Supporting information

Figure S1: Multivariate analysis of fruit metabolomic profiles according to growing season and orchard of origin. PCA score plots were visualized according to growing season (a) and orchard (d). Supervised OPLS‐DA was used to evaluate year‐associated variation, showing the score scatter plot (b) and the corresponding S‐loading plot (c). Supervised O2PLS‐DA was used to evaluate orchard‐associated variation, showing the score scatter plot (e) and the corresponding loading plot (f).

Figure S2: Multivariate analysis of leaf metabolomic profiles according to growing season and orchard of origin. PCA score plots were visualized according to growing season (a) and orchard (d). Supervised OPLS‐DA was used to evaluate year‐associated variation, showing the score scatter plot (b) and the corresponding S‐loading plot (c). Supervised O2PLS‐DA was used to evaluate orchard‐associated variation, showing the score scatter plot (e) and the corresponding loading plot (f).

File S1: Datamatrix including the list of putatively identified metabolites.

File S2: Heat map data showing the relative abundance of metabolites selected from the O2PLS–DA loading plots for fruits and leaves.

Table S1: Agronomical data.

Table S2: Meteorological data from the three weather stations representative of the sampled production area during the 2014 and 2015 growing seasons.

Table S3: Metabolite values (mg/100 g fresh weight) per growing season.

Table S4: Metabolite values (mg/fruit or leaf) per growing season.