2007-2009 | Piaui, Brazil

Bearded capuchin (Sapijus libidinosus) stone digging

Wild monkeys and apes use stone tools to break nuts. Exploring the global pattern of that simple fact made up a large part of the Primate Archaeology program that I ran at Oxford University. And one thing we learned is that while nuts aren’t everything, most primates ignore all the other things they could be hitting with stones, and stay narrowly focused on food.

Even us humans, confronted with something that needs to be split or cracked, often turn to a sharpened object like a saw or knife or axe or chisel, rather than blunt force. Which makes the bearded capuchins of Sierra da Capivara National Park (SCNP), in northeast Brazil, something of an anomaly.

Today’s post relies mostly on the work of one diligent researcher: Tiago Falótico. Tiago was a postdoctoral fellow on my Oxford project, but when we first met he was already one of the few people in the world dedicating their life to understanding wild capuchin monkey tool use. Armed with an eye for detail, tireless fieldwork, and an ever-growing array of recording equipment, he uncovered perhaps the most diverse stone tool repertoire of any non-human animal.

Learning more

From mid-2007 until 2009, Tiago followed two wild groups of bearded capuchins through the semi-arid caatinga scrub forests of SCNP. It’s difficult and hot work, with sheer sandstone cliffs towering over dense, thorn-riddled bush, and the monkeys able to move over and across them both with surprising speed. But Tiago couldn’t give up. He was gathering data for his doctoral thesis at the University of São Paulo, under the supervision of fellow capuchin expert, Eduardo Ottoni.

Capuchin monkeys are widespread and well known throughout Piaui, the state that contains SCNP. Capuchins are even represented in local rock art, painted prior to European invasion and genocide. That artistic and archaeological record contributed to the park’s listing as a UNESCO World Heritage site in 1991. Unfortunately, we currently don’t have any rock art images of capuchins using tools, although our excavations at SCNP have shown that capuchins were using stones there for at least the past 3000 years.

This rock art panel (photographed by Tiago) accurately shows an older monkey carrying a baby on its back, along with the hooked, prehensile tails found in capuchins and some related New World primates:

At the time of Tiago’s doctoral work, reports were emerging of the creativity and intelligence that these cat-sized monkeys employed when selecting and using stone hammers in the wild. But little was known of the diversity of their behaviour.

Captive capuchins—held in zoos or laboratories—were already recognised problem solvers, and the prominent US-Italian duo of Dorothy Fragaszy and Elisabetta Visalberghi had pioneered the use of field experiments with wild tool-using monkeys at another site in Piaui (called Fazenda Boa Vista) just a few years earlier. However, like with most primates, that stone tool use was directed at local foods that needed to be broken open, especially hard nuts and seeds. Tiago found the same behaviour at SCNP, but he found more.

Trapdoors and tubers

One of the most surprising things he saw was monkeys using a stone to dig into the sandy soil. Capuchins already have quite sturdy hands and nails that help them tear into and extract objects, and they are adept at smashing open fruits and branches directly on other surfaces. But here were monkeys finding and then using hand-held stones to directly pound the earth. Why?

It must have been a valuable resource that the monkeys were after—they spent minutes at a time, and considerable energy, hammering away at the same spot, sometimes ending up elbow-deep in a hole. That meant it was most likely food of some kind they were after. But the speed of monkey handy-work didn’t help. After carefully watching a monkey repeatedly drive the stone into the ground, and then shovel away loose dirt, over and over again, Tiago often had less than a second to try and see whatever it was that got shoved into a capuchin’s mouth.

Here’s one of Tiago’s videos from that fieldwork, in which the same adult male capuchin digs two separate holes. See how you go identifying the food it eats:

Helpfully, that video carries a title: ‘Digging a trapdoor spider with stone tool’. Did you spot the spiders? At the end of the video, Tiago points out the telltale silk-lined tunnel of a trapdoor spider—likely the same clue that the monkey followed when it began its excavation.

These spiders turned out to be a favourite digging target. Of the 1702 times that Tiago saw them digging with a stone tool, spiders were the prey just under 300 times. Despite the tricky issue of identifying a spider as it gets thrown into a monkey’s mouth, the main species seemed to be Actinopus or Margulla, the Brazilian tarantula.

Two other primary targets were less likely to escape a digging capuchin, and easier for Tiago to see. These were the root of the Ocotea or louro tree, which the monkeys would peel before eating the soft core, and tubers formed by the shrub Thiloa (or Combretum) glaucocarpa. The tubers are rich in carbohydrates, and again the monkeys would skin them before eating, using either their hands and teeth, or by smashing them with a second stone tool for good measure.

Cracking the planet

In hindsight, perhaps this activity should have been less of a surprise. After all, digging into the hard earth is roughly the same as striking a nut with a stone. It’s just that the stone is hand-sized—weighing around 100-150g—while the ‘nut’ is planet sized.

In both cases, the aim is to break open the outer crust, using a directed force that can’t be generated by the monkey’s muscles alone. That force is higher for the larger capuchin males, who performed over three-quarters of the stone-assisted digging that Tiago saw. However, almost all capuchins in the two studied groups dug at some point, and males and females had about the same success rate in terms of actually finding and getting food using the tool technique.

Underground storage organs (or USOs) aren’t just for monkeys, either. They may have played an important role in human evolution. Their starchy reserves and nutrients can sustain populations during times of other food scarcity, and even today, potatoes, cassava and sweet potato rank in the top ten human food crops. Yet the ease with which many of us now find these foods—laid out in a market, or handed to us as a fried side dish—can blind us to the difficulty of finding and retrieving these resources in a wider landscape.

Some animals—notably pigs—are adept at digging for these hidden treasures. But capuchins are not alone in using tools for the task. Wild eastern chimpanzees in Tanzania can use digging sticks to reach roots or tubers up to 25cm underground, concentrating this activity around their nesting sites. And although it doesn’t involve tubers, chimpanzees in the Republic of Congo use strong, straight sticks to puncture underground termite nests, before inserting more flexible probes to catch the insects. Elsewhere, wild chimpanzees simply use their hands to dig up a meal.

The SCNP capuchin digging tools also aren’t a species-wide trait. Recent work has shown that monkeys at the (now long-running) Brazilian site established by Fragaszy and Visalberghi don’t currently use any tools to get at roots. They do eat USOs, and they go through a time-consuming sequence of excavation, transport, cleaning and breaking to do so, but digging technology is absent.

Here’s an example of the non-tool use sequence, from the Fazenda Boa Vista (FBV) site:

The scientists suggest three possible reasons for the absence of digging tools at FBV, all of which reference the environment rather than any natural monkey abilities:

(a) the soil may be more cohesive (e.g., contain more silt and/or clay) in SCNP than in FBV, making digging with hands alone more difficult at the former location, and thus tools more useful; (b) the greater abundance of stones hard enough to be effective for digging provides more opportunities to excavate using stones in SCNP than in FBV, where most stones are friable sandstones; (c) in FBV these plants grow in sandy plains, where stones, even sandstones, are rare.

This idea—that the soil or stone characteristics are different enough between the two Piaui sites that tools are only needed or available at SCNP—holds merit. There are certainly a lot more loose quartz-rich stones at Serra da Capivara, eroding from nearby conglomerate cliffs. This may be a relatively clear-cut case where reliable tool/task co-occurrence has led to a learned behaviour specific to a particular group: a cultural trait, in other words.

We’ll dig into the somewhat contentious topic of animal cultures later; it’ll take more than a stone tool to get to the bottom of that rabbit hole. Instead, let’s leave the SCNP capuchins happily burrowing away, hunting a tarantula or a tasty tree root. And the next time you wander past a leaf-strewn patch of ground, have a quick look to see if there’s a hand-sized stone close by, and maybe join in the fun yourself.

Sources: Falótico, T. et al. (2017) Digging up food: excavation stone tool use by wild capuchin monkeys. Scientific Reports 7: 6278. || Falótico, T, et al. (2019) Three thousand years of wild capuchin stone tool use. Nature Ecology & Evolution 3: 1034–1038. || Fragaszy, D. et al. (2004) Wild Capuchin Monkeys (Cebus libidinosus) Use Anvils and Stone Pounding Tools. American Journal of Primatology 64: 359–366. || Hernandez-Aguilar, A. et al. (2007) Savanna chimpanzees use tools to harvest the underground storage organs of plants. PNAS 104: 19210–19213. || Sanz, C. & D. Morgan (2007) Chimpanzee tool technology in the Goualougo Triangle, Republic of Congo. Journal of Human Evolution 52: 420-433. || Truppa, V. et al. (2019) Manual skills for processing plant underground storage organs by wild bearded capuchins. American Journal of Physical Anthropology 170: 48-64.

Main image credit: Michael Haslam, Serra da Capivara National Park, Brazil || Rock art image credit: Tiago Falótico || Third image credit: Truppa et al. (2019) Fig. 3