Fig. 6: Piecewise structural equation modeling (piecewiseSEM) accounting for the direct and indirect effects of the parameters of the leaf economics spectrum and abiotic factors on the effect size of phosphorus (P) addition on plant biomass (lnRR).
a The effects of PC1, climate, soil properties, and fertilization regimes on the effect size (lnRR) as revealed by piecewiseSEM. PC1 was derived from the principal component analysis (PCA) using the eight leaf functional traits as Fig. 4, and depicted the leaf economics spectrum, which can be characterized by a single acquisitive-conservative axis. Dark and blue arrows indicate positive and negative correlations, respectively. Solid and dashed lines indicate significant and insignificant correlations, respectively (* p < 0.05, ** p < 0.01, *** p < 0.001). Arrow width is proportional to the strength of the relationship. The numbers adjacent to arrows are path coefficients, which indicate the standardized effect size of the relationship. The R2 values beneath each response variable are the proportion of variance explained by both observed and latent variables. Global goodness-of-fit tests, including AIC, Fisher’s C, and tests of directed separation (p > 0.05) are shown below the model. All statistical tests are conducted as two-sided. b Standardized direct, indirect, and total effects derived from the piecewiseSEM. AIC Akaike information criterion, df degree of freedom, duration fertilization duration, soil [TN] soil total nitrogen concentration, soil [TP] soil total phosphorus concentration, MAT mean annual temperature, MAP mean annual precipitation. Source data are provided as a Source Data file.
