Figure 3
Linear regression of tiller number (TN; a–c) and root biomass (RB; d,e) based on the contribution of the Rondo allele for any of the QTLs identified for TN and RB, respectively, from the 2017GH (a,d), 2018GH (b,e), or 2017F study (c). The QTLs that are significantly associated with TN or RB (p < 0.05) regardless of environments or developmental stage are marked as ‘a’ while the QTLs that are significantly associated with TN under only one developmental stage, i.e. maximum tillering stage only or harvest stage only, are marked as ‘b’ or ‘c’, respectively. The multiple-regression model expresses the phenotype of TN and RB (y-axis) as a linear function of Rondo allele at TN-QTLs and RB-QTLs identified (x-axis). The proportion of the total phenotypic variation (y-axis) explained by each QTL (x-axis) was calculated as an R2 value, from the regressions of each marker/phenotype combination.
