Fig. 3: The relationships between plant species diversity and spatial variability of productivity across 83 globally distributed grasslands sites of the Nutrient Network.

Both a alpha (slope and 95% CIs = −0.026 (−0.038 to −0.015)) and b gamma (−0.007 (−0.013 to −0.001)) diversity were negatively associated with the spatial variability. c Beta-diversity, in contrast, was positively associated with spatial variability (1.57 (0.59 to 2.54)); d alpha (−0.06 (−0.08 to −0.04)); and e gamma (−0.03 (−0.04 to −0.02)) diversity were negatively associated with species covariation. f Beta-diversity, in contrast, was not associated with species covariation (−1.27 (−2.92 to 0.38)). g Species covariation, in turn, was positively associated with spatial variability of productivity (0.48 (0.44 to 0.53)). a–g Different colors represent different sites (see Fig. 2 for site color key assignment), major lines (in turquoise) represent the fixed-effect linear regression slopes among sites and small colored lines show patterns within sites. h Structural equation model (SEM) analysis showing the direct and indirect pathways through which different scales of diversity determine the spatial variability of biomass. Model fit was assessed using Shipley’s test of d-separation (Fisher’s C = 8.82, df = 6, P = 0.2). Solid blue arrows and solid orange arrows represent significant (P ≤ 0.05, no multiple comparison adjustments made) positive and negative paths, respectively, and light-gray arrows represent non-significant paths that were included in the initial model. Test of significance of path coefficients are two-sided for a difference from 0. Bidirectional arrows represent paths that were modeled as correlated errors (i.e., bidirectional relationships instead of causal and unidirectional relationships). Numbers next to the arrows are averaged effect sizes as standardized path coefficients; arrow widths reflect these standardized effect sizes. For spatial variability of biomass and species covariation, the marginal (i.e., explained by the fixed factors alone) and conditional (i.e., explained by both the fixed and the random factors; in parentheses) percent of variance explained is shown below and to the right of the variable name (see Supplementary Table 6 for non-standardized coefficient values and exact P values of individual paths).