Single or multiple trajectories for the spread of maize cultivation into South America?

Maize is the cultivated form of a grass called teosinte, which grows wild in parts of southern and western Mexico. The prehistoric Americans first cultivated maize about 9000 years ago, probably close to the Rio Balsas river in western Mexico. By the time that Europeans arrived in the 15th century AD, maize was being grown throughout the Americas, and was the staple crop for many indigenous civilisations such as the Maya and the Inca. How did maize spread from its centre of origin in Mexico throughout North and South America? The first model for the spread of maize into South America arose from a large study of chromosome structures in different varieties from all over the continent, carried out in the 1970s. Subtle variations in chromosome structure, in particular the positions of small knobs, enabled the genetic relationships between the varieties to be deduced. The pattern of spread was then inferred by comparing these genetics relationships with the geographic location of each variety. The resulting model suggested that maize was initially introduced into the central Andes and from this location spread extensively throughout the highland and lowland regions of the continent, not being significantly supplemented by other types of maize until new varieties spread southwards along the east coast of Brazil in relatively recent times. This model was held to be generally correct until 2002 when it was replaced by the results of a more detailed study of variable DNA sequences called microsatellites, again in different varieties of maize grown throughout South America. This new model holds that the southwards movement of maize from Mexico led first into the lowlands of South America, and only then into the Andes Mountains. Unfortunately, there is a problem with both these models. The geographical distribution of the varieties being studied is a critical component of the analysis, as this distribution, when compared with the genetic relationships between different plants, is what reveals the pattern of spread. This means that in order for the analysis to work, the present-day distribution of the maize varieties must be similar to the pattern set up when maize first spread into South America. But this is unlikely to be the case, because Europeans have had a major impact on South American agriculture, and in particular have been responsible for the extensive movement of maize varieties from one place to another. This movement has created 'noise', obscuring the prehistoric geographic signal and disguising the original pattern of spread. We therefore need to strip off the modern noise to get closer to the prehistoric geographic signal. In a previous project we started to do this by examining archaeological maize specimens that predate the arrival of Europeans, as well as indigenous crops collected from remote native villages which until recently were unaffected by European influences. Our genetic analysis of these archaeological and indigenous specimens suggested that both previous models might be incorrect, and that instead there were two independent routes for the spread of maize cultivation into South America, one along the Andes on the western side of South America, and the second along the lowlands of the northeast coast. Our project was very limited in scope, involving examination of a single short DNA sequence in just eleven indigenous varieties and ten archaeological specimens, most of these from Brazil, and so we cannot yet be sure if our model is accurate. The purpose of the proposed project is to test our model by obtaining substantially more genetic data from a greater number of archaeological specimens and indigenous varieties, from a much broader geographical range throughout South America.