Summary
Camera-trap footage from Senegal shows wild chimpanzees transferring tools during algae fishing, ant dipping and termite fishing. The study found that recipients used tools more after transfers, while initial possessors generally reduced their tool use.
Wild chimpanzees in Senegal sometimes allow, facilitate or actively arrange the transfer of tools to less experienced animals, creating opportunities to learn complex foraging skills. A camera-trap study at Dindefelo found that chimpanzees receiving tools increased their tool use immediately afterwards, while the animals that gave them up generally used tools less.
The study, published in Frontiers in Psychology, examined western chimpanzees (Pan troglodytes verus) living in the Dindefelo Community Nature Reserve. These chimpanzees are not provisioned and are not habituated to human observers, so the observations were collected indirectly with camera traps.
Between 2017 and 2025, the researchers obtained 14 hours 58 minutes of footage from 14 locations where chimpanzees used tools for algae fishing, ant dipping and termite fishing. They also collected 30 minutes of footage from 11 locations where chimpanzees carried baobab fruits and pounded them against fixed stone or wooden surfaces. Baobab pounding is considered proto-tool use because the chimpanzees manipulate the fruit but do not modify or carry the anvil.
In total, the footage contained 33 tool transfers and four baobab-fruit transfers. Most tool transfers involved females, and many occurred between mothers and offspring or other related animals. The most common events were recoveries, in which a chimpanzee took a tool that another had just abandoned, and tolerated takes, in which a recipient grabbed a tool and the initial possessor allowed it.
Tool transfers changed behaviour on both sides
To assess whether a transfer created a learning opportunity, the researchers compared each animal’s behaviour for up to 30 seconds before and after the event. They used linear mixed-effects models and excluded recoveries from the main analysis because the original tool possessor had not interacted with the recipient before leaving the tool.
After giving up a tool, initial possessors spent a smaller proportion of their visible time using tools and made fewer tool insertions per second. Many stopped using tools altogether, although some reused a tool found on the ground, made a new one or resumed using the tool after the recipient stopped.
Recipients showed the opposite pattern. After obtaining a tool, they spent more time using it, made more insertions per second and, in the models with longer observation windows, took more mouthfuls per second. In practical terms, the transfer shifted access to both the object and the foraging activity towards the less experienced animal.
The researchers interpret non-recovery transfers as a form of teaching or facilitation because the initial possessor often incurred a cost while the recipient gained a chance to practise. However, the process was usually initiated by the recipient rather than by a deliberate handover: all but one of the transfers began when the recipient sought or took the tool. The study recorded one clearly proactive transfer.
A mother and a juvenile provided closer examples of teaching
In the clearest example, an adult female and her infant were ant dipping at an army-ant nest. The mother inserted a tool into the nest, left it in place and brought the infant back after the infant began to move away. The infant then retrieved the tool, collected ants and continued dipping while the mother watched. This behaviour actively arranged the conditions for the infant to perform the technique.
A second case involved three baobab fruits, an adult female, a male juvenile and a male infant. The infant repeatedly tried to pound a fruit against a rocky surface while the adult held it. The juvenile slapped and stamped the rock, later moved the fruit with the infant’s hand and eventually cracked open another fruit while the infant watched. The infant ate a piece and then resumed trying to pound a separate fruit.
The authors suggest that the juvenile’s movements and sounds may have focused the infant’s attention on the pounding action. They also note that the juvenile could have been reacting to the infant’s movements or attempting to take the fruit, making the purpose of this interaction less certain than in the mother–infant ant-dipping case.
Together, the observations support a role for social learning in the spread of chimpanzee technological behaviours. They also suggest that learning may involve more than copying skilled adults: mothers can create practice opportunities, while juveniles may sometimes serve as models for infants. The findings extend this possibility to proto-tool use such as baobab pounding, although more observations are needed to determine how common these teaching patterns are across chimpanzee communities and foraging behaviours.