Endangered Brazilian capuchins filmed using stones to crack nuts for first time

A camera trap in Brazil's Caatinga dry forest caught a moment no researcher had documented before: a blonde capuchin monkey lifting a stone overhead and bringing it down against a nut wedged into a natural anvil. The shell cracked. The monkey ate what was inside. For a species that had never been recorded doing this, it was not a small thing.

The blonde capuchin, Sapajus flavius, lives only in northeastern Brazil, scattered across fragments of Atlantic Forest and a more recently discovered population in the Caatinga, the semi-arid scrubland that stretches across much of the country's interior. Researchers had long known these monkeys used tools in one way—poking sticks into termite mounds to extract insects. Stone-assisted foraging was different, and the team led by Geovana C. B. Lima documented it for the first time in the species anywhere. The cameras were simply running when the monkeys went about their ordinary day. Nobody had set up nuts for them to crack. The behaviour was passively observed, not induced.

The technique itself was familiar from other capuchin populations in Brazil: a hammerstone brought down on a nut or seed pod resting on a flatter stone anvil. The researchers measured the stones involved. Hammers averaged around 476 grams—roughly the weight of a full soft drink can—striking anvils about 48 centimetres high with a working surface just over a square metre. Six instances were recorded in total. The researchers framed this carefully: a first documented case, not yet a settled account of how common or widespread the habit might be across the species. The food items were nuts and seed pods from native Caatinga plants, including the licuri palm and favela, a spiny native shrub. Both grow in the dry forest the monkeys inhabit, and neither yields easily to hand alone.

A follow-up study published in 2025, led by Maria Gabriella Rufino and João Pedro Souza-Alves, suggested the behaviour was more than a single lucky sighting. Working in Caatinga forest in the state of Alagoas, the researchers mapped 296 stone tool sites on the forest floor. Nearly 16 per cent of them showed clear signs of repeated reuse, with 72 separate reuse events logged over ten months of fieldwork. This second study did not observe monkeys mid-strike as often as the original paper did. Instead, it worked backwards from physical evidence: worn anvils, scattered hammerstones, cracked shells left where the animals had been feeding. The pattern of reuse tracked closely with the size of nearby trees, measured by trunk cross-section, which the researchers treated as a rough proxy for how much food a patch of forest could supply. Sites near larger, more productive trees were reused more often. The monkeys were not choosing anvils at random. They were returning to spots that kept paying off.

Sapajus flavius is listed as Endangered on the IUCN Red List. Its Atlantic Forest range has been fragmented for decades by sugarcane agriculture, and the Caatinga population was only formally documented in recent years, following local reports of sightings stretching back around four decades. A behaviour going unrecorded for this long does not necessarily mean it is new. It may simply reflect how little sustained observation this particular group of monkeys, in this particular habitat, had received before now. Stone tool use is not unique among Brazilian capuchins. Bearded capuchins at Serra da Capivara National Park have been studied for years using stones to crack nuts. Other Sapajus species, including the critically endangered yellow-breasted capuchin, have shown similar behaviour. What the new observation does is extend that pattern to a species where it had not previously been confirmed, in a habitat drier and more open than the environments where capuchin tool use is best studied.

The researchers note that a single study cannot establish how the behaviour arose in this population—whether through independent invention, social learning within the group, or some combination of both. Answering that would require longer-term observation across more sites. For a species already fighting habitat loss on two fronts, the finding adds a concrete detail to what is known about how blonde capuchins actually live day to day. It also carries practical weight for conservation planning. If particular trees and anvil sites are used repeatedly across generations, then protecting a patch of Caatinga for this species means protecting more than open ground and cover. It means protecting the specific trees and stones the monkeys have already learned to rely on.