Fig. 4: Structural equation models (SEMs) of hypothesised causal relationships among genomic traits and their ecological drivers. | Communications Biology

Fig. 4: Structural equation models (SEMs) of hypothesised causal relationships among genomic traits and their ecological drivers.

From: The MetaInvert soil invertebrate genome resource provides insights into below-ground biodiversity and evolution

Fig. 4

a Initial SEM with hypothesised links; b–f SEMs fitted to all taxa, and to major taxonomic groups. Arrows indicate hypothesised or modelled relationships, positive (black) or negative (red). Links marked with grey arrows were not statistically significant in the SEM. Fisher’s C evaluates conditional independence claims among nodes and indicates model fit, with p-values showing whether discrepancies between the model and the data are statistically significant. Degrees of freedom are marked with df. Values next to arrows show standardised estimates, with asterisk indicating the statistical significance of the relationship (*p < 0.05, **p < 0.01, ***p < 0.001). Animal silhouettes originate from phylopic.org, and they can be reused under Creative Common licences.

Back to article page