Primates

, Volume 50, Issue 4, pp 333–341 | Cite as

Pirouettes: the rotational play of wild chimpanzees

Original Article

Abstract

A pirouette is a locomotor-rotational movement in which a young chimpanzee spins around in a (mostly) quadruped posture while advancing forward in a straight line. We addressed whether this behavior evolved as a practice of general athletic ability or of sexual selection. The former hypothesis would predict no sex differences in skill or the developmental process, while the latter would predict the opposite. Chimpanzees most likely master the pirouette around the time of weaning. We found no conspicuous sex differences in the developmental process or the number of rotations per bout of pirouettes, so the pirouette’s main function may be to facilitate general athletic ability. Infants pirouetted regardless of the context of rest or travel, whereas juveniles and adolescents pirouetted primarily during travel. This is consistent with the survival strategy hypothesis, because juveniles and adolescents would be expected to display pirouettes to many watchers if this practice were sexually selected. However, the fact that males tend to pirouette faster than females and to pirouette even during adolescence suggests that sexual selection has some influence in shaping the evolution of the practice. Despite this, no conspicuous tendency was found for juveniles or adolescent chimpanzees to display pirouettes to opposite-sex individuals. More data on adolescent individuals are needed to definitively determine the role of sex differences in pirouetting behavior.

Keywords

Chimpanzee Play Pirouette Locomotor-rotational play Mahale 

Introduction

This paper forms a link in a chain of systematic studies of chimpanzee play that began in 1999 (Nishida and Wallauer 2003; Matsusaka 2004; Nishida 2004; Matsusaka et al. 2006). Despite almost half a century of field research on chimpanzees, few studies have examined play behavior in general and locomotor-rotational play in particular. In Animal Play Behavior, Fagen (1981) wrote, “no monograph on chimpanzee play now exists, but a full-length study of chimpanzee play would be an immensely valuable document” (p. 112). This statement holds true a quarter of a century later. Our research ultimately addresses the question of the origins and evolution of human play, which constitutes one of the most important characteristics of humans. Huisinga (1950) suggested that play, ranging from combat, games, gambling, sports, and war to law, poetry, philosophy, art, science, and politics, appears to be the foundation of human culture.

Humans perform various kinds of rotational gymnastics, such as somersaults, spinning, swinging, and synchronized swimming. They even enjoy tool-assisted rotational entertainment, such as ice skating and roller coasters. In Man, Play and Games, Caillois (1961) included “ilinx,” or rotational play, as one of the four basic elements of human play.

Rotation is apparently also important for arm-swinging apes living in trees, as one might imagine from seeing the acrobatics that go on in the “gibbon houses” of many zoos. Tumbling and hanging upside down are important locomotor-rotational movements of gibbons (Ibscher 1967, cited in Fagen 1981). Gibbons also dangle to incite a chase (Aldis 1975, p. 14). But why do humans retain interest in such acrobatics when they live primarily on the ground? Because erect bipedalism is a habitual human locomotion pattern, modern humans’ enthusiasm for rotational play poses an interesting question with regard to human evolution.

Chimpanzee locomotor-rotational play

Chimpanzees engage in various kinds of locomotor-rotational play (see Burghardt 2005, p. 84 for this term). Rotational play above the ground includes one-arm hanging, jumping, swinging, and forward and upward swinging. Chimpanzees swing from a vertical vine by kicking the stem of a tree trunk with one or both feet.

Rotational play on the ground includes forward somersaults, backward somersaults, sideways somersaults, turning around, and pirouetting. One or two chimpanzees may turn around a tree or another individual or may turn freely, without any visible axis (Goodall 1968; Hayaki 1985; Nishida 2004). This can be done in a quadruped or a bipedal manner.

A typical pirouette is executed by spinning around with one or both hands on the ground while advancing forward in a straight line (Fig. 1). The first and only unambiguous description of a pirouette in wild chimpanzees was given by Goodall (1968), who reported that two young chimpanzees in Gombe performed a pirouette. Aldis (1975, p. 59) showed a photograph of a captive chimpanzee pirouetting. Studying young bonobos, Pika et al. (2005) described “ice skating” as a gesture of “making a pirouette with hands on the ground or in the air” (p. 49), which solicits play. Perhaps this is different from the chimpanzee pirouette, which contains the element of advancing forward in a straight line and does not function as play solicitation. Palagi and Paoli’s (2007) definition of pirouetting in adult bonobos, “one or more animals together turn, somersault, or roll over either on the ground or on vertical supports” (p. 220), is ambiguous, and it is not clear whether their understanding of pirouetting includes the same pattern described by Goodall above.
Fig. 1

An 8-year-old male (XM) pirouettes

Rotational locomotion is likely useful in food acquisition, avoidance of predators, and competition among conspecifics. The aim of this study is to collect basic information on rotational play patterns and to examine the evolutionary pressure (i.e., survival strategy or sexual selection) primarily responsible for the development of locomotor-rotational play in chimpanzees.

Hypotheses

With regard to the function of rotational play that involves performing a pirouette, two hypotheses are possible based on the practice theory of play (Groos 1898). The first is the survival strategy hypothesis, which is consistent with the “motor training” hypothesis (Byers 1998) or the “training for the unexpected” hypothesis (Spinka et al. 2001). The pirouette and other patterns of locomotor-rotational play function to facilitate quick and versatile movement during feeding and during escape from predators. Aldis (1975, p. 52) considered the function of rotation in vestibular play to be the encouragement of graduated risk-taking and the acquisition of skills for emergencies. This would lead to both immediate and later benefits. Spinka et al. (2001) suggested that, “a major ancestral function of play is to rehearse behavioral sequences in which animals lose full control over their locomotion, position, or sensory/spatial input and need to regain these faculties quickly” (p. 143). The second is the sexual selection hypothesis: The pirouette may be effective during conflict between conspecifics. The ability to make smooth rotational movements enables a fighter to assume a more intimidating threat posture or to more quickly strike a devastating blow against a rival during conflict. Sex differences in the frequency and type of social play and/or play partner preferences have been reported in some primates (e.g., Aldis 1975; Hayaki 1985; Fry 2005; Maestripieri and Ross 2004; Paukner and Suomi 2008). However, Fagen (1993) reported that little is known about possible sex differences in nonsocial or adult play in any nonhuman primate.

Here, we limit our discussion to the pirouette, one type of rotational play done on the ground. This is because the pirouette is the most remarkable locomotor-rotational play in wild chimpanzees, it can be clearly differentiated from other behavioral patterns and thus readily analyzed in video recordings, and unlike many other rotational acrobatics, such as turnarounds or somersaults, this pattern is not mingled with social play nor is it a communicatory gesture, although it can occasionally occur immediately after social play.

Predictions for pirouette development

If the survival strategy is more important in shaping the evolution of the pirouette, this play pattern, being vital to survival, should be established before weaning, when an individual becomes independent from its mother. Thus, the frequency of pirouetting should increase from birth to 4–5 years of age (weaning period) and then decrease (inverse U curve). The speed of the pirouette (number of rotations per unit time) should also increase from birth and reach a plateau by the beginning of the juvenile years. Because the pirouette would no longer be expected after the weaning age, an adult or adolescent individual should seldom engage in this type of play. No sex difference in growth pattern would be expected, because food acquisition and escape from predators are equally important tasks for males and females.

If sexual selection is more important, this play pattern should continue developing even during adolescence, and sex differences in frequency and speed would be expected: Males would not only engage in pirouette play more frequently and for longer periods up to adolescence, but would have faster rotational speeds than females.

Moreover, males would be expected to display or show off the pirouette to other chimpanzees, in particular, to nonmother females. If the pirouette is a practice for a survival strategy, we would expect neither showing-off elements nor pirouetting directed to individuals of any particular age/sex class, even if the chimpanzees seem to display their pirouettes.

Chimpanzees pirouette in both travel and rest. If sexual selection is important, older individuals would be expected to perform pirouettes more often during rest rather than during travel because more watchers are available during rest. If this is a survival strategy, younger individuals would be expected to pirouette more often during rest, because of lower risk of predation during rest.

Methods

Subjects

Subjects were habituated chimpanzees of the M group, living in the Mahale Mountains National Park, Tanzania (Nishida 1990). The M group consists of 50–60 chimpanzees observed during the 7-year study period from 1999 through 2005. Fieldwork was conducted from August to October 1999, from September to October 2000, from September to October 2001, from September to November 2002, from August to September 2003, from August to September 2004, and from September to October 2005. T.N. selected immature (<12 years) chimpanzees, nine males and ten females, as the targets of observation. In addition, ANC Production provided video clips of four immature males (one bout each for LT, IW, CD, and OR). In total, the numbers of individuals recorded pirouetting were 11 males and 11 females.

T.N. followed a target for as long as possible, and if he missed a target, another chimpanzee was selected. T.N. always carried a video recorder and continued videotaping the behavior of a target unless the target was sleeping or eating high above the ground. Thus, T.N. recorded all behavior patterns including locomotor-rotational play. The recording time ranged from 70 to 90 h each year, amounting to 569 h in total. A.I. checked all of the descriptive data drawn by T.N. from the video images and analyzed the data. When our counts differed, we both checked the video clips again until we agreed (Table 1).
Table 1

Targets, observation minutes, and characteristics of pirouette movements

Age in months

Sex

Name

Obs. mins

No. of episodes

No. of bouts

No. spinning

Total bout length (s)

Median no. spinning/s

Max no. spinning/s

Min.

Max.

Median

17

IM

492

1

1

 

1.00

1.00

2

0.50

0.50

17

ZH

383

1

1

 

1.00

1.00

1

1.00

1.00

21

IH

818

1

1

 

0.50

0.50

1

0.25

0.25

21

TD

749

2

4

0.50

1.00

0.50

5

0.50

0.50

22

AQ

356

1

1

 

0.50

0.50

1

0.50

0.50

24

CE

1038

1

2

0.50

1.00

0.75

2

0.50

0.50

Subtotal

  

(3836)

(7)

(10)

      

27

IM

512

6

13

0.40

1.00

0.70

14

0.50

1.00

27

ME

985

1

1

 

0.70

0.70

1

0.70

0.70

27

OS

608

5

5

1.00

2.00

2.00

11

0.68

1.50

32

TD

1276

3

6

0.50

1.00

0.50

9

0.50

0.50

34

AQ

243

3

15

1.00

3.50

1.00

35

0.55

1.00

34

XP

704

1

1

 

0.70

0.70

1

0.70

0.70

35

CE

430

5

6

0.50

2.00

1.00

11

0.50

1.00

35

EM

201

3

6

1.00

4.50

3.75

26

0.81

1.00

Subtotal

  

(4959)

(27)

(53)

      

39

IM

441

1

2

1.00

2.00

1.50

5

0.67

1.00

39

OS

413

2

3

1.00

1.00

1.00

3

0.75

1.00

41

ZH

383

1

2

1.00

1.00

1.00

4

0.67

0.67

42

PF

455

4

4

1.00

2.50

1.25

9

0.63

0.83

45

XP

592

1

1

 

1.50

1.50

2

0.75

0.75

46

TD

na

2

2

1.50

2.00

1.75

3

1.25

1.50

47

AQ

590

7

9

1.00

4.00

2.00

21

1.00

1.00

48

CE

1146

5

5[3]

2.00

3.50

2.00

14

1.00

1.00

Subtotal

  

(4020)

(23)

(28)[26]

      

50

OS

820

2

8

1.00

5.00

2.44

23

0.85

1.50

51

MC

1722

8

11

0.50

6.00

2.50

35

0.84

1.00

52

IM

513

9

17[14]

0.50

9.00

2.50

62

0.90

1.00

52

PF

725

1

2

2.00

5.00

3.50

9

0.78

0.83

53

AC

621

3

3[2]

1.00

1.00

1.00

6

0.50

0.75

53

XM

979

8

8

1.50

10.00

4.50

45

0.83

1.50

58

AQ

573

1

1

 

5.00

5.00

7

0.71

0.71

58

CE

955

10

19

1.00

12.00

3.00

85

0.82

1.00

59

EM

na

2

8

0.75

5.00

3.00

28

0.72

1.00

Subtotal

  

(6908)

(44)[42]

(77)[73]

      

62

OS

857

8

11(10)

0.50

7.50

1.50

28

0.67

1.13

62

CR

289

1

1

 

1.00

1.00

1

1.00

1.00

62

FV

514

1

3

0.50

3.00

1.50

7

0.75

0.75

62

MC

672

2

2

3.00

4.00

3.50

9

0.78

1.00

63

IM

804

2

3

0.50

4.50

2.00

7

1.00

1.13

64

AC

593

1

1

 

4.50

4.50

7

0.64

0.64

66

PF

na

1

1

 

1.00

1.00

1

1.00

1.00

70

AQ

516

5

5

1.00

9.50

3.00

27

0.60

0.79

Subtotal

  

(4245)

(21)

(27)[26]

      

73

OS

836

2

2

0.50

1.00

0.75

3

0.50

0.50

76

IV

558

4

13

0.50

3.00

1.25

25

0.84

1.00

77

XM

548

1

1

 

2.00

2.00

2

1.00

1.00

85

LT

na

1

1

 

2.00

2.00

2

1.00

1.00

86

MC

994

3

4

2.00

7.00

5.50

18

1.00

1.67

86

OS

na

2

2

 

6.00

6.00

7

0.86

0.86

88

IV

358

1

1

 

2.00

2.00

2

1.00

1.00

89

AC

na

1

1

 

1.00

1.00

2

0.50

0.50

90

XM

866

2

3

1.00

14.50

4.00

28

0.78

1.25

92

CD

567

2

2

 

3.00

3.00

4

0.75

0.75

Subtotal

  

(4727)

(19)

(30)

      

99

IV

672

1

4

2.00

6.00

4.00

17

0.92

1.00

99

OR

983

2

2

0.50

2.00

1.25

4

0.63

0.67

100

XM

776

1

1

      

105

CD

468

4

4

3.00

5.50

5.25

15

1.38

1.83

117

CD

417

2

2

1.50

2.00

1.75

5

0.70

0.75

125

IW

na

1

1

 

6.00

6.00

6

1.00

1.00

130

CD

755

2

2

1.50

4.00

2.75

7

0.78

0.80

136

OR

670

1

2

1.00

2.50

1.75

5

0.70

0.83

249

TZ

686

1

1

 

1.00

1.00

2

0.50

0.50

Subtotal

  

(5247)

(15)

(19)

      

Figures in Gothic type indicate the maximum number of spinnings or spinning rate achieved during the career of each individual who was observed in the major part of juvenility

Definition of episode, bout, bout length, and number of spins

One bout of a pirouette began when a chimpanzee stopped walking or resting and assumed the particular posture for spinning. A bout continued until the spinning ended and the chimpanzee sat or fell down. If the animal got up and performed another pirouette movement after a delay of 5 s, it was counted as a new bout. An episode consisted of one or several bouts that occurred within a 5-min period. The number of spins was counted with a one-quarter spin as the unit, determined by change in the direction of the ventral side from the original position. Time was measured to the nearest 0.5 s. Data were obtained by watching video images repeatedly. To investigate sex differences, we compared the maximum number of spins per bout and the maximum spin velocity (number of spins/s) for the chimpanzees that were sampled during the third year of life or later (ten males, ten females).

Context of pirouetting

The context of travel was divided into five categories. “Long rest” means mid-day leisure time, which sometimes spans 2 hours or more at a stretch and is at least more than 10 min. “Brief rest during procession” means a resting time of less than 10 min. “Travel” refers to a long period of travel not accompanied by feeding, but by brief (<20 s) social or solitary play and slowly moving travel during which chimpanzees were foraging as they moved along a path. “Onset of travel” means the time immediately after chimpanzees start to travel.

Individuals who could watch a performance

Chimpanzees pirouette during travel and rest. During travel, only individuals walking behind a player can observe a performance well. We operationally defined the watcher of a pirouette as the individual who immediately followed the performer during travel or the individual nearest to the performer during rest. We also noted social stimuli that apparently incited a pirouette, such as the arrival or departure of an individual.

Laterality of spinning

The laterality of spinning cannot be predicted on the basis of either hypothesis, because both hypotheses predict the same trend. We assumed that, in both survival and fighting, a lack of laterality would be more advantageous because an enemy, a rival, or food would not be limited to either a chimpanzee’s right or left side (Marchant and McGrew 2007).

To investigate the validity of this hypothesis, we counted the number of clockwise and counterclockwise rotations. We defined our laterality index (LI) as (A − B)/(A + B), where A is the number of clockwise rotations and B is the number of counterclockwise rotations. If LI = 0, this indicates that the individual is ambilateral in the rotation of its pirouette.

Statistics

Age and sex differences in the frequency of episodes, maximum number of spins, and velocity of spinning were examined by the Mann–Whitney U test and Fisher’s exact probability test. Laterality of rotation and context of pirouettes were examined by a binomial test. All tests were two-tailed.

Results

Pirouettes

A total of 152 episodes (240 bouts) were observed during the study period and in an additional four episodes (four bouts) of video footage provided by ANC production (Table 1). Eight bouts lacked essential elements, such as number of rotations, because of delayed video shooting.

The context during which pirouetting occurred was investigated. The majority of pirouettes were observed when the chimpanzees were traveling in procession (Table 2). Since chimpanzees spend more time during rest than during travel (Nishida 1989; Wrangham 1977), they are expected to perform pirouettes during rest much more often than during travel. Therefore, p = 1/2 is a conservative figure to calculate the possibility of doing pirouettes during rest. However, even using this figure, we can obtain a significant result showing that chimps pirouette during travel. The role of context differed among chimpanzees by age. Among the three age classes of infancy (1–2, 2–3, and 3–4 years of age), the number of individuals who pirouetted more during travel than during rest and those who showed the opposite tendency were not different. In contrast, among the four juvenile or older (≥49 months old) age classes, most individuals pirouetted more often during travel than during rest, and the differences were statistically significant for two of the four age classes (Table 3). Chimpanzees in the four juvenile or older age classes pirouetted significantly more often during travel than did those in the three infant age classes (p = 0.03, Fisher’s exact probability test).
Table 2

Contexts of pirouette

 

No. of episodes

%

No. of bouts

%

Long rest

31

20.5

50

20.5

Brief rest during procession

7

4.6

8

3.3

Onset of travel

45

29.8

59

24.2

Travel

65

43.0

123

50.4

Travel while feed

3

2.0

4

1.6

Total

151

99.9

244

100.0

Table 3

Differences in context by age class

Age class

No. of individuals

Travel ≥ resta

Rest > travelb

Binomial test

12–24

6

4

2

n.s.

25–36

8

4

4

n.s.

37–48

8

6

2

n.s.

49–60

9

9

0

p = 0.004

61–72

8

7

1

n.s.

73–92

6

6

0

p = 0.032

93≤

5

5

0

n.s.

aNo. of individuals who pirouetted more often (or as often) during travel than (as) during rest

bNo. of individuals who pirouetted more often during rest than during travel

Developmental trends

The small sample size permits us to characterize only general trends in the development of the pirouette. Thus, we compared median number of episodes, median value of maximum number of spins, and median value of maximum velocity (number of spins/s) among age classes (Table 1, Fig. 2).
Fig. 2

Trends in the development of pirouettes in terms of median frequency of episodes, median maximum number of spins, and median maximum pirouette velocity

Frequency of pirouette episodes

The youngest infants (<17 months) were never seen to perform pirouettes. The median number of episodes increased from the second year of life to the third year and then decreased (Fig. 2). The number of pirouette episodes increased between 1–2 and 2–4 years of age (Mann–Whitney U test, n1 = 6, n2 = 15, U = 21, p < 0.05). Older infants (2–5 years old) pirouetted more often than did juveniles (5–9 years old; Mann–Whitney U test, n1 = 23, n2 = 18, U = 140, z = 1.76, p < 0.039). Late adolescent and adult males and females were not seen to pirouette, except on one occasion.

Maximum number of rotations

Until the beginning of the third year of life, chimpanzees did not spin more than one cycle; after that, they gradually spun more and more, and between the fifth and sixth years of life, they usually attained a maximum number of spins (Table 1). This corresponded to the end of weaning or to the beginning of the juvenile period. The median values of maximum numbers of spins per bout increased from the second year of life through the fourth to fifth year (Fig. 2). Four-year-old infants or weanlings showed greater maximum spins per bout than did 3-year-olds (Mann–Whitney, U = 7, n1 = 9, n2 = 8, p < 0.05), and likewise, maximum spins in 3-year-olds exceeded those in yearlings (U = 4, n1 = 8, n2 = 6, p < 0.05). The maximum number of spins did not differ significantly between 4-year-olds and older age classes. However, the highest number of spins recorded was for a 7-year-old male (XM), who once pirouetted 14.5 times consecutively.

Rotational velocity

Until the third year of life, chimpanzees usually spun at a velocity of 0.5 cycles/s (Table 1). Then, this velocity doubled (however, a statistical difference was only found between 3- and 1-year-olds; Mann–Whitney U = 6, n1 = 6, n2 = 8, p < 0.05), and the 1 cycle/s velocity continued throughout the remaining years, as if it were the “basic” pirouette speed (Fig. 2). However, maximum spin velocity was achieved by the eighth and ninth years in juvenile males (Table 1). This degree of skill disappeared when the youngsters entered adolescence, except in a 9-year-old adolescent male (CD) who could pirouette very quickly without falling down.

Sex differences

We examined whether sex differences existed in pirouette behavior. Because there were insufficient data for statistical tests, we can only show general trends.

The three youngest males who pirouetted were 21, 21, and 24 months old. In contrast, the three youngest females who pirouetted were 17, 17, and 22 months old. These young males and females spun at 0.5 or 1 cycle/s, respectively. This suggests that females mature more quickly and are able to pirouette at an earlier age than males.

Adults never pirouetted except on one occasion, when TZ, a 20-year-old female, spun once with her son, TD, on her back during her travel. This could hardly be considered play. The scarcity of data on adolescent individuals makes it impossible to show sex differences in this period.

When the comparison was limited to the 20 individuals who were observed during the third year of life or later, males tended to show a greater maximum number of spins per bout than females, but this was not statistically significant (Mann–Whitney, n1 = 10, n2 = 10, U = 31, ns; Table 1).

Males achieved a higher maximum velocity (spin rate) during their juvenile years than did their female counterparts. This was statistically significant (Mann–Whitney, n1 = 10, n2 = 10, U = 17, p < 0.05; Table 1).

Laterality of rotation

Only nine individuals pirouetted more than ten times (Table 4), and six of the nine individuals approached LI = 0 as they matured.
Table 4

Laterality of rotation

Individual

No. of bouts

Clockwise

Counterclockwise

Index

AQ (♀)

31

22

9

0.42*

CE (♂)

32

18

14

0.13

EM (♂)

14

10

4

0.43

IM (♀)

35

20

15

0.14

IV (♀)

18

4

14

−0.56*

MC (♂)

17

15

2

0.76*

OS (♂)

31

25

6

0.61*

XM (♂)

13

7

6

0.08

TD (♂)

13

3

10

−0.53

p < 0.05

Of the nine individuals who pirouetted ten or more times, two males (MC and OS) and one female (AQ) pirouetted significantly more often in a clockwise direction, whereas one female (IV) did so significantly more often in a counterclockwise direction. The other five individuals were ambilateral in their spinning.

Individuals nearest to performers

In 156 pirouette episodes, we could determine the presence or absence of individuals who were in a position to watch the pirouette performance (Table 5). In 33 cases (21.2%), performers were not followed by other chimpanzees. In 58 (37.7%) episodes, only the mother of the performer was able to watch her offspring’s performance. Additionally, no tendency for performers to display the pirouette to a particular age/sex class was observed.
Table 5

Individuals following immediately after performer of pirouette

Sex

Age class

Mother/foster mother

Same-sex sibling

Opposite-sex sibling

None

Same-sex play partner

Opposite-sex play partner

Mature/adolescent male

Mature/adolescent female

Total

Male

1–5

(%)

24

(39.3)

0

(0)

3

(4.9)

16

(26.2)

7

(11.5)

3

(4.9)

3

(4.9)

5

(8.2)

61

(99.9)

5≤

(%)

4

(11.4)

0

(0)

1

(2.9)

12

(34.3)

3

(8.6)

6

(17.1)

5

(14.3)

4

(11.4)

35

(100)

Female

1–5

(%)

25

(59.5)

1

(2.4)

1

(2.4)

2

(4.8)

6

(14.3)

1

(2.4)

4

(9.5)

2

(4.8)

42

(100.1)

5≤

5

(27.8)

1

(5.6)

0

(0)

3

(16.7)

1

(5.6)

2

(11.1)

1

(5.6)

5

(27.8)

18

(100.2)

Total

(%)

58

(37.7)

2

(1.3)

5

(3.2)

33

(21.2)

17

(10.9)

12

(7.7)

13

(8.3)

16

(18.3)

156

(100)

Chimpanzees often performed pirouettes in playful situations: 37 episodes occurred after social play, such as wrestling, and 20 episodes after solitary play, such as somersaults. Additionally, the arrival or departure of a particular individual (mother, sister, playmate, adult male, etc.) seemed to elicit pirouettes in 14 and 2 episodes, respectively. Moreover, chimpanzees performed pirouettes immediately after they displayed against T.N. (slapped the ground, charged, threw a stone) in five episodes, after they touched T.N.’s or a colleague’s trousers gently in two episodes, and after T.N. recovered to an appropriate distance (5 m) from a targeted chimpanzee who had moved away in seven episodes. No clear episode occurred in which a pirouette functioned as the sole visual gesture to solicit social play.

Discussion

We hypothesized that the frequency, rotational velocity, and number of spins in chimpanzee pirouettes would increase as the individual matured. Also, if sexual selection were a dominant factor in pirouetting, males’ performance would be better than females’ and would peak in adolescence.

The median frequency of pirouette episodes increased from early infancy, reached a plateau during the weaning period, and virtually disappeared after late adolescence. This inverse U curve tends to support the survival strategy hypothesis rather than the sexual selection hypothesis.

If sexual selection were a major factor, juveniles and adolescents would be expected to pirouette more often during rest than during travel, when the number of watchers is limited to one or two individuals. Actually, they pirouetted significantly more often during travel than at rest. Thus, the context of pirouetting is consistent with the survival strategy hypothesis.

Our results indicate that sex differences were negligible in general developmental trends. Therefore, the survival strategy hypothesis, or that of selection for general athletic ability, is supported. However, male supremacy in velocity and number of spins indicates that sexual selection plays a part, although the numbers of adolescent males and females studied were too small to draw any definite conclusions. Despite this, pirouettes were not obviously direct social displays to the opposite sex or to same-sex competitors. The fact that, in almost 60% of pirouette episodes, only the mother or no apparent watcher was present suggests that the performance is basically a self-centered practice.

When T.N. observed chimpanzee youngsters in the wild, he often felt that they pirouetted to show him their favorite moves. In other cases, youngsters seemed to perform for other chimpanzees, such as their mothers, adult females or playmates, because they were gazing at these individuals while they were pirouetting. Thus, chimpanzees seem to be interested in attracting attention from others. Consider the origins of human spectator games such as the Olympics. No one has ever asked why humans are so attracted to performances by other persons. Human performers likely show off for the crowd, and here we hit upon the common roots in locomotor-rotational play patterns between humans and chimpanzees. Perhaps humans continue to have a propensity toward locomotor-rotational play, even in the world of two-dimensional movement, because the development of the cerebellum was useful in their new savanna environment, and males with excellent acrobatic skills attracted females and vice versa.

Finally, in this study, only nine individuals pirouetted more than ten times (Table 4), and six of the nine individuals approached LI = 0 as they matured. Although this result is ambiguous, it is possible that pirouettes lose laterality as chimpanzees mature.

Notes

Acknowledgments

We thank the Commission for Science and Technology, Tanzania Wildlife Research Institute, and Tanzania National Parks for permission to work at the Mahale Mountains National Park; and the Mahale Mountains Wildlife Research Centre and Mahale Mountains National Park for logistical support. We thank Miho Nakamura for providing video footage (four video clips in 1991, 1992, and 1999). This research was funded by a MEXT grant-in-aid for basic scientific research A (#12375003, #16255007, and #19255008 to T.N.) and the L.S.B. Leakey Fund. We are indebted to Bill McGrew, Simone Pika, and two anonymous reviewers for valuable comments.

References

  1. Aldis O (1975) Play fighting. Academic Press, New YorkGoogle Scholar
  2. Burghardt GM (2005) The genesis of animal play. MIT Press, CambridgeGoogle Scholar
  3. Byers JA (1998) Biological effects of locomtor play: getting into shape, or something more specific? In: Bekoff M, Byers JA (eds) Animal play. Cambridge University Press, Cambridge, pp 205–220Google Scholar
  4. Caillois R (1961) Man, play and games. Free Press, New YorkGoogle Scholar
  5. Fagen R (1981) Animal play behavior. Oxford University Press, OxfordGoogle Scholar
  6. Fagen R (1993) Primate juveniles and primate play. In: Pereira ME, Fairbanks LA (eds) Juvenile primates. Oxford University Press, New York, pp 182–196Google Scholar
  7. Fry DP (2005) Rough-and-tumble social play in humans. In: Pellegrini A, Smith PK (eds) The nature of play. Guilford Press, New York, pp 54–85Google Scholar
  8. Goodall J (1968) Behavior of free-living chimpanzees of the Gombe Stream area. Anim Behav Monogr 1:163–311Google Scholar
  9. Groos K (1898) The play of animals. D. Appleton and Company, New YorkGoogle Scholar
  10. Hayaki H (1985) Social play of juvenile and adolescent chimpanzees in the Mahale Mountains National Park, Tanzania. Primates 26:343–360CrossRefGoogle Scholar
  11. Huisinga J (1950) Homo ludens. Beacon, BostonGoogle Scholar
  12. Ibscher L (1967) Geburt und fruhe Entwicklung zweier Gibbons (Hylobates lar L.). Folia Primatol 5:43–69PubMedCrossRefGoogle Scholar
  13. Maestripieri D, Ross SR (2004) Sex differences in play among western lowland gorilla (Gorilla gorilla gorilla) infants: implications for adult behavior and social structure. Am J Phys Anthropol 123:52–61PubMedCrossRefGoogle Scholar
  14. Marchant LF, McGrew WC (2007) Ant fishing by wild chimpanzees is not lateralized. Primates 48:22–26PubMedCrossRefGoogle Scholar
  15. Matsusaka T (2004) When does play panting occur during social play in wild chimpanzees? Primates 45:221–230PubMedCrossRefGoogle Scholar
  16. Matsusaka T, Nishie H, Shimada M, Kutsukake N, Zamma K, Nakamura M, Nishida T (2006) Tool use for drinking water by immature chimpanzees of Mahale: prevalence of an unessential behavior. Primates 47:113–122PubMedCrossRefGoogle Scholar
  17. Nishida T (1989) Social structure and dynamics of the chimpanzee: a review. In: Seth PK, Seth S (eds) Perspectives in primate biology. Today & Tomorrow’s Printers and Publishers, New Delhi, pp 157–172Google Scholar
  18. Nishida T (1990) The chimpanzees of the Mahale Mountains. University of Tokyo Press, TokyoGoogle Scholar
  19. Nishida T (2004) Lack of “group play” in wild chimpanzees. Pan Afr News 11:2–3Google Scholar
  20. Nishida T, Wallauer W (2003) Leaf pile pulling: an unusual play pattern in wild chimpanzees. Am J Primatol 60:167–173PubMedCrossRefGoogle Scholar
  21. Palagi E, Paoli T (2007) Play in adult bonobos (Pan paniscus): modality and potential meaning. Am J Phys Anthropol 134:219–225PubMedCrossRefGoogle Scholar
  22. Paukner A, Suomi SJ (2008) Sex differences in play behavior in juvenile tufted capuchin monkeys (Cebus apella). Primates 49:288–291PubMedCrossRefGoogle Scholar
  23. Pika S, Liebal K, Tomasello M (2005) Gestural communication in subadult bonobos (Pan paniscus): repertoire and use. Am J Primatol 65:39–61PubMedCrossRefGoogle Scholar
  24. Spinka M, Newberry RC, Bekoff M (2001) Mammalian play: training for the unexpected. Quart Rev Biol 76:141–168PubMedCrossRefGoogle Scholar
  25. Wrangham RW (1977) Feeding behaviour of chimpanzees in Gombe National Park, Tanzania. In: Clutton-Brock TH (ed) Primate ecology. Academic Press, London, pp 503–538Google Scholar

Copyright information

© Japan Monkey Centre and Springer 2009

Authors and Affiliations

  1. 1.Japan Monkey CentreInuyamaJapan

Personalised recommendations