Tropical forests face the greatest combined pressure from land use and climate change, so predicting their resilience requires understanding how human management shapes biodiversity, including at the genetic level. The Maya managed Central American forests for around 3,500 years, providing a natural experiment for studying these legacies over centuries rather than decades. As part of the interdisciplinary Bladen Legacy project, which combines genomics with archaeological and palaeobotanical evidence from a major Maya site in Belize, this project compares two species at opposite ends of the domestication continuum. Cacao (Theobroma cacao) is an economically important crop that the Maya intensively selected and traded, and fewer than 30 trees remain at the field site. The cohune palm (Attalea cohune) was never selected in the same way, yet it is abundant and unusually common at Maya sites, suggesting deliberate management whose lasting influence is unresolved. We will generate chromosome-level reference genomes for both species and use whole-genome resequencing and target-capture data to characterise genetic diversity across their ranges in relation to gradients of past Maya land use. This will reveal how sustained human management has shaped genetic diversity and whether, and how quickly, populations recover after abandonment. Because genetic variation underpins species’ ability to adapt to a changing planet, these insights will help predict the long-term consequences of today’s land use.