Authors: Sabrina Krief, Hugo Magaldi, Raymond Katumba, Robert Kajobe, Julia Dif, Pierre Poquin, Sarah Bortolamiol, John Justice Tibesigwa, Colin A. Chapman, David P. Watts
Categories: Article, Pan troglodytes schweinfurthii, Culture, Bee foraging, Toolkits, Ecology, Evolution, Zoology
Source: Scientific Reports
Authors: Sabrina Krief, Hugo Magaldi, Raymond Katumba, Robert Kajobe, Julia Dif, Pierre Poquin, Sarah Bortolamiol, John Justice Tibesigwa, Colin A. Chapman, David P. Watts
Chimpanzee behavior, including tool use, varies widely among communities and populations. Tools made by chimpanzees for extracting products from the underground nests of stingless bees are among the most complex used by the species. They have rarely been described in East Africa and have never been observed in three chimpanzee communities in the Kibale National Park, Uganda—two at Ngogo and one at Kanyawara community—that have been studied for over 30 years. In the current study, we present the results of a 15-year study of a fourth community of chimpanzees ranging at Sebitoli, in the northern part of Kibale, and the insects they consume. We identified the stingless bees and carpenter bee species on which they feed. In addition, we collected the tools used by the Sebitoli chimpanzees. Of the 443 tools used in 152 episodes to extract products from insect nests, 332 were used by chimpanzees to explore or exploit underground or arboreal nests of Meliponula sp.. In addition, individuals sometimes left sticks vertically inserted into the entrances of underground bee nests. We discuss the implications for the transmission of the tool use behavior for subterranean extractive task within this social group, given that it appears to be absent in the other three Kibale Forest communities being studied. Our results highlight the importance of taking small-scale cultural variation into account in understanding chimpanzee behavioral repertoires as well as planning and implementing conservation strategies. We dedicate this article to the chimpanzee Hugo, who loved honey the most. He was slaughtered with machetes by poachers on April 3, 2026.
The online version contains supplementary material available at 10.1038/s41598-026-50645-5.
Chimpanzees (Pan troglodytes) manufacture and use a wide variety of tools in the wild. Patterns of tool use vary widely between populations, and some evidence exists of widespread “behavioral realms”^1^. Many of the differences are consistent with culture, defined as “a community-specific set of behaviors that an individual is exposed to and can socially learn from others”^2^. For widely-separated populations, it is difficult to rule out possible genetic influences on chimpanzee behavior^3,4^. Thus, to achieve a clearer understanding of the diversification of chimpanzee traditions, it is important to compare behaviors groups living in proximity in shared habitats and those in populations that are now separated but were recently contiguous^2,5,6^. In such situations, genetic differences among communities in the same population should be minimal and are unlikely to have any impact on behavior. Likewise, although ecological variation over short distances can lead to considerable differences in diet and demography between chimpanzee communities^7–11^, such ecological differences should be lower within than among populations^2,5,6,12^. Thus, intra-population comparisons potentially provide a better understanding of the social effects driving behavioral diversity than comparisons made over large geographic scales.
Insectivory targeting social insects, such as termites, ants, and bees, is widespread among chimpanzees, but tool-assisted consumption is not universal. Extractive foraging with sticks, twigs, or grass blades is one of the most common forms of chimpanzee tool use, and chimpanzees in many communities use stick tools to obtain honey and/or other bee products, such as larvae (e.g. Cameroon^13^; Central African Republic: Bai Hokou^14^ and Ngotto^15^; Democratic Republic of Congo: Kahuzi-Biega^16^, Bili-Uéré^1,17^; Gabon: Lopé^18^; Ivory Coast: Taï^19^; Republic of Congo: Goualougo^20^; Tanzania: Mahale^21^ and Gombe^22^; Uganda, Bulindi^23^, Bwindi^24,25^, Kibale^26^. In some populations, chimpanzees harvest honey produced by multiple types of bees that differ in behavior, sociality, ecology, nesting location and aggressiveness. Honeybees (Apis spp.) are aggressive and their stings are painful, while stingless bees in the genus Meliponula deliver relatively gentle bites. Meliponula spp. and other stingless bees often make ground nests in addition to tree nests. Carpenter bees make nests by tunneling into dead wood, making it necessary for chimpanzees to use tools to access brood, pollen and nectar stored in the nests. Carpenter bee males are harmless, and females rarely sting. Chimpanzees can harvest honey from nests of Apis spp. without using tools, but extracting carpenter bee products and honey from stingless bee ground nests requires tools^1,27,28^.
Tool use for extractive foraging is uncommon in Kibale compared to some other sites (e.g., Taï^19^; Bili-Uéré^1,17^; Loango^29^; Fongoli^30,31^; Mahale^32^. Chimpanzees use tools in four kinds of extractive activities in the Ngogo communities^26^ and two in the Kanyawara community^33^. The Ngogo chimpanzees occasionally use sticks to extract honey from bee nests in fallen tree trunks and from arboreal nests in standing trees [26; also see below]. No published records exist of chimpanzees consuming honey from ground nests of stingless bees in either community and use of tools to fish for ants or termites is rare at Kanyawara and Ngogo^34^. Watts^26^ suggested that the absence of Macrotermes termite mounds in Kibale^35^ explains the rarity of termite fishing there. However, at Kanyawara and Ngogo, Dorylus ants are abundant, but not consumed, unlike at other sites across Africa (Fongoli^30,36^; Gashaka^37^; Gombe^38^; Goualougo^39,40^; Mahale^41^; Mt Assirik^42^; Tai^43^; Bili-Uéré^1^. Chimpanzees at several other Ugandan sites also regularly encounter ant swarms, but do not consume the ants with sticks (Semliki^44^; Bulindi and Budongo^45^. The chimpanzees in this region may not recognize these ants as food. Evidence from Budongo, also in western Uganda, is chimpanzees in the Sonso apparently do not eat Dorylus, but those in the Waibira community do, although using leaves and not sticks^46^. In contrast, chimpanzees at the southwest Ugandan sites of Kalinzu^47^ and Bwindi^48^ use sticks to feed on Dorylus.
Watts^26^ proposed the “necessity hypothesis”^49^ as an explanation for the low rate of tool use at Ngogo compared to other sites in Africa in the chimpanzees have little need to use tools to exploit inaccessible resources because nutritionally suitable alternatives are readily available. Notably, meat consumption at Ngogo is relatively high^50,51^. However, the hypothesis of meat consumption cannot as easily explain the rarity of tool-assisted extractive foraging at Kanyawara, where hunting is much less common x^50^. Furthermore, meat is presumably important for chimpanzees as a source of protein and micronutrients^52,53^, while honey is 80–90% carbohydrates, even if chimpanzees also eat bee larvae that provide protein. In their comparison of two sites in the Budongo Forest Reserve, Uganda, Gruber et al.^4^ reported only one kind of tool used in extractive foraging and found no evidence that chimpanzees there used sticks to harvest honey. These authors argued that the diversity of extractive tool types was higher at Ngogo compared to Sonso or Kanyawara because reliance on fruit was highest at Ngogo, and dietary diversity was lowest. They argued that the Ngogo chimpanzees may face irregular periods of fruit scarcity and thus, it may pay off for them to use tools when they do not have access to their favorite fruits. Long-term data from Ngogo, however, are inconsistent with this argument because periods of fruit scarcity are relatively infrequent and short at Ngogo, at least compared to Kanyawara^54^.
The lack of a clear ecological explanation for the differences in tool use at these sites suggests that social learning best explains both tool kit diversity and tool use frequency. Toolkits of Western Ugandan P. t. schweinfurthii are smaller than those of P. t. schweinfurthii at Gombe, Bili-Uéré and Mahale^1,38,41^, P. t. verus at Fongoli^55^, Taï^43^ and Bossou^56^, and those of P. t. troglodytes at Goualougo^20^, Loango^29^, and Ngotto^15^. Van Schaik et al.^57^, citing data from Kanyawara, proposed that relatively low social tolerance in eastern chimpanzees might account for this difference. This hypothesis cannot, however, account for the relatively small toolkit at Ngogo^26^, where gregariousness is high, particularly for females^58,59^.
To increase our understanding of the factors driving the observed geographical variation in chimpanzee tool use, we use new data from a fourth chimpanzee community in Kibale (Sebitoli), plus published and updated information from Ngogo and Kanyawara. The home range of the Sebitoli community borders that of Kanyawara, facilitating female immigration. Female immigration between Ngogo and Kanyawara has already occurred. We address the following three How frequently do chimpanzees at Sebitoli use tools to collect insects or their products?Do Sebitoli chimpanzees harvest different insect products from different insect species than those harvested by Ngogo and Kanyawara chimpanzees?Do ecological differences rather than social factors explain variation in tool-assisted extractive foraging among chimpanzee communities in Kibale National Park, and what does this imply for species-wide variation in chimpanzee tool use?
Kibale National Park lies in southwestern Uganda between 0°13–0°41 N and 30°19 − 30°32 E. The 795-km^2^ park is mostly covered by a mix of moist evergreen and semi-deciduous forest, much of it at various stages of regeneration from past human disturbance^60,61^. Sebitoli is found in the northern sector of the park, Kanyawara is located between Sebitoli and Ngogo. Sebitoli and Kanyawara chimpanzee ranges are contiguous and female migrations occur, while Sebitoli and Ngogo core areas are 17 km apart (Fig. 1;^62^. The Sebitoli chimpanzee community contains approximately 80 individuals, who range over an approximately 20 km^2^ area. Sebitoli’s forest was commercially logged in the 1970s, leading to loss of about 50% of the trees^63^. Today, 70% of the forest in the Sebitoli community home range is regenerating and only 14% is old growth^62^. More than 80% of the community’s perimeter borders farmland and tea estates, compared to 47% at Kanyawara and none at Ngogo^62^. The human population density in areas adjacent to Kibale is high (~ 300 inhabitants/km^2^;^64,65^, and a heavily-traveled tarmac road crosses the Sebitoli community’s home range^66,67^. Of the three sites, Ngogo, which was never subjected to commercial logging, has the largest proportion of old growth forest^63^. Part of the Kanyawara community’s home range was subjected to commercial timber extraction that included both selective logging of natural trees before the 1960s and extraction of pines and other exotics which had been planted in the grasslands in the 1960s. Both Kanyawara and Ngogo comprise mixes of vegetation types that include regenerating forest of varying ages.
Fig. 1locations of the habituated chimpanzees communities where chimpanzees have been studied in the Kibale National Park of Uganda, including the Sebitoli area, where we conducted the current tool use survey.
Habituation of chimpanzees at Sebitoli began in 2008 under the auspices of the Sebitoli Chimpanzee Project (SCP), supported by the Muséum National d’Histoire Naturelle in Paris and the Great Ape Conservation Project (GACP). Sixty of the estimated 80 community members (15 adult males, 18 adult females, 7 adolescents, 15 juveniles, 5 infants) are easily recognized by the researchers and regularly observed. Although many of the chimpanzees have injuries due to snares or show congenital and/or disease-induced deformities^68–71^, population density is high (> 3 individuals/km^2^) compared to Kanyawara, although lower than that at Ngogo. SCP teams made up of researchers, student assistants, and local field assistants monitor the chimpanzees daily. Data reported here come from direct observation and camera-trapping between February 2009 and January 2023 (see below).
Observations at Ngogo have been continuous since June 1995^72^. The original single Ngogo chimpanzee community included at least 205 members (35 adult males and 66 adult females) in 2016, before undergoing a permanent fission into two communities, “Ngogo Central” and “Ngogo West”, that was compete by 2018^73^. Ngogo data come from ad lib sampling of tool use by D. Watts between 1998 and 2007^26^ and from^34^ plus unpublished ad lib sampling by D. Watts between 2008 and 2025.
We identified the bee species whose honey or products are consumed at Sebitoli, but did not try to identify the ants and termites there because we did not observe chimpanzees using tools to consume either. To identify the stingless bee species in Sebitoli and to confirm whether they are also present elsewhere in Kibale, we collected at least 10 adult individuals from each of three nests (Table 1). Samples #1, #2, and #4 were collected in the Sebitoli chimpanzee home range, and sample #3 was collected at Kanyawara. This allowed us to take into account individual phenotypical variation. Eleven adult stingless bees were collected at a location where chimpanzees used tools to get honey from a dead tree on December 20th, 2016 (sample #1). We also collected 11 adult stingless bees (sample #2) from a nest located in a tree at 2 m from ground on January 15th, 2017. Sample #3, comprising 12 stingless bees, was collected at Kanyawara on February 22nd, 2017. We collected an adult carpenter bee from a dead branch of Carapa sp. after a female chimpanzee had explored its tunnel with a tool on February 4th, 2017 (sample #4). Specimens were stored in 50% ethanol. Identification of the bee species was performed by R. Kajobe based on Eardley^74^, Eardley et al.^75^, and Pauly and Vereecken^76^.
We were able to determine the bee genus for each nest that the chimpanzees exploited during the study period, but could not always identify the species.
In 2008, SK discovered indirect evidence of chimpanzee tool use at Sebitoli, apparently to collect honey from a stingless bee ground nest. She saw a stick inserted into the small entry hole of a bee nest. Upon removing the stick, she noticed that one end resembled a brush. She then discarded the stick by the bee nest entrance. The next day, while following fresh chimpanzee signs in the same area, she found a stick, likely the same one as the day before, with the same brush end planted into the same bee nest. From the footprints, it appeared that one or more chimpanzees had stopped at the nest. Subsequently, the Sebitoli research team began to pay attention to ground bee nests. Later observations confirmed that the co-occurrence of footprints, inserted sticks, and bee nests was not coincidental. We then set up camera-traps at locations where stingless bee nests and chimpanzee prints were present.
We also directly recorded data on tool ad libitum, filming the chimpanzees whenever we observed them consuming honey or other bee products (larvae, combs, nectar, pollen). We geolocated sites where chimpanzees were observed consuming honey or using tools to get bee products using a Garmin 64CS. We considered a tool use event to have occurred when we either (1) saw or filmed chimpanzees using sticks to explore bee nests, regardless of whether they extracted and consumed honey, comb, or larvae, or (2) we found sticks with clear signs of modification or/and use (e.g., brush tips; peeled bark, removed or modified branches, sharpened ends, mud at one or both ends) within a horizontal distance of less than 5 m from a ground-level bee nest or under an arboreal nest. We recorded the type of insect nest (tree cavity, tunnel in a branch, underground tunnel) and its location (ground/fallen dead branch/dead trunk/live tree). Some indirectly inferred events might have grouped several cases of tool use carried out at different times. At locations where we observed chimpanzees collecting honey or probing insect nests, and at nests where we suspected chimpanzees might visit, we set up camera traps focusing on the nest entrances. We used HD cameras (Bushnell Trophy Cam HD Max™ and Reconyx™) with day/night auto-sensors (motion sensors capable of recording to a distance of 18 m or beyond and using an LED night vision flash (IR) and an adjustable Passive Infra-Red (PIR) motion detector, with no-glow black LEDs) and sound recording capabilities. We set the cameras to record definition videos of 1280 × 720 pixels, with a video length of 30 s, a trigger interval of 1 s, and low PIR. Video-traps were strapped to trees at a height of 1 m and were operational 24 h per day.
Our camera traps recorded 111,135 clips. Chimpanzees were detected in 7114 of them. 684 videos recorded behaviors involving tool manufacture and/or use, stingless bee nest inspection, and interactions between chimpanzees in proximity to bee nests. We will analyze this dataset, including social interactions at bee nests, in detail elsewhere. Here we focus on tool users and tool characteristics. Of the 152 clips in which at least one chimpanzee could be observed using tools, we either identified the individuals (N = 15) for 138 clips or, when this was not possible, recorded their sex and age classes. “Immatures” refer to infants and juveniles less than 10 years of age, while “mature” individuals include subadults and adults over 10 years of age. In some cases, we could match specific tools with their users, but this was not always possible.
During direct observations, we collected sticks used to explore nests of carpenter bees, Apis and stingless bees. We also collected any tools we found close to stingless bee nests during weekly collection of SD-cards from the camera traps. A stick found near the bee colony was considered a tool if it had been broken off from nearby trees, had leaves removed, and showed signs of use such as bark stripping or a blunted or a sharp end. These usually occurred in clusters of several tools.
We measured the length and diameter of each of the sticks recovered and noted features related to tool manufacture as well as direct or indirect modification.
For each tool collected around the insect nest, we tried to identify the origin of the broken branch so we could identify the plant species from which it came. Most species used to prepare tools were also commonly used for nest or for food. SCP and KCP field assistants could identify most species based on their long-term knowledge. We also conducted a plant collection campaign between 2000 and 2010, within the framework of Memorandum of Understanding / Cooperation between the Centre National de la Recherche Scientifique (CNRS), the Muséum National d’Histoire Naturelle (MNHN), Makerere University (MU) and the Uganda Wildlife Authority (UWA). This involved collecting three specimens each of 24 species identified by research assistants in the field and comparing these to herbarium specimens at the herbarium of Makerere University, Kampala (Uganda) and that of the Museum National d’Histoire Naturelle in Paris, France^77–79^. L. Allorge and A. Hladik compared one set of voucher specimens (SK1 to SK24^77^, to specimens at the MNHN and J. Kasenene compared a second set to at specimens at MU. We deposited these SK1-SK24 specimens at the respective herbaria and kept the third sample of the triplicate the research station. The specimens collected in KNP can be reviewed at the following https://science.mnhn.fr/institution/mnhn/collection/^77,78,80^.
The sample sizes and the variances of the tools used for Apis mellifera (n = 25) and for Meliponula sp (n = 331) were unequal, so we used Mann-Whitney tests to compare their lengths and diameters. We used Student’s t-tests to compare the mean length and diameter of tools between female and male tool users. We compared observed frequencies of tool users’ age and sex classes recorded on clips to those expected based on male-female ratios in other video clips during the same period.
Data collection was strictly observational, and no contact was ever made with the chimpanzees. We adhered to the research protocols defined by the guidelines of Kibale National Park, under a Memorandum of Understanding (Muséum National d’Histoire Naturelle, Uganda Wildlife Authority, Makerere University; SJ 445-12) and following the ethical guidelines of The International Primatological Society and The American Society of Primatologists.
Samples #1–3, collected in different types of nests (dead tree, live tree, ground nest) and in two chimpanzee home ranges (Sebitoli and Kanyawara), corresponded respectively to three species of stingless Meliponula (Meliplebeia) lendliana (Friese 1900); Meliponula (Meliponula) bocandei (Spinola 1853) and Meliponula (Axestotrigona) ferruginea (Lepeletier 1841). The carpenter bee (sample #4) was identified as Xylocopa -(Mesotrichia) torrida (Westwood 1838) (Table 1).
Table 1Identification of bees collected in Kibale National Park, Uganda.Date of collectionChimpanzee home rangeBee nest locationVernacular nameNumber of individualsIdentification20/12/2016SebitoliDead treeStingless bees11Meliponula (Meliplebeia) lendliana (Friese 1900)15/01/2017SebitoliLive treeStingless bees11Meliponula (Meliponula) bocandei (Spinola 185322/02/2017KanyawaraUnderground nestStingless bees12Meliponula (Axestotrigona) ferruginea (Lepeletier 1841)04/02/2017SebitoliDead logCarpenter bee1Xylocopa (Mesotrichia) torrida (Westwood 1838)
We observed or recorded chimpanzees using stick tools to obtain and consume honey, comb, and larvae from the arboreal nests of stinging bees, Apis mellifera, from the arboreal and terrestrial nests of stingless bees (Meliponula spp.) and from tunnels in dead wood drilled by carpenter bees (X. torrida). They also used sticks to extract larvae of beetles belonging to the Cerambycidae from the soil. A female chimpanzee was observed on one occasion using a tool to collect termites from a ground nest. In most cases of tool-assisted honey collecting, the chimpanzees also used their hands to try to collect honey.
We measured 443 sticks collected in 152 events between February 2009 and January 2023 (Table 2). Of these, the chimpanzees used 332 tools to explore and extract insect products during 28 events for stingless bee nests, 28 during nine events for Apis nests, 12 for carpenter bees during five events, nine for beetle larvae, and four for termites. For 58 tools, the targeted insect was not determined, including 10 having clear signs of use by chimpanzees for which the nest/dead log with carpenter bee could not be found. They used the majority of tools (381 out of 443, used in 28 events) to explore terrestrial insect nests. We collected 168 tools after directly observing their use and 91 after observing video records. We also collected another 165 sticks apparently used as tools from bee nest sites. These included 96 next to stingless bee nests and nine inserted in holes made by stingless bees, one close to a log with a carpenter bee tunnel in it, six under Apis nests, five near a tunnel of a Coleopteran and two inserted into the tunnel.
Table 2number of tools used by Sebitoli chimpanzees between 2009 and 2023 at different insect nests (Kibale National Park). Identification of M. lendliana was confirmed by R. Kajobe.Location of the bee nestApis melliferaMeliponula lendliana*Meliponula sp.Xylocopa torridaCerambycidaetermitesNATotalTerrestrial40268129448381Arboreal282452NA1010Total2840292129458443
The 443 tools used by Sebitoli chimpanzees for bees, Coleopterans and termite nests had an average length of 40.10 cm (range = 1.00–137.50) and an average diameter of 0.94 cm (range = 0.20–3.50). Tools used for Apis were significantly shorter than those used for stingless bees. Diameter values were similar (Table 3).
Twelve tools used for getting carpenter bees products during seven tool use events had an average length of 33.02 cm (range = 11.1–97.2) and an average diameter of 0.90 cm (range = 0.30–1.50).
Table 3Comparison of diameter and length of the tools used for Apis mellifera and Meliponula sp. Using direct observations and camera trap videos, we identified 41 individual chimpanzees who had used tools which we later collected (20 males, 21 females). Males used 45 of these tools and the females used 69. We could not identify the users in another 10 cases of tool use by females, although we include them in the totals of females observed. Tools used by males were significantly longer and have larger mean diameters than tools used by females (Table 4).VariableSpecies n Mean ± SD (cm)Median (IQR) (cm)Mann–Whitney U testp-valueLength (cm) Apis mellifera 2535.0 ± 16.934.0 (25.5–47.0)23800.02Meliponula sp.33142.0 ± 20.142.0 (31.0–58.0)Diameter (cm) Apis mellifera 251.01 ± 0.750.80 (0.60–1.10)35600.41Meliponula sp.3310.96 ± 0.470.90 (0.60–1.20)
Table 4Comparison by sex of length and diameter of chimpanzee tools used for Meliponula sp. at Sebitoli, Kibale National Park, Uganda.VariableSex n MeanStandard Deviationtp-valueLength (cm)Females6935.1117.58− 4.44< 0.001Males4857.0230.83––Diameter (cm)Females690.770.28− 3.000.004Males481.160.87––
The most frequently identified tool users were three individuals from the same family (adult female KE and her two offspring, KW and KJ). We collected more Apis tools from males (7) than from females (2), but more Meliponula tools from females (69) than from males (48).
We identified 44 plant species used to manufacture 340 of the tools (Appendix 2). Those used most often were Brillantaisia cicatricosa (90), Mimulopsis arborescens (n = 32), Alangium chinense (n = 23) and Strombosia scheffleri (n = 14). All of the 23 stick tools except one made from A. chinense were peeled, most of them along their entirety. The chimpanzees also peeled some tools made of Mimulopsis arborescens (14/32), Brillantaisia cicatricosa (16/90), Strombosia scheffleri (5/14), and Solanecio manii (4/20). Out of 258 tools, one or both ends (70 tools) had brush-like shapes and 14 tools had ends which have been used for different purposes (i.e., mud on one side and brush without mud the other side).
We used all recorded camera trap clips, including those for which tools were not collected, to investigate if there was a sex bias for stingless bee tool use. The sex and age classes of the tool user could be identified in 150 out of 152 clips recorded at stingless bee nest sites where at least one chimpanzee was recorded using a tool. In 115 clips (76.6% of the clips), females were recorded handling a tool (mature females in 98 clips, immature females in 17). In 35 clips, males were recorded using a tool (mature males in 25 clips, immature males in 10 clips). The observed frequencies of tool users by age and sex classes differed significantly from the frequencies expected on the basis of age-sex representation in the full video sample (total number of clips with immature 9%, immature 1.8%, mature males 56.75% and mature females 32.36%; Chi 2 = 113, df = 3, p value < 0.001).
In one instance, we observed a novel use of tools in the context of honey eating, for a purpose other than honey extraction or consumption. On July 24th, 2013, we observed an elderly adult female chimpanzee collecting honey from a fallen tree with her hands. When attacked by bees, she stripped a handful of leaves from the closest tree and placed them in the hole, thereby blocking the bees inside the trunk. Immediately after this, she rapidly left the area.
Four camera trap videos filmed at the same subterranean stingless bee nest provided clear evidence that chimpanzees inserted sticks into underground bee nests (see Videos 1–4 in Supporting Information). This camera trap was installed on April 20th, 2015, perpendicular to a trail commonly used by humans and nonhuman animals and where chimpanzees were observed exploring a bee nest. Here we describe in detail the behaviors recorded by this
Video April 28th, 2015, 34 am (Clip 62): An adult female missing her right hand (KG) picked up a stick 20 cm from the nest entry with her left hand, inserted it vertically into the ground at the nest, dropped it on the ground, scraped the ground with her hand and inserted the stick again at the same location just before leaving the bee nest. A second adult female (PP) arrived, also scraped the ground with her right hand, manipulated the stick left by KG with her right hand, then dropped it and scraped the ground once again before leaving. Neither female extracted honey or even smelled the stick. A third old female (HI) followed KG and PP and she passed to the right of PP, who was manipulating the stick, brushing lightly against her and slightly turning her head toward her after passing her.
Video May 1st, 2015, 03 pm (clip 80): None of the next 17 clips recorded by the same camera showed chimpanzees, although they recorded field assistants, three baboons, a bushbuck, several guinea fowls and a golden cat. Both the baboons and the golden cat stopped and sat next to the bee nest. On clip 80, the old female (HI) who was on clip 62 walked by the bee nest without stopping. The female following her, PP, who had manipulated the stick on 28th April, stopped at the nest; she picked up a stick and planted it in a hole, but walked off without removing it and checking it for the presence of honey.
Video May 2nd, 2015, 54 am (Clip 81): There was no clip recorded since PP planted the stick in the ground. On the next morning, adult female GL arrived at the site and smelled the end of the stick that PP had inserted. She dropped it and scraped the ground, then forcefully inserted a finger of her left hand into the hole. She turned her head to look at juvenile male KI who was not looking at her. Next, she removed her finger without smelling it and gently inserted the stick vertically into the hole. However, she left the nest with the stick planted in the ground, without smelling or inspecting it. KI is not facing the stick, he is smelling and eating something in his hand. At the end of the clip (8:55:26), KI scanned the area, but seemed to look into the distance and not immediately to look at the stick.
Video May 2nd, 2015, 55 am (clip 82): Eight seconds after the end of video 3, however, juvenile male KI pushed the stick deeper into the hole with both hands and his left foot. He then removed it, smelled the end that was in the ground, but dropped the stick and left.
Since 2006, ad lib observations at Ngogo by D. Watts have added 35 cases of Apis bee nest raids. During 11 of these raids, a total of 28 individual chimpanzees used 40 tools to extract honey. Six adults used two or more tools during single honey-feeding bouts and three juveniles re-used tools made by their mothers. At least 58 individuals obtained some honey during these 11 bouts, not always via tool use. Some of those who used tools did so after using their hands/arms to reach easily-extractable honey. Tool use seems to be more common at partly or largely abandoned Apis nests; when the chimpanzees were observed to raid active nests with many bees, they made quick manual grabs for honeycombs and then fled to escape the defending bees, which precludes tool use. Chimpanzees at Ngogo, however, have never been seen trying to extract stingless bees from underground or tree nests, nor have researchers found any indirect indications that they do this.
Were identified at least three species of stingless bees of the genus Meliponula, Apis mellifera and one carpenter bee species (Xylocopa torrida) in Kibale forest. Chimpanzees at Sebitoli use tools to extract bee products from nests of bees belonging to these five bee species, as well as to extract underground Coleopteran larvae and very rarely termites. The Sebitoli chimpanzees were not completely habituated when we began collecting data, but direct and indirect observations documented use of 443 tools in 152 events to explore nests or extract honey. Most of the tools (72.6%) and the events (76.6%) targeted stingless bees, indicating that this behavior is a common part of the Sebitoli chimpanzee repertoire.
Watts^26^ reported 10 cases of the use of stick tools to extract honey from Apis nests between 1995 and 2006. Since then, 28 individual chimpanzees used 40 tools during 11 bouts to extract honey that were located either in standing trees or fallen tree trunks at Ngogo. Chimpanzees at Ngogo mostly extract honey and comb with their hands; some of those who used tools did so after reaching into nests with their hands/arms. Tool use seems to be more common at partly or largely abandoned Apis nests; when the chimpanzees raided active nests with many bees, they made quick manual grabs for honeycombs and then fled to escape the defending bees, which precludes tool use. Chimpanzees at Ngogo, however, have never been seen trying to extract stingless bees from underground or tree nests, nor have researchers found any indirect indications that they do this. In contrast, D. Watts has never seen Ngogo chimpanzees use tools to extract honey from underground stingless bee nests in over 18,000 h of direct observations between 1995 and 2025, D. Watts has never seen Ngogo chimpanzees use tools to extract honey from underground stingless bee nests, nor have any other researchers or field assistants observed such behavior.
Kanyawara chimpanzees commonly use stick tools to extract ‘bee-bread’ from nests of carpenter bees (Xylocopa caffra L., Wrangham pers. comm.). They also occasionally use stick tools to extract honey from Apis spp. nests, but they have never been seen to probe or extract honey from the underground nests of stingless bees^4^.
Chimpanzees at all three long-term research sites in Kibale use tools to extract honey from Apis nests and to extract carpenter bee products, but only those at Sebitoli use tools to excavate the ground nests of Meliponula and extract Coleopteran beetle larvae, to collect the honey and other products of stingless bees from dead trees, and to eat termites. The 152 episodes of tool use recorded at Sebitoli over 14 years, including the habituation period, far exceed the total number of cases (n = 21) of tool use at Apis nests documented at Ngogo, although we caution the Ngogo sample includes only observations made by D. Watts. Extraction of honey and other products from the underground nests of stingless bees is a very complex task, given the difficulty of spotting and accessing these hidden resources^27,28^. Underground bee nests are difficult to detect, as the only signs of their presence are small wax tubes protruding from the ground and that the entry tunnels that these mark rarely descend vertically to the nests. Estienne et al.^27,28^ found that tubes of M. lendliana nests at Loango, Gabon, were located at a mean distance of 21.5 cm from nest entrances and up to a 52.2 cm horizontal distance from the entrance. Because the nests are out-of-sight and deep in the soil (49.8 cm on average and up to 117 cm^27,28^, chimpanzees must make prolonged efforts to access them. This requires much more effort than most other forms of tool-assisted extractive foraging in chimpanzees^20,27,28^.
Hicks et al.^1^ propose that the use of stick tools to excavate honey of underground stingless bees was mostly confined to the forests of the Congo Basin (i.e., Loango:^27,28^; Goualougo :^20^, Bili-Uéré:^1,17^ and rare or absent in the far west and east of chimpanzees’ range in Africa. However, McLennan^23^ recorded its occurrence at Bulindi in western Uganda. Stingless bees produce honey popularly known for its distinct flavor and aroma, with a more fluid texture and slower crystallization, which is appreciated by human consumers worldwide^81^. Stingless bee honey is also highly sought in Africa for its medicinal uses and has been used by indigenous peoples for millennia^82^. It is used in traditional medicine against colds and throat inflammation^83^ and has known antimicrobial properties^84^. Such attributes, as well as the absence of risk of being stung, may explain the efforts produced by some Central and East African chimpanzees to acquire the honey.
The Sebitoli chimpanzees used 44 plant species as raw material to make extraction tools. These included herbaceous vegetation and shrubs (e.g., Solanecio manii, Piper umbellatum, Piper capense, Aframomum sp.). They peeled the tools in 22.8% of the cases and appeared to modify their tools differently depending on the plant materials. They peeled all sticks for Alangium chinense, while they partially peeled 43.8% of tools made of Mimulopsis arborescens, 35.7% of Strombosia scheffleri and 20% of Solanecio manii (20%). A. chinense is a very hard wood traditionally used by local people for pestles and mortars^85^, but its bark and its roots are highly toxic to humans^86^. Chimpanzees may select A. chinense branches because of their hardness and durability but remove the bark near the tips of the branches to avoid the toxins. The Sebitoli chimpanzees also manufactured multifunctional tools, some apparently having been used at each end with different goals (brush at one end, mud on the other one), as previously reported by Bessa et al.^87^. Camera trap clips showed mature females less often than mature males, but mature females were recorded using tools four times as often as mature males (65.3% vs. 16.6%) to extract Meliponula honey. We documented males (7 cases) using tools more often than females (2 cases) to get Apis honey, but we could not determine the tool user in 19 of 28 cases and thus are unsure whether a sex bias existed. Female-biased tool use in chimpanzees is known at several other sites^43,88,89^. The bias at Sebitoli towards mature females and their offspring using tools at stingless bee nests may be explained by the fact that raiding Meliponula nests is a risk-free activity because the bees do not sting. In addition, it provides excellent learning opportunities for immature individuals, who need to develop their manual dexterity and be able to calmly observe tolerant models. Such training sessions may also help them to develop stick tool skills that they will be needed for the nests of aggressive Apis bees^90,91^.
Sebitoli chimpanzees appear to recognize the locations of underground bee/beetles’ nests and check them regularly. Besides using tools to obtain insect products, the Sebitoli chimpanzees sometimes inserted stick tools into nest entrances (14 cases were observed), but then left these in the holes without smelling the ends of the sticks or using them to dig into the nests. This might have been (1) incidental (i.e., the chimpanzees started the process and stopped it for some reason), (2) a way to close the exits and thereby prevent the insects from leaving the nests, (3) a way to avoid the destruction of potential entry-points by trampling by other animals including humans, or (4) to create a signal to mark the locations of nest entries for later use by themselves or others. We have insufficient data to determine which, if any, of these possibilities is/are correct. The videos presented in this article were recorded over a period of 5 days. In them, we see three different females (KG, PP, GL) independently planting a stick vertically into a bee nest entrance, then leaving calmly, with no evidence of being startled or distracted, thus ruling out Hypothesis 1. The second hypothesis does not seem likely because video 2 shows one of the adult females (PP) placing the stick tool in the ground with apparent deliberation, while she was walking past the bee nest and without inspecting the ground. The third hypothesis seems unlikely as a stick placed in the entry of the nest would be unlikely to stop honey badgers or forest elephants from excavating bee nests (see^27^ for data on predation on stingless bee nests by these animals). However, in the few days separating the clip 62 to 80, several animal species including humans walked and stopped near the bee nests and could have trampled the hole and masked the location of the nest entrance. The fourth hypothesis involves intentional signaling using objects, which has not yet been described in wild chimpanzees, although intentional communication has been documented in chimpanzees via vocalizations^92–94^. Underground insect nests are very difficult to locate and easily masked, and leaving sticks in them would make sense for chimpanzees to prevent trampling and mark them to aid themselves or even groupmates to locate the nests at a later date. The fact that the old female GL had planted the stick while the juvenile KI was present but not observing, then passed by the site without looking, and eventually discovered the planted stick, lent support to this hypothesis. However, we cannot rule out a lower-level that the chimpanzees learn to recognize nest entrances; at least some of them learn to use sticks to extract insect honey/product from nests; when they have finished extracting insect products or have probed nests without finding anything, individuals sometimes unintentionally leave their tools in the entrances. Others who see the sticks might then be attracted to manipulate them via stimulus enhancement (and perhaps sometimes by social enhancement, as in video 4 above). However, in all cases presented in the videos here, the females did not extract any honey either before or after inserting the stick so the function of “stimulus enhancement” of this hypothesis is at least consistent with our current hypothesis.
The cases involving fallen trees in Sebitoli and Ngogo (appendix 1) are not consistent with Gruber et al.’s^2^ assertion that “honey is easily accessible [from nests in fallen trees] and chimpanzees do not use sticks on such trees.” Chimpanzees at Ngogo and Kanyawara occasionally use stick tools for both easily and not easily accessible products : they have been observed to probe for Apis, carpenter bee products or for unknown material in the hollows of dead branches^26,34^; also Wrangham RW, pers. comm.; Krief pers. obs.).In addition, Sebitoli chimpanzees also often extract honey from Meliponula nests from fallen trees with tools.
Chimpanzees at Sonso, in the Budongo Forest Reserve, Uganda, eat honey from Apis nests^95^, but only use their arms and hands to obtain the honey; they have never been observed to use stick tools for anything^2^. Chimpanzees in forest fragments of Bulindi, adjacent to Budongo, also have not been observed to use tools to harvest honey from Apis nests, but do use tools to obtain honey from subterranean nests of stingless bees^23^. Given the relatively recent connections between Kibale and Budongo, it seems unlikely that the populations in these two forests have diverged genetically sufficiently to account for these differences in toolkits and extractive activities^2,4,96^. A genetic basis for differences in tool use between communities within Kibale or between Sonso and Bulindi is even less plausible.
Stingless bees occur in many tropical habitats, and their distribution includes a wide range of temperature regimes and altitudes^97^. They are present at Sebitoli and Kanyawara (this study), as well as at Ngogo (Watts and Mitani, pers. comm.). Occurrence of tool-assisted honey extraction at Sebitoli, but not the two other sites within Kibale, resembles the presence of honey tools at Bulindi vs. their absence in Budongo, and points to a possible social learning explanation. Tool-assisted consumption of Apis and Meliponula honey and other bee products has also been described at Bwindi 150 km to the south of Kibale, but the Meliponula nests were in tree cavities and the chimpanzees were not known to eat bee products from subterranean nests^24^. The main hypothesis is that stingless bee honey, either from arboreal of terrestrial nests, is not recognised as food by Ngogo and Kanyawara chimpanzees.
Why chimpanzees at Kanyawara and Ngogo do not excavate Meliponula nests remains unknown. We cannot rule out the possibility that Meliponula bees are not sufficiently common at Kanyawara and Ngogo to make exploiting them profitable, whereas exploitation is profitable at Sebitoli. Comparing chimpanzee diet, health and presence and abundancy of Meliponula nests in the four communities would potentially improve our understand of the factors affecting their exploitation. Forest composition varies considerably accross a small geographic scale in Kibale, as do chimpanzee diets and energetics^7–9,54,62,98,99^, and we cannot assume that a potential food type is equally profitable at all sites. However, the explanation may simply be that chimpanzees at Ngogo and Kanyawara do not know how to find the bee nests or do not recognize them as sources of food. Given that access to subterranean stingless bee honey requires tool use, and the probable importance of social learning in the acquisition of tool use repertoires (especially for the complex task of underground nest exploitation), it is at least equally plausible that the chimpanzees at Kanyawara and Ngogo do not engage in this activity because they do not know how to do so – i.e., the differences illustrate “cultural diversification” among nearby communities in a single continuous population (for Tai Forest, see Luncz and Boesch^100^, and Boesch and Boesch-Achermann^19^; for Mahale, see Nishida and Uehara^21^; for Gombe, Goodall^22^; and for Kahuzi-Biega, Yamagiwa et al.^16^. Interestingly, while Hicks et al.^1^ found abundant honey digging tools north of the Uele River in DR Congo, they failed to find a single one in the extensive forests to the south of that river despite the high density of chimpanzees. As with general variation in chimpanzee tool use^88^, differential use of stick tools for honey dipping does not follow geographic boundaries and probably has appeared independently in many distinct areas. Females who know how to perform the behavior likely sometimes immigrate into communities from which such knowledge is absent. They may also enter forest patches where chimpanzees were absent before because of anthropogenic disturbances like warfare, logging, agriculture, or mining. Whether immigrants would continue to engage in the behavior or members of their new communities would follow their example is an important question for future research. Recently, a knowledgeable female (KJ) who was among the most frequent tool users at Sebitoli transferred to Kanyawara. Whether she will introduce the behavior to her new community or/and continue to practice and transmit the behavior to her offspring remains to be seen.
Finally, our results highlight the fact that conserving non-humans is not just a question of conserving individuals. We must strive to preserve the diversity of non-human cultures as well, and their means of transmission, namely habitats and migration routes that allow individuals to exchange socially learned behavioral variants between social groups.
Below is the link to the electronic supplementary material.
Supplementary Material 1
Supplementary Material 2
Supplementary Material 3
Supplementary Material 4