Ancient genomes reveal early Andean farmers selected common beans while preserving diversity

Abstract

All crops are the product of a domestication process that started less than 12,000 years ago from one or more wild populations [1, 2]. Farmers selected desirable phenotypic traits, such as improved energy accumulation, palatability of seeds or reduced natural shattering [3], while leading domesticated populations through several more or less gradual demographic contractions [2, 4]. As a consequence, erosion of wild genetic variation [5] is typical of modern cultivars making them highly susceptible to pathogens, pests and environmental change [6,7]. The loss of genetic diversity hampers further crop improvement programs to increase food production in a changing world, posing serious threats to food security [8,9]. Using both ancient and modern seeds, we analyzed the temporal dynamic of genetic variation and selection during the domestication process of the common bean (Phaseolus vulgaris) that occurred in the Southern Andes. Here we show that most domestic traits were selected for prior to 2,500 years ago, with no or only minor loss of whole-genome variation. In fact, i) all ancient domestic genomes dated between 600 and 2,500 years ago are highly variable - at least as variable as a modern genome from the wild; the genetic erosion that we observe in modern cultivars is therefore a recent process that occurred in the last centuries; ii) the majority of changes at coding genes that differentiate wild and domestic genomes are already present in the ancient genomes analyzed here. Considering that most desirable phenotypic traits are likely controlled by multiple polymorphic genes [10], a likely explanation of this decoupling of selection and genomic erosion is that early farmers applied a relatively weak selection pressure [2] by using many phenotypically similar but genomically diverse individuals as breeders. Selection strategies during the last few centuries were probably less sustainable and produced further improvements focusing on few plants carrying the traits of interest, at the cost of marked genetic erosion.