The probability of retrieving well-preserved DNA from cranial fragments, other than the petrous bone, dated to ~37,000 years ago and exhumed from non-permafrost sites is, in our experience, small. But when we got the opportunity to perform palaeogenetic research on skull fragments from the Buran Kaya III rock shelter, located on the eastern part of Crimea, we did not hesitate to fully commit to it! Indeed, this is a key site for the understanding of the arrival and dispersal of Anatomical Modern Humans (AMHs) in Europe as well as their potential biological and/or cultural interactions with Neanderthals and the transition between the Middle and Early Upper Palaeolithic. 
Buran Kaya III was discovered in 1990 by Alexandr Yanevich from the Institute of Archaeology, National Academy of Sciences of Ukraine in Kyiv, Ukraine. He directed the excavation of the Neolithic to Palaeolithic layers together with Masayoshi Yamada (Tokyo Metropolitan University) while the excavation of the Middle and Early Upper Palaeolithic layers were excavated under the direction of Anthony E. Marks. Between 2009 and 2011, new excavations were directed by Alexandr Yanevich and Stéphane Péan, National History Museum, Paris, France. These revealed that the site had been occupied by humans producing Middle, Early Upper Palaeolithic, Aurignacian and Gravettian industries. The cultural artefacts and the radiochronological data imply that a major climatic crisis is recorded in this site: the 2,000-year-long stadial Heinrich 4 cold period that had been reinforced by the Campanian Ignimbrite (CI) super-eruption ca. 39,847-39,136 before present (BP), which occurred in southern Italy and blanketed south-eastern Europe in ashes, resulting in a catastrophic impact on the ecosystem in this area. Culturally, there was a “before” and an “after” of this climatic crisis at Buran Kaya III evidenced by the absence of Neanderthal-associated artefacts in the more recent stratigraphic layers following this event. The depositional succession at Buran Kaya III between 40,000 and 34,000 BP is characterized by Aurignacian and then Gravettian assemblages attributed to AMHs. The latter industry might represent an early or proto-state of the Gravettian, which led to some discussion among scholars. The zooarchaeological analysis of the bone tools found in the layers corresponding to this period, performed by Stéphane Péan and Laurent Crépin from the National History Museum in Paris, showed that the site was a seasonal camp and butchery site. Subsequently, the site was abandoned for ~20,000 years. Large quantities of lithic tools (such as burins, end-scrapers, and backed microliths) have been found in the layers that also preserved the skull fragments. Moreover, these layers also revealed the artistic capacities of the population inhabiting this area who produced body ornaments such as perforated mollusk shells, fox and red deer teeth, as well as an engraved blade of mammoth tusk ivory and beautifully shaped, perforated and polished mammoth ivory pendants (Fig.1).

The human skull fragments were found in two consecutive layers testifying to a cold but not harsh climate, with increasingly dry conditions. These human remains are among the oldest direct evidence of anatomically modern humans in Europe in a well-documented archaeological context, which is not the case for many finds of this period. The anthropological analysis done by Sandrine Prat, Musée de l’Homme in Paris, revealed on some of the skeletal remains signs of post-mortem treatments of the corpses that could be attributed to ritual cannibalism or a specific mortuary practice. Isotopic analyses showed that the individuals were meat eaters, preferentially of mammoths and saiga antelopes.
The Epigenomics & Paleogenomics team of the Institut Jacques Monod (EP-IJM), University Paris-CIté, CNRS, wanted to know more about these individuals and the populations to whom they belonged. Owing to a close, friendly relationship with the archaeologists A. Yanevich and Stéphane Péan, as well as with the paleoanthropologist Sandrine Prat and zooarchaeologist Laurent Crépin who carried out the excavation between 2009 and 2011, we launched together a research programme that should allow us to perform a palaeogenomic analysis. It was not difficult for me to convince my colleagues to excavate and sample the human remains dedicated to a palaeogenomic analysis in the very best manner using precautionary measures whenever possible, comparable to those used by the police at a crime scene, to avoid contamination with DNA from present-day humans, i.e., the excavators and analysts (Fig. 2).

The bone fragments were kept in new plastic bags and transferred directly to my lab upon arrival of the French archaeologists in Paris where I stored them in the freezer. Before subjecting them to our analysis, which implies the grinding of small bone pieces, I took photographs under the binocular (Fig. 3) and Sandrine Prat carried out the anthropological analysis under the binocular in the contained laboratory of the Jacques Monod Institute dedicated to ancient DNA research (Fig. 4). 

The first experimental series of our palaeogenetic analysis was performed in 2010 together with E. Andrew Bennett, a post-doc in the lab at the time. At the end of the numerous experiments, we had collected enough data to perform statistical analyses. These allowed us to get a rough idea about the genetic relationship between these individuals and roughly contemporaneous ones at Kostenki-Borshchevo, Russia, and Peştera cu Oase, Romania, and others. Yet, the data were too scarce to yield robust results.
Instead of defending our analyses and “fighting” with the reviewers, I decided a few years later to restart from scratch using experimental methods of molecular biology that Thierry Grange and myself had improved in the meantime (Fig. 5). 

It was a good decision as we succeeded to get much better and more data on the two bone pieces exhumed from layers dated to ~700 years apart, in particular of the one that had been excavated using special precautions (Fig. 3C). DNA in the third fragment (Fig. 3B) that had been recovered from the sediment through sieving in the Borulcha river was poorly preserved and not amenable to analysis. This result is consistent with our previous analyses showing that the excavation, treatment and storage conditions can have a detrimental effect on DNA preservation in bone. Most importantly, however, the increase in power and resolution that we achieved during the second experimental series, in addition to the inclusion into the comparative analysis of several key Paleolithic genomes that had been published since our earlier work, allowed us to perform more elaborate population genomics analyses than were possible with the previous dataset. These analyses were performed by a graduate student, Oğuzhan Parasayan, who had joined our lab (Fig. 6), the results interpreted by the EP-IJM team and the paper was written by E. Andrew Bennett with the input of the team. 

Our genomic analysis of these two early ~36,000 and 37,000-years-old European individuals from the site Buran-Kaya III in Crimea now shows that the populations to which they belonged were associated with the oldest appearance of a cultural assemblage anywhere displaying features reminiscent of the later Gravettian assemblages. Furthermore, the dates of these genomes are important because they correspond to the period that just followed the major climatic changes and ecosystem crisis caused by the stadial Heinrich 4 cold period and the CI volcanic eruption period that coincided with human population turnover in Europe. The combined analysis of these genomes with other recent ancient genomic findings from Europe allowed us to create a broad and updated model of population movements, interactions and turnovers during the peopling of Europe in the Early Upper Paleolithic.  Incorporating evidence from genomic, climatic and cultural sources and fitting them into an updated context of other recent discoveries, our study adds several compelling new facts to our understanding of the early peopling of Europe. 
These results were obtained from a key archaeological site in Crimea whose excavation came to an abrupt halt due to the annexation of the Crimean Peninsula through Russia and that is now in a war zone, inaccessible to the Ukrainian excavators, as well as to us and the colleagues who are committed to democracy and freedom – while at the time when these individuals lived, they were mainly struggling for survival in a rather hostile environment but succeeded to establish a network of exchange that finally embraced all of Europe. This led to the unfolding of the Gravettian culture known to many of us through small ivory female figurines, such as the venuses of Dolni Věstonice in Czechia and Willendorf in Austria and the “Dame de Brassempouy” in France. Our Ukrainian colleagues now also struggle for survival and I hope that they soon can join us again in the international scientific exchange network in peaceful and friendly interaction, just as more than 30,000 years ago.