Excavation in Cameroon
© ISABELLE RIBOTExcavation of a double burial at the Shum Laka rock shelter containing the remains of two boys who lived around 8000 years ago.
Remains from an ancient gravesite in Cameroon have opened a window into the world of the people who lived in western central Africa before farming and herding became widespread.

Africa is the ancient homeland of our species, and the genetic diversity across the continent is unmatched anywhere else on the planet, yet only a handful of sites bearing human fossils have successfully yielded ancient DNA, which is essential for grasping the genetic make-up of prehistoric Africans.

DNA rapidly degrades in tropical conditions and burials were rare prior to the Iron Age, which started around 1500 BCE in Africa.

"There's a tremendous amount of genetic diversity in the past in Africa that we have not been able to see by sampling modern populations," says archaeologist Mary Pendergrast from Saint Louis University, Spain, one of the lead authors of a paper in the journal Nature.

That's where the Shum Laka rock shelter comes in. The site, which lies in the Grassfields region of western Cameroon and bears clear signs of use by human foragers from as long as 30,000 years ago, is unique, says Pendergrast.

Eighteen people, mostly children, were buried there during two periods - 8000 and 3000 years ago respectively. Both communities were hunter-gatherers, and by the later time period they were using pottery and relying heavily on fruit picked from nearby forest trees.

These time points "book-end" an important transition period from the late Stone Age to the Iron Age, says Pendergrast, during which humans took up farming and herding.

Linguists have also long thought that the Grassfields region could be the birthplace of the Bantu language group. Bantu languages spread across sub-Saharan Africa from about 4000 years ago and are today spoken by more than a third of Africans.

"There's been a century of discussion about where Bantu languages originated and how they spread," says Pendergrast.

Mark Lipson, from Harvard Medical School, and colleagues extracted then analysed ancient DNA from the DNA-rich inner ears of four children buried at Shum Laka - two from each period - to see whether they were the ancestors of modern-day Bantu speakers.

But that turned out not to be the case. None of the children - who were all related to each other - were related to present-day Bantu speakers.

Instead, they were part of a population that has almost been completely replaced. About two-thirds of the Shum Laka ancestry is from a previously unknown line that is distantly related to present-day West Africans. The other third is from a lineage related to present-day central African hunter-gatherers.

Bantu languages may still have their origins in the Grassfields region, says Pendergrast, but the Shum Laka children aren't part of that ancestral population.

Lipson and his colleagues also used the ancient genomes of the four children, along with genomes from modern-day Africans, to cast their focus much further back in time.

An adolescent male from the rock shelter carried a rare genetic variant of his Y chromosome that is today found almost exclusively in western Cameroon, close to Shum Laka.

A previous study proposed southern Africa as the homeland of the oldest human lineage, but the rare Y chromosome suggests that a lineage contributing to central African hunter-gatherers is just as ancient, cleaving off from other African branches some 250,000-200,000 years ago.

At the time of the split, four lineages emerged. Three are ancestral to present-day central African hunter-gatherers, southern African hunter-gatherers, and all other modern humans. The fourth, a previously unknown ghost population, contributed a small amount of ancestry to both western and eastern Africans.

Another split occurred around 80,000-60,000 years ago, leading to genetic lines that are present in the majority of present-day eastern and western African and all non-Africans.

Compared to Europe or Asia, African ancient DNA studies are in their infancy.

"It's incredible what we've learned in the past few years," says Pendergrast. But, she adds, "we're just scratching the surface of what the human landscape would have looked like before the spread of food production".