henhaobio
Sign in

NaturePlant BiologyPeer-reviewed journal article

A haplotype-based evolutionary history of barley domestication

Yu GuoiD, Murukarthick JayakodiiD, Axel HimmelbachiD, Erez Ben‐YosefiD, Uri DavidovichiD, Michal DavidiD, Anat HartmanniD, Mordechai E. KisleviD, Tzion FahimaiD, Verena J. SchuenemanniD, Ella ReiteriD, Johannes KrauseiD, Brian Joel SteffensoniD, Nils SteiniD, Ehud WeissiD, Martin MascheriD

PDFDOIPreprintCodeData
In-depth full-text digest available · Jump to digest ↓

Abstract

Barley is an old crop with a complex history. Its evolution and the molecular basis of domestication have been intensely studied. This research has ruled out a single origin of the crop and motivated a model of mosaic genomics ancestry. As more and better genome sequences have become available, this concept can be refined: where do the building blocks of the mosaic, that is, the haplotypes come from? Were all cultivated barleys created equal or did some wild barley populations contribute more to some domesticated lineages? To answer these and other questions, we pursue a haplotype-based approach to connect diversity and population structure in wild and domesticated barley. We sequence the genomes of 628 genebank accessions and 23 archaeological specimens. Using these data, we infer the spatiotemporal origins of haplotypes and map the contributions of different wild barley populations either during the initial phase of domestication or through later gene flow. Ancient DNA sequences corroborate our genome-wide analysis in present-day samples. Our results indicate that an early domesticated founder population formed in the Fertile Crescent during an extended period of pre-domestication cultivation. A practical implication of our findings is that the high haplotype differentiation between barley populations, which arose possibly without, and most likely on top of, selective forces, complicates the mapping of adaptive loci.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (21 Dec 2024) · CC BY 4.0

All five wild populations fed an early barley founder population in the Fertile Crescent during extended pre-domestication cultivation; about 91% of wild-matched haplotypes split 32 to 8 ka BP.

Earlier work ruled out a single barley origin and proposed mosaic ancestry, but relied on crop-ascertained markers or reduced-representation sequencing. This work groups whole-genome sequences of wild and domesticated barleys into ancestral haplotype groups (AHGs), so the nearest wild relative and approximate entry date can be assigned to individual genomic windows rather than whole plants, and checks the picture against ancient DNA from 23 archaeological grains.

Key results

  1. ShownDomesticated barley is a mosaic of wild haplotypes. Over half of its haplotypes have a wild match, but few wild haplotypes reached the crop, and most of the shared ones are rare in the wild today.

    55.9% of domesticated haplotypes occur in at least one wild sample vs 7.0% of wild haplotypes found in a domesticate; 79% of wild-matched domesticated haplotypes are below 5% frequency in the wild. Domesticated divergence is bimodal, peaking near 98 SNPs per Mb (about 8000 years) and 6,500 SNPs per Mb (530,000 years); the haplotype-group cutoff was 400 SNPs per Mb (32,000 years). · A haplotype-based view of barley evolution; Definition of ancestral haplotype groups

  2. ShownAll five wild populations contributed to domesticated barley, mostly while the progenitor formed and domestication began. Later wild gene flow is minor overall but notable in some regions; the authors prefer it over a second, eastern domestication centre as the explanation.

    About 91% of wild-matched domesticated haplotypes split from their wild counterpart 32 to 8 ka BP, fewer than 9% later. Southern Levant, Northern Levant and Central Asia each supplied 20 to 27% of haplotypes; Syrian Desert 16.3% and 'Western Asia' 12.9%. Post-8 ka BP wild introgression makes up 12.8% of Syrian SYR-THM haplotypes. Poets et al. (5,000 SNPs) gave the Southern Levant > 50%. · The origins of domesticated haplotypes in time…; Analysis of the AHG matrix

  3. SuggestedDomestication alleles at BTR1/2, VRS1 and NUD trace to different wild populations and arose at different times. The btr1 haplotype is dated well before the earliest domesticated archaeobotanical remains.

    btr1 haplotype estimated at 25 ka BP, some 15,000 years before the earliest domesticated barley remains; inferred order btr1, btr2, vrs1.a1, nud, vrs1.a3, vrs1.a2, vrs1.a4. Closest wild matches: Southern Levant (btr1, nud, vrs1.a3), Northern Levant (btr2, vrs1.a2), Northern Mesopotamia and Central Asia (vrs1.a1, vrs1.a4). Dates use μ = 6.13×10^-9 (from Brachypodium) and 1-year generations. · A single-gene view of mosaic ancestry; Demographic history of domesticated barley; Evolutionary analyses in wild barley

  4. SuggestedGrains from the Southern Levant dated 6000 to 2000 cal BP were already domesticated types. Their two-rowed and six-rowed forms were genetically distinct and fit within modern barley lineages.

    23 grains from three sites in Israel all group with cultivated barley in PCA and carry btr1Btr2. Two-rowed grains (Yoram Cave, Timna) are closest to extant Western Asian barleys; six-rowed Abi'or cave grains (about 2 kya) resemble Mediterranean ones. All but TU1120 share the extant crop's 1H pericentromeric haplotype, probably of Central Asian origin. · Persistent population structure revealed by…; Archaeological excavations; Ancient DNA sequencing and analysis

  5. SuggestedDomesticated populations are strongly differentiated by haplotype, which the authors suspect reflects local lineage sorting. This could make outlier scans for selective sweeps unreliable, and the known domestication genes were not found by scans for low diversity.

    Seven regions had a haplotype Shannon index of 1 or less in domesticated but not wild barley; Btr1/2, Vrs1 and Nud were not among the low-diversity outliers. Pairs of domesticated populations share no haplotype across 7.5% of the genome on average. Ethiopian ETH-MNS and Central Asian CA-SNS hulless barleys carry the same 17-kb NUD deletion yet share no haplotype across 44.8% of the genome. · A haplotype-based view of barley evolution; A single-gene view of mosaic ancestry; Haplotype-based genetic diversity and selective…

Methods, data and reproducibility

Whole genomes of wild and domesticated barley were mapped to MorexV3, and coverage-normalised pairwise SNP distances among 367 high-coverage samples fed IntroBlocker to define ancestral haplotype groups in 100 kb windows. Rerunning at 400 to 24 SNPs per Mb thresholds dated when wild haplotypes entered the crop, and the averaged ADMIXTURE ancestry of wild carriers assigned their source.

  • Wild barley accessions (285 from the Wild Barley Diversity Collection, 95 mainly from the Russell et al. panel), Illumina WGS at about 10-fold · barley (Hordeum vulgare subsp. spontaneum) · n = 380 sequenced; 251 unadmixed high-coverage samples analysed · ENA PRJEB65046, PRJEB56087, PRJEB53924 (study-wide; per-genotype codes in Supplementary Table 1)
  • Domesticated barley accessions from 15 populations, selected from the Milner et al. GBS diversity space; seeds from the IPK Gatersleben genebank · barley (Hordeum vulgare) · n = 302 (116 at about 10-fold, 186 at about 3-fold) · ENA PRJEB65046, PRJEB56087, PRJEB53924 (study-wide; per-genotype codes in Supplementary Table 1)
  • Genotyping-by-sequencing (GBS) data of a global domesticated barley panel (Milner et al.), used for sample selection and IBS comparison with ancient grains · barley (Hordeum vulgare) · n = 19,778
  • Ancient barley grains from Yoram Cave, Timna Valley Site 34 and Abi'or Cave (Israel), radiocarbon dated 6000 to 2000 cal BP; includes 5 previously published Yoram Cave grains · barley (Hordeum vulgare) · n = 23 (7 high-coverage)
  • Hordeum pubiflorum whole-genome data used as outgroup · Hordeum pubiflorum
  • Barley pangenome assemblies (Jayakodi et al.) used for chromosome 5H alignments and as a wild-barley reference for sweep scans · barley (Hordeum vulgare) · n = eight assemblies aligned on 5H
Methods & tools
IntroBlocker, Minimap2, bcftools, SAMtools, PLINK2, ADMIXTURE, smartpca, PSMC, ADMIXTOOLS (qpDstat), mapDamage2.0, SnpEff, Illumina NovaSeq 6000
Software & environment
Versions given: bcftools 1.15.1, SAMtools 1.16.1, PLINK 1.9 and 2.00a3.3LM, ADMIXTURE 1.23, smartpca 7.2.1, novosort 3.06.05, PopLDdecay 3.42, SnpEff 4.3t, MUMmer 4.0.0, ADMIXTOOLS 3.0, R 3.5.1; no container or dependency file is mentioned.
Reproducibility gaps
The repository holds shell and Perl scripts, but several steps cite unnamed custom or published Perl scripts and no IntroBlocker version is given. The EVA variant matrix (PRJEB79752) is described as forthcoming, and the AHG matrix link is a temporary reviewer link.
Limitations (6)
  • Molecular dates are approximate. They rest on an uncertain mutation rate (6.13×10^-9, taken from Brachypodium distachyon), and allele ages mark only the earliest split among extant domesticates because the donor wild lineage is unsampled, so true origins should be older. · stated by the authors
  • The ancient grains come from a narrow area of the Southern Levant and say nothing about Eastern or Ethiopian barley. · stated by the authors
  • The panel does not saturate haplotype diversity in distal chromosome regions, so sampling more wild barleys could reveal more haplotypes shared with the crop. · stated by the authors
  • The seven putative sweep regions are too large to point to any single candidate gene. · stated by the authors
  • Assigning a wild source assumes that present-day wild population structure reflects the structure when cultivation began. Haplotypes from extinct or unsampled wild lineages cannot be placed. · noted in this summary
  • The threshold-based dating and source assignment are not tested on simulated data with known introgression histories. Sample counts also disagree: the abstract gives 628 genebank accessions, but the methods give 380 wild plus 302 domesticated. · noted in this summary

Read this if you work on crop domestication or wild-to-crop introgression and want a window-level haplotype approach (IntroBlocker plus divergence-time thresholds), or barley whole-genome and ancient DNA data to reuse.

Citations
32
FWCI
14.94
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISPankin '16Tan '25Bedford '26Liu '26Poets '15Abrouk '18
85% semantic matchPlant Biology
A. Pankin, J. Altmüller, C. Becker+1 more·bioRxiv·Aug 19, 2016