A new discovery from Bianfu Cave in southwestern China is transforming what we know about the Denisovans. Ancient proteins recovered from human teeth and bones reveal that Denisovans occupied this mountainous region between approximately 167,000 and 134,000 years ago. They hunted large animals, butchered carcasses, extracted marrow, produced bone tools and manufactured a surprisingly straightforward stone-tool industry.
Located in Yunnan Province, Bianfu Cave fills a major geographic gap between the Denisovans of Siberia and Tibet and a mysterious Denisovan-like molar found at Cobra Cave in Laos. The two sites are only around 770 kilometres apart, and their dates overlap almost perfectly. Together, they point toward a southern Denisovan population corridor extending through southwestern China and mainland Southeast Asia.
The discovery also raises a much larger question. If fragmentary teeth and bones from Bianfu could only be recognized as Denisovan through ancient proteins, how many other Denisovans remain unidentified in Chinese museum collections? Fossils from Hualongdong, Tongzi, Panxian Dadong and Xujiayao possess unusual dental characteristics that may deserve renewed comparison with securely identified Denisovan remains.
Even older protein evidence from approximately 400,000-year-old fossils traditionally assigned to East Asian Homo erectus could indicate relationships with the later Denisovan lineage. This does not prove that every archaic Chinese fossil was Denisovan, but it challenges the old picture of Denisovans as a rare Siberian population known only from a few fragments. They may instead have represented one of the most geographically widespread and successful human populations in Pleistocene Asia.
In this video, we examine the Denisovan remains from Bianfu Cave, their tools and hunting behavior, the connection with Cobra Cave in Laos, and the possibility that some of China’s most mysterious fossils have been hiding under the wrong names.
TIMESTAMPS
00:00 Denisovans Were Hiding in Plain Sight 00:55 The Discovery at Bianfu Cave 02:10 Teeth and Bones Identified Through Proteins 03:42 The Protein Link with East Asian Homo erectus 05:10 Surprisingly Simple Denisovan Stone Tools 06:35 Skilled Hunters of Southwestern China 08:02 A Denisovan Population Spanning Asia 09:22 The Connection with Cobra Cave in Laos 10:43 China’s Unidentified Denisovan Teeth 12:21 Hualongdong, Tongzi and Panxian Dadong 13:43 Were Some “Homo erectus” Early Denisovans? 15:05 Rewriting the Human History of Asia
Research discussed includes the 2026 Bianfu Cave studies published in *Nature*, research on the Denisovan-like molar from Tam Ngu Hao 2 in Laos, and recent paleoproteomic work examining Middle Pleistocene human fossils from eastern Asia.
For more than a century, the story of human evolution in Eastern Asia
was dominated by one familiar character, Homo erectus.
Almost every archaic skull, jaw, or tooth discovered between China and Southeast
Asia seemed to be interpreted through that single species.
The famous fossils from Jokudian became Peking Man,
the remains from Java became Java Man, and many younger and more unusual
fossils were treated as regional descendants.
Denisovans barely entered the story at all.
When their existence was first revealed in 2010,
it was not through a recognizable skull or skeleton,
but through DNA extracted from a tiny finger bone found in a Siberian cave.
Yet a remarkable discovery from southwestern China now suggests
that Denisovans were neither rare nor confined to the frozen margins of
northern Asia. At Bianfu Cave, researchers have
identified Denisovan teeth and bones alongside thousands of stone artifacts
and the remains of hunted animals. The discovery does more than place
Denisovans in another location. It gives us one of the clearest pictures
yet of how they lived and raises a much larger question.
If these anonymous fragments turned out to be Denisovan,
how many other fossils in China have been hiding under the wrong names?
Bianfu Cave, also known as Bianfudong, lies in Heqing County in Yunnan Province
at an elevation of approximately 2,370 meters or 7,775 feet.
The cave occupies a mountainous region of southwestern China
close to the eastern edge of the Tibetan Plateau and the Great River corridors
leading towards Southeast Asia. Excavations began after cave sediments
were exposed along the wall of an industrial quarry.
What archaeologists discovered was not simply an isolated tooth
or a few displaced bones. Bianfu contained a long cultural sequence
extending from approximately 190,000 to 70,000 years ago,
including more than 1,300 stone artifacts and tens of thousands of animal fossils.
Among this material were several fragmentary human remains,
teeth, pieces of the skull and a fragment of an arm bone.
None of them looked dramatic enough to transform our understanding of Asia.
Without biomolecular testing,
they might have remained unidentified fragments in a museum collection.
But proteins preserved within the enamel and bone
revealed that at least five of the tested remains belonged to Denisovans.
The human fossils came from two archaeological layers
separated by tens of thousands of years.
The older remains date to approximately 167,000 to 161,000 years ago
while the younger fossils date to approximately 142,000 to 134,000 years ago.
This means Bianfu was not the scene of one brief Denisovan visit.
Denisovan groups returned to, or continuously occupied,
this landscape over a substantial period.
The identified specimens include two teeth,
a lower fourth premolar and a lower second molar,
as well as three bone fragments.
Both tested teeth belonged to males,
according to sex-specific proteins detected in their enamel.
Another molar could not be molecularly identified
but closely resembles the unusual dental anatomy associated with other Denisovan fossils.
Together, the remains provide a small but important sample
of a population living in southwestern China during the later Middle Pleistocene.
The identification depended on ancient proteins rather than DNA.
DNA rarely survives well in warm or humid environments,
which is one reason the fossil record of southern Asia has remained so difficult to interpret.
Proteins are more durable.
They cannot provide a complete genome
but certain amino acid variants can identify biological relationships
between ancient populations.
All five tested Bianfu fossils carried a collagen variant
which has become one of the strongest molecular markers associated with Denisovans.
The enamel proteins contained additional information.
One variant had previously been found in Denisovans,
in the Harbin skull,
in the Penghu mandible from Taiwan,
and in fossils traditionally classified as East Asian Homo erectus.
The researchers stated that this variant was also present in East Asian Homo erectus.
That does not mean the Bianfu remains contain some separate erectus protein,
nor does it prove that Denisovans descended directly
from the individuals represented by those fossils.
It does, however, reveal a molecular connection
across a series of archaic East Asian populations
that were once classified in completely different ways.
The stone tools from Bianfu make the discovery even more interesting.
The Denisovans living there did not manufacture
an elaborate European-style musterian industry
dominated by prepared Levallois cores.
Most of their tools were produced through relatively direct methods.
River cobbles were struck to produce useful flakes
and many of those flakes were used without extensive reshaping.
Some discoidal cores and more organized reduction strategies were present,
but the assemblage was generally expedient.
If the tools had been discovered without the human fossils,
they might easily have been attributed
to a persistent East Asian technological tradition
associated with late Homo erectus.
The proteins now tell us
that at least some of their makers were Denisovans.
This creates an important warning about judging human intelligence
from the apparent sophistication of a stone tool assemblage.
The Bianfu tools may look simple,
but the behavior recorded in the cave was not.
The animal bones show that these people were effective hunters
with primary access to large carcasses.
They targeted animals including seeker deer,
red deer, gore, and water buffalo
with a noticeable representation of healthy, prime-aged adults.
These were not merely the weakest animals
conveniently found after dying from natural causes.
Many bones carry cut marks and signs of marrow extraction,
while comparatively few show evidence of carnivore gnawing.
Denisovans appear to have acquired the animals themselves,
butchered them efficiently,
and transported valuable portions back to the cave.
They also used bones as tools,
including retouches employed to shape
or maintain the edges of stone artifacts.
This combination is one of the most revealing aspects of Bianfu.
The stone technology was comparatively straightforward,
yet the inhabitants possessed the cooperation,
planning, and ecological knowledge
required to hunt large mammals in a mountainous environment.
Their tools were not simple because their minds were simple.
They were appropriate to the available raw materials
and to the tasks people needed to perform.
At Denisova Cave in Siberia,
archaeological layers associated with Denisovan occupation
contain more elaborate Middle Paleolithic reduction strategies.
The difference demonstrates that Denisovan identity
cannot be recognized through one particular style of stone tool.
Denisovans adjusted their technology to local conditions,
just as modern human and Neanderthal populations did.
Biology did not dictate a single culture.
The position of Bianfu Cave is also crucial.
Before this discovery, confirmed Denisovan sites
formed a vast but incomplete pattern across Asia.
Denisova Cave lay in southern Siberia.
Baishiya Cast Cave, where the Xiahe mandible was found,
lay high on the Tibetan plateau.
The Harbin Skull came from northeastern China.
Denisovan proteins were also recovered from the Penghu mandible
and other remains dredged from the seabed
between Taiwan and mainland China.
Bianfu now anchors the southwestern corner
of this enormous distribution.
Draw a perimeter around these sites
and it encloses much of China's
Middle Pleistocene archaeological record.
This boundary should not be presented
as a proven migration route,
but it illustrates the extraordinary geographic extent
of populations now identified as Denisovan.
Bianfu also closes part of the distance between China
and a particularly mysterious tooth
discovered in northern Laos.
That tooth came from Tamngu Hao,
commonly known as Cobra Cave.
It belonged to a young girl who lived
approximately 164,000 to 131,000 years ago.
No ancient DNA was recovered
and the available proteins could confirm
that the tooth came from a hominin
but could not conclusively identify its population.
Its internal structure and massive crown,
however, closely resembled Denisovan molars.
For that reason, its discoverers concluded
that the girl was most likely Denisovan,
although a Neanderthal relationship
could not be completely excluded.
Bianfu Cave lies approximately 770 kilometers
or 478 miles from Cobra Cave in a straight line.
More importantly, their dates overlap almost perfectly.
The older Bianfu fossils are between
approximately 167,000 and 161,000 years old,
while the Laos tooth may be as old as 164,000 years.
These were not sites separated
by hundreds of thousands of years.
Denisovan populations may have been living
simultaneously across the highlands of southwestern China
and the warmer landscapes of mainland Southeast Asia.
Bianfu, therefore, provides the strongest confirmed anchor
for what could have been a southern Denisovan corridor
extending through Yunnan, Guizhou, and northern Laos.
Such a corridor would have contained
sharply different environments.
Bianfu stood in an elevated landscape
with conifer-dominated forests and open forest steppe.
Cobra Cave lay much farther south
within a tropical or subtropical ecosystem.
Baishakast Cave was higher and colder still
on the Tibetan Plateau.
Other Denisovans occupied Siberia, northeastern China,
and the exposed continental shelf near Taiwan.
This was not a narrowly specialized,
cold-adapted population trapped in northern mountains.
Denisovans, or multiple related Denisovan populations,
survived from Siberia to Southeast Asia
and from high plateaus to coastal plains.
Their geographic range in latitude
exceeded the currently documented range of Neanderthals.
The new discovery, therefore, forces us to look again
at the numerous unidentified teeth scattered across China.
Hualongdong is an especially important comparison.
The site produced a partial skull, a jaw, and teeth
from approximately 300,000 years ago.
Its fossils possess an unusual combination
of archaic and more derived features,
and the Bianfu researchers found similarities
between the molars from the two sites.
Hualongdong remains much older than Bianfu
and has not been molecularly identified as Denisovan,
but it now becomes an obvious candidate
for future protein analysis.
The same is true of the four teeth found at Tongzhi
in Guizhou Province,
dated to approximately 240,000 to 172,000 years ago.
Researchers have struggled to fit them comfortably
within Homo erectus, Neanderthals, or modern humans.
Some features resemble East Asian archaic populations,
while others recall the large and complex molars
associated with Denisovans.
Pangxian Dadong, also in Guizhou,
produced additional middle Pleistocene teeth
with a similarly puzzling mixture of characteristics.
Farther north, the enormous and complex teeth
from Sujiyao have frequently been compared
with Denisovan dental anatomy.
None of these comparisons proves
that their owners were Denisovan,
but Bianfu demonstrates that Denisovans
really were living in the region,
making what once seemed speculative
far more plausible.
The implications may reach even deeper
into the Chinese fossil record.
Protein studies of approximately 400,000-year-old material
from Jokudian and Hexian
have identified amino acid variants
that appear Denisovan-like.
These fossils have traditionally been assigned
to Homo erectus,
but some researchers,
including paleoanthropologist John Hawkes,
now suspect that at least part of this material
may represent an early Denisovan-related population.
Hawkes has argued
that these middle Pleistocene individuals
could have a relationship to later Denisovans
comparable to the relationship
between the 430,000-year-old humans
from Cima de los Huesos and later Neanderthals.
They may not have been Denisovans
in the narrowest genetic sense,
but they could belong near the base
of the Denisovan evolutionary lineage.
That remains a hypothesis
rather than an established reclassification.
Shared protein variants can reflect common ancestry
extending far back in time,
and the available protein sequences
provide much less resolution than complete genomes.
It would be premature simply to rename
every Chinese Homo erectus fossil as Denisovan.
Yet the traditional categories
are becoming increasingly difficult to defend.
For decades,
fossils such as Dali, Jinyushan, Suchang,
Maba, and Hualongdong
were treated as disconnected regional forms,
transitional populations,
or taxonomic puzzles.
Molecular evidence is now revealing
that several apparently different fossils,
including Harbin, Siahi, Penghu, and Bianfu,
belong to populations connected
through Denisovan ancestry.
The morphological diversity among them
may represent variation across an enormous
and long-lasting population network
rather than a succession of unrelated species.
This is why Bianfu matters
far beyond the number of bones recovered.
It supplies a securely identified reference population
in a region filled with mysterious fossils.
It shows what southwestern Denisovan teeth looked like,
what protein variants they carried,
which animals they hunted,
and which technologies they used.
It also proves
that relatively unremarkable fragments
can overturn classifications
that once appeared secure.
The next Denisovan discovery
may not come from another spectacular excavation.
It may come from a tooth collected decades ago
sitting in a museum drawer
under a label such as archaic Homo sapiens,
late Homo erectus,
or simply unidentified hominin.
The older picture of Denisovans
as an obscure Siberian group
was created by the accident of preservation.
Cold caves preserve DNA
while the warmer regions
in which many Denisovans actually lived
destroyed it.
Proteins are now allowing scientists
to move beyond that northern preservation zone.
With every successful analysis,
Denisovans appear less like a peripheral population
and more like one of the dominant human groups
of Pleistocene Asia.
Their remains have now been recognized
across thousands of kilometers
while their DNA survives in living people
from Asia, Australia, and Oceania.
Bianfu Cave may therefore mark
the beginning of a much larger reassessment.
The great question is no longer
whether Denisovans reached China.
We now know that they occupied
its northeast, northwest, southwest,
and southeastern continental margin.
The real question is how much
of China's archaic fossil record
belongs to them.
Were the strange teeth from Tongzhi
and Pangtzean Denisovan?
Was the child from Hualongdong
part of an early Denisovan population?
Were some fossils called Asian Homo erectus
actually ancestral relatives of the Denisovans
who later spread across the continent?
And if a Denisovan girl lived in Laos
at almost exactly the same time
the Denisovan hunters occupied Bianfu Cave,
how far south did their world extend?
The answers are likely waiting
inside ancient enamel.
For more than a century,
paleoanthropologists attempted
to reconstruct Asia's human history
almost entirely from the visible shapes
of skulls and teeth.
Now, microscopic protein sequences
are exposing relationships
that anatomy alone could not resolve.
Bianfu has placed Denisovans
firmly within the mountains
of southwestern China.
But its greater importance
may be the new lens it provides
for examining everything around it.
The Denisovans were not absent
from the Asian fossil record.
We simply did not yet know
how to recognize them.
And if the emerging molecular evidence
is correct,
some of the most famous
mystery humans in China
may not be mysteries for much longer.
Thanks for watching.