Crop wild relatives that have experienced multiple and independent domestication events provide an excellent model for understanding adaptation processes in crop populations and a first and relevant aspect to investigate is the geographic origin of landraces. The aim of this research was to establish the origin of Mesoamerican and Andean Lima bean (Phaseolus lunatus L.) landraces by analyzing chloroplast DNA and ITS polymorphisms in a sample of 59 wild and 50 landrace accessions. According to seed size, genetic distance analyses, and haplotype networks, at least two independent domestication events are proposed. The first one would have taken place in the Andes of southern Ecuador–northwestern Peru and would have given rise to the large‐seeded landraces collectively known as the “Big Lima” cultivars. The second one would have taken place in central–western Mexico, more likely in the area to the north and northwest of the Isthmus of Tehuantepec. This event, along with post‐domestication migrations of landraces in South America, would have given rise to the great variety of small‐seeded Mesoamerican landraces that exist today. We did not find any evidence supporting the existence of two discrete groups within Mesoamerican landraces that might correspond to the previously proposed “Sieva” and “Potato” cultigroups. A severe reduction in genetic diversity because of domestication, known as the “founder effect”, was detected, which may have implications for the conservation of genetic resources in this species.

Moche art, deer and bean warriors on a pottery, Lambayeque, Peru, 700-800 years A.D.