(i) Changes of abiotic and biotic factors along elevational gradients
We conducted principal component analysis (PCA) to summarize soil properties (W, pH, C, NO3- and NH4+) using the “vegan” package (v. 2.5.7; Oksanen et al., 2020) (Fig. S1.2). We then calculated the Spearman rank-order correlation between Soil PCA1 and each of the soil properties (W, pH, C, NO3- and NH4+) (Fig. S1.2). We also calculated the plant species richness (SR ) and Pielou’s evenness index (Evenness ) using the “vegan” package for each plot. To test correlations among various variables, we plotted the correlation matrix for all biotic or abiotic variables (Elevation , SR ,Evenness , Proneness , AGB, BGB, MDT, MDH andSoil PCA1 ), and we then calculated the Pearson’s correlation between these variables and PL /various soil pathogen indices (Fig. S1.3a). We also conducted Mantel tests based on “Bray-Curtis” distance between sfpOTUs and biotic or abiotic variables using the “ggcor” package (v. 0.9.8.1; Huang et al., 2020) (Fig. S1.3b).
To minimize the influence of the potential spatial autocorrelation on the results, we used the respective coordinates of each sample plot to generate a spatial matrix. Specifically, given the sample plots were distributed along a cambered mountain slope which approximated to the spherical surface, we calculated the spatial matrix based on spherical correlation structure (i.e. “corSpher” class) using the “nlme” package (v. 3.1-152; Pinheiro et al., 2021). We then introduced the spatial matrix into a series of linear mixed-effects models with five elevations as a random effect in following analyses, using the “nlme” package. We set Elevation as independent variables in a series of linear mixed-effects models to test its associations with various community-level indices (SR , Evenness , Proneness ,AGB and BGB ) and soil properties (SoilPCA1 ), respectively. At the plant community level, we set Elevation as the independent variable and PL and soil pathogen indices (sfpOTUs andsfpRA ) as response variables in a series of linear mixed-effects models to test the direct correlations between Elevation and above- and belowground plant pathogens.
For the soil fungal pathogen community, we conducted permutational multivariate analysis of variance (PERMANOVA) to test the compositional difference of soil fungal pathogens along the elevational gradient. A significant result of PERMANOVA supports the hypothesis that pathogen communities change along elevational gradients.