New Study Reveals Homo floresiensis Was a Scavenger, Not a Hunter

The classic picture of Homo floresiensis — a tiny, big-game-hunting, fire-using human relative — has just been redrawn. A new study published in Science Advances on July 3, 2026, led by Dr. Elizabeth Grace Veatch of the Smithsonian National Museum of Natural History, presents evidence that the 'hobbits' of Flores were passive scavengers who survived on raw scraps left by Komodo dragons . The research not only rewrites their daily life but also raises fundamental questions about where they belong on the human family tree.

The Komodo Dragon Experiment That Changed Everything

The central evidence came from an ingenious experiment. The research team fed a goat carcass to a Komodo dragon named Rinca at Zoo Atlanta. After the dragon finished feeding, researchers collected and 3D-scanned the 72 remaining bones, creating a reference library of Komodo tooth-mark characteristics. These marks — shallow, wide grooves often ending in a fan of striations from the dragon’s serrated teeth — became the key to interpreting ancient bones .

Using high-powered microscopy, the team compared these known Komodo tooth marks against marks on Stegodon (a dwarf elephant relative) bones excavated from Liang Bua cave, the same site where H. floresiensis was discovered in 2003. The result was decisive. Veatch said she was surprised that 'almost all of them turned out to be Komodo dragon tooth scores' .

Crucially, the distribution of marks told a story of who ate first. Komodo dragon tooth marks clustered on the meatiest body parts — shoulders and hips. In contrast, stone-tool cut marks made by the hobbits were found only on the less choice parts, such as feet and ribs. This spatial pattern is a classic signature of scavenging: the apex predator eats first, and smaller scavengers move in for the leftovers .

No Fire in the Cave

For years, charred bones found in Liang Bua were thought to be evidence that H. floresiensis controlled fire. The new study systematically dismantles that case. The team analyzed approximately 4,500 rodent bones deposited in the cave by roosting owls over thousands of years. Not a single bone showed charring or burning. If hearths had been built in the cave, the team argued, the underlying bones would show clear thermal alteration .

No Stegodon bones showed char marks either . Earlier burned bones once attributed to H. floresiensis were re-dated to younger cave layers occupied by Homo sapiens — around 46,000 years ago, long after the hobbits and Stegodon had vanished. The fire evidence was a case of mistaken stratigraphy .

What They Actually Ate

Without fire for cooking, the hobbits’ diet was necessarily different from what had been assumed. While tooth-wear analysis has long indicated a tough, fibrous diet requiring forceful chewing — suggesting they consumed plants — the meat component of their diet was raw and scavenged . Komodo venom proteins, which can be dangerous, are broken down by stomach enzymes, so scavenging posed no poisoning risk to the hobbits . They likely also supplemented their diet with small animals, insects, and local plant foods .

Living Alongside Dragons

The study clarifies the hobbits’ place in the Flores ecosystem. Komodo dragons (Varanus komodoensis) were the undisputed apex predators, using their venomous bite to take down Stegodon . H. floresiensis was just one member of a broader scavenging guild on the island, alongside vultures, giant storks, and crows, all feeding on Stegodon carcasses .

Despite the dragons’ dominance, the hobbits were not typically preyed upon. Modern Komodo dragons rarely attack humans unprovoked, and the authors suggest that living in groups with wariness was likely sufficient protection .

A Simpler Evolutionary Path

Perhaps the most profound implication of the study is what it suggests about H. floresiensis's place in the hominin family tree. The absence of big-game hunting and controlled fire indicates a significantly less advanced behavioral toolkit than previously assumed. This directly undercuts the idea that they were a dwarfed form of Homo erectus, which is known to have used fire and hunted large game .

Instead, the study reinforces a minority view: that H. floresiensis may descend from a more primitive hominin — possibly Homo habilis-like or even Australopithecus-like — that reached Flores more than 1 million years ago, before the evolution of advanced behaviors in later Homo species .

Lead author Veatch stated: 'A more simplistic behavioral repertoire may indicate an ancestry that separated from the Homo lineage prior to these more advanced behavioral adaptations' . Dr. Chris Stringer of the Natural History Museum in London, who was not an author of the study, noted that it 'reinforces the minority view that floresiensis does not really belong in the genus Homo and should be redesignated' .

Caveats and Ongoing Questions

Not all experts are fully convinced. Paleoecologist Kay Behrensmeyer of the Smithsonian said it is 'very hard to distinguish hunting from scavenging based on bone modification features' and is not yet fully sold on a 'scavenging only' hypothesis — though she did call the team’s case for identifying Komodo versus tool marks convincing .

The study is also specific in scope. It does not rule out the possibility that H. floresiensis occasionally caught small game for themselves. What it does overturn, definitively and with new experimental evidence, is the case for large-game hunting and controlled fire use .