face signal is used, and long play contexts. Although the characterization of Tibetan macaque boutsaretypicallymoreintenseandcontainmoreplayfaces, play signaling differs from that of rhesus macaques, there especially when ambiguous behaviors are involved (such as does appear to be some cross-species similarity in both wrestle and gnaw) (Spijkerman et al., 1996). This increased play behavior and signaling repertoires. The similarities and signalingmaybecriticalformaintainingplayfulinteraction, as differences in their play style may reflect social organization additional evidence indicates that short play bouts are often (Palagi,2006);however,morecomparativeresearchisneeded. marked by a misinterpretation of signals, thus halting play. Relative to audience members, we observed the OurstudydemonstratedthatcandidateplaysignalsinTibetan combination of slap (a play behavior) and play face (a play macaquesmaybeusedinsimilarways, withthelongestplay signal) more than other play signals, such as gamboling, boutscontainingthemostplaysignals. However,furtherstudy hide-and-peek, and look-back (Figure 1). It is possible that isneededtoexaminethecausalrelationshipbetweenplaybout the playful intention of a slap behavior needs to be clarified lengthandthefrequencyanddistributionofplaysignals. asplaybecauseofitsassociationwithaggression(Burghardt, 1999;Pellis&Pellis,1996). Inthisway,theslapbehaviorand CONCLUSIONS play face signal would contradict each other and, therefore, mightincreasethelikelihoodofthereceiverunderstandingthe Juvenile Tibetan macaques maintained playful interactions situationasnon-threatening. using multiple candidate play signals, which combined body Although we expected that more play signals would be andfacialgestures,tobeginplay,encouragethecontinuationof observed in polyadic play bouts (at least one actor and two play,andendplay.Acrossallaudiencemembernumbers,play audience members) (Spijkerman et al., 1996), we did not face was the most frequently observed play signal, possibly find support for this prediction. Our results indicated that two indicating its salient and general nature. Tibetan macaque audiencemembers(threeplayersintotal)includedthelargest juveniles utilized play behaviors and signals similar to rhesus numberofplaysignalsemittedforthisboutcomposition. The macaques;however,thedifferencesobservedintheirplaystyle playsignalsobservedinplaygroupslargerthanthreemembers andsignalvarietymaybeinfluencedbythedominancestyleof werelessfrequent,andthenumberofplaysignalsusedwithin Tibetanmacaques. larger group compositions declined as more members were added. Thisresultimpliesapossiblethresholdforthesaliency COMPETINGINTERESTS of play signals past a certain audience member number. It is possible that play bouts with more players increase the Theauthorsdeclarethattheyhavenocompetinginterests. complexityofthebout,makingthemhardertomanage(Bekoff, 1972). Furthermore, the play signal message may lose AUTHORS’CONTRIBUTIONS salienceasthecomplexityandsizeoftheplaygroupincrease beyond three players. This breakdown of communication K.R.W., J.A.M., and L.K.S. designed the study. L.K.S. (P.I.) and R.W.S. may be more readily observed in despotic macaques due to (co-P.I.)acquiredfundingthatpartiallysupportedthisresearch(NSF-OISE their strict social organization and therefore more risky and 1065589). K.R.W.andR.S.Wcollecteddata, andJ.H.L.superviseddata uncertainsocialinteractions. Itisalsoreasonabletoconclude collectioninthefield.K.R.W.andJ.A.F.performedstatisticalanalysis,under that play bouts with fewer members reinforce affiliative bonds the supervision of J.A.M, L.K.S., and L.S. K.R.W. and J.A.M. wrote the betweengroupmembers,whereaslargerplaygroupsmaybe manuscriptwiththeotherauthors’input. Allauthorsreadandapprovedthe used to assess individuals’ physical strengths (Pellis & Pellis, finalmanuscript. 1996).Therefore,signalingmaybecrucialinasmallplaygroup toreaffirmthesocialcontext: affiliationratherthanaggression. ACKNOWLEDGEMENTS Ourresultsalsoshowedthatplaysignalswerelesslikelytobe emittedwhennewaudiencememberswereaddedtoaboutor We thank the Huangshan Garden District Bureau staff for permission to whenaplayerleftabout. Thisindicatesthatwhenplaysignals conductresearchatthefieldsite,andXiWangandtheChenfamilyfortheir areusedbyjuvenileTibetanmacaques,theyareusedgenerally operationalsupportandencouragementinthefield. ZoologicalResearch39(4):272–283,2018 281

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.