Taxonomic and functional characteristics of communities
Based on the fuzzy-weighted species composition, the grassland communities were classified into five groups, which form a functional gradient from groups G1 to G5 (Figure 2). At one end group G1 comprises communities with high LDMC, LA and predominantly C4species, whereas at the other end group G5 represents communities with high SLA, leaf N and predominantly C3. Most of the variation across communities is related to the leaf economic spectrum, represented in the PCA ordination. The first principal component explained 94.97% (P-value < 0.0001) of the variation and was positively correlated to SLA, leaf N and predominance of C3 species in the communities, and negatively correlated to LDMC and LA. The second principal component explained only 2.91% (P-value = 0.001) and was positively correlated with high SLA, and leaf N, while negatively correlated with LDMC, LA and C3species.
The correlations between biodiversity descriptors varied across community functional groups (Figure 3). The correlations between species richness with species diversity, and between species diversity with FR were positive for all community groups. For G1 and G5 communities, all relationships were positive, however, they were stronger for G1. For G2 communities, the correlations between species richness were negative with both FR and functional response diversity. Moreover, species diversity and functional response diversity were also negatively correlated. Regarding G3 and G4 communities, the correlations between functional response diversity were negative for both, species diversity and FR.
In general, anomalous climatic events triggered more positive than negative effects of biodiversity on both communities’ resistance and resilience, and their estimated effects varied across community groups and climatic events (Figures 4, 5, S2, and Tables S1 and S2).
Biodiversity effects on resistance
We found analogous taxonomic components effects on communities’ resistance. Such as a gradient in the effects of species richness and species diversity on the resistance of the communities under all anomalous events, ranging from strong and positive in group G1 to weaker and negative in group G5 (Figure 4a-b). Species diversity (Gini-Simpson) effects on resistance were positive in G2 and G3 communities under both extreme dry and wet events, while in G4 and G5, its effects were negative under both moderate and extreme wet events (Figure 4b).
Regarding the functional components of biodiversity, we also found similar effects on communities’ resistance. For instance, both functional redundancy and functional response diversity effects on the resistance of G1 communities were strong and positive under all anomalous events, whereas variable and weaker in the other community groups (Figure 4c-d). In addition, functional redundancy effects were positive in G3 and G5 under extreme dry anomalous events, while under extreme wet they were positive in communities G2 and negative in G4 (Figure 4c). The effect of functional response diversity was negative for G2, G3 and G4 communities under extreme dry anomalous events (Figure 4d), yet under moderate dry events, it was positive in G3 and G5 (Figure 4d).