Primates

, Volume 47, Issue 2, pp 113–122

Tool-use for drinking water by immature chimpanzees of Mahale: prevalence of an unessential behavior

Authors

    • Laboratory of Human Evolution Studies, Graduate School of ScienceKyoto University
  • Hitonaru Nishie
    • Laboratory of Human Evolution Studies, Graduate School of ScienceKyoto University
  • Masaki Shimada
    • Laboratory of Human Evolution Studies, Graduate School of ScienceKyoto University
  • Nobuyuki Kutsukake
    • Department of Cognitive and Behavioral Science, Graduate School of Arts and SciencesThe University of Tokyo
  • Koichiro Zamma
    • Primate Research InstituteKyoto University
  • Michio Nakamura
    • Laboratory of Human Evolution Studies, Graduate School of ScienceKyoto University
  • Toshisada Nishida
    • Japan Monkey Centre
Original Article

DOI: 10.1007/s10329-005-0158-4

Cite this article as:
Matsusaka, T., Nishie, H., Shimada, M. et al. Primates (2006) 47: 113. doi:10.1007/s10329-005-0158-4

Abstract

Use of leaves or sticks for drinking water has only rarely been observed during long-term study of wild chimpanzees (Pan troglodytes schweinfurthii) at Mahale. Recently, however, we observed 42 episodes of tool-use for drinking water (73 tools and two cases of using “tool-sets”) between 1999 and 2004. Interestingly, all of the performers were immature chimpanzees aged from 2 to 10 years. Immature chimpanzees sometimes observed the tool-using performance of others and subsequently reproduced the behavior, while adults usually paid no attention to the performance. This tool-use did not seem to occur out of necessity: (1) chimpanzees often used tools along streams where they could drink water without tools, (2) they used tools for drinking water from tree holes during the wet season when they could easily obtain water from many streams, and (3) the tool-using performance sometimes contained playful aspects. Between-site comparisons revealed that chimpanzees at drier habitats used tools for drinking water more frequently and in a more “conventional” manner. However, some variations could not be explained by ecological conditions. Such variations and the increase in this tool-use in recent years at Mahale strongly suggest that social learning plays an important role in the process of acquiring the behavior. We should note here that such behaviors that lack obvious benefits or necessity can be prevalent in a group.

Keywords

Chimpanzees (Pan troglodytes schweinfurthii)CultureMahaleTool-setsTool-use to drink water

Introduction

Wild chimpanzees (Pan troglodytes) show vast inter-group variations in feeding repertoires (Nishida et al. 1983), social behaviors (Nakamura 2002), acoustic signals (Mitani et al. 1992; Clark Arcadi et al. 2004), tool-use (McGrew 1992), and so on. Among these, tool-using behaviors have been studied the most, and outstanding variations have been detected so far. Extensive comparison of behavioral variations from nine long-term study sites listed 39 candidates for “cultural” variations (Whiten et al. 1999, 2001), most of which were tool-using behaviors.

Tool-use for drinking water was reported as customary, as it occurs in all or most able-bodied members of at least one age–sex class (sensu Whiten et al. 1999), among chimpanzees of most long-term study sites: Bossou, Taï, Gombe, Kibale, and Budongo (Whiten et al. 2001). In contrast, it has only rarely been observed in the chimpanzees at Mahale, where long-term extensive studies have been conducted for 40 years. Nishida (1980) presented detailed descriptions of water-contact behaviors of chimpanzees at Mahale, but at that time he had never seen tool-use for drinking water. Subsequently, a few cases of drinking tool-use were observed in some females, although the frequency was very low: two adult females (CH and GW) were each seen to once use a leaf sponge to drink water from a tree hole in 1987 (H. Takasaki, T. Tsukahara and M. Bunengwa, personal communication); an immature female (MG) pushed a large leafy twig into a tree hole and licked the leaves on two occasions in 1996 (L.F. Marchant and W.C. McGrew, personal communication); an immature male (PR) was seen to use a leaf, a stick, and a crumpled vine to drink water from a tree hole in 1996 (M. Nakamura, unpublished). However, in recent years, we have observed many immature chimpanzees using tools for drinking water in habitual manners (see also Matsusaka and Kutsukake 2002), and the frequency of this tool-use has increased.

Tool-use for drinking water was first reported in the earliest work at Gombe (Goodall 1964). In later years, some authors gave detailed descriptions at Gombe (McGrew 1977) and at Bossou (Sugiyama 1995; Tonooka 2001). Chimpanzees use crumpled leaves (leaf sponges), uncrumpled leaves (leaf spoons), or sticks to drink water from tree holes or streams. They soak the tool in water and then suck on or lick it, and they often repeat this series of soaking and sucking/licking. These studies also showed various important aspects of this tool-use: modification of leaf sponges before use, selectivity for a particular kind of leaf (at Bossou, but not at Gombe), and the behavior’s seasonality due to the high probability of tool-use occurring during the dry season in order to slake thirst. Tonooka (2001) suggested that Bossou chimpanzees used “leaf folding,” in which the leaves were not chewed as “leaf sponges” but neatly folded in the mouth. Tool-use for drinking water has also been reported from other study sites in recent years. Lanjouw (2002) reported that chimpanzees at Tongo, eastern Democratic Republic of Congo, used “moss sponges” to drink water on a frequent basis. Hunt and McGrew (2002) reported that chimpanzees at Semliki, Uganda, frequently dug wells in dry, sandy riverbeds and used leaf sponges for gathering water from the wells. Tool-use for drinking water has also been observed in other wild primates such as the capuchin (Cebus albifrons trinitatis) (Phillips 1998) and the orangutan (Pongo pygmaeus) (van Schaik et al. 2003).

In this paper, we report increased tool-use for drinking water among immature chimpanzees at Mahale. The aims of this study are (1) to describe the special characteristics of the tool-using behavior that is performed only by immature chimpanzees, and (2) to compare these data with those from other study sites.

Methods

Subjects

The subjects were chimpanzees of the M group living in the Mahale Mountains National Park, Tanzania (see Nishida 1990). The M group has been habituated to human observers since 1968, and it has been the major study group at Mahale since 1981, when the previous major study group, the K group, became nearly extinct (Nishida et al. 1985). Almost all members of the M group were identified around 1981 (Hiraiwa-Hasegawa et al. 1984), although some females and their offspring remained relatively shy until recent years. The range of the M group is covered with semi-deciduous forests and woodlands. The climate of this range is characterized by a dry season (from mid-May to mid-October) and a wet season (from mid-October to mid-May). There are several rivers or streams that have pools even at the height of the dry season (Nishida 1980). Lake Tanganyika also provides drinking water for the chimpanzees (Nishida 1980).

Data collection and analyses

Data were collected by seven authors between 1999 and 2004. We mainly conducted focal individual observation. Table 1 summarizes the study periods, observation hours, and main focal targets of each researcher. Since sampling methods varied between observers and between years, we cannot examine the fluctuation in the frequency of the studied tool-use between our study periods. For age–sex classes of chimpanzees, we defined infants as 0–4 years old, juveniles as 5–8 years old, adolescent males as 9–15 years old, adolescent females as 9–12 years old, adult males as 16 years old or older, and adult females as 13 years old or older, following Nishida et al. (2003).
Table 1

Study periods on chimpanzees (Pan troglodytes schweinfurthii) of each observer

Observer

Study period

Main targets

Observation hours

Number of tool-use episodesb

Nishida

Aug 1999–Oct 1999

Immature

94

0

Zamma

Aug 1999–Jun 2000

Adult

499

1

Nakamura

Aug 2000–Sep 2000

Adult

149

0

Nishida

Sep 2000–Oct 2000

Immature

89

1

Matsusaka

Sep 2000–Jun 2001

Immature

647

8

Kutsukake

Sep 2000–Sep 2001

Adult

1,100

2

Nishida

Sep 2001–Oct 2001

Immature

67

0

Shimada

Sep 2001–Oct 2002

Adult

1,164

3

Nakamura

Dec 2001–Jan 2002

Adulta

96

0

Nishie

Aug 2002–Nov 2002

Adulta

534

5

Nishida

Sep 2002–Nov 2002

Immature

104

8

Matsusaka

Nov 2002–Oct 2003

Immature

997

7

Zamma

Aug 2003–Nov 2003

Immature

189

1

Nishie

Nov 2003–Sep 2004

Adulta

1,381

5

Nakamura

Dec 2003–Feb 2004

Adulta

153

1

Total

7,263

42

aFocal individuals were adults, but immature chimpanzees were also observed to collect data on the development of cultural behaviors

bNumber of observation of the tool-using episodes

Terms are defined as follows:

Tool-use for drinking water is defined as the behavior of chimpanzees to suck or lick drenched objects that have been immersed in water or simply pulled out of water. Once, a 5-year-old female (IM) picked up a broad leaf from the ground and drank rainwater collected on it as if she was drinking tea from a “saucer.” However, we excluded such cases from our analyses in this study.

An episode of tool-use for drinking water is defined as beginning when a chimpanzee starts to drink water with objects and as ending when the individual leaves the site for another feeding or resting place. When tool-using performance of one chimpanzee was intermitted by another’s tool-using performance, we counted only one episode for each performer. For example, when three chimpanzees (A, B, and C) successively performed the tool-use at a tree hole in the order “A-B-C-B-C,” we counted three episodes. Social or solitary play and grooming was sometimes sandwiched between the tool-using performances. Unless otherwise noted, we analyzed even those episodes that were performed by individuals other than the focal chimpanzees or that were observed at a time outside of the focal observations. Some episodes were incomplete, since they had already begun when we started to observe them, or we quit observing them in order to follow the focal chimpanzees.

The wet season is defined as the period from 16 October to 15 May, while the dry season is defined as the period from 16 May to 15 October. We compared the frequency of the tool-use episodes between wet seasons and dry seasons.

Use of leaf sponges: Chimpanzees sucked one or more crumpled leaves that were drenched with water. Since they usually sucked uncrumpled leaves on the first drinking action, use of leaf sponges usually can be recognized only when they used leaves more than twice repeatedly (see “Tool materials and modification”).

Use of uncrumpled leaves: Chimpanzees sometimes stuffed one or more uncrumpled leaves in their mouth to suck water and then discarded them. If they kept on using them and the leaves became crumpled, we recognized use of leaf sponges. In a few cases, we quit observing the tool-use at the moment of the first drinking action, in order to follow the focal chimpanzee. In such cases, we recognized use of uncrumpled leaves, although chimpanzees might use leaf sponges after we quit observing them.

Use of leaf spoons: Chimpanzees licked water off an uncrumpled leaf or poured water with it into their mouth.

Observation: During tool-using episodes, other chimpanzees occasionally came and watched the performance of the tool-users at a close proximity (Fig. 1). We analyzed the direction of these observations in terms of the ages of the tool-user and the observer.
Fig. 1

Tool-use for drinking water by immature chimpanzees (Pan troglodytes schweinfurthii): FV (right) observes the tool-using performance of PF (left)

Results

Frequency of tool-use for drinking water and the performers

In total, 42 episodes of tool-use for drinking water were directly observed between 1999 and 2004 (Table 1). Table 2 shows the number of tool-use episodes by each performer in each year. All of the 13 performers were immature chimpanzees aged from 2.3 to 10.8 years old. They made up 59% of the 22 immature chimpanzees that were born between 1990 and 2001 and were still alive in the M group in 2004. No sex difference in performance was shown: 5 out of 9 males (56%) and 8 out of 13 females (62%) performed the tool-use (Fisher’s exact probability test, P>0.05).
Table 2

Number of episodes of tool-use for drinking water by each performer

Performera

CD

JD

XM

AT

MC

CR

AC

IM

CE

FV

PF

EM

ME

Total

Cadmus

Jidda

Xmas

Athena

Michio

Carmen

Acadia

Imani

Caesar

Flavia

Puffy

Emory

Mitsue

Year of birth

1991

1995

1995

1996

1996

1997

1998

1998

1998

1999

2000

2000

2001

Year of observation

1999

   

1

      

b

b

b

1

2000

1

2

1

1

        

b

5

2001

 

3

1

3

         

7

2002

1

 

2

 

9

1

1

1

1

1

   

17

2003

   

1

3

 

1

  

1

1

1

2

10

2004

          

1

 

1

2

Underlined numbers include the use of “leaf-sponge”

aNames of performers and their abbreviations. Abbreviations in italics indicate the individual is female

bNot yet born

Fluctuations in demographic and environmental conditions

The frequency of the tool-use was much higher than that reported at Mahale before 1999. We examined whether this increase could be explained by the fluctuations in demographic or environmental conditions. Since all of the tool-users were immature individuals, one might expect that the increase in the tool-use is due to the increase in the number of immature individuals. The number of immature chimpanzees in the M group between 1999 and 2003 ranged from 24 to 30. This range was, however, much lower than the range of 35–55 between 1980 and 1995 (Nishida et al. 2003).

One might also expect that the increase in the tool-use might be the result of individual adaptations in response to water scarcity in recent years. Mean annual rainfall between 1999 and 2003 was 1,648 mm (n=4, range=1,400–1,995 mm). These values were, in fact, less than the 1,889 mm mean annual rainfall between 1976 and 1988 (n=9, range=1,458–2,338 mm) (Takasaki et al. 1990), although the difference was not statistically significant (two-tailed Mann–Whitney’s U test, P>0.05). However, this difference still does not explain why only immature chimpanzees came to use tools for drinking water frequently.

Location of the water

Among 42 episodes of the tool-use, water was obtained from tree holes in 20 episodes and from streams in 22 episodes. Drinking water from tree holes was generally an arboreal activity: only one episode occurred on the ground. Almost all of the tree holes were located higher than 2 m above the ground. Five species of host trees could be identified: Cordia africana, Croton sylvaticus, Ficus exasperata, Myrianthus arboreus, and Pseudospondias microcarpa. Drinking water from streams occurred at standing pools or at the spots where the water flow was gentle.

Seasonality of tool-use for drinking water

Figure 2 compares mean frequency of the tool-use during wet seasons and dry seasons. Tool-use for drinking water from tree holes was only observed during wet seasons. The difference in frequency between wet seasons and dry seasons was statistically significant (n=6 years, two-tailed Wilcoxon’s signed rank paired test, T=0, P<0.05). In contrast, tool-use for drinking water from streams did not show such seasonality (n=6 years, T=6, P>0.05).
Fig. 2

Mean frequency of the tool-use (/100 h) during dry seasons and wet seasons. Two episodes were excluded from the analysis because they were observed at times outside the observation hours listed in Table 1. Error bars show the SDs. *P<0.05

Drinking methods

Chimpanzees often used several objects successively during one episode of tool-use for drinking water. In total, 73 tools were used, involving a variety of materials: leaves (31 leaf sponges, 10 leaf spoons, and 5 uncrumpled leaves), sticks (11 twigs, 9 vines, 2 grass stems, a peel of vine, and a mid-rib of a leaf), a stone, a fruit, and an unknown object (these values were underestimated since some episodes were not completely observed). In some cases, several leaves were used simultaneously as a “leaf sponge.” A 6-year-old male (MC) dipped a fruit of Ficus thonningii in a water puddle, sucked it once, and discarded it [cf. a similar behavioral pattern, wadge-dipping, was reported in Taï chimpanzees (Boesch 1991)]. In addition, use of a “tool-set” (sensu Sugiyama 1997) was observed twice: a 5-year-old female (JD) and a 6-year-old male (MC) used a stick (or a leafstalk) to pick a leaf sponge out of a tree hole.

Use of a leaf sponge may be regarded as the most efficient tool-use to drink water, since leaf sponges are supposedly more absorptive than leaf spoons or sticks. Indeed, use of a leaf sponge occurred most frequently and was observed in most of the performers (11 out of 13). However, chimpanzees did not only use leaf sponges, even after they acquired the skills; drinking with sticks or fingers were often observed immediately after use of leaf sponges or sometimes between use of leaf sponges, and in episodes on later days.

The duration of the tool-use was generally less than 2 min, although it exceeded 10 min in some cases. In the longest incident, three female chimpanzees (AC, FV, and PF) performed the tool-use one after another for 48 min at a tree hole. In total, seven objects were used for drinking water, and a 2.8-year-old female (PF) drank water with a leaf sponge repeatedly more than 30 times.

This tool-use seemed to involve playful aspects in some cases, especially when it occurred at streams. Chimpanzees often made contact with water in playful ways before or after the tool-use: splashing or stirring water with a hand, walking around cautiously in streams, and so on. Occasionally, they even stirred water with a hand while holding a leaf sponge. Once, a 10-year-old male (CD) stirred a leaf sponge in the water while holding it in his mouth.

Tool materials and modification

Chimpanzees always picked nearby materials, within 3 m from the water source, after arriving at the site. They occasionally used leaves or sticks that were already in the water, e.g. materials that remained in tree holes and that floated or sunk in streams. In four cases, they also used materials that were previously used by other individuals.

Raw materials for the leaf sponges were identified in nine tools: Ficus vallis-choudae (four tools), Saba comorensis (three tools), Landolphia owariensis (one tool), and Grewia flavescens (one tool). We could not draw any conclusions regarding selectivity for the particular materials from these values because the availabilities of F. vallis-choudae and S. comorensis seem to be very high; the former is the dominant tree species at riverbeds (Turner 2000), and the latter is the dominant woody-vine species in the forest (Itoh 2002).

Leaf sponges were rarely modified before use. Immature chimpanzees at Mahale usually soaked an uncrumpled leaf in water and then stuffed the drenched leaf into the mouth. The leaf became crumpled during the first sucking action. Although chimpanzees sometimes held leaves between the lips after picking them from vegetation, they did not chew the leaves before they soaked them in water. Only once, a 7.4-year-old female (AT), who was a well-experienced user of leaf sponges, chewed a leaf before using it. In four cases, chimpanzees stripped leafstalks from the leaves after stuffing the leaves into the mouth. Leaf spoons were never modified, but they were sometimes used in a “U” shape. Stick tools were sometimes modified before use in a manner similar to the modifications made before performing wood-boring ant-fishing (cf. Nishida and Hiraiwa 1982): cutting a stick with the teeth and shortening it, stripping leaves off of twigs, removing leaf-blades from a leaf to use the mid-rib, and so on. Chimpanzees tried to enlarge the tree holes by biting the brim in two cases, seemingly in vain.

Social interactions: observations and some conflicts

The tool-using performance was observed carefully by other individuals at close proximity (Fig. 1) in 12 episodes. Since several chimpanzees occasionally gathered and observed the tool-use simultaneously or successively, the number of dyadic combinations of “a tool-user and an observer” amounted to 16. Table 3 shows the directions of relationship among participants in the observations in terms of the ages of tool-users and observers. Adults rarely paid attention to the tool-using performances by immature chimpanzees. And, if they did so, adults never performed the tool-use by themselves after observation. In contrast, immature chimpanzees were more likely to observe the tool-using performance of others, especially the older or same-aged companions. Furthermore, an immature observer often performed the tool-use himself/herself after the previous tool-user went away from the water source. Most of the observations (except three) occurred within non-kin dyads.
Table 3

Direction of observations

Observer → tool-user

Number of combinations

Tool-use after observationa

Younger → older

7

4

Between same ages

5

3

Older → younger

2

1

Adult → immature

2

0

Total

16

8

Between same ages = combinations in which the age-difference was less than 1 year; younger = younger immature chimpanzee; older = older immature chimpanzee

aNumber of cases in which observers performed the tool-use themselves after observation

The tool-users were generally tolerant toward the observers. However, agonistic interactions occurred between user and observer in some cases. The tool-users sometimes threatened the observers by shaking one hand in a rapid motion, by raising one hand with the mouth open, by shaking branches, by emitting cough-bark, and so on. After the threat, observers went away from the tool-users for a while, but often came back and observed the tool-using performance again.

Observers took the tools that the tool-user had just immersed into water in two cases and attempted to do so in three other cases. In one case, after the “theft,” the tool-user (AC) took back the tool from the “thief” (IM). In the other case, the “thief” (PF) ran away and sucked the tool while the tool-user (AC) threatened another observer sitting next to her.

On one occasion, an observer (FV, 3 years old) soaked a leaf in a tree hole in front of a tool-user (PF, 2 years old) after PF quit using a stick and started to drink water with the hand. Then, PF used the leaf as a leaf sponge to drink water repeatedly and FV observed the tool-using performance of PF without agonistic interaction. As a result of FV’s behavior, PF got a more “efficient” tool. Thus, this case may be regarded as a kind of “teaching,” although we cannot confirm that FV had such intention.

Between-site comparisons

Table 4 summarizes site variations in the tool-using performance and the environmental conditions. The sites were arranged in the order of aridity level in terms of annual rainfall and water availability from streams or lakes. We regarded water availability as a more essential condition than the amount of rainfall. Tool-use for drinking water is very frequent at Tongo and Semliki where water is a scarce resource. Moss sponges or leaf sponges, which were assumed to be more efficient tools, were used more exclusively at the drier sites. Furthermore, preparation of the tools before arrival at the water source and modification of the leaf sponges before use were more likely to be observed at the drier sites. Selectivity for particular tool materials was not observed at the wetter sites. In summary, chimpanzees at drier habitats perform the tool-use more frequently and in a more “conventional” manner. Chimpanzees of most habitats use tools for drinking water during the dry season, supposedly to compensate for the scarcity of water at streams. However, chimpanzees at Tongo and Mahale showed the tool-use more frequently during the wet season, which may reflect the water availability in tree holes during this period of the year.
Table 4

Between-site comparisons of tool-use for drinking water

Sitesa

Environmental conditions

Variations in the tool-using performance

Annual rainfall

Distribution of streams

Frequencyb

Age classc

Tool typed

Locatione

Tool preparationf

Sponge makingg

Material selectionh

Seasoni

Tongo

1,753 mm

Noj

+++

A IM

MS

TR

Yes

Wet

Semliki

1,206 mm

Few?

+++

?

LG

WL STR

?

?

?

Dry

Bossou

2,230 mm

Few

++

A IM

LG (LN)

TR

Yesk

Yes

High

Dry

Gombe

1,600 mm

Many

++

A IM

LG (LN TW OT)

TR STR

No

Yes

Low

Dry

Mahale

1,889 mm

Many

+

IM

LG LN TW (OT)

TR STR

No

Rare

Low

Wetl

aTaï, Kibale, Budongo, and Lopé chimpanzees also show the tool-use, but they were omitted because detailed descriptions were not available. Most of the data at Semliki were obtained from the remnants of the tools

bFrequency of the tool-use: + = habitual, ++ = customary but not frequent, +++ = customary and frequent

cAge classes of the performers: A = adults, IM = immature individuals

dTool types: LG = leaf sponge, LN = leaf spoon, TW = twig or vine, MS = moss, OT = others (stone or fruit). Low-frequency tool types are in parentheses. “Leaf folding” was reported only from Bossou, but was omitted (see Discussion)

eLocations: TR = at tree holes, STR = at streams, WL = at dug wells

fCollection of tool materials before arriving at the site of the water source

gModification of leaf sponges before use by chewing leaves with the mouth

hSelectivity for particular materials of leaf sponges

iSeasonality: season in which the tool-use is observed most frequently

jIn spite of high rainfall, Tongo forest is very dry because it grows on a lava flow where rain drains rapidly into the ground and there are no pools or streams

kSome chimpanzees were observed to prepare tools only in the latter part of the field experiment, in which a specific tree hole was filled with water every morning (Tonooka 2001)

lSeasonality was shown only in the tool-use at tree holes

For references on the tool-use, see Introduction. References on rainfall and environment: Bossou (Sugiyama 1989; Yamakoshi 1998), Gombe (Appendix of McGrew et al. 1996; McGrew 1977), Mahale (Nishida 1980; Takasaki et al. 1990), Tongo (Lanjouw 2002), and Semliki (Hunt and McGrew 2002)

Discussion

Increase in the frequency of the tool-use for drinking water

The majority of immature chimpanzees at Mahale have performed tool-use for drinking water in recent years. Since this tool-use was only rarely observed previously at Mahale, it is clear that we observed many more cases than before. However, some observational biases remain to be examined. First, the observational conditions of immature chimpanzees seem to have improved: some of the previously shy females and their offspring have become better habituated to human observers in recent years. However, at least concerning the tool-use at tree holes, this factor does not seem to be significant because chimpanzees were not so shy in the trees (cf. Hunt and McGrew 2002, p. 48), and such tool-use would have been more frequently observed if it had occurred in the trees. Second, during this study, researchers may have paid more attention to immature chimpanzees than before, since they were investigating play behaviors (Matsusaka 2004) and the development of cultural behaviors (e.g., Nishida 2003). This may cause some bias, but this factor also appears to be not so significant: although some previous studies also focused on immature chimpanzees, investigating play behaviors (Hayaki 1985) and development of mother–infant relationships (e.g., Nishida 1988; Hiraiwa-Hasegawa 1990), those researchers seldom saw tool-use for drinking water.

Thus, we can conclude that the frequency of this tool-using behavior has increased at Mahale in recent years. This increase cannot be attributed to genetical change because it occurred over a short term. Demographic parameters also fail to explain the increase. Annual rainfall in recent years has been slightly less than in previous years, so some might argue that the increase in this behavior was the result of individual adaptations in response to water scarcity. However, this explanation is unlikely because the tool-using performance often contained playful aspects and did not seem to occur out of necessity (for detailed discussion, see the next section).

Here, it should be noted that this type of tool-use might not be so difficult for chimpanzees to invent. The drinking tool-use was also sporadically observed in the previous years at Mahale, suggesting that individual invention occasionally occurred. Perhaps because of the relative ease of inventing a drinking tool, this kind of tool-use for drinking water (or juice) is observed almost universally among wild chimpanzees (Whiten et al. 1999) and captive chimpanzees (Kitahara-Frisch and Norikoshi 1982; Takeshita and van Hooff 1996; Tonooka et al. 1997; Morimura 2003). Hayashi and Matsuzawa (2003) suggested that the chimpanzee has a strong behavioral tendency to “insert sticks into holes.” This behavioral tendency may be an innate propensity of chimpanzees, and it is reasonable to assume that it can explain the origins of this tool-use to some extent.

However, it is impossible to explain all of our observations by individual invention alone. First, the increase in the tool-use at Mahale in recent years cannot be explained in terms of individual inventions. Second, individual innovation in the use of “leaf sponges” is assumed to be relatively difficult compared to the use of “sticks” (cf. Hayashi and Matsuzawa 2003). Moreover, since Mahale chimpanzees did not have sponge-like tools in their repertoire, their use of leaf sponges could not have been an application of another tool-using skill that was already acquired. These observations strongly suggest that social learning played an important role in the process of disseminating these tool-using behaviors. This notion is further supported by the results showing that young chimpanzees sometimes carefully observed the tool-using performance of others and then practiced it themselves (Table 3).

Prevalence limited to immature chimpanzees

A striking characteristic of this tool-using behavior at Mahale was that all of the performers were found to be immature chimpanzees aged from 2 to 10. Adults rarely paid attention to the tool-using performance by immature chimpanzees, while immature chimpanzees often observed the tool-using of others (Table 3). It is highly probable that this tool-using behavior was transmitted among immature individuals, especially from the older to the younger chimpanzees. This result contrasts with the results of the previous studies on the development of tool-using behaviors, in which the importance of adult models, in particular the mothers, was emphasized (e.g., Tonooka 2001; Boesch and Boesch-Achermann 2000). Transmission of behaviors among immature chimpanzees was also suggested for a play pattern called leaf-pile pulling (Nishida and Wallauer 2003), but this study presents the first example for tool-using behaviors.

The tool-use of immature chimpanzees in this study, which sometimes involved playful aspects, did not seem to occur out of necessity. First, immature chimpanzees often used tools to drink water from streams, where they could easily drink water without tools. This tool-use at streams was also observed at some other sites (Goodall 1986; Hunt and McGrew 2002), but adults rarely engaged in it (Goodall 1986, p. 542). Second, immature chimpanzees at Mahale used tools to drink water from tree holes only during the wet season (Fig. 2), when they could easily obtain water from streams. They did not use tools to compensate for the water scarcity at streams in the dry season. The observed seasonality may simply reflect the tendency for tree holes to collect rainwater during the wet season. Third, immature chimpanzees sometimes showed play behaviors before or after the tool-use, and the tool-using performance itself was clearly playful in some cases.

It might be possible, however, to imagine some “benefits” of the tool-use. First, use of leaf sponges may be a more efficient way to drink water than drinking with the hand, as suggested by Goodall (1964) from an experiment based on trying the methods herself. However, chimpanzees should be able to drink water from streams more efficiently with the mouth than with tools. Second, the tool-use may enable chimpanzees to maintain a sitting posture, in which they can remain vigilant of the surrounding conditions, since drinking with the mouth usually prevents chimpanzees from watching out. However, the tool-users were not cautious but playful, generally concentrating on the tool-use itself. Thus, the benefits of the tool-use were not so remarkable in this study.

Tool-use is often thought to fulfill some specific requirements that are hard to achieve without tools or to gain some advantage for survival by providing additional resources. Some studies have demonstrated such aspects of tool-use. Chimpanzees at Bossou depend strongly on tools for consuming oil palm nuts and pith when fruit is scarce (Yamakoshi 1998). Chimpanzees at Ndoki, Republic of Congo, were suggested to use sticks to perforate termite mounds and obtain termites even in the seasons when termites were located in deeper sections of the mound (Suzuki et al. 1995). However, the results of our study contrast with the results of these previous studies. In a sense, the activity we observed may be called “a non-adaptive” tool-use.

It is important to note that such behaviors that lack obvious benefits or necessity can prevail in a group. Benefits or necessity of a behavior are not prerequisites of chimpanzee culture. Indeed, cultural play behaviors such as leaf-pile pulling at Mahale (Nishida and Wallauer 2003) and use of self-tickling sticks at Gombe (Goodall 1986) have been reported. A similar suggestion was also made for Japanese macaques concerning “stone-play” behavior (Huffman 1984), although a recent study questioned its playful aspect by presenting that the behavior occurred under a stressful situation of food competition at a provisioning site (Nishie 2002).

Human cultures show similar characteristics: it is often difficult to explain different traditions only in terms of necessity or direct benefits. For example, Japanese people use chopsticks to eat almost any food, although they have the choice of eating by hand as Muslim people do. Moreover, the use of chopsticks is sometimes not an efficient or easy way to eat food: Japanese people often struggle to use chopsticks for eating slippery beans, a fragile tofu, or a chicken wing, and sometimes eventually use hands to eat these foods. However, this tool-using behavior is transmitted from generation to generation, as an “elegant” tradition. There are many other examples of “non-adaptive” human cultures in the areas of play, art, music, etc.

Between-site comparisons

Regional variations can be explained, in part, by environmental conditions. Chimpanzees in drier habitats use tools to drink water more frequently and in more conventional manners: they are more likely to prepare tools before arrival at the water source, to modify the leaf sponges before use, to show selectivity for particular tool materials, and to use “efficient” sponge-shaped tools exclusively. At Mahale, however, these tool-using performances were not so refined, in marked contrast with another tool-using behavior, wood-boring ant fishing (Nishida and Hiraiwa 1982). The range of the M group has much rainfall, many streams, and Lake Tanganyika, and thus it is one of the wettest sites inhabited by chimpanzees. Mahale chimpanzees can drink water from streams or the lake even at the height of dry season (Nishida 1980). Therefore, the necessity for drinking water in tree holes, and thus the necessity of the tool-use, is assumed to be low. This may be the reason why adult chimpanzees at Mahale show the tool-use only very rarely.

However, environmental conditions cannot explain all of the variations between sites. Variations in the types of tools used are difficult to explain solely in terms of ecological conditions. Only Kibale chimpanzees used “stem sponges (chewed-stem)” (Wrangham et al. 1994, p. 10), although twigs are no doubt also available at every other site. Tongo chimpanzees use moss sponges and seem not to use leaf sponges (Lanjouw 2002), without apparent ecological reasons. Similarly, the use of leaf sponges has never been observed in captive chimpanzees, even when leaves were available (Kitahara-Frisch and Norikoshi 1982; Tonooka et al. 1997). These variations can be understood as the results of cultural transmission.

Furthermore, the timing of modification of leaf sponges may also be the result of cultural transmission. Bossou chimpanzees, even infants, always modify leaf sponges before use (Tonooka 2001), and Gombe chimpanzees also usually do so (van Lawick-Goodall 1968, p. 207). In contrast, at Mahale, only one case of modification before use was observed in a 7-year-old female (AT), who was a well-experienced user of leaf sponges. Modification before use would seem to require chimpanzees to possess both an image of the modified tool and knowledge of the effect of the modification. Therefore, it is plausible that it occurs only in well-experienced individuals (see also Kitahara-Frisch and Norikoshi 1982). Keeping this in mind, it is surprising that even infants always stuffed leaves into the mouth before use at Bossou. This comparison suggests that the presence of experienced models can facilitate the earlier acquisition of this technique.

Tonooka (2001) suggested that Bossou chimpanzees often used “leaf folding,” which was different from leaf sponges in that leaves were neatly folded in the mouth but not chewed. We found no tools like “leaf folding” at Mahale. However, we still doubt that this is a regional variation in the “technique” of processing leaves. Another possible explanation is that this difference should be a by-product of leaf-selectivity in Bossou chimpanzees. According to Tonooka (2001), Bossou chimpanzees mostly used leaves of a THV species (Hybophrynium braunianum: family Marantaceae), which seem to have outstanding flexibility. This special property of the leaves may prevent the leaves from being crumpled. At Mahale, chimpanzees always used more rigid leaves of trees or woody-vines, which must be crumpled after being sucked. Furthermore, detailed inspection of video-recorded episodes revealed that, at least in some cases, chimpanzees at Mahale did not chew leaves by the teeth but “folded” leaves in the shape of bellows with the lips, but the leaves became crumpled after being repeatedly sucked by chimpanzees. This behavioral pattern resembles “leaf-swallowing” (Wrangham and Nishida 1983) but is different in that the leaves were neither rough nor hispid. Similarly, McGrew (1977) stated that Gombe chimpanzees did not chew leaves with the teeth but sucked the leaf sponge “apparently by compression between tongue and palate.”

Although tool-use for drinking water has been recognized as almost universal among wild chimpanzees, the results of this study strongly suggest that cultural transmission played an important role in the process of the behavior’s becoming prevalent recently at Mahale, as well as in the development of some regional differences. After the work on chimpanzee cultures by Whiten et al. (1999), studies on regional variations have kept on stepping forward by adding new candidates of cultural behaviors (e.g., Nakamura et al. 2000), by examining the effects of ecological conditions (e.g., Humle and Matsuzawa 2002), by conducting field experiments (e.g., Biro et al. 2003), and by presenting more detailed comparisons (e.g., Yamakoshi and Myowa-Yamakoshi 2004; Nishida et al. 2004; Nakamura and Uehara 2004; Assersohn et al. 2004). Tool-use for drinking water also merits further careful comparisons across many study sites, including Budongo, Kibale, Taï, and Lopé.

Scenarios for the future course of the tool-use

We present here five scenarios for the future course of the tool-use for drinking water in the M group chimpanzees at Mahale: (1) It will also be transmitted among adult chimpanzees; (2) the immature chimpanzees will continue to perform the tool-use when they become adults, although adults do not learn from immature chimpanzees; (3) the immature performers will cease using the tools as they grow up, and the tool-use will be retained as “subculture” of immature individuals; (4) all of the immature performers will cease using the tools, and the tool-use will disappear as a “temporary fashion”; or (5) other “sponge”-using behaviors, which are absent at present, will be differentiated from the tool-use, e.g., Leaf-wipe in Whiten et al. (2001). We still do not have enough data to derive any firm conclusions on this issue, but the results of this study concerning the necessity of the tool-use suggest that scenarios (1) and (2) are unlikely and that scenarios (3) or (4) are more likely. We will continue to observe the course of the tool-use.

Acknowledgments

We thank the Tanzania Commission for Science and Technology, Tanzania Wildlife Research Institute, Mahale Mountains Wildlife Research Centre, and Tanzania National Parks for permitting us to do this research at Mahale and for their support while we were in Tanzania; L.F. Marchant, W.C. McGrew, H. Takasaki, T. Tsukahara, and M. Bunengwa for providing information about the previous observations of the studied tool-use; T. Humle, K.D. Hunt, and an anonymous referee, and the associate editor, W.C. McGrew, for useful comments and suggestions that have improved the manuscript; J. Yamagiwa, Y. Takenoshita, and other members of the Laboratory of Human Evolution Studies, Kyoto University for their useful comments and support. This study was financially supported by MEXT Grant-in-Aid for Scientific Research (A1) (#12375003 and #16255007 to T.N.), MEXT COE Research (#10CE2005 to Osamu Takenaka), MEXT 21st century COE Research Kyoto University (#A14), and JSPS Research Fellowships (to N.K.).

Copyright information

© Japan Monkey Centre and Springer-Verlag 2005