Repeated global adaptation across plant species

Abstract

Global adaptation occurs when all populations of a species undergo selection toward a common optimum. This can occur by a hard selective sweep with the emergence of a new globally advantageous allele that spreads throughout a species' natural range until reaching fixation. This evolutionary process leaves a temporary trace in the region affected, which is detectable using population genomic methods. While selective sweeps have been identified in many species, there have been few comparative and systematic studies of the genes involved in global adaptation. Building upon recent findings showing repeated genetic basis of local adaptation across independent populations and species, we asked whether certain genes play a more significant role in driving global adaptation across plant species. To address this question, we scanned the genomes of 17 plant species to identify signals of repeated global selective sweeps. Despite the substantial evolutionary distance between the species analyzed, we identified several gene families with strong evidence of repeated positive selection. These gene families tend to be enriched for reduced pleiotropy, consistent with predictions from Fisher's evolutionary model and the cost of complexity hypothesis. We also found that genes with repeated sweeps exhibit elevated levels of gene duplication. Our findings contrast with recent observations of increased pleiotropy in genes driving local adaptation, consistent with predictions based on the theory of migration-selection balance.

Keywords: global adaptation; plant genomics; pleiotropy; repeated adaptation; selective sweep.

Fig. 1.

Geographical distribution and relatedness within and among the study species. (A) Sampling locations of the 17 datasets included in the study. (B) Time-calibrated phylogenetic tree (retrieved from https://timetree.org/) of the 17 datasets, based on 12 reference species. (C) fastSTRUCTURE ancestry pie plot (K = 5) of the Helianthus argophyllus dataset in Texas (USA), showing substantial substructuring. (D) Relatedness filtering summary bar plot by dataset. Datasets labeled by corresponding color from B.

Fig. 2.

Orthology assignment summary of the final set of 13,268 orthogroups. (A) Bar plot of the number of paralogs per orthogroup for each species. (B) Bar plot of the number of species included in each orthogroup, for the orthogroups of each species. (C) Distribution of the 13,268 tested orthogroups across species number. (D) Comparison between the TimeTree and Orthofinder2 phylogenies, each based on 12 reference species.

Fig. 3.

Evidence for repeated selective sweeps across multiple species. (A) Distribution of PicMin FDR (−log10 PicMin FDR on the Y axis) for the 13,268 tested orthogroups, ordered by number of putative driving genes on the X axis (number of OmegaPlus emp-P < 0.1). Points above the red line have FDR < 0.1. (B) Species contribution to PicMin top candidates (FDR < 0.1), calculated for each species as: [number of empirical P-values < 0.1 in the PicMin FDR < 0.1 orthogroups]/[total number of orthogroups with PicMin FDR < 0.1 tested]. (C) Heatmap of OmegaPlus empirical P-values for the 33 candidate orthogroups (PicMin FDR < 0.1); driving genes cells (emp-P < 0.1 – black cells) are outlined in red, species are ordered by phylogenetic distance along the Y axis. A white cell indicates that an orthogroup was not tested in a species.

Fig. 4.

Orthogroups with repeated sweeps tend to be less pleiotropic and harbor more duplications. (A) Pleiotropy assessment of the relaxed set of 107 orthogroups with FDR < 0.2, against 10,000 random draws. Red triangles represent the average pleiotropy of the candidate set, while black circles represent the mean of 10,000 random draws. Black lines represent the 95% interval of the random draws means. Each row represents a different pleiotropy measure, labeled on the Left. (B) Pleiotropy Stouffer’s Z of the top 33 orthogroups (FDR < 0.1), calculated using different pleiotropy measures labeled on the Left. Dotted lines delimit the 95% CI. (C) Pleiotropy Stouffer’s Z for the more conservative set of 15 orthogroups identified by intersecting the 33 candidate OGs of the main analysis (FDR < 0.1) with results derived from nine additional PicMin omitting closely related Eucalyptus and Helianthus species. Dotted lines delimit the 95% CI. (D) Duplication bootstrapping assessment of the top 33 orthogroups (FDR < 0.1) against 10,000 random draws. The red triangle represents the average duplication number in the candidate set of orthogroups, while the black circle and line represent mean and 95% interval of 10,000 random draws respectively.