Original Article: Apple fruit texture QTLs: year and cold storage effects on sensory and instrumental traits: Ben Sadok et al. 2015
Supplementary Data
Table. S1 Paired sample t-test for texture trait mean values over storage periods in 2011: decreasing values at S2 are indicated in bold whereas increasing values are underlined.
Table. S2 PCA analysis for Instrumental traits: (a) Eigenvalues and Percentage of Explained Inertia by the first three components (b) Correlations among variables and PCs: significant values are indicated in bold and changes over storage are underlined.
Year / PC / Eigen value / %Variance / % Cumulative VarianceS0 / S2 / S0 / S2 / S0 / S2
PC 1 / 7.23 / 7.23 / 60.22 / 60.32 / 60.22 / 60.32
2007 / PC 2 / 2.57 / 2.86 / 21.48 / 23.86 / 81.70 / 84.19
PC 3 / 1.39 / 1.15 / 11.55 / 9.59 / 93.26 / 93.79
PC 1 / 6.16 / 6.63 / 51.33 / 55.32 / 51.33 / 55.32
2011 / PC 2 / 3.42 / 3.70 / 28.51 / 30.85 / 79.85 / 86.17
PC 3 / 1.81 / 1.01 / 15.13 / 8.43 / 94.99 / 94.61
(a)
Variables / S0 / S2Penetromety / PC1 / PC2 / PC3 / PC1 / PC2 / PC3
Fs_07 / 0.67 / 0.73 / 0.06 / 0.40 / 0.89 / -0.05
Dp_07 / 0.15 / 0.17 / 0.96 / -0.32 / 0.83 / 0.38
Stiff_07 / 0.59 / 0.69 / -0.38 / 0.83 / 0.26 / -0.43
Ws_07 / 0.62 / 0.66 / 0.41 / 0.15 / 0.96 / 0.15
FF_07 / 0.59 / 0.72 / -0.29 / 0.54 / 0.49 / -0.57
Compression
H1_07 / 0.94 / -0.32 / -0.03 / 0.98 / -0.14 / 0.14
H2_07 / 0.95 / -0.30 / 0.01 / 0.98 / -0.11 / 0.15
Grad1_07 / 0.91 / -0.14 / -0.15 / 0.91 / -0.05 / -0.24
Grad2_07 / 0.75 / -0.29 / -0.14 / 0.82 / -0.29 / -0.24
WH1_07 / 0.89 / -0.36 / 0.07 / 0.93 / -0.14 / 0.31
WH2_07 / 0.93 / -0.24 / 0.10 / 0.91 / 0.02 / 0.34
Cohesiv_07 / 0.92 / -0.31 / 0.08 / 0.93 / -0.07 / 0.33
Penetromety
Fs_11 / 0.35 / 0.93 / 0.07 / 0,21 / 0.96 / 0.09
Dp_11 / -0.19 / 0.85 / -0.44 / -0.22 / 0.86 / 0.36
Stiff_11 / 0.68 / 0.37 / 0,57 / 0.63 / 0.59 / -0.40
Ws_11 / 0.14 / 0.96 / -0.14 / 0.03 / 0.96 / 0.23
FF_11 / 0.40 / 0.67 / 0.51 / 0.23 / 0.77 / -0.52
Compression
H1_11 / 0.97 / -0.20 / -0.15 / 0.99 / -0.12 / 0.06
H2_11 / 0.98 / -0.13 / -0.15 / 0.99 / -0.10 / 0.07
Grad1_11 / 0.85 / -0.01 / 0.45 / 0.92 / 0.05 / -0.29
Grad2_11 / 0.56 / -0.48 / 0.54 / 0.79 / -0.36 / -0.26
WH1_11 / 0.88 / -0.19 / -0,42 / 0.95 / -0.15 / 0.27
WH2_11 / 0.91 / 0.04 / -0.37 / 0.93 / 0.07 / 0.28
Cohesiv_11 / 0.91 / -0.09 / -0.41 / 0.95 / -0.05 / 0.28
(b)
Table. S3 PCA analysis for sensory traits: (a) Eigenvalues and Percentage of Explained Inertia by the first three components (b) Correlations among variables and PCs: significant values are indicated in bold.
Year / PC / Eigen value / %Variance / % Cumulative VarianceS0 / S2 / S0 / S2 / S0 / S2
PC 1 / 4.42 / 4.37 / 63.18 / 62.53 / 63.18 / 62.53
2007 / PC 2 / 1.16 / 1.06 / 16.65 / 15.21 / 79.83 / 77.74
PC 3 / 0.46 / 0.55 / 6.59 / 7.86 / 86.43 / 85.60
(a)
Variables / S0 / S2Sensory / PC1 / PC2 / PC3 / PC1 / PC2 / PC3
CRUNCH / 0.82 / 0.28 / 0.30 / 0.88 / 0.00 / 0.22
FIRM / 0.87 / -0.08 / 0.22 / 0.84 / -0.26 / -0.09
JUIC / 0.68 / 0.59 / 0.06 / 0.71 / 0.58 / 0.01
MELT / -0.52 / 0.73 / -0.32 / -0.58 / 0.69 / 0.32
GRAIN / -0.91 / 0.09 / 0.31 / -0.90 / -0.05 / -0.20
FIBER / 0.82 / -0.37 / -0.34 / 0.88 / -0.15 / 0.28
MEALI / -0.86 / -0.24 / 0.10 / -0.70 / -0.41 / 0.53
(b)
Table. S4 Correlations between sensory and instrumental traits on the basis of the mean phenotypic values per genotype: significant correlation values are underlined between groups of texture descriptors and indicated in bold within each group of sensory, penetrometry and compression traits
Table. S5 Summary of the QTLs detected for the storage index (SI) of texture traits
Figure S1. Temperature records during the two studied harvest years 2007 and 2011: (a) minimal temperature (b) Maximal temperature.
GLOBAL MODEL RESULTS
Instrumental traits
Regarding traits related to flesh and skin firmness (i.e. FF, Fs and Ws), three QTLs were detected on LG5, LG6 and LG12 for FF in 2007 (FF_07; Table 5, Figure 4). A global model was built for these QTLs and showed no significant interaction between cofactors. Together, these QTLs explained 15.9% of the trait variability (Table 5). Two other QTLs were detected for FF at S0 on LG6 and LG12 (FF_07_S0; Table 5, Figure 4). The global model did not include any interaction between QTLs and explained 13.4% of the variability (Table 5). Similarly, two QTLs were detected on LG6 and LG12 for Fs at S0 and explained together 13.8% of the variability (Fs_07_S0; Table 5). Two other QTLs were detected for Fs at S2 on LG4 and LG6 (Fs_07_S2; Table 5). The global model did not include any interaction between QTLs and explained 25.9% of the variability (Table 5). Three QTLs were detected for S2-specific BLUPs of Ws on LG2, LG6 and LG16 (Ws_07_S2; Table 5, Figure 4). Together, these QTLs explained 30.69% of the variability (Table 5).
In 2011, two QTLs were mapped on LG1 and LG15 for FF (FF_11; Table 5, Figure 4). Together, the two QTLs explained 22.9% of the variability (Table 5). Moreover, three QTLs were mapped for Fs at S2 on LG1, LG5 and LG11 (Fs_11_S2; Table 5). The global model did not include any interaction between QTLs and explained 37.5% of the variability (Table 5).
Considering traits describing tissue elasticity (i.e. Dp and Stif), three QTLs were detected on LG2, LG12 and LG16 for Dp in 2007 (Dp_07; Table 5, Figure 4). A global model built for these QTLs showed significant interaction between the QTLs on LG2 and LG16 and explained 31.14% of the trait variability (Table 5). Moreover, two QTLs were detected on LG2 and LG16 for Dp at S2 (Dp_07_S2; Table 5, Figure 4). Together, the two QTLs showed significant interaction effect and explained 35.75% of the variability (Table 5). Regarding the fruit stiffness, co-localizing QTLs were detected for Stif_07 and Stif_07_S0 on LG6 and LG12 (Figure 4). The global model did not include any interaction between QTLs and explained 11.7% and 13.8% of the variability, respectively (Table 5). In 2011, two QTLs were detected for Dp at S2 on LG5 and LG11 (Dp_07_S2, Figure 4). Together, the two QTLs explained 13.4% of the variability (Table 5).
For traits describing global fruit hardness and elasticity (i.e. H1&2, WH1&2, Cohesiv and Grad1&2), three QTLs were detected on LG2, LG4 and LG12 for H2 in 2007 (H2_07; Figure 4). The global model did not include any interaction between QTLs and explained 28.73% of the variability (Table 5). Considering the positive compression force area, two QTLs were mapped for WH1 in 2007 on LG1 and LG12 (WH1_07, Figure 4). The global model did not include any interaction between QTLs and explained 16.21% of the variability (Table 5).
In 2011, five QTLs were detected for H1 on LG1, LG2, LG3, LG5 and LG8 (H1_11, Figure 4). Together, these QTLs explained 39.7% of the variability and showed significant interaction effect (P<0.05). Moreover, two QTLs were mapped for H1 at S0 on LG5 and LG15 (H1_11_S0, Table 5, Figure 4). The global model did not include any interaction between QTLs and explained 18.9% of the variability (Table 5). Likewise, three QTLs were detected for H2 at S0 on LG5, LG8 and LG15 (H2_11_S0; Table 5, Figure 4). Together, these QTLs explained 28.52% of the variability (Table 5). Again, three QTLs were mapped for WH1 in 2011 on LG1, LG5 and LG8 (WH1_11; Table 5, Figure 4). The global model did not include any interaction between QTLs and explained 37.68% of the variability (Table 5). Similarly, four QTLs were detected for WH2_11 on LG1, LG5, LG8 and LG15 (Figure 4). Together, these QTLs explained 40.31% of the variability and showed significant interaction effect (P<0.05), (Table 5). In addition, two QTLs were detected for the coefficient of elasticity Grad 1 on LG1 and LG8 (Grad1_11; Table 5, Figure 4). The global model did not include any interaction between QTLs and explained 27.53% of the variability (Table 5). Two other QTLs were mapped for Grad1 at S0 on LG3 and LG5 (Grad1_11_S0; Table 5, Figure 4). Together, these QTLs explained 11.11% of the variability and showed significant interaction effect (P<0.05), (Table 5).
Sensory traits
Two QTLs were detected for fruit meltiness in 2007 on LG5 and LG7 (Melt_07, Figure 4). The global model did not include any interaction between QTLs and explained 9.93% of the variability (Table 5). In 2011, four QTLs were mapped for Meali at S2 on LG2, LG5, LG7 and LG12. The global model did not include any interaction between QTLs and explained 20.69% of the variability (Table 5). Finally, two QTLs were detected for Melt at S2 on LG6 and LG7 (Melt_11_S2, Figure 4). Together, these QTLs explained 20.69% of the variability (Table 5).
8