Back

Palaeogenomics reveals 1,500 years of population history of the peoples of the Chonos Archipelago, Chile.

Davidson, R.; Reyes, O.; Roca-Rada, X.; Arango-Isaza, E.; Kassadjikova, K.; Barbieri, C.; Llamas, B.; Fehren-Schmitz, L.

2025-11-28 genomics
10.1101/2025.11.26.690513 bioRxiv
Show abstract

The Chonos archipelago, located in northwestern Chilean Patagonia, was historically inhabited by an Indigenous sea-faring population known to the European colonists as the "Chono". Previous research has contextualized the human occupation of this region with radiocarbon dates and ancient mitochondrial genomes, offering a partial perspective on the history of its inhabitants. Here we present a paleogenomic analysis of 20 ancient human individuals from 6 archaeological sites, dated between 1,600-50 years before present (BP). We successfully captured over 100,000 SNPs from 15 of 20 individuals and recovered 12 full mitogenomes. All individuals presented unadmixed Indigenous South American ancestry and formed a separated genetic cluster relative to other ancient and modern South American genomes, indicating a unique Chonos ancestry. This Chonos ancestry falls within a broader cluster of late-Holocene Patagonian ancestry, most similar to the Kawesqar peoples who neighbour to the south. Within the Chonos, we distinguish a northern and southern ancestry cluster. The northern Chonos cluster exhibits some genetic connections to the present-time inhabitants of the neighbouring island of Chiloe, who are connected to Huilliche Indigenous history. Our findings reveal a point of contact between southern Chonos/Patagonian ancestry to the south and Mapuche ancestry to the north, and confirm that sea-faring subsistence was a knowledge transferred between Patagonian peoples.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.