Figure 2
Simultaneous representations of principal component map and hierarchical genotype clustering. Factorial plan is represented with the first two dimensions of nine, defined based on average variables (detailed in Table7) on two replicates for each of the 43 rice genotypes studied. Dimension 1 is significantly (p < 0.01) and positively correlated to (in decreasing effect order): SDW ΔDC/C, NBL ΔDC/C, NBT ΔDC/C, DRonset ΔDC/C, and SOURSTAc and negatively to: SDWc, NBLc, NBTc, DRonsetc, SOURSTA ΔDC/C, SINKHEX ΔDC/C. Dimension 2 is significantly (p < 0.01) and positively correlated to (in decreasing effect order): SOURSUC ΔDC/C, SINKSUC ΔDC/C, SINKSTA ΔDC/C, DRonset ΔDC/C, SOURSTA ΔDC/C, SDW ΔDC/Cand negatively to: SINKSTAc, SINKHEXc, LDIMc, SOURHEXc, SOURSTAc, SINKSUCc. Resulting genotype clusters are delimited with dashed lines. Black squares represent the center of gravity for each cluster ( i.e. highest nodes of the hierarchical clustering). Since clusters were defined base on nine dimensions, two varieties close together can be in the same cluster or not (if they differ on dimensions apart from 1 & 2).