Sumidouro: Lagoa Santa, 10,100 BP

Beneath the limestone roof of Caverna do Sumidouro, the earth keeps a quiet archive of a cooling world at the end of the Pleistocene. Archaeological data indicates human presence in the Lagoa Santa region during the Early Holocene; the five dated individuals here fall between 8612 and 7607 BCE. These remains sit within a landscape of gallery forests, savanna patches, and freshwater lakes that framed human lifeways after the last glacial retreat.

Genetically, the assemblage ties to pan-American founding lineages: Y-chromosome haplogroup Q appears in four of the five males, while mitochondrial haplogroups D1 and D are represented across the sample. This pattern is consistent with broader models in which haplogroup Q dominates male lineages in many early and later Native American populations, and D1 is one of the canonical maternal founder clades. Archaeological contexts—fragmentary burials, stone tools, and faunal remains—suggest mobile hunter-gatherer lifeways, though preservation and excavation histories vary.

Limited evidence suggests population continuity in eastern Brazil across the Early Holocene, but with only five genetic samples the picture is provisional. Archaeological signatures and aDNA together paint an evocative origin story: small groups moving through postglacial environments, carrying genetic variants that would persist across the continent. Further sampling is required to resolve migration routes, local continuity, and possible interactions with contemporaneous groups.

The recovered aDNA from five individuals at Caverna do Sumidouro provides a rare window into early-Holocene genetics in eastern South America. Y-chromosome data show haplogroup Q in four male samples — a lineage widely observed among Native American populations and interpreted as part of the initial male-mediated colonizing pool. Maternal lineages are dominated by mitochondrial haplogroup D: three samples assigned to D1 and two to broader D clades. Haplogroup D1 is one of the established founder mtDNA lineages in the Americas.

These results align with continental patterns where Q and D-derived mtDNA lineages are common, supporting models of rapid dispersal and subsequent local differentiation. Nevertheless, low sample count (n=5) and potential limits in sequencing coverage constrain resolution: subclade assignments (e.g., Q sublineages or specific D1 variants) may be unresolved, and genetic diversity estimates are sensitive to sampling error. Admixture signals, if present, would require larger datasets and genome-wide SNP data to detect and date reliably.

Archaeogenetics thus offers a complementary narrative to the material record: it suggests continuity with broader Native American ancestry yet leaves open questions about regional structure, migration routes into eastern South America, and demographic changes through the Holocene. In short: suggestive, consistent with expectations, but preliminary.