Ramos-Madrigal, J.,Smith, B. D.,Moreno-Mayar, J. V.,Gopalakrishnan, S.,Ross-Ibarra, J.,Gilbert, M. T. P.,Wales, N.

The complex evolutionary history of maize (Zea mays L. ssp. mays) has been clarified with genomic-level data from modern landraces and wild teosinte grasses [1, 2], augmenting archaeological findings that suggest domestication occurred between 10,000 and 6,250 years ago in southern Mexico [3, 4]. Maize rapidly evolved under human selection, leading to conspicuous phenotypic transformations, as well as adaptations to varied environments [5]. Still, many questions about the domestication process remain unanswered because modern specimens do not represent the full range of past diversity due to abandonment of unproductive lineages, genetic drift, on-going natural selection, and recent breeding activity. To more fully understand the history and spread of maize, we characterized the draft genome of a 5,310-year-old archaeological cob excavated in the Tehuacan Valley of Mexico. We compare this ancient sample against a reference panel of modern landraces and teosinte grasses using D statistics, model-based clustering algorithms, and multidimensional scaling analyses, demonstrating the specimen derives from the same source population that gave rise to modern maize. We find that 5,310 years ago, maize in the Tehuacan Valley was on the whole genetically closer to modern maize than to its wild counterpart. However, many genes associated with key domestication traits existed in the ancestral state, sharply contrasting with the ubiquity of derived alleles in living landraces. These findings suggest much of the evolution during domestication may have been gradual and encourage further paleogenomic research to address provocative questions about the world's most produced cereal.