Fig. 2: Effects of plant species richness on community productivity and community overyielding over time.
A, B Effects on community productivity and community overyielding (log response ratio (lnRR) of productivity of a mixture divided by the mean productivity of all monocultures of the component species) in terms of total aboveground biomass (log-10 scale) in grasslands (n = 39). Effects on (C, D) annual basal area increment and (E, F) total accumulated basal area (log-10 scale) in forests (n = 26). Points are community-level values for each plot in the respective year. Lines are mixed effect model fits across all experiments. For grasslands, species richness (F1,1317 = 157.8, P < 0.001), year (F1,5976 = 21.4, P < 0.001), and the species richness × year interaction (F1,5976 = 8.0, P = 0.005) significantly affected aboveground biomass, while species richness (F1,716 = 78.1, P < 0.001) and the species richness × year interaction (F1,3482 = 40.0, P < 0.001) significantly affected community overyielding. In forests, annual basal area increment was significantly affected by year (F1,3474 = 31.9, P < 0.001), and the species richness × year interaction (F1,3474 = 7.6, P = 0.006), while community overyielding was significantly affected by the species richness × year interaction (F1,1934 = 12.6, P < 0.001). Forest accumulated basal area was significantly affected by species richness (F1,1348 = 13.6, P < 0.001), year (F1,4023 = 2229, P < 0.001) and the species richness × year interaction (F1,4023 = 12.6, P < 0.001), while community overyielding was significantly affected by species richness (F1,781 = 4.9, P = 0.027), and the species richness × year interaction (F1,2611 = 19.4, P < 0.001). Reported P values were calculated from one-sided F-tests. Refer to Supplementary Table 2 for more details. Y-axis was trimmed to enhance resolution comparing model fit lines (5% extreme values are not visible).
