Large haploblocks underlie rapid adaptation in the invasive weed Ambrosia artemisiifolia

Abstract

Adaptation is the central feature and leading explanation for the evolutionary diversification of life. Adaptation is also notoriously difficult to study in nature, owing to its complexity and logistically prohibitive timescale. Here, we leverage extensive contemporary and historical collections of Ambrosia artemisiifolia-an aggressively invasive weed and primary cause of pollen-induced hayfever-to track the phenotypic and genetic causes of recent local adaptation across its native and invasive ranges in North America and Europe, respectively. Large haploblocks-indicative of chromosomal inversions-contain a disproportionate share (26%) of genomic regions conferring parallel adaptation to local climates between ranges, are associated with rapidly adapting traits, and exhibit dramatic frequency shifts over space and time. These results highlight the importance of large-effect standing variants in rapid adaptation, which have been critical to A. artemisiifolia's global spread across vast climatic gradients.

Fig. 1. A phased diploid genome assembly of Ambrosia artemisiifolia.

A The distribution of gene and repeat density in 1 Mbp windows across the 18 chromosomes of each haplotype. B Alignments of the 18 chromosomes of each haplotype. C Benchmarking Universal Single-Copy Orthologs (BUSCO) results for each haplotype assembly (h1; h2), each chromosome-only assembly (h1-chr; h2-chr), and the gene annotations for haplotype 1 (h1-annot) using the eukaryota odb10 gene set. Source data are provided as a Source data file.

Fig. 2. Signatures of climate adaptation in Ambrosia artemisiifolia.

A GWAS −log₁₀p-value of mixed model association against genomic location for flowering onset (solid line indicates a Bonferroni-corrected p-value of 0.05). B Distribution of a strongly-associated non-synonymous SNP in ELF3 among modern A. artemisiifolia populations used in this study. C Genome-wide XtX scans between sampling locations within each range separately. Solid lines indicate Bonferroni-corrected significance derived from XtX p-values; dashed lines indicate the top 1% of genome-wide XtX values. Green highlights represent the top 5% of 10 kbp WZA windows for each scan that are also among the top 5% of EAA WZA windows for at least one environmental variable, with dark green indicating outlier windows shared between North America and Europe. Pale blue bars indicate the location of 15 haploblocks (putative chromosomal inversions) that overlap shared outlier windows. Source data are provided as a Source data file.

Fig. 3. Temporal signatures of selective sweeps in Europe.

A Distributions of F_ST between historic and modern samples and the ratio of historic to modern nucleotide diversity (θ_πH/θ_πM) from Berlin and Bordeaux, and F_ST against genomic location. Red points indicate putative selective sweep windows, which are in the top one percent of per-window F_ST and θ_πH/θ_πM (dashed lines). B Strong evidence for a selective sweep on chromosome 2 in European populations corresponds with local divergent population structure (MDS1), indicating the presence of a haploblock (putative chromosomal inversion; pale blue) in this region. C A standardized measure of allele frequency change, y_t (calculated according to Eq. 1) for shifts between historic and modern populations across putatively neutral SNPs (histograms) and selective sweep candidates (red lines). Source data are provided as a Source data file.

Fig. 4. Population-genomic signatures of inversions correspond to inversions segregating in the diploid reference.

A Divergent local population structure identifies putative inversions (pale blue regions) from population-genomic data. B Alignments of homologous chromosomes reveal inversion polymorphisms segregating in the diploid reference genome. Source data are provided as a Source data file.

Fig. 5. Haploblock distributions and trait associations.

A Logistic regression models with error bands representing 95% CI of least-squares regressions (see Supplementary Table 9 and Supplementary Data 5–9 for model details) of haploblock frequency (allele 1) against latitude for four haploblocks across five time bins ranging from most historic (purple) to most modern (green). B Examples of significant associations between haploblock alleles and phenotypes (n = 121 biologically independent samples; boxes denote mean +/− SEM for each genotype). C hb-chr5b allele frequency in modern A. artemisiifolia populations. Source data are provided as a Source data file.