Figure legends
Figure 1  Schematic diagram showing vertical root-placement patterns of individual species within the 0-30 cm soil zone. Each number in the box indicates a unique species with the solid black lines indicating species-specific roots. For each soil cube, species root-placement patterns were assigned with one of the seven possible rooting types based on their root depths. In the scenario of vertical root aggregation, species with similar root-placement patterns (a) or relative root distributions (b) are more likely to co-occur. In the scenario of vertical root segregation, species with different root-placement patterns (c) or relative root distributions (d) are more likely to co-occur. The width of the shapes in panel (b) and (d) is proportional to the species’ relative root abundance.
Figure 2  Root species richness (a) and relative root abundance (b) along the 0-100 cm soil profile. Different letters indicate significant difference among soil zones.
Figure 3  Relative root distributions for the 53 most common species within the 0-30 cm soil zone. Panel (a) shows the frequencies of species occurrences in the 0-10, 10-20 and 20-30 cm soil zones. Panel (b) shows the frequency distribution of species relative root abundance in the 0-10 cm soil zone. Species with asters suggest that soil zone had a significant effect on their relative root distributions, based on which the 53 species were grouped into five groups. Group A, relative root abundance in the 0-10 cm soil zone was significantly higher than the deeper two zones, while the deeper two showed no significant difference; Group B, relative root abundance in the 0-10 cm zone was the highest and it decreased significantly toward the deeper zones; Group C, relative root abundance did not differ significantly among soil zones; Group D, relative root abundance was the highest in the 0-10 cm soil zone while for some species, relative root abundance in the 20-30 cm zone was higher than the 10-20 cm zone; Group E, relative root abundance was the highest in the 10-20 soil zone; ns, nonsignificant. *P< 0.1, **P < 0.05, ***P < 0.01.
Figure 4  Vertical root segregation evaluated based on root-placement patterns (a) within the 0-30 cm soil zone and relative root distributions (b-d) in the 0-10 cm soil zone. The similarity in relative root abundance of co-occurring species was evaluated with mean pairwise functional distances (MPD; b), standard deviation (SD; c) and variance (Variance; d). The histograms illustrate the frequency distribution of the standardized effect size of the corresponding parameters across 604 root neighbourhoods. In panel (a), a positive value indicates that co-occurring species in the 0-30 cm soil zone tended to avoid co-occurring in the same soil zone, whereas a negative value indicates that co-occurring species were more likely to co-occur in the same soil zone than expected. In panels (b)-(d), a positive value indicates that co-occurring species tended to avoid placing similar relative root abundance in the 0-10 cm soil zone, whereas a negative value indicates that co-occurring species tended to place more similar relative root abundance than expected in the 0-10 cm soil zone. The dashed red line in each panel indicates the mean value of the standardized effect size of the corresponding parameters.
Figure 5  Pairwise correlations of relative root abundance for the 29 most common species in the 0-10 (a), 10-20 (b) and 20-30 cm (c) soil zone. Black solid lines link two species with significant positive correlations, whereas blue dashed lines link two species with significant negative correlations. The size of the lines is proportional to the strength of the correlations.