Extended Data Fig. 3: Precipitation legacy shapes the genetic differentiation of bacterial lineages within taxa.

To assess genetic differences among bacterial lineages along the precipitation gradient, we selected 33 bacterial species (15 of the bacterial biomarkers of precipitation, plus 18 additional abundant and prevalent taxa) that exhibited high genome coverage in our metagenomic dataset as proxies for the broader bacterial communities. Reference genomes for each species were retrieved from the NCBI Genome database, and filtered shotgun metagenomic reads were mapped to these genomes to identify high-quality biallelic single nucleotide polymorphisms (SNPs). Genetic distances between the bacterial lineages were calculated based on the identified SNPs and PCoA plots were generated and coloured by a. soil collection sites and b. mean annual precipitation. The variance explained by each axis is indicated. c. The Manhattan plot illustrates significant SNPs associated with precipitation, derived from the genetic-environment association (GEA) analysis. The GEA was conducted using a two-sided test of coefficients from a general linear model, with precipitation at each sampling location as the environmental variable. The significance threshold was calculated using the permutation method to account for multiple association testing. The x-axis of the plot represents the SNP positions along the genomes of the selected bacterial species, while the y-axis displays the -log₁₀ p-values from the association model. The horizontal line indicates the statistical significance threshold, as determined by the permutation test. SNPs above this threshold, highlighted in red, were significantly associated with the precipitation gradient. Bacterial taxa with fewer than 1,000 high-quality biallelic SNPs after filtering, and with no significant SNPs detected from the GEA analysis, are not shown in the plot.