Fig. 3.

Principal component analysis of PDMS TD-GC–MS data of all melon cultivars. A PCA model was produced for the PDMS TD-GC–MS VOC data where all five melon cultivars were included, each group was based upon n = 9 (3 biological replicates × 3 technical replicates). The PCA model was based upon the first 5 PCs. PC1 (42.1% TEV) was plotted against PC2 (20.4% TEV) to produce a PC scores plot (a) and a PC loadings plot (b). The identifications of each of the PC reference numbers applied in the PC loadings plot (b) are given in Table 1.