Breakdown: Two New Genetics Papers On The Upper Palaeolithic And Mesolithic

Posth et al 2023 and Villalba-Mouco et al 2023: Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers & A 23,000-year-old southern Iberian individual links human groups...

As a one-off interruption from my usual monthly round-up of interesting archaeological finds and papers, I want instead to take a deep dive into two papers, both of which came out yesterday - 01/03/23:

• Posth, C., Yu, H., Ghalichi, A. et al. Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers. Nature 615, 117–126 (2023)

• Villalba-Mouco, V., van de Loosdrecht, M.S., Rohrlach, A.B. et al. A 23,000-year-old southern Iberian individual links human groups that lived in Western Europe before and after the Last Glacial Maximum. Nat Ecol Evol (2023)

This is going to be a fairly dense article, heavy on genetics terminology, but I will try to make it all as simple as possible, and explain the significance of the findings. Publications are the way academics talk to one another, and those not familiar with the topic can easily get lost.

So the subject in front of us is the genetic structure through time of European populations from the earliest Homo sapiens to after the end of the Ice Age. For reference we can roughly break down this time period as follows:

• Aurignacian (43 - 28 kya) - the Cro-Magnons, the earliest modern humans in Europe, who bred with the Neanderthals and became the sole human species on the continent.

• Gravettian (33 - 21 kya) - the next group of humans to occupy Europe during its coldest period (Last Glacial Maximum - LGM - 26-20kya).

• Solutrean (22 - 17kya) - as the ice expanded, some humans and animals were pushed into the Franco-Iberian refugium and developed a new culture.

• Epigravettian (21 - 10kya) - simultaneously, other humans were pushed into Italy and the Balkans as the glaciers expanded southwards.

• Magdalenian (17 - 12kya) - as the ice retreated, Solutrean derived groups moved out of Iberia and followed the vast reindeer herds into the continent. This culture produced much of the famous cave art.

• Mesolithic (12 - 7kya) - the end of the Ice Age produces warmer weather, multiple human groups derived from the earlier peoples split and colonised Europe, including Ireland and Scandinavia.

As we go back in time, the availability of well-preserved human remains decreases. This means that our knowledge of exactly how all these different groups developed and changed is in its infancy. Up until recently these cultures were defined on the basis of stone tool typology, artefacts, burial styles and physical anthropology. Now we are seeing enormous advances in the genetics sphere.

So let’s add some of these genetics now into the chronology. The earliest European genomes which contributed to modern populations are: Kostenki 14 (37kya, western Russia), Goyet Q116-1 (35 kya, Aurignacian, Belgium) and Bacho Kiro 1653 (35 kya, Aurignacian, Bulgaria). The Kostenki 14 genome forms what geneticists call a cluster (a group with greater internal relatedness than external). This group then migrates west and appears to form the central European Gravettian signature, itself defined by the Věstonice cluster, named after the hugely important Gravettian site of Dolní Věstonice. The Gravettians lived through the extremes of the LGM, where people moved into different refuge regions, as mentioned above. We lack any good genetics for the LGM period.

After the LGM we get a new cluster called GoyetQ2, a 15,000 year old sample from the Goyet Caves in Belgium. What is interesting is that the original Goyet Q116-1 cluster from 35ya shares some genetic relatedness with GoyetQ2, ancestors separated by 20,000 years being buried in the same caves. GoyetQ2 is associated with the new post-LGM Magdalenian culture in our chronology. Meanwhile, the other great refuge in Italy and the Balkans becomes defined by its own cluster - Villabruna - which roughly corresponds to the Epigravettian culture.

So this is a simplified overview of where we stand before we look at the papers. Clearly there’s a lot missing from this story, in particular what exactly happened during the LGM?

Paper One: (Posth et al, 2023): Upper Palaeolithic - Early Mesolithic

In this huge paper the researchers generated genome-wide DNA sequences for 356 individuals, dated from 35 - 5kya. 116 of these were new, which adds an enormous amount of fresh data to the archives.

What did they find?

1. A new Gravettian cluster. Previously the Gravettians were believed to be a biologically homogenous pan-European culture, but the new samples show a distinctive cluster, named Fournol, based in western and southwestern sites.

   1. This new Fournol cluster is closely associated with the Aurignacian Goyet Q116-1 cluster, including sharing the mitochondrial haplogroup M, which disappears during the LGM.

   2. In fact, the Gravettians are split into two distinct camps - the Fournol and Věstonice, divided geographically. The Goyet Q116-1 individuals show affinity to both groups.

   3. When modelled differently, the Fournol and Goyet Q116-1 clusters become sister lineages, and the Věstonice group sits between Goyet Q116-1 and more eastern hunter-gatherers (Sunghir).

   4. These genetic distinction match different burial traditions. The Fournol cluster areas made use of caves, whereas the Věstonice group buried their dead in open-air sites with elaborate grave goods.

2. Evidence for direct continuity between the Gravettian and the Magdalenian through the Solutrean period. The Solutrean individual Le Piage II from southern France has close links to the Fournol cluster and some affinity to the later Magdalenian cluster El Mirón. This indicates that the Solutreans arose from the western Gravettian people and maintained their population through the LGM.

   

3. It has long been assumed that the Italian/Balkan Epigravettians were a direct ancestor to the earlier Gravettians, hence the name. However, by sequencing and comparing several new genomes the researchers showed the Epigravettians were a different, unrelated population.

   1. All Epigravettians so far belong to the Villabruna cluster, and show no affinity to any Gravettian cluster. The oldest basal individual is the 13 kya Pradis 1 sample from north-east Italy. But given that Villabruna genetics appear at El Mirón some 19kya, the turnover must have been much earlier.

   2. The Villabruna cluster shows a relationship with Near East hunter-gatherers, suggesting that they arrived in Italy from the Balkans.

   3. The genetic diversity within the Epigravettian population was very low and branched out twice within Italy. The Sicilian branch was probably founded by only 70 people.

   4. Given that Villabruna affinities appear in both Iberia and the Balkans, it seems the LGM created a land and likely sea corridor from east to west, running underneath the glacial extremes.

4. The post-LGM Magdalenian culture has been characterised by both the El Mirón and GoyetQ2 clusters, suggesting that the Solutreans and Epigravettians mixed in some proportion to form the Magdalenians, which spread out north and north-east across Europe. New genomes confirm this, but El Mirón looks to be an outlier, with 43% Villabruna ancestry, whereas all other Magdalenian samples can be placed into the GoyetQ2 cluster.

Obviously this is a lot of information, and is difficult to piece together into a graspable narrative. Very roughly it looks like the Gravettians spread into Aurignacian Europe around 35kya and split into two distinct groups. As the glaciers expanded, the south-western peoples were forced into Iberia. Meanwhile a new group from the Balkans arrived in Italy and colonised the peninsula from north to south, as well as moving west to connect with the Solutreans. As the ice retreated, the Solutrean-Epigravettian mixture formed the Magdalenians, who swept northwards and east as far as Poland.

Paper One: (Posth et al, 2023): The Mesolithic and Neolithic

Next up, the researchers sequenced a number of younger genomes to try and understand the complex population formations and movements after the LGM. Readers who follow any popular literature on this time period might recognise the terms ‘Western Hunter-Gatherer’ (WHG) and ‘Eastern Hunter-Gatherer’ (EHG), which refer to the two broad population groups in Europe after the Ice Age, but before the arrival of agriculture. Generally the WHGs occupied most of Europe, with the EHG based in western Russia and extending their influence down the eastern Baltic, Belarus, Ukraine and into Fennoscandia. The ‘Scandinavian Hunter-Gatherers’ (SHG) are defined as a mixture of the WHG and EHG groups.

The researchers here do not radically alter this picture, but they do broaden and clarify it. We already knew that WHGs were characterised largely by the Villabruna cluster, and that EHGs had both Villabruna and a more eastern ancestry known as ‘Ancient North Eurasian’ or ANE. But what does the paper say?

1. The WHG lineage can be better defined as the Oberkassel cluster, based on two 14kya newly sequenced individuals. This cluster shows the strongest affinity with a particular Epigravettian/Villabruna genome called Arene Candide16.

   1. The Oberkassel cluster is made up of a large contribution from the Villabruna group and smaller amount from GoyetQ2 . All individuals within the WHG-Oberkassel group can be modelled as approx. 75% Arene Candide16 and 25% GoyetQ2.

   2. The presence of 100% Villabruna individuals in post LGM Britain suggests that there was various population movements before the Oberkassel group formed - note that Cheddar Man matched the Oberkassel admixture.

   3. Unlike Britain, Iberian hunter-gatherers show repeated mixtures of Villabruna/Oberkassel and GoyetQ2 ancestry, which reflects the deep Solutrean presence. Even today Y-haplogroup C-M130 can be found amongst Catalonian males with the surname ‘Llach’, whereas C-M130 largely vanished after the LGM elsewhere in Europe.

2. Defining the EHG genetic group came down to looking for the group with the lowest Villabruna contribution. This was found at the Mesolithic Oleni’Ostrov cemetery in Karelia. The oldest individual which matched this profile came from an 11kya burial in Samara, western Russia, and the EHG cluster is now called the Sidelkino cluster.

   1. The Oberkassel cluster shows high amounts of mitochondrial haplogroup U5 and Y-haplogroup I.

   2. The Sidelkino cluster shows higher amounts of mitochondrial haplogroups U2, U4 and R1b, and Y-haplogroups Q, R and J

3. Modelling population change through time using the GoyetQ2, Oberkassel, Sidelkino and Anatolian Neolithic Farmer clusters between 14-5kya, we see the following observations:

   1. Between 14-8kya, everyone in western and central Europe came from the Oberkassel lineage

   2. Sidelkino ancestry was noted however in the Baltics, Scandinavia, the Balkans and Ukraine. These groups also had some level of Anatolian Farmer admixture.

   3. The earliest Sidelkino individuals showed greater affinity to the Oberkassel lineage than their descendants. This puts the formation of EHGs possibly around 15-13kya.

   4. Around 8kya, the Sidelkino/EHGs began expanding into central Europe. This is first detected at Gross Fredenwalde, and then around 10% of most WHGs had this ancestry.

   5. Possible additional Oberkassel ancestry was noted in samples from the upper Volga, but radiocarbon dating has not confirmed when this took place.

   6. After 7.5kya the Neolithic farming populations restricted the WHG/EHGs to the northern fringes of Europe. More Oberkassel ancestry spread east reaching Samara by 6.5kya, and Sidelkino ancestry increased in the Baltic, matching the transition from Narva to Comb Ceramic culture.

   7. The youngest known predominantly-WHG individual dates to 3200 BC in Ostorf, Germany, indicating that the hunter-gatherer populations managed to survive in small pockets up to the beginning of the Bronze Age.

4. Looking at interesting physiological characteristics, the researchers noted:

   1. No assessed populations carried the rs4988235 lactase mutation.

   2. Based on the genetic markers for skin and eye colour, its possible that the WHGs had darker skin and light eyes, while the EHGs had lighter skin and darker eyes. Personally I’m very cautious about these types of results, since skin and eye colour are notoriously difficult to model accurately.

   

This was a huge paper and I haven’t included every detail, otherwise I would be simply replicating it. The main takeaways are: the Gravettians were not a homogenous population; the Epigravettians were an entirely separate population; the Solutreans maintained genetic continuity between the western Gravettians and Magdalenians; the Magdalenians were ultimately replaced by the WHGS; the WHGS and EHGs were largely isolated from one another until around 8kya.

Paper Two: (Villalba-Mouco et al, 2023) - A New Solutrean Genome

We’ve seen already that the question of genetic continuity in the Iberian refugia has long been a concern for archaeologists trying to understand the effects of the LGM. By luck or design we got a second paper on the same day, presenting new Solutrean data.

The researchers analysed genomes from a Solutrean site approx. 22.5kyo (Cueva del Malalmuerzo), two early Neolithic individuals from Cueva de Ardales/Las Aguilillas and two Chalcolithic individuals from Cueva de Ardales and Los Caserones. The Solutrean individual is referred to as MLZ in the paper for ease.

So what did they find after assessing the Solutrean genome?

1. The individual carried mitochondrial haplogroup U2’3’4’7’8’9. The U2 haplogroup was also found at Kostenki and in Epigravettian Italy, which is interesting, although mitochondrial haplogroups tend to be much broader than the Y.

2. Y-haplogroup was C1, which was also found in much older samples, including Goyet Q116-1 and Sunghir. Although this paper doesn’t say it, we know from the first that this matches the western Gravettian Fournol cluster and confirms the previous findings.

3. When modelled alongside the Věstonice, Goyet Q116-1, El Miron and GoyetQ2 clusters, the results fall perfectly into line with the first paper. MLZ sits between Goyet Q116-1 and GoyetQ2, which accounts for genetic drift and the emergence of the Magdalenians.

4. The researchers come to the same conclusion as the former with regards to the population structure of the Gravettians, which is impressive given that they were using just one new genome: “This implies that at least two genetically distinct groups must have existed in Europe when the Gravettian technocomplex prevailed: one in western Europe, represented by Goyet Q116-1 and a second in central (and perhaps eastern) Europe, described as the Věstonice cluster”

5. Comparing MLZ to the earliest Eurasian genomes (Tianyuan, Bacho Kiro, Ust’Ishim) also confirmed that the Solutreans derived part of their ancestry from the Aurignacians, rather than exclusively from the Gravettians.

6. MLZ also shows a small contribution from the Villabruna cluster, and the researchers noted the Near East contribution from the Epigravettians by way of a Natufian signal.

7. Where this paper differs from the previous is in suggesting that the Solutreans shared with the Natufians an excess of Near East ancestry which cannot be explained purely by admixture with the Epigravettian/Villabruna. This means that we need many more samples from the Balkans, Greece, Anatolia and North Africa to make sense of the Near Eastern - Villabruna connection.

8. Testing MLZ against North African Iberomausurians and sub-Saharan African populations showed no sign of any admixture across the Gibraltar Straits.

9. Southern Iberia retained higher levels of LGM hunter-gatherer ancestry compared to the rest of the region after the introduction of Anatolian farming populations. Taken together with the other paper, this infers that southern Iberia possessed a distinctive genetic population compared to the rest of Europe?

This was of course a summary of the paper and more details can be found there, but overall the results indicate that the first paper was broadly correct. Questions remain though about the relationship between the Epigravettians, the Near East hunter-gatherers, the Natufians and other pre and post LGM Eurasian populations. Now that we know the Epigravettians were descended from a different biological group, we need samples from those regions to unravel the formation of the Epigravettians, the Natufians, the Anatolian HGs and the Iberomaurusians.

We are also still working through the formation and expansion of both the EHG and WHG groups. The turnover from the Magdalenian to the WHGs seems extraordinary and the movements of different groups during the 8.2kya climate event need to be better understood to make sense of the EHG signal in western Europe - perhaps the introduction of pottery also came with a small number of exogamous marriages from east to west?

Hopefully this summary was not too difficult to follow and proves useful to anyone interested in the prehistory of Europe, please do ask questions in the comments for any clarifications or issues.