SEARCH

SEARCH BY CITATION

Fig. S1 Visualization of the correlation matrix for 62 climate and geographical variables reveals strong correlations among variables. Labels on the bottom axis denote groups of climate variables, which are listed on the right vertical axis. Single monthly names refer to the aridity index. The image was generated using the HEATMAP function in the base distribution of R (R Development Core Team 2007).

Fig. S2 Scree plots summarizing the application of PCA on 62 climate and geographical variables. (A). Scree plot based upon eigenvalues. (B). Scree plot based on PVE.

Fig. S3 Loadings for PCs 1–10 for 62 climate and geographical variables. Types of variables are differentiated by colour (black: latitude and longitude; blue: minimum monthly temperatures; red: maximum monthly temperatures; orange: monthly aridity index; green: monthly precipitation; yellow: growing-degree days above 5 °C). The PVE is listed for each PC.

Fig. S4 A summary of GO categories nested under the term molecular function for all 1730 genes. (A) The fraction of 1730 genes assigned a term nested within the molecular function category. (B). The distribution of level three GO terms for those genes assigned a GO term for a molecular function.

Fig. S5 A summary of GO categories nested under the term biological process for all 1730 genes. (A) The fraction of 1730 genes assigned a term nested within the molecular function category. (B). The distribution of level three GO terms for those genes assigned a GO term for a biological process.

Fig. S6 Distributions of diversity metrics for the 1730 SNPs genotyped using the Illumina Infinium platform are consistent with expectations for conifers and previous applications of this technology (Eckert et al. accepted). The distributions shown in panels A, B and D were smoothed using the DENSITY function in the base distribution of R (R Development Core Team 2007). (A). The distributions of expected (HEXP) and observed (HOBS) heterozygosity. Vertical dashed lines denote averages. (B). The distribution of Wright’s inbreeding coefficient (FIS). The vertical solid line denotes the average, while the vertical dashed red lines define the 99% bootstrap confidence interval (= 10 000 replicates) for FIS. (C). The allele frequency spectrum for the minor allele visualized as a set of 50 bins. The vertical red line denotes the minimum frequency threshold of 2.5% used for inclusion of SNPs. (D). The distributions of hierarchical fixation indices across loci reveal moderate differentiation among 54 populations across the genome of loblolly pine (FRT: hierarchical fixation index for regions relative to total; FCR: hierarchical fixation index for counties nested within regions; FCT: hierarchical fixation index for counties relative to the total). Further information is available in the Materials and Methods section. Vertical dashed lines denote averages.

Fig. S7 The correlation matrix estimated from a single draw from the posterior distribution of covariance matrices estimated using MCMC reveals broad patterns of population structure consistent with three genetic groups. Draws from the posterior distribution of Ω were similar, with coefficients of variation typically < 5% for most elements (Fig. S8). Colours denote correlations among counties reordered with hierarchical clustering (Euclidean distance and Ward’s linkage method). The colour scale ranges from yellow (higher correlations) to red (lower correlations). Labels refer to the genetic clusters identified previously (WMC: west of the Mississippi cluster, GCC: Gulf Coast cluster and ACC: Atlantic Coast cluster).

Fig. S8 Variation among draws of the variance-covariance matrix from its posterior distribution was small. Most elements of this matrix had a coefficient of variation less than 5%. Colours denote values ranging from a maximum of 6.9% (yellow) to a minimum of 0.001% (red).

Fig. S9 Relationship between climate variation and minor allele frequencies for the top two SNPs per climate PC as determined using Bayesian geographical analysis. Colours denote regional populations (black = WMC, green = GCC, red = ACC). Regional populations are defined in the Material and methods.

Table S1 Factor loadings for the top five PCs resulting from PCA on 62 geographical and climate variables

Table S2 Summary of SNPs (= 6) with very strong support (BF > 100) for PC3. This PC is comprised largely of winter and summer precipitation and summer aridity. Dashes indicate that the EST contig had no significant similarity to known gene models in Arabidopsis

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
MEC_4698_sm_TableS1-2-FigS1-9.pdf1182KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.