Independent research libraryEvidence first · sources openHow claims are graded
SCRIBEThe research behind the world’s biggest questions
SCR-085 · Human Origins · WALLACEA · ROTI ISLAND · 30,000-YEAR OCCUPATION

New Archaeology Strengthens Roti Island’s 30,000-Year Human Record

Share this investigation

A new open-access study used ground-penetrating radar, augering and laboratory analyses across six caves and rockshelters on Roti Island, Indonesia. At Lua Meko, a newly dated marine-shell sample returned 30,257–29,657 calibrated years before present. Cultural material extends deeper, but those deeper deposits are not yet directly dated into Marine Isotope Stage 3.

Last reviewed 8 September 20264 sourcesVerdict version 1
Evidence rule: KNOWN: a new marine-shell sample from Lua Meko on Roti dates to 30,257–29,657 cal BP, and the survey recovered cultural material including stone artefacts and marine remains within the deposit. METHOD: researchers combined GPR, augering, artefact analysis, radiocarbon dating, magnetic susceptibility and sediment geochemistry across six sites. INTERPRETATION: deeper cultural deposits may preserve an occupation record older than the dated shell. OPEN: those deeper deposits still need direct dating, and the marine-shell calibration lacks a local ΔR correction because none is currently available for Wallacea.

What the new study actually did

The research was designed less as a conventional excavation than as a rapid archaeological triage system. Across six Roti caves and rockshelters, the team combined ground-penetrating radar with narrow auger sampling, microscopic artefact identification, radiocarbon dating, magnetic susceptibility and sediment chemistry. The aim was to identify where cultural deposits occur, how deep they run and which sites deserve larger future excavations.

Lua Meko produced the key Pleistocene result

At Lua Meko in north-central Roti, radar helped guide an auger into a low-reflection part of the rockshelter deposit. The core reached about 130 centimetres before an impermeable barrier. Cultural indicators included marine resources, stone artefacts and enhanced magnetic susceptibility consistent with heating, with a marked archaeological signal appearing around 100 centimetres depth.

A new shell date reaches about 30,000 years ago

A marine-shell sample from 60 centimetres depth returned a calibrated range of 30,257–29,657 years before present at the 95.4% interval. Earlier work at Lua Meko had already shown late-Pleistocene occupation; the new result strengthens that chronology with a fresh sample collected during the rapid-prospection programme.

The deeper deposit is promising — but it is not yet a date

The most important restraint is also the most interesting lead. Cultural material extends below the level associated with the roughly 30,000-year shell date. The authors therefore say Lua Meko has the potential to extend Roti’s occupation record into Marine Isotope Stage 3, which spans roughly 57,000–30,000 years ago. Potential is not confirmation: the deeper cultural deposit itself has not yet been directly dated into that interval.

The marine calibration caveat matters

The paper states that its marine-shell ages were calibrated without a local ΔR value because such a correction is not currently available for Wallacea. ΔR accounts for local variation in the marine reservoir effect. That does not erase the radiocarbon evidence, but it means the precise calendar placement of marine-shell dates carries an additional local uncertainty that should travel with the headline.

The six-site survey also found reasons not to excavate everywhere

Three sampled sites contained deep sedimentary sequences but proved very young and culturally sparse. At Pia Hudale, the team found a complex deep sequence but lacked suitable old charcoal for direct dating. At Lua Nggalana, marine shell from the lowest auger sample dated to 9,419–9,096 cal BP. These mixed results are part of the method’s value: it can identify promising targets and rule out expensive dead ends before large excavations begin.

Why this matters beyond one island

Wallacea is central to questions about early maritime movement between Sunda and Sahul, yet many of its more than 1,600 islands remain archaeologically under-explored. Other evidence from the wider region already demonstrates sophisticated seafaring, pelagic fishing, ancient surgery and very early figurative art. Roti adds another piece to that landscape, but this paper is primarily a demonstration of how to find and prioritise the buried record without destroying large areas of it first.

What would strengthen the case for older occupation

The clearest next test is direct dating of the deeper cultural horizon at Lua Meko using material demonstrably associated with human activity. Where radiocarbon is unavailable or approaches its practical limits, the authors suggest approaches such as optical dating from carefully recovered sediment cores. A securely dated deeper layer would move the story from ‘potentially older than 30,000 years’ to a defensible earlier occupation chronology.

The evidence-led conclusion

The new Roti work strengthens the case that people were using Lua Meko around 30,000 years ago and shows that more archaeological material lies deeper in the deposit. It does not yet prove a substantially earlier occupation. The strongest result is therefore twofold: a fresh late-Pleistocene date, and a clearly defined deeper target that future dating can either confirm or reject.

Research record

Norman et al. — Rapid minimally invasive subsurface archaeological site prospection on Roti Island, IndonesiaQuaternary International 775 · 1 September 2026 · open access CC BY 4.0 · primary study · open sourceNorman et al. — An early colonisation pathway into northwest Australia 70–60,000 years agoQuaternary Science Reviews 180 · regional Sahul migration context · open sourceO’Connor, Ono & Clarkson — Pelagic fishing at 42,000 years before the present and the maritime skills of modern humansScience 334 · Wallacean maritime-behaviour context · open sourceOktaviana et al. — Narrative cave art in Indonesia by 51,200 years agoNature 631 · wider Island Southeast Asian behavioural context · open source
SCRIBE VERDICT

New Archaeology Strengthens Roti Island’s 30,000-Year Human Record

Direct answer: The new study strengthens evidence that people were using Lua Meko on Roti Island around 30,000 years ago. A marine-shell sample dates to 30,257–29,657 cal BP, while cultural material extends deeper in the deposit. Those deeper deposits may preserve an older record, but they have not yet been directly dated into Marine Isotope Stage 3. The paper also notes that local marine ΔR correction data are unavailable for Wallacea, adding uncertainty to the exact calendar placement of shell dates.

KNOWN: a newly analysed marine-shell sample from Lua Meko dates to 30,257–29,657 cal BP and archaeological material occurs within the deposit. HISTORY: earlier work had already established late-Pleistocene occupation at Lua Meko; the new programme adds a fresh date and maps previously uncharacterised deeper deposits. INTERPRETATION: cultural material below the dated level could extend the Roti occupation record into Marine Isotope Stage 3. UNKNOWN: the direct age of those deeper cultural deposits and the local marine-reservoir ΔR correction for Roti/Wallacea. NOT DEMONSTRATED: a securely dated human occupation on Roti substantially earlier than 30,000 years ago, or that depth alone proves greater antiquity.

Share this investigation

Open research · SCR-085

Add to this investigation

Found a stronger source, correction, counter-evidence or new lead? Send it to SCRIBE. Contributions are reviewed against the evidence policy and do not change a published verdict automatically.

The investigation ID and page address are prefilled. Public credit is optional.