Hominins voyaged to Sulawesi before one million years ago

New report of stone artifacts from Calio place human relatives in Wallacea more broadly and earlier than anyone knew.

Stone artifacts from Calio, Sulawesi. Images by Emma Watt and Mary-Anne Stone, CC-BY 4.0

In southern Sulawesi near the Walanae River is a small museum dedicated to display of local fossils of animals and stone tools. The few display cases of the Calio Paleolithic Museum only hint at the rich ancient heritage nearby.

This stretch of the Walanae has been an area for fossil exploration from at least the 1940s when the Dutch archaeologist Robert Van Heekeren came looking for evidence of ancient humans. Van Heekeren and subsequent Dutch-led expeditions found stone tools, some of which seemed to be Pleistocene in age—possibly old indeed. Yet the true antiquity of such artifacts was hard to establish with the tools of the era.

That changed in 2016 with the description of ancient artifacts from Talepu, in the Walanae valley. There, Gerrit van den Bergh and collaborators dug a deep trench and established that stone artifacts from the site date to at least 118,000 years ago. Sulawesi—like Flores, and Luzon—had been occupied by hominins long before modern people arrived in the region.

Now, our knowledge of the length of the Pleistocene occupation of Sulawesi has grown by an order of magnitude.

In 2019 archaeologists identified a stone artifact from a field just a half kilometer from the Calio Museum. They worked for several years, finding several additional artifacts in excavated contexts together with teeth from fossil animals that could be dated. This week the team reports its results in an article led by Budianto Hakim, establishing hominin presence on the island in association with a date of 1.26 ± 0.22  million years.

“After confirming the stone fragments were indeed artefacts, the field team celebrated in the usual way of the Bugis-Makasar people — with lots of ballo (palm wine).”—Budianto Hakim

The discovery is as simple as they come. The stone artifacts are simple flake tools, some retouched. There are no hominin bones or teeth, and the team does not report cut marks or other evidence of hominin activity on any animal bones. Still, the presence of stone tools here says much about how early human relatives in this region could adapt and disperse across natural barriers.

Into Wallacea

Sulawesi is just one of many large islands that comprise modern Indonesia, part of the great island arc that stretches from Malaysia to New Guinea. Shaped roughly like a Gothic letter “k”, the valley of the Walanae River is in the island’s southwest leg.

In the mid-nineteenth century, the naturalist Alfred Russel Wallace noted that Sulawesi, Flores, and other nearby islands have a unique combination of animal species. Large islands in the western reaches of the archipelago, like Java, Borneo, and Sumatra, share species with mainland Southeast Asia, including rhinoceros, tigers, and gibbons. Papua New Guinea and Australia have a very different set of animals, including a wide variety of marsupial mammals.

Sulawesi and Flores have few terrestrial mammals, with a handful related to Asian mammals, and—in the easternmost islands like Seram—some marsupial relatives of species from Papua or Australia. Wallace considered it probable that a deep sea channel had formed a barrier to the dispersal of animals from the Asian mainland into Lombok, Flores, and Sulawesi. Wallace recognized the Philippine Islands, with many endemic species, to be a separate situation, but his contemporary Alfred Henry Huxley thought these islands also were part of the same biogeographic history.

Locations of fossil hominin and archaeological sites in island Southeast Asia, mostly from Middle and Early Pleistocene times.

Today biologists call these and nearby islands Wallacea. Pleistocene sea level fluctuated to levels as much as 130 meters below today, connecting many land masses that are islands today. Wallace’s deep water channel really did exist, separating Sulawesi, Lombok, and Flores from Asia. This kept most terrestrial mammals from invading these islands, except for species that could swim or raft occasionally across the barrier.

A history of discovery

Archaeologists who investigated these islands starting in the early 1900s found artifacts made from river cobbles, choppers and flake tools, in contexts with fossils of Pleistocene animals that are now extinct. Van Heekeren’s explorations on Sulawesi after the Second World War contributed to this research. He was building on an earlier record of excavation of cave sites in southern Sulawesi, starting in the first decade of the 1900s, and a later Dutch expedition in 1970 would add substantially to the paleontological knowledge of the Walanae valley.

In the Philippines, parallel investigations were carried out by H. Otley Beyer, starting in the 1920s and 1930s. Beyer worked widely across the region including the Cagayan valley of northern Luzon, known today for the 700,000-year-old Kalinga paleontological site. He suspected that many of the simple choppers, chopping tools, and unretouched flakes that he found represented Middle Pleistocene habitation of the island.

The most prescient of these early investigations were carried out by the archaeologist and missionary Theodor Verhoeven. Verhoeven worked in the Soa Basin of Flores during the 1950s, finding stone tools and extinct species including the fossil elephant relative named Stegodon. He also initiated work in Liang Bua during the early 1960s.

Some of these finds were happening at the same time as major hominin fossil discoveries on nearby Java. By the 1960s, anthropologists were grouping most early Javan fossils into the species Homo erectus. Archaeologists working in the region did not assume that early humans like H. erectus were incapable of reaching Sulawesi, Flores, or the Philippines. Indeed they assumed that occasional crossing of Wallace’s Line must explain their findings. But none of the work unearthed clear evidence of H. erectus or any other early human relatives.

By the 1970s and 1980s, many archaeologists based in North America and Europe took on a hardened attitude of skepticism about the dispersal abilities of H. erectus and other Pleistocene hominins. This attitude epitomizes the “Culture Last” assumption that I wrote about last month. Following this assumption, ancestral humans should have had the same limitations as most large Asian mammals. They couldn’t swim across the straits that separated Wallacea from Java and Borneo. They had no boats until modern humans appeared in the region, sometime after 70,000 years ago. Following this way of thinking, for hominins like Homo erectus, Flores and Sulawesi may as well have been the Moon.

The discoveries from Liang Bua, Flores—including the fossil bones attributed to H. floresiensis—shifted perceptions by showing that other hominins were likely present in Wallacea. At first, the true age of the Liang Bua evidence was not recognized, with the most complete finds argued to be less than 35,000 years old. That would change. More important, teams of archaeologists and paleontologists began to reassess evidence from a wide range of sites across the region, from Luzon and Sulawesi as well as Flores.

From Liang Bua and the Soa Basin, to Callao Cave and Kalinga, to Calio and Talepu, recent archaeologists are revisiting regions where discoveries were made and published more than sixty years ago. They keep finding that there is vastly more to find, particularly with today’s improved geochronological methods. These current discoveries are emerging because of the interest and investment three decades ago by researchers who wanted to know if Verhoeven’s discoveries could be replicated or exceeded. Those investments keep paying off for the entire field.

The importance of Calio

I’ve seen a few people asking if Sulawesi is going to produce another species of small hominins, similar to what is observed of H. floresiensis and suspected of H. luzonensis. I doubt it.

Sulawesi is larger in area than Flores or Luzon, and it sustained a more diverse community of large mammals, including two extant species of anoa and a possible fossil ancestor; a dwarf stegodont, the extant babirusa, warty pig and an extinct pig, Celebochoerus. The continued survival of most of these species after the arrival of modern humans may suggest that they coadaptated within a faunal community that included hominin predators more formidable than Homo floresiensis.

The greater diversity of large mammals may also owe something to geography. Sulawesi also shares a much longer coastline with the Strait of Makassar, looking across at Borneo. The crossing is longer than the Bali-Lombok crossing toward Flores to the South, but with prevailing currents someone who set out toward Sulawesi may be less likely to miss the island entirely. Island-hopping from Mindanao to the north may not be out of the question. This geography makes it plausible that hominins entered Sulawesi from different geographic sources, repeatedly.

The geography is not new, of course, and it is why I still expect Denisovans likely occupied the island. I interpret later Middle Pleistocene or Late Pleistocene record, including Talepu, with that idea in mind. But Denisovans did not exist a million years ago. The Denisovans, Neanderthals, and African Idaltu branched from each other around 700,000 years ago, making the earliest possible Denisovan presence in the region much later.

In fact, the date that Hakim and coworkers present for the Calio material has an upper error bar stretching to 1.48 million years. If these tools are actually close to this upper extreme, they are within error of the oldest known hominin fossils from Java. The geochronology of Sangiran and Mojokerto has always been disputable, but recent results suggest that the first appearance date of early H. erectus on Java is less than 1.5 million years ago.

Based on those numbers, Calio could be even earlier than H. erectus on Java.

Was Java was first colonized from the east? Although it would be far from the first time a half-million-year error range was the hanging point for a theory, it’s a thin hook to support one. I wouldn’t go there.

Instead here’s what I would suggest as a shift in perspective. Hominins in island Southeast Asia had behavioral familiarity with shoreline resources and crossing deep rivers from the start—a set of strategies that Carlos Neto de Carvalho and collaborators have recently called coasteering. The early hominins in this region found ways again and again to exploit rich coastal ecosystems including mangrove forest, tidal pools, riversides, and near-shore islands. It is not possible to develop this array of foraging strategies without becoming familiar with swimming, interacting with floating materials like logs and vegetation mats, and constructing simple watercraft.

Sea level fluctuations in island Southeast Asia sometimes happened on decadal scales, creating new near-shore islands from once-connected landmasses. The dispersal potential of hominin groups with the ability to employ simple watercraft was greater at every water crossing compared to land-limited groups. Geography selected for water familiarity throughout this region, if not long before.

I expect that Sulawesi saw dozens or hundreds of hominin arrivals once they arrived in the region. Every newcomer would have had a crack at the island at some point, from early H. erectus to Denisovans, to the first modern people, to Mesolithic and Neolithic immigrants, to historic times. The question in my mind is whether, during some long span of the Pleistocene, these newcomers had enough numbers to constrain the genetic and cultural evolution of the Sulawesi hominins. Or whether instead Sulawesi, as is possible on Flores and Luzon, may have developed a population with sustained numbers and adaptations that differed from new arrivals.

It will take many more discoveries to find out.

References

Anonymous. (2025). Stone tools suggest that hominins arrived on Indonesian island much earlier than thought. Nature. https://doi.org/10.1038/d41586-025-02386-0

de Carvalho, C. N., Cunha, P. P., Belo, J., Muñiz, F., Baucon, A., Cachão, M., Figueiredo, S., Buylaert, J.-P., Galán, J. M., Belaústegui, Z., Cáceres, L. M., Zhang, Y., Ferreira, C., Rodríguez-Vidal, J., Finlayson, S., Finlayson, G., & Finlayson, C. (2025). Neanderthal coasteering and the first Portuguese hominin tracksites. Scientific Reports, 15(1), 23785. https://doi.org/10.1038/s41598-025-06089-4

Hakim, B., Wibowo, U. P., van den Bergh, G. D., Yurnaldi, D., Joannes-Boyau, R., Duli, A., Suryatman, Sardi, R., Nurani, I. A., Puspaningrum, M. R., Mahmud, I., Haris, A., Anshari, K. A., Saiful, A. M., Arman Bungaran, P., Adhityatama, S., Muhammad, P. H., Akib, A., Somba, N., … Brumm, A. (2025). Hominins on Sulawesi during the Early Pleistocene. Nature, 1–6. https://doi.org/10.1038/s41586-025-09348-6

Matsu’ura, S., Kondo, M., Danhara, T., Sakata, S., Iwano, H., Hirata, T., Kurniawan, I., Setiyabudi, E., Takeshita, Y., Hyodo, M., Kitaba, I., Sudo, M., Danhara, Y., & Aziz, F. (2020). Age control of the first appearance datum for Javanese Homo erectus in the Sangiran area. Science, 367(6474), 210–214. https://doi.org/10.1126/science.aau8556