Reconstruction
Liang Bua, a limestone cave in western Flores, became one of paleoanthropology’s most remarkable sites when excavators uncovered a partial hominin skeleton there in 2003. Announced as Homo floresiensis in 2004, the individual known as LB1 combined a stature of roughly one meter with a very small brain and a distinctive mixture of anatomical traits. Additional bones and teeth showed that LB1 was not alone. Although the species’ ancestry remains debated, the fossils are widely treated as evidence of a separate, long-isolated hominin population rather than simply unusually small modern humans.
Revised stratigraphic work substantially changed the site’s chronology. The skeletal remains and their containing deposits date to approximately 100,000–60,000 years ago, while stone artifacts attributed to Homo floresiensis occur in deposits extending from roughly 190,000 to 50,000 years ago. The card’s 50,000 BCE endpoint therefore fits the youngest associated artifact horizon, not the youngest securely dated fossils. Modern human activity becomes visible at Liang Bua around 46,000 years ago, and direct overlap between the two populations has not been demonstrated.
Food evidence is narrower than early reconstructions suggested. Thousands of bones from the dwarf proboscidean Stegodon florensis insularis occur with stone artifacts, and a small number bear hominin cut marks. A systematic taphonomic study published in 2026 found that Komodo dragon tooth marks were concentrated on high-value portions, whereas hominin marks occurred mainly on lower-utility elements. This supports secondary access to carcasses—probably passive scavenging—rather than organized hunting of Stegodon. The same study found no evidence of intentional fire use in layers associated exclusively with Homo floresiensis. The most defensible reconstruction is consequently that at least some Stegodon tissue was cut from carcass remains and eaten raw. Giant rats, birds, reptiles and other animals also inhabited or entered the cave, but their presence alone does not establish that Homo floresiensis ate them.
Liang Bua’s food web may have become increasingly stressed near the population’s disappearance. A 2025 reconstruction indicates that mean annual rainfall declined markedly between about 76,000 and 61,000 years ago, followed by exceptional summer dryness from approximately 61,000 to 55,000 years ago. Stegodon remains declined at the same time. This correlation makes reduced water and carrion availability a plausible contributor to abandonment of the cave, but it does not prove the species became extinct then or identify a single cause.
The site has no demonstrable culinary continuity with food today. Its importance lies instead in showing how an island hominin obtained food within a restricted ecosystem—and how careful analysis can overturn vivid but weakly supported images of hunters cooking dwarf elephants over fires.
Historical context
Sea levels repeatedly stood below those of today, joining many western Indonesian islands to mainland Southeast Asia as the exposed Sunda Shelf. Flores nevertheless remained within Wallacea, separated from Sunda and Sahul by deep-water channels. Hominins had reached Flores hundreds of thousands of years earlier, although whether the crossings were deliberate remains unknown. Elsewhere, Homo sapiens lived in Africa and was dispersing into Eurasia, while Neanderthals occupied western Eurasia and Denisovan-related populations were present in Asia. By roughly 50,000 years ago, modern humans had crossed maritime Southeast Asia and reached Sahul.
Evidence
Archaeological
StrongControlled excavations at Liang Bua recovered Homo floresiensis skeletal remains, stone artifacts and faunal remains within a documented cave stratigraphy.
Zooarchaeological
StrongSystematic analysis of Stegodon remains documents both Komodo dragon tooth marks and less frequent hominin cut marks, permitting reconstruction of the likely order of carcass access.
Food identification
ProbableCut-marked Stegodon bones support hominin access to proboscidean tissue, but the amount and specific tissues consumed cannot be established. Other fauna found in the cave should not automatically be classified as food.
Dating
StrongRevised stratigraphic and radiometric analyses place the skeletal remains at approximately 100,000–60,000 years ago and attributed stone artifacts at approximately 190,000–50,000 years ago.
Preparation method
ModerateThe absence of intentional burning in Homo floresiensis-associated units, combined with cut marks, supports raw consumption as a scholarly inference. No meal or preparation event was directly observed.
Geographic attribution
StrongThe fossils and associated archaeological deposits were excavated at Liang Bua Cave in western Flores, Indonesia.
Historical interpretation
ModerateProgressive aridification and declining Stegodon abundance plausibly contributed to abandonment of Liang Bua, but neither extinction causation nor interaction with Homo sapiens has been demonstrated.
Visual reconstruction
LimitedSkeletal anatomy supports reconstruction of small body size, but clothing, group organization, food sharing and the appearance of daily subsistence activities remain unknown.
Sources
- 1.Peter Brown, Thomas Sutikna, Michael J. Morwood, et al. (2004). A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia. Nature 431, 1055–1061. doi:10.1038/nature02999Primary archaeology
- 2.Thomas Sutikna et al. (2016). Revised stratigraphy and chronology for Homo floresiensis at Liang Bua in Indonesia. Nature 532, 366–369. doi:10.1038/nature17179Scientific literature
- 3.Thomas Sutikna, Matthew W. Tocheri, J. Tyler Faith, et al. (2018). The spatio-temporal distribution of archaeological and faunal finds at Liang Bua (Flores, Indonesia) in light of the revised chronology for Homo floresiensis. Journal of Human Evolution 124, 52–74. doi:10.1016/j.jhevol.2018.07.001Scientific literature
- 4.E. Grace Veatch, Nico Alamsyah, Michael Pante, et al. (2026). Taphonomic analysis at Liang Bua reveals the behavioral and technological capabilities of Homo floresiensis. Science Advances 12(27), eaeb7219. doi:10.1126/sciadv.aeb7219Scientific literature
- 5.Michael K. Gagan et al. (2025). Onset of summer aridification and the decline of Homo floresiensis at Liang Bua 61,000 years ago. Communications Earth & Environment 6, 992. doi:10.1038/s43247-025-02961-3Scientific literature
Limitations
- moderatedate end
The approximately 50,000 BCE endpoint corresponds to the youngest stone artifacts attributed to Homo floresiensis, not to securely dated skeletal remains. Revised dating places the fossils at approximately 100,000–60,000 years ago, with a 2025 refinement concentrating documented remains around 87,000–61,000 years ago.
