(iii) Plant community-, environment- and soil-mediated
relationships between elevation and above- and belowground plant
pathogens
To explore how soil properties, plant community and environment
conditions mediated the relationship between elevation and above- and
belowground plant pathogens, we used the “piecewiseSEM” package (v.
2.1.2; Lefcheck, 2016) to build a structural equation model (SEM) based
on a series of linear mixed- effects models to test the relationship
between elevation and PL and sfpOTUs including paths that
were mediated by the plant community and environment (i.e. elevation may
first influence plant community characters, soil properties and
environmental conditions, which then indirectly change plant fungal
pathogen communities) (Halliday et al., 2021; Fig. 2a). We did not
include MDT in the SEM, given that there was a strong
collinearity between MDT and elevation (Pearson’ r= 0.985). We calculated the standardized path coefficients (scaled by
their mean and standard deviation) and corresponding significance
(P values) for each path of the final models. Statistically
significant path(s) (P < 0.1) highlight potentially
important mechanisms generating relationships between elevation and
above- and belowground plant pathogens. We used the Fisher’s C test to
test the goodness-of-fit of the SEM. We then used partial residual plots
to interpret the relationships highlighted in the model (Grace, 2006).
In brief, we extracted the residuals of each variable in SEM and then
set the residuals of plant pathogen related variables (i.e. PLand sfpOTUs ) as functions of residuals of explanatory variable
(i.e. Elevation , Soil PCA1 , Evenness andProneness ) in a series of linear models (Grace, 2006).
Significant results of residual analysis correspond to significant paths
in the SEM.