Figure 3

(a) Ordination plot of sampling locations generated by principal component analysis (PCA) of 31 environmental variables. (b) Environmental variables involved in the discrimination of the study locations defined by the first two principal components (PC). Only the top four contributing variables per PC are labelled, others are shown in grey. For more information see Supplementary Tables S1 and S11. (c) Neutral genetic population structure of the studied Arabidopsis halleri populations based on PCA. We used the allele frequencies of 500 000 randomly selected SNPs. (a,c) Black circles and grey triangles represent metalliferous (M) and non-metalliferous (NM) sites, respectively. (b) PrecSMAA – precipitation sum maximum annual amplitude; PrecSSu – precipitation sum summer; PrecSWM – precipitation sum wettest month; C_org – organic carbon content; C_tot – total carbon content; Ca_tot – total calcium content; P_Olsen – phosphorous content (Olsen method); Zn_tot – total zinc content.