Deep inside Teixoneres Cave, beneath the hills near Moià in the province of Barcelona, the archaeological record has been quietly keeping a secret: not every small-animal bone buried in a Neanderthal-associated layer was put there by Neanderthals. A new taphonomic study of the avian and leporid remains from the cave’s sub-unit IIIb shows that this bone accumulation was assembled by at least three very different agents—Middle Palaeolithic humans, red foxes and Eurasian eagle owls—whose overlapping signatures had to be painstakingly disentangled one fossilized fragment at a time. The research, published in the journal Archaeological and Anthropological Sciences, concludes that carnivores were quantitatively the dominant contributors to the assemblage, but that Neanderthals unmistakably handled small prey and used fire at the site, adding fresh nuance to the long-running debate over how varied the Neanderthal diet really was.

Teixoneres Cave is one of the key Middle Palaeolithic localities of the northeastern Iberian Peninsula, and its story is one of alternating tenants. During Marine Isotope Stage 3, broadly between about 57,000 and 29,000 years ago, its galleries were visited in turn by Neanderthal foragers and by large carnivores, including hyaenas and bears, which used the cave as a den. The site has a well-defined chronology established through radiometric dating, and previous multidisciplinary work—from pollen records to tooth microwear—has painted a picture of short but repeated human visits to a resilient, resource-rich landscape rather than permanent settlement. Sub-unit IIIb, the focus of the new analysis, lies within Unit III and has already yielded faunal evidence consistent with brief Neanderthal occupations. That history makes it an ideal test case for a deceptively simple question that haunts every cave excavator: when you find heaps of rabbit and bird bones in a Palaeolithic layer, who actually put them there?

The study was carried out by Goizane Alonso of the Universidad de Burgos together with Anna Rufà, affiliated with the Interdisciplinary Center for Archaeology and the Evolution of Human Behaviour at the Universidade do Algarve and the PACEA laboratory of the University of Bordeaux, and Florent Rivals, Jordi Rosell and Ruth Blasco of the Institut Català de Paleoecologia Humana i Evolució Social (IPHES-CERCA) and the Universitat Rovira i Virgili. The team analysed the avian and leporid assemblage recovered from sub-unit IIIb—bones of leporids such as rabbits and hares and of a range of bird species. Such small animals are notoriously difficult to attribute to a single accumulator. They are prey to mammalian carnivores, birds of prey and humans alike, so their bones can enter a cave through hunting, scavenging, denning, pellet-casting or natural mortality, and each route leaves its own, partially overlapping, trace.

The analytical weapon of choice was taphonomy, the discipline that reconstructs everything that happens to an organism’s remains between death and discovery. Under magnification, each bone surface acts as a crime scene. Carnivores leave tooth pits, scores and punctures whose dimensions reflect the size of the biter, along with crenulated, scalloped breakage edges produced by gnawing; they also swallow bone, and gastric acid etches it into rounded, corroded surfaces that range from light digestion to near-total destruction. Owls, which swallow prey whole, regurgitate the skeleton in pellets, so assemblages accumulated by eagle owls display characteristic digestive etching, distinctive fragmentation patterns and an over-representation of robust elements such as wing and leg bones. Foxes gnaw and digest bone in their own recognisable way and, as denning animals, drag carcasses into sheltered cavities. Rodent incisors can add their parallel grooves, and roots, trampling and sediment pressure each leave further marks—all of which must be told apart from the modifications produced tens of thousands of years earlier.

Human activity writes a different script. Stone tool edges cut fine, V-shaped grooves—cut marks—typically concentrated at anatomical locations where skin, meat or tendons are removed, and hammerstone percussion leaves conchoidal scars and flaking on shafts where marrow has been extracted. Fire produces a graduated thermal signature: bone darkens to black as organic collagen chars, then turns grey and chalky white as temperatures climb toward calcination, accompanied by shrinkage, cracking and changes in crystalline structure. To interpret the Teixoneres material, the researchers drew on the rich body of actualistic and experimental reference data: bone assemblages from modern fox dens, pellets and nests of the Eurasian eagle owl, and experimentally butchered, cooked and burned rabbit remains. Comparing archaeological patterns against these known benchmarks is what allows a taphonomist to move from “something gnawed this bone” to naming, with reasonable confidence, the species that did the gnawing.

On that evidence, the verdict was clear on one point: non-human predators shaped this assemblage more than people did. The damage attributable to mammalian carnivores—identified through both mechanical tooth marks and digestive corrosion—proved quantitatively more significant than the anthropogenic evidence. The most likely mammalian contributor is the red fox (Vulpes vulpes), a species well known to shelter and den in caves and to carry in, gnaw and partially digest the carcasses of small prey. For the bird remains, the accumulated evidence points to the Eurasian eagle owl (Bubo bubo), a powerful nocturnal predator that roosts and nests at cave mouths and disgorges pellets dense with identifiable bones. Foxes also prey on birds, and the combination of gnawing, digestion and skeletal-part patterning in the assemblage is consistent with both animals using the cave—very possibly at moments when Neanderthals were elsewhere.

Yet the human signal, though numerically subordinate, is real and revealing. Cut marks on the small-prey bones and the presence of burnt specimens confirm that Neanderthals processed small animals in the cave and that fire played some role in their activities there. The burning pattern is the intriguing part: most of the charred bones appear to result from waste-cleaning activities—food refuse thrown onto or swept into hearths after the meat was gone—or from accidental exposure to open flames, rather than from deliberate roasting. That distinction is exactly what the team has explored experimentally. In earlier work, Alonso, Rufà and Blasco burned rabbit bones under controlled conditions and showed that the distribution and character of thermal alteration can help separate bones burned as cooking accidents from those burned as garbage. Applied to Teixoneres, the evidence leans toward refuse disposal and accidental burning—although, the authors caution, cooking activities cannot be definitively ruled out.

The findings feed directly into one of the liveliest debates in Palaeolithic archaeology: whether Neanderthals were narrow big-game specialists or broad-spectrum foragers. Under classic diet-breadth models from behavioural ecology, prey are ranked by the calories they return relative to the cost of capturing and processing them; small, fast animals such as rabbits and birds carry high handling costs, so heavy reliance on them is often read as a sign of demographic pressure or depleted large game. For decades, systematic exploitation of such small prey was considered a hallmark of anatomically modern humans. That picture has been steadily eroding: evidence from across Iberia and France now documents Neanderthal consumption of tortoises, rabbits and birds; pigeons and choughs were taken at sites in Gibraltar; wing bones at Fumane Cave in Italy bear traces of feather removal; and eagle talons from Krapina in Croatia have been interpreted as personal ornaments. Exploitation of the resources provided by small animals is today considered an important aspect of understanding the complex relationships between these hominins and their surrounding environment, and the hunting and processing practices of Middle Palaeolithic groups have attracted considerable attention in recent years.

Teixoneres adds a crucial caution to that narrative. A rich leporid and bird assemblage in a Neanderthal-bearing layer is not, by itself, proof of Neanderthal hunting. Foxes den; owls roost; both defecate, regurgitate and die in caves, and each activity can seed the sediment with small-prey bones that look, at first glance, remarkably like the residue of a hunter’s camp. The sub-unit IIIb assemblage is a palimpsest—a mixed contribution of non-human agents together with human activity—in which the same layer records, superimposed, the visits of foragers, the denning of foxes and the night hunts of eagle owls. Disentangling those threads required quantifying every category of surface modification, weighing the relative frequency of carnivore damage against cut marks and burning, and tracking which skeletal elements survived and which vanished. The result is a more honest—and arguably more interesting—portrait of the cave as a contested space shared, across time, by Neanderthals, foxes and eagle owls.

The authors are the first to stress that the questions raised by the study outweigh those it settles. The exploitation of small prey by Neanderthals and their use of fire are, they conclude, undoubtedly important topics that deserve further investigation, and future research will help clarify these issues and expand our understanding of the implications of small game exploitation for dietary diversification and the hunting strategies employed by Neanderthal groups. Until then, the bones of sub-unit IIIb stand as a reminder that the archaeological record is never authored by a single hand. The Neanderthals who sheltered at Teixoneres belonged to an ecosystem of rival predators, and the bone pile they left behind is, in equal measure, the record of their meals, the foxes’ larders and the owls’ night hunts.