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

Abstract

Introductory paragraphAll 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 and 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 heterozygosity. In fact, i) the majority of changes at coding genes and linked regions that differentiate wild and domestic genomes are already present in the ancient genomes analyzed here; ii) all ancient domestic genomes dated between 600 and 2,500 years ago are highly variable -at least as variable as modern genomes from the wild, and single seeds from modern cultivars show reduced variation when compared to ancient seeds, indicating that intensive selection within cultivars in the last centuries likely partitioned ancestral variation within different genetically homogenous cultivars. When cultivars from different Andean regions are pooled, genomic variation of the pool is higher than that observed in the pool of ancient seeds from north and central western Argentina. Considering that most desirable phenotypic traits are likely controlled by multiple polymorphic genes [10], a plausible explanation of this decoupling of selection and genetic erosion is that early farmers applied a relatively weak selection pressure [2] by using many phenotypically similar but genetically diverse individuals as parents. Our results imply that selection strategies during the last few centuries, as compared to earlier times, more intensively reduced genetic variation within cultivars, and produced further improvements focusing on few plants carrying the traits of interest at the cost of marked genetic erosion within Andean landraces. Main textThe onset of domestication by early farmers has been suggested as the period during which the most intense genetic bottleneck affecting genome-wide diversity occurred [4,11]. Yet, artificial selection, possibly at different rates in different times, was likely a continuous process that produced both landraces and, more recently under modern breeding programs, high-yielding and more resistant elite cultivars [2,12,13]. Understanding when the majority of genetic diversity was lost, and how such loss is related with the intensity o...

Common bean as a potential crop for future food security: an overview of past, current and future contributions in genomics, transcriptomics, transgenics and proteomics

“…There was three-fold greater genetic diversity reduction in the Mesoamerican gene pool compared to the Andean gene pool [33]. Very recently Trucchi et al [34], investigated the temporal dynamics of genetic diversity and selection throughout the domestication process of the common bean in the southern Andes. It was reported that the genomic variation of the pool of cultivars from different Andean regions is higher than the pool of ancient seeds from the central-western and north Argentina.…”

Harnessing Crop Wild Diversity for Climate Change Adaptation

“…In short, ML [138,[176][177][178][179] and deep learning approaches [133,170,180,181] promise assisting the conservation [102,[182][183][184], managing [185,186], prioritization [187][188][189], and introgression [190,191] of crop wild variation from genebanks (Figure 2). ML may be particularly useful in unexplored isolated pockets of diversity, which contain allelic variants otherwise eroded from modern genotypes [192,193].…”

Multivariate genomic analysis and optimal contributions selection predicts high genetic gains in cooking time, iron, zinc, and grain yield in common beans in East Africa

“…Genomic information is an important breeding tool to improve selection accuracy and to accelerate genetic gain in beans (Delfini et al., 2021; Diaz et al., 2021; Keller et al., 2020). However, selection for desirable traits also results in erosion of genetic diversity (Esquinas‐Alcázar, 2005; Trucchi et al., 2021) and inefficient selection procedures in crop breeding may lead to premature plateauing of genetic gain (Cowling et al., 2017). The rate of erosion of genetic diversity in breeding programs may be reduced and selection efficiency improved through optimal contributions selection (OCS), which balances genetic gain against loss in genetic variation (Cowling et al., 2017, 2019; Gorjanc et al., 2018; Kinghorn, 2011; Kinghorn & Kinghorn, 2021; Woolliams et al., 2015).…”