Fig. 4

Bayesian cluster analysis of 124 rutabaga accessions from Norway, Sweden, Finland, Denmark and Iceland estimated using the software STRUCTURE based on 6861 SNP markers. The optimal value of K, determined by the method of Evanno et al. (48) with populations unassigned (A) or assigned (B) to their respective countries, as well as by the method of Puechmaille [37] and Li and Liu [38] (C), suggested that the 124 rutabaga accessions could be placed into 3 or 4 clusters (K = 3 most likely). Detailed Bayesian clustering of 124 rutabaga accessions by the CLUMPAK program [39] (D-E). Each column or rectangular bar represents the individual rutabaga accessions used in this study, while each colour represents one gene pool and the stacked bars with different colours represent admixtures with their shared ancestry components (D). Simplified view suggests three ancestral populations (E)