Recent growth changes in Western European forests are driven by climate warming and structured across tree species climatic habitats

Table 7 Main statistics for the PLS regression models of BAI against historical environmental predictors for the eight species under study

	N comp	Q ² (%)	R ² (%)	Number of selected variables	Cumulated partial R ² of selected variables (%)
P. abies	3	31	34	10	11
A. alba	3	22	23	8	10
P. sylvestris	4	11	13	9	4
F. sylvatica	2	26	26	7	18
Q. robur	3	8	10	9	2
Q. petraea	5	13	16	13	4
Q. pubescens	2	5	10	6	3
P. halepensis	4	9	19	7	4

See Eq. 3. We reported the number of selected components (N comp), the predictive power assessed by cross-validation (Q ²), the goodness-of-fit (R ²), the number of predictors retained in the PLS regressions after aggregating variables showing seasonal patterns and the partial R ² associated to the finally selected predictors

ISSN: 1297-966X