Data and code for: Global Hotspots of Mycorrhizal Fungal Richness are Poorly Protected

Data and code for: Global Hotspots of Mycorrhizal Fungal Richness are Poorly Protected Folder structure 1_rarefaction: this folder contains code and data used to generate rarefied richness values for each location. 2_geospatial: this folder contains code and data used to train the geospatial RF models used to predict mycorrhizal fungal richness. 3_visualisations: this folder contains code used to generate the figures and tables in the manuscript. 99_exploratory_analyses: this folder contains the code used to perform exploratory spatial analyses. 99_misc: various files. The biomes subdirectory contains .tif files to reproduce the hotspot x biome spatial analysis. Usage To reproduce the results presented in the publication, please run scripts in sequential order. That is, first execute the rarefaction scripts in folder 1_rarefaction/code, when these are completed proceed to use the output data to run the scripts in 2_geospatial/code. Make sure to run scripts in sequential order, as indicated by the file names. To create the stack of covariate layers (refered to as "composite" subsequent scripts) in Google Earth Engine, start by downloading all .tif files from source; refer to 2_geospatial/data/ExtendedDataTable1.csv for the download URLs. Next, upload these to GEE and create a composite image using 2_geospatial/code/functions/create_composite.js. Required packages Make sure to install the following R packages: data.table, RColorBrewer, iNEXT, janitor, doParallel, caret, sf, tidyverse. In our analysis, R version 4.3.2 was used. To run the Python scripts, make sure to install: pandas, geopandas, numpy, subprocess, time, datetime, earthengine-api, shap, fasttreeshap, matplotlib, pathlib, multiprocessing, functools, contextlib, scipy, scikit-learn, itertools. In our analysis, Python 3.12.2 was used. Although technically not strictly required, uploading files in an automated fashion is easier when the Google Cloud CLI is installed. Citation Van Nuland et al. Global hotspots of mycorrhizal fungal richness are poorly protected. Nature (2025). https://doi.org/10.1038/s41586-025-09277-4