Table S1 Vegetation map units (Salas et  al. 2010) and reclassification for the present study (e.g. Fig. 3).

Table S2 Analysis of variance summary for soil nutrient and quadrat data, which were averaged prior to analysis.

Table S3 Summary statistics for 16 Helianthus microsatellite loci and analysis of molecular variance (amova) results.

Table S4 Locus summary statistics averaged across sunflower subpopulations, grouped by habitat type.

Table S5 Multiple regression of distance matrices (MRDM) to identify the optimal ratio of landscape conductance between barren sand and vegetated sunflower habitat.

Table S6 Multiple regression of distance matrices (MRDM) to identify environmental variables that are associated with genetic distance.

Fig. S1 Comparison of variables among habitat types.

Fig. S2 Summary of Structure analysis of microsatellite data.

Fig. S3 Non-heirarchical clustering of individuals.

Fig. S4 Mean likelihood of the data given K clusters within (a) dune and (b) non-dune samples, with 10 replicates at each value of K.

Fig. S5 Subpopulation estimates of Shannon’s diversity index (Ha) negatively related to the proportion of dune ancestry, Q, (a) and positively related to the distance outside the dunes (b).

Fig. S6 Summary of individual-level spatial autocorrelation analysis with 1000 m distance classes.

Fig. S7 Genetic distances between pairs of populations in each habitat type.

Fig. S8 Relationship of genetic distance matrix with geographic distance, landscape distance and with differences in soil nitrogen and vegetation cover.

MEC_5454_sm_TableS1-S6-FigS1-S8.pdf340KSupporting 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.