Genomic and Phenotypic Divergence in Wild Barley Driven by Microgeographic Adaptation

Abstract

Microgeographic adaptation is a fundamental driving force of evolution, but the underlying causes remain undetermined. Here, the phenotypic, genomic and transcriptomic variations of two wild barley populations collected from sharply divergent and adjacent micro-geographic sites to identify candidate genes associated with edaphic local adaptation are investigated. Common garden and reciprocal transplant studies show that large phenotypic differentiation and local adaptation to soils occur between these populations. Genetic, phylogenetic and admixture analyses based on population resequencing show that significant genetic divergences occur between basalt and chalk populations. These divergences are consistent with the phenotypic variations observed in the field. Genome sweep analyses reveal 162.7 Mb of selected regions driven by edaphic local adaptation, in which 445 genes identified, including genes associated with root architecture, metal transport/detoxification, and ABA signaling. When the phenotypic, genomic and transcriptomic data are combined, HvMOR, encoding an LBD transcription factor, is determined to be the vital candidate for regulating the root architecture to adapt to edaphic conditions at the microgeographic scale. This study provides new insights into the genetic basis of edaphic adaptation and demonstrates that edaphic factors may contribute to the evolution and speciation of barley.

Keywords: adaptive evolution; edaphic adaptation; genetic diversity; whole genome resequencing; wild barley.

Figure 1

Comparison of the phenotypic values of basalt and chalk wild barley populations and the edaphic composition of basalt and chalk soils. A) Elements in basalt and chalk soils. B) Agronomic performance of basalt and chalk wild barley during a common garden study. C,D) Phenotypic values from the reciprocal transplant experiments. C) Phenotypic value and plasticity of plants of the basalt wild barley population growing in basalt soil (BB) and chalk soil (BC). D) Phenotypic value and plasticity of plants of the chalk wild barley population growing in chalk soil (CC) and basalt soil (CB). *,p< 0.05; **,p< 0.01.

Figure 2

Summary of resequencing data of 33 wild barley plants. A) SNP density, B) INDEL density, C) Fst, D) Selected sweep regions (red bar, basalt; green bar, chalk), E) Tajimas’ D‐basalt (blue), Tajimas’ D‐chalk (purple), F) Pi‐basalt (light green), Pi‐chalk (light red).

Figure 3

Genetic diversity and population divergence of 33 wild barley accessions. A) Neighbor‐joining phylogenetic tree based on all of the SNPs. Red, basalt individuals; blue, chalk individuals. The reliability of each branch was evaluated by bootstrapping with 1000 replicates. B) Principal component analysis of these wild barley plants. The first principal component clearly separated the two populations. C) Population structure of two wild barley populations whenK= 2, 3, and 4. D) LD decay of the two wild barley populations. The x axis stands for physical distances (bp), whereas theyaxis stands forr ²values. E)δaδi analysis showing the demographic history of basalt and chalk wild barley population. They were split by a single divergence event with two different stages (T ₁≅ 0.0227 Mya andT ₂≅ 0.0098 Mya, respectively). Effective population size (Ne) remained unchanged across the two divergence stages, while different migration rate during (T ₁) and after (T ₂) last glacial maximum (LGM). The average number of migrants between them is shown between the black arrows.

Figure 4

Venn diagrams and GO enrichment of DEGs in wild barley. A) Overlapping and unique differentially expressed genes (DEGs) in the leaf and root organs of basalt and chalk accessions. B) Overlap of upregulated differentially expressed genes (DEGs) and select genes in the basalt and chalk populations. C) Overlap of downregulated differentially expressed genes (DEGs) and select genes in basalt and chalk populations. Upleaf, upregulated in the leaves; Downleaf, downregulated in the leaves; Uproot, upregulated in the roots; Downroot, downregulated in the roots. D–G) GO enrichment of differentially expressed genes in the leaves and roots.

Figure 5

HORVU4Hr1G056890 is the vital adaptive gene regulating root architecture for local adaptation. A) Homologous rice genes of selected genes between the basalt population and chalk population. B) Structure of HORVU4Hr1G056890. Orange rectangles, ORFs; gray rectangles, 5′ and 3′ UTRs; orange arrow, nonsynonymous nucleotide substitution in the basalt population compared with the chalk population; RE1, RE2, RE3, RE4, RE5, RE6, and RE7, AuxREs in the promoter region of HORVU4Hr1G056890 (the numbers in parentheses show the position of each element relative to the start of the 5′ UTR). C) Analysis of RPKMs of roots and leaves between basalt and chalk populations (blue, basalt; orange, chalk). D) Difference in root architecture of the basalt and chalk wild barley populations.