Latte Age Saipan: Stones and Seafarers

Rising from the coral shores and volcanic ridges of the Mariana archipelago, the Latte Period communities of Saipan crafted a landscape of memory in stone. From around 1250 CE, builders raised paired latte pillars — upright megalithic supports — and arranged them into house platforms that anchored family groups to place. Archaeological excavations at Garapan, including the Best Sunshine Casino site, reveal dense midden deposits, stone-working debris, and house-floor traces that mark continuous occupation through the late precontact centuries.

These monumental forms are part of a longer story of Oceanic voyaging and settlement that links the Marianas with Island Southeast Asia and Near Oceania. Radiocarbon dates from charcoal and shell place intensive latte construction and occupation between the 13th and 18th centuries CE. Limited evidence suggests regional interaction in pottery styles, exotic tool materials, and botanical remains, but the full networks of exchange remain incompletely mapped. The dramatic appearance of latte architecture signals social investment in lineage houses and territorial markers — an island choreography of kinship, labor, and seafaring memory.

Ancient DNA from 46 individuals associated with the Latte context at Garapan provides one of the clearer genetic windows into late precontact Mariana Island populations. The maternal lineages are dominated by haplogroup E subclades: E2a (27 individuals), E (13), E2 (5), and a single B lineage. This distribution suggests strong matrilineal continuity with broader Austronesian-derived populations of Island Southeast Asia and parts of Near Oceania, where mtDNA E variants are known to occur. The presence of a B lineage is not unexpected given regional mtDNA diversity and possible interactions across ocean networks.

Crucially, common Y-chromosome haplogroups were not reported for these samples, so paternal ancestry patterns remain unresolved; without Y-DNA or genome-wide autosomal summaries, inferences about sex-biased migration, admixture, or later colonial-era male-mediated gene flow are limited. With 46 samples, these mtDNA patterns are statistically informative compared with very small series, but interpretation still requires caution: sampling is concentrated at Garapan, temporal spans cover several centuries, and post-contact demographic disruptions could have reshaped genetic landscapes. Archaeogenetic evidence aligns with archaeological signals of Austronesian cultural roots while leaving open questions about the timing and directionality of specific gene flow events.