We show that cultural transmission is essential for orangs to learn basic subsistence information in the wild. The repertoire of cultural information possessed by apes is likely to be far more expansive than social customs and highly technical skills, including simple info about 'what to eat'.

Even when orangutans were presented with over 148,000 opportunities to interact with different food items over long-term development, orangs only developed adult-like diets (at least 90% adult repertoire) by adulthood if they also engaged in social learning.

Every single parameter of our ABM was estimated from wild data. When simulated orangs could engage in forms of social learning seen in the wild, the timings of diet development matched those observed in wild individuals (validating that our ABM captured wild orangs' learning from real-world inputs).

Leveraging over a decade of behavioral data collected in the wild, we constructed an agent-based simulation (ABM) of orangutan diet development between birth and adulthood (~15 years), focusing on individual exploration + social influences on exploration, e.g. exposure, enhancement and peering.

To answer these questions, we investigated whether social learning has important, long-term effects on the rate and outcomes of diet development in wild Sumatran orangutans. Moreover, we circumvent these methodological difficulties through running biologically relevant experiments ‘in-sillico’.

Whilst many animal species exhibit the capacity to socially learn, it is unclear whether animals’ social learning similarly leads to the development of knowledge repertoires which are broader than animals could construct individually.

Many thanks to all researchers and field assistants involved with data collection over the years at Bossou, and in curating the Bossou video archive. (📸 credit T Matsuzawa).

Further research is required to precisely characterize how these structural features of chimpanzee actions are generated, including further categorizing the predictable 'chunks' of actions chimpanzees use in tool-use sequence. (Photo Credit: TMatsuzawa)

Half of adults produced NADs at distances which could not be reproduced by non-hierarchical, null models. MI decay dynamics often reflected those predicted by 'composite' hierarchical action organization, providing objective support for the organization of actions through hierarchical 'chunks'.

We collected data from chimpanzees at Bossou cracking nuts using stone tools (>8000 actions, cracking > 300 individual nuts). We used this data to run statistical models which can objectively identify signatures of NADs and possible hierarchical organization ('chunking'). (Photo Credit: T.Matsuzawa)

Tool-use behaviours have previously been hypothesized to involve hierarchical organization, but it has been hard to evidence this objectively (humans are biased to infer 'chunks'). This may also suggest that these behaviours could contain NADs, though never investigated. (Photo Credit: D.Biro)

We used video footage of wild chimpanzees at Bossou cracking nuts (>1600 nuts) across a 17 year period, to quantify whether aging influences tool use. We found reductions in engagement with aging, and also for some individuals, reductions in efficiency.

We studied how aging influences a naturally-occurring tool-use behavior of wild west-African chimpanzees (nut cracking). Tool use draws on many physical & cognitive traits which can senesce in captive primates (see manuscript); yet, senescence of tool use has never been studied.

Given their close phylogenetic relationship with humans, apes are extremely useful models for understanding the human aging process. Yet, given their long lives, it is challenging to observe how aging influences individuals over time...