Fig. 3 Relative importance of structural diversity, functional
dominance, multiple- and single-trait functional diversity metrics used
as predictors in the multiple linear model. Statistical fit summaries
are presented in Table 1
We analysed the direct and indirect response of AGB to diversity metrics
by considering only independent important variables for AGB, as retained
in the previous selected model. The outputs of the SEM testing for
species richness effect on AGB via functional evenness and structural
diversity revealed chi-square values ranging from 0.08 to 1.72 andP > 0.05 (Table 2), indicating good fit of the SEM
to the data. The values of cfi (close to 1), and rsmea (<0.16)
also demonstrate good fits of the models (Table 2).
Table 2. Summary of fit statistics for the SEMs relating
species richness, functional evenness and functional diversity metrics
to AGB; cfi: comparative fit index; and rmsea: root mean square error of
approximation