*2.6. Statistical Analysis*

Data were firstly tested for normality and homogeneity of variance using Shapiro– Wilk and Bartlett's tests, respectively. Thereafter, data were subjected to two- or three-way ANOVA according to the experimental design using SPSS v. 25.0 programme. When

the means differed significantly, they were separated by Student's or Tukey HSD tests (α = 0.05), when the factors were applied at two or more levels, respectively. A PCA was performed with the object principal normalization method on data collected from 2016 to 2018 regarding total phenols and nutrient concentration in cones. The principal components were retained considering the eigenvalues superior to 1 and the scree plot. Internal consistency was measured with Cronbach's alpha. In addition, the scores of each one of the PCA components were calculated as a function of plant vigour, foliar treatment, limestone treatment, cultivars and year, and subjected to analysis of variance, using the Tukey–Kramer HSD test (α = 0.05) to compare averages for each trial and year. A correlation analysis was applied to the same data as the PCA analysis with the Spearman coefficient.
