To simplify the interpretation of the different dimensions of the
PCA, the factors were orthogonally rotated following the VARIMAX
transformation with Kaiser normalisation method (Marchi et al.,
2012, pp. 1e34) (Fig. 1A and B). A new distribution of the total
variance explained by each component was obtained (D1¼46.36%,
D2¼21.7% and D3¼20.46%), maintaining the accumulated variance for thefirst three components (88.52%). Although the variance
accumulated by thefirst three components remained the same, the
distribution among the components was more homogeneous
(Bonany et al., 2014).