Extended Data Fig. 2: (a) The relationship between directly measured species-level absolute biomass and species-level absolute cover for the 7 sites that simultaneously collected these data; and (b-g) patterns of these 7 sites based on species level biomass data (black) and species level cover data (purple).

All data were natural log-transformed. Each regression curve in Fig. a represents one year (in different colors). Relationships between mean richness and (b) mean community biomass, and (c) mean absolute biomass of the two most dominant species, and (d) mean relative biomass of the two most dominant species; and (e) between mean relative biomass and mean absolute biomass of the two most dominant species; and between mean community biomass and (f) mean absolute biomass of the two most dominant species, and (g) mean relative biomass of the two most dominant species, of these 7 sites. The black dots and lines in Fig. b to g are the results of each pattern by directly calculating the absolute and relative biomass of the two most dominant species based on the species level biomass data. The purple dots and lines in Fig. b to g are the results of each pattern by indirect calculations of the absolute and relative biomass of the two dominant species based on species level cover data. The dashed and solid lines indicate that the overall relationship is not significant (P > 0.05) and significant (P < 0.05), respectively, with shaded areas indicating 95% confidence intervals. Different sites are represented by points of different shapes. Slopes in Fig. d and f are reported as the mean ± SEM. All statistical tests are conducted as two-sided.