Fig. 5: Effects of functional diversity and identity on productivity and stability in wetlands under different levels of anthropogenic disturbance.
a–d, f–i Predicted productivity (a–d) and stability (f–i) as functions of FDis of resource economic traits (a, f), FDis of plant size traits (b, g), CWMPC1 (c, h), and CWMPC2 (d, i) across anthropogenic disturbance levels (DIS; least (n = 269), intermediate (n = 573), and most (n = 293) disturbed wetlands). The predictions are based on multiple linear regression models with interactions between functional indices and DIS (n = 1135). Lines represent mean predicted values, with nonfocal predictors fixed at their means. The shaded areas represent 95% confidence intervals. Points represent partial residuals for the focal functional indices in each panel. The F and P-values of the interaction terms (from ANOVAs of the multiple linear regression models, two-sided F-tests) are shown in each panel. e, j Standardized regression coefficients for productivity (e) and stability (j) for each predictor across DIS levels, derived from the same models (n = 1135). Points indicate coefficient estimates, with error bars showing 95% confidence intervals. The dashed line at zero denotes no effect; error bars crossing this line indicate that the coefficient is not statistically significant (at P < 0.05). The adjusted R2 and P-values of the multiple linear regression models (two-sided F-tests) are given. Detailed ANOVA results of the multiple linear regression models and the most parsimonious models are given in Supplementary Table 6. High CWMPC1 values reflect more acquisitive traits, and high CWMPC2 values indicate larger plant size. FDis_RE, functional dispersion of resource economics traits; FDis_PS, functional dispersion of plant size traits. All variables were scaled. Source data are provided as a Source Data file.
