In September 2018, China Daily ran a story that found its way to my desk. It featured the photograph of a beautifully preserved fossil skull with huge brow-ridges, with a caption speculating – wrongly, as it would turn out – that it belonged to a species of human known as Homo heidelbergensis.
The article told a remarkable story about the discovery of this fossil some 85 years earlier. Japanese forces had been occupying the city of Harbin in north-eastern China in 1933. They ordered a team of local workers to help build a railway bridge across the Songhua River.
The workers uncovered the skull in river deposits, and one of them hid it from the Japanese by lowering it down a disused well. Feeling ashamed of having worked for the enemy, he kept quiet about the find for many years. Finally some 80 years later he told his son and grandson as he lay dying, which led to the almost intact skull being recovered and later donated to Hebei Geo University near Beijing.
That origin story is difficult to verify, and there were rumours that the fossil was actually a fake, but I was able to confirm first hand that it was real when I was invited to join Chinese researchers in studying this astonishing relic at the end of 2018. I jumped at this wonderful opportunity to help put one of the most complete human fossils ever found in its proper place in human evolution.
So many ways to be human
Establishing that place has got a lot trickier in the last 20 years or so. We think that humans (or hominins) split from the ancestors of our closest living relatives, the chimpanzees, about 7 million years ago. For the next 5 million years or so, the hominins continued their evolution in Africa, and about 2 million years ago, began to disperse across Asia and Europe.
It used to be thought that human evolution then progressed from more primitive forms through to modern humans (Homo sapiens), either in just one region, most likely east Africa, or across the whole inhabited world, following the early exit from Africa. Now, we know it wasn’t so simple.
Welcome to our series on the great mysteries of archaeology. Viking explorers, Amazonian cities, artefacts from before civilisation. Archaeology may be all about the past, but it’s constantly shifting with every scientific discovery. This series will dig into some of the most fascinating debates in the field today.
There was a profusion of different kinds of humans, sometimes living at the same time and even in the same regions, as nature conducted multiple experiments over millions of years in how to become human. We are the sole survivors of all those experiments.
Even during the last 300,000 years the number of human species has been startling. The ancient Homo heidelbergensis, with their heavy-set physiques and large faces and jaws, was surviving in Europe – along with similar people in Africa – 300,000 years ago. The even more ancient Homo erectus, which had a smaller brain and angular flat-topped skull, was clinging on in parts of Indonesia until at least 100,000 years ago.
Homo neanderthalensis (Neanderthal human), with a broad thick-set body and protruding midface around a large nose, was living in western Eurasia from at least 400,000 to 40,000 years ago. We Homo sapiens – first given that name in 1758, meaning “wise human” – existed at least 300,000 years ago, first in Africa and then globally within the last 50,000 years.
In addition, there are three enigmatic small-bodied forms (and also known to be small-brained in the first two cases) from the same recent period: Homo naledi (star human), Homo floresiensis (Flores human) and Homo luzonensis (Luzon human). These lived respectively in South Africa about 300,000 years ago; on the island of Flores, eastern Indonesia, from at least 190,000 to 50,000 years ago; and on the island of Luzon, the Philippines, about 130,000 years ago. All three seem to be long-lasting relics of even more ancient stages of human evolution.
The leading human species
Our study of the Harbin skull added yet another species to the list. When my Chinese colleagues sent me a full-size 3D print of this skull, I couldn’t believe it. I’ve seen hundreds of fossil human skulls, and this was the biggest I had ever set eyes on.
Our analyses showed it was clearly different from fossils of , , and . It showed a distinctive combination of a very large brain size, a massive brow-ridge and a large, low but rather smooth braincase without the other bony ridges found in and skulls. It also had a big but modern-looking face with delicate forward-facing cheekbones but very wide nose and large eye sockets.
This warranted a new species name, and in 2021 my Chinese colleagues chose Homo longi, named after the Black Dragon River (Heilongjiang). Ever since, the skull has been known as Dragon Man. It appears that this species was in eastern Asia from at least 300,000 until about 45,000 years ago. All these human timespans are show on the chart below (click to make it bigger).
Human timespans
Enter the Denisovans
Not only did the skull turn out to be a new species of ancient human, it also helped to solve several mysteries about other discoveries. One of the keys was its single preserved tooth – a very large second molar, with no sign of a socket behind it for a third molar. In other words, Dragon Man did not have wisdom teeth.
These dental features provided a link with other less complete fossils from China that also had large molars but no wisdom teeth. This and other resemblances indicated that these were also Homo longi, overturning previous possibilities that they might have belonged to an Asian branch of Homo heidelbergensis or an interim species between Homo erectus and modern Asian humans.
It also shed light on the mysterious Denisovans, named after Denisova Cave in the Russian part of the Altai Mountains. Archaeologists had been digging there for decades, finding thousands of animal bones and stone tools, but only scraps of human bones and some large molar teeth.
In 2010 one of those bone scraps – part of a little finger – yielded ancient DNA matching the quality of modern bones. This was despite being about 60,000 years old, showing the astonishing preservation potential of this cold and dry cave in southern Siberia.
Thanks to DNA yielded from other fossils and even sediments from the site, we know the Denisovans occupied the cave intermittently from more than 200,000 until about 45,000 years ago, when Homo sapiens took over.
Placing the Denisovan genome on the human evolutionary tree was difficult, but also in 2010 a high quality Neanderthal genome was reconstructed from European fossils. The Denisovan genome was quite different both to this and to modern humans.
Fast-forward a decade and our analyses suggested that the molars found in Denisova Cave were also from the same species as the Dragon skull. More recently the Harbin skull has been analysed for DNA and fossil proteins, confirming both the connection to Denisova Cave and various other fossils found in China. These include a jawbone from a cave high up on the Tibetan plateau, bones dredged up by fishermen from off the coast of Taiwan, and remains recently described from a cave in the south-west of China.
The skull therefore turns out to be the most complete example of a species that palaeoanthropologists have been circling for some time. It’s now known by both names: Denisovan and the more formal Homo longi.
The great river of life
If people today had evolved from Neanderthals, we might have expected that modern Europeans would show the highest levels of Neanderthal DNA in their genomes, since Europe is where most Neanderthal fossils have been found over the last two centuries. It turns out they do have Neanderthal DNA, at a level of about 2%. But that level of DNA is also found throughout the world today, apart from Africa, where the level is generally much lower.
When similar comparisons were made for Denisovan DNA there were even more surprises. The highest levels were not in modern Siberians or Chinese, but in people from places like the Philippines, Papua New Guinea and Australia – in some cases at levels of around 4%.
How can we explain these unexpected results? Well, we know that the vast majority of our modern human genome – more than 90% – derives from Africa. About 50,000 years ago, having evolved there, the main ancestral wave of early Homo sapiens migrated to western Asia. There they encountered local Neanderthals and did a bit of interbreeding.
That may seem impossible for different species, but today we know from thousands of genome studies that closely related species can often breed hybrids successfully. Examples include jackals and wolves, polar bears and brown bears, and Canada and Greylag geese. It can take millions of years for enough genetic differences to accumulate to make such interbreeding impossible, and that hadn’t happened in the cases of modern humans, Neanderthals and Denisovans.
Those intermixed Homo sapiens in western Asia then spread out to every corner of the world, taking their Neanderthal DNA with them. A lot of it was selected away since it was potentially harmful, causing genetic incompatibilities in regions like the brain and reproductive organs. But other segments were kept as they helped survival, for example protecting against diseases outside of Africa. Over time, the Neanderthal DNA settled down to about the 2% of the modern genome that we see outside of Africa today.
As for the Denisovans, they too interbred with early Homo sapiens groups reaching east Asia. This happened at a lower level in mainland Asia, and seemingly at a much higher level as the Homo sapiens spread across islands of south-east Asia like Borneo and Sulawesi. So far there is no evidence that those other late-surviving relatives such as Homo floresiensis and Homo luzonensis contributed detectable amounts of their DNA to our genomes today. Either there was no interbreeding, or the species were too different to produce viable hybrids.
We don’t yet have fossil remains from those islands to prove the Denisovans were there before Homo sapiens, only archaeological evidence such as butchered animal bones and stone tools. But assuming this idea is correct, the Denisovans must have been a very successful and adaptable species, surviving freezing winters in northern China but also thriving in tropical grasslands in places like Sumatra. As previously mentioned, they were even living on the Tibetan plateau at altitudes above 3,000 metres.
And by the way, the same Russian cave provides evidence of another dimension to this story. DNA shows that Neanderthals were also there some of the time, and that the two species interbred. We know this from another bone fragment, whose owner, a young girl, was the offspring of a Neanderthal mother and Denisovan father.
What we got right and wrong
Most of what I’ve talked about here, we’ve only learnt in the last 16 years. For my PhD more than 50 years ago I was testing the then widespread idea that modern humans evolved directly from Neanderthals. I spent my time measuring and comparing fossil skulls, using the primitive measuring techniques and limited computing power available at the time.
I concluded that we did not evolve from them. I could see no sign that Neanderthal fossils in Europe were gradually changing towards a more modern anatomy. They weren’t developing a more rounded braincase, a smaller browridge or a more pronounced chin.
Yet now, with genetic powers that I could never have imagined, we can see that my PhD quest only gave part of the answer. Certainly we did not evolve from Neanderthals but yes, they are partly our ancestors because of that ancient interbreeding. Twenty years ago I would have told you that any interbreeding with Neanderthals was trivial in the big picture.
Together with the findings about interbreeding between modern humans and Denisovans, I have to concede that our more recent origins are only mostly out of Africa, not 100%. It means the Neanderthals and Denisovans didn’t die out completely 30,000 or 40,000 years ago: bits of their DNA live on in us today, and some of it is still active.
As for our African beginnings, it was in the 1980s and 1990s that these became clear as DNA sequences from recent human populations accumulated, as well as better fossil material. However, recent evidence suggests that our African origins were quite complex, with several lineages evolving in parallel in different parts of the continent and mixing now and again. For instance, some geneticists think ancient interbreeding with Homo heidelbergensis could have contributed to our evolution, so the human origin story keeps getting more complex.
What’s next
It’s difficult to imagine what even the next ten years will bring, as work on ancient DNA is developing so fast – along with the newer field of ancient proteomics, which studies fossil proteins. Scientists are turning to these because DNA survival is patchy, especially in warm conditions and deep time.
For example, scientists are looking for traces of the connective tissue collagen, which is widespread in our bodies, skeleton and teeth. Collagen is much more resilient than DNA, and has already been identified in fossils that are millions of years old. As the work progresses and we can sequence the different amino acids in these ancient proteins, we will start to gather knowledge of fossil relationships where DNA cannot reach, as has already been achieved for the Dragon Man fossil.
The fossil record itself will undoubtedly continue to surprise us with unexpected finds, hopefully including eagerly awaited Denisovans from the islands of south-east Asia. Also waiting to be found somewhere are fossils of first-generation hybrids between our species and the Neanderthals, and our species and the Denisovans.
Meanwhile I’m still pursuing the biggest puzzle of them all. Why are we the only surviving species out of dozens that preceded us, and the half-dozen that were still living alongside us 100,000 years ago? Did we overcome them through our better brains or better adaptations or did we just get lucky, time after time, as first the Neanderthals went under, then all the others?
I suspect competition for often limited resources was an important part of the story, where we may have been that bit more efficient than our surviving relatives, but we simply don’t know the full solution yet. More discoveries from archaeology, fossils and DNA will lead us closer to the answers.
For you: more from our Insights series:
To hear about new Insights articles, join the hundreds of thousands of people who value The Conversation’s evidence-based news. Subscribe to our newsletter.