Table 1.

Multiple linear stepwise regressions explaining synchrony (â_C) patterns from changing growth–climate relationships (correlation coefficients; r) and climate variability (standard deviations of climate factors; SD) over 1930–2009

Synchrony	Initial variable	Initial r2	Initial MSE	Final stepwise model	Final r2	Final MSE
Siberia	Correlation with climate factors (rTempAPR, rTempJUN)
	rTempAPR	0.63**	0.020	âC = 0.26 ‒ 0.29 rTempAPR	0.63**	0.020
	Correlation with climate factors and variability (rTempAPR, rTempJUN, SDTempAPR, SDTempJUN)
	SDTempJUN	0.63**	0.020	âC = ‒0.47 + 0.28 SDTempJUN + 0.10 SDTempAPR	0.75***	0.013
Spain	Correlation with climate factors (rTempFEB, rPrecMAY, rPrecJUN, rSPEIMAY, rSPEIJUN)
	rSPEIJUN	0.43*	0.003	âC = 0.18 + 1.03 rSPEIJUN ‒ 0.93 rPrecJUN + 0.08 rTempFEB	0.75***	0.002

Codes for the variables are as in Figs. 4 and 5 (Upper) for Siberia and Spain, respectively. MSE, mean squared error.

^*P < 0.05; **P < 0.01; ***P < 0.001.