Data collection
Our empirical approach to study the relationship between elevation and plant pathogens might be sensitive to the relatively small number of replicates across a single environmental gradient in a single year, as trophic interactions can vary across space and time, and are often context dependent (Roslin et al., 2017; Liu, Chen, et al., 2020). Therefore, to test the generality of our main results from the field survey, we conducted a systematic literature search in ISI Web of Science and China National Knowledge Infrastructure (www.cnki.net). We searched for research on foliar fungal pathogen [(fungal OTU* OR fungi OTU* OR fung* operational taxonomic unit OR fung* abundance OR fung* richness) AND (elevation* OR altitud*) AND (folia* OR leaf OR leaves)], foliar fungal diseases [(plant disease* OR pathogen* OR infect* OR epidemic*) AND (inciden* OR prevalen* OR load* OR severity OR occur* OR abundance) AND (elevation* OR altitud*) AND (folia* OR leaf OR leaves)] and soil plant pathogens [(fungal OTU* OR fungi OTU* OR fung* operational taxonomic unit OR fung* abundance OR fung* richness) AND (elevation* OR altitud*) AND (soil OR belowground OR underground)]. We finally identified 41 papers (providing a total of 62 effect sizes) that met our criteria (Fig. S1.1; Table S1.1): (i ) focused on the relationship between elevation and foliar fungal diseases and foliar/soil plant pathogens in nonagricultural ecosystems; and (ii ) reported sample sizes greater than three. Detailed process for literature screening and basic information of 41 papers were provided in Fig. S1.4 and Table S1.1.
We collected the OTU table for studies on foliar and soil plant pathogens, identified the putative plant pathogens according to the aforementioned methods, and calculated ffpOTUs (i.e. foliar fungal pathogen OTU richness), sfpOTUs and sfpRA for each study. All plant pathogenic genera identified in the meta-analysis are listed in the Table S2.1. We extracted the sample sizes and Pearson’s correlation coefficients (r ) from the main text, tables, figures (using WebPlotDigitizer v. 4.4; Rohatgi, 2020), or raw data. We also recorded background information on location and the elevation range of sampling (as highest sampling elevation minus lowest sampling elevation) from original papers, then we extracted the mean annual temperature and mean annual precipitation of the lowest elevation location for each study based on the WorldClim database (Fick and Hijmans, 2017).