Fig. 4: Effects of GPPmax and climate variables on start of season (SOS) based upon FLUXNET data.
a Regression coefficients (Slope) of normalized anomalies of SOS in current year in response to normalized anomalies of GPPmax in previous growing season (N = 28). b Piecewise structural equation model (SEM) considering both GPPmax and climate variables. The GPPmax refers to the maximum daily gross primary productivity (GPP) in each year. In a, the black dash lines indicate when slope is equal to zero, the box spans from the first to the third quartile, with intermediate values marked as the black line in the middle of the box, and the gray points represent the outliers whose values exceed 1.5 times the length of the box. In b, both climate factors (temperature, radiation, soil water content, precipitation, and CO2) and GPPmax were incorporated into the SEM to explore the direct (arrows from each climate factor directly point to the SOS) or indirect (arrows from each climate factor firstly directly point to GPPmax then to the SOS) effects of climate factors on spring phenology, with green lines indicating a negative effect and orange lines indicating a positive effect. The solid lines represent significant relationships (P < 0.05) between variables, while dashed lines represent no significant relationships between variables (P > 0.05). The calculated P values based on two-sided test and other statistics were listed in Supplementary Table 4.
