Complex patterns of local adaptation in teosinte

Abstract

Populations of widely distributed species encounter and must adapt to local environmental conditions. However, comprehensive characterization of the genetic basis of adaptation is demanding, requiring genome-wide genotype data, multiple sampled populations, and an understanding of population structure and potential selection pressures. Here, we used single-nucleotide polymorphism genotyping and data on numerous environmental variables to describe the genetic basis of local adaptation in 21 populations of teosinte, the wild ancestor of maize. We found complex hierarchical genetic structure created by altitude, dispersal events, and admixture among subspecies, which complicated identification of locally beneficial alleles. Patterns of linkage disequilibrium revealed four large putative inversion polymorphisms showing clinal patterns of frequency. Population differentiation and environmental correlations suggest that both inversions and intergenic polymorphisms are involved in local adaptation.

Keywords: Zea mays; admixture; inversion; mexicana; parviglumis; population structure.

Fig. 1.—

Map of sampled Zea mays ssp. parviglumis and ssp. mexicana populations.

Fig. 2.—

Diversity statistics. (A) Proportion of SNPs deviating from Hardy–Weinberg Equilibrium (HWE), proportion of polymorphic SNPs, and mean inbreeding coefficient F_IS. (B) Length and number of ROH and average pairwise length of genomic segments IBS.

Fig. 3.—

LD reveals structural rearrangements in teosinte. Shown are LD (r², red) and permutation P value (black) for pairs of SNPs across chromosomes 1, 4, and 9. Dashed black lines delineate the likely boundaries of structural variants discussed in the text. P value and r² cutoffs were chosen to visualize the LD blocks against the background level of LD.

Fig. 4.—

Population structure in teosinte. (A) Principal component analysis of all individuals, labeled according to the sampled population. (B) STRUCTURE results for all individuals. Individuals are grouped by population and populations ordered by increasing altitude.

Fig. 5.—

Differentiation and environmental correlation in inversions compared with genome-wide distribution of test statistics. Joint distribution of F_CT (A) and F_ST (B) versus heterozygosity under a hierarchical island model for all SNPs. The black line indicates the 1% tail based on simulations. Values for each inversion when treated as a single locus indicated with yellow. Extreme outlier loci indicated with red. (C) Bayes factors for PC1 in mexicana. Values are plotted across all 10 chromosomes, with each chromosome in a different color. Black dots represent outlier SNPs, and black horizontal bars below the plots indicate the positions of inversions.

Fig. 6.—

Fold enrichment of observed ratios of genic/nongenic SNPs among candidates. Fold of enrichment for each set of candidates (red line) from BAYENV, F_ST, and F_CT, SNPs significant in any of the population-based F_FT and PHS analysis, and phenotypic association analysis (GWAS). Expected distribution of fold of enrichment ratios (gray dots) was obtained by random sampling of the same amount of SNPs as in the candidate set 1000 times and calculating the ratio and fold enrichment. SNPs within inversions were excluded.