UNDERGROUND ATLAS

Technical specs

Native resolution: 30 arc-seconds (1 km)

Spatial extent: Global

Masked area: Non-vegetated landcover from remote sensing datasets (e.g., rock, ice, and desert habitats) and highly urban/built-up landcover

Units: Richness = number of species / 100 m2; Endemism = rarity-weighted richness (relative importance, unitless); Uncertainty = coefficient of variation (unitless); Model Extrapolation = %

Description

The mycorrhizal mapping data products shown here are high-resolution spatial interpolations from ensemble machine-learning models. These predictive models are trained on mycorrhizal fungal diversity metrics from a curated global database of soil fungi, encompassing 2.8 billion fungal DNA sequences from 25,000 soil samples across 130 countries, and dozens of open-source environmental layers on climate conditions, vegetation, topography, soil properties, and human factors (e.g., percent of human-modified landcover). Models were built as k-fold cross-validated random forest regression models, with final predictions calculated as an ensemble average of the top 10 highest performing models over 100 bootstrapped runs.

Each map pixel represents a prediction of mycorrhizal fungal diversity per 100 m2 — in other words, the expected mycorrhizal diversity (combined from multiple sub-samples covering a 100 m2 area) within each 1-km pixel. The ‘richness’ predictions come from sample-level calculations of the total number of unique fungal species using a CHAO rarefaction/extrapolation estimator. The ‘endemism’ predictions are based on rarity-weighted richness calculations using a sample-level sum of species rarity scores. Note that mycorrhizal fungal ‘species’ here refer to 97% similar clustered Operational Taxonomic Units, which is a standard method of delineating fungal taxa in eDNA sequencing datasets.

Each pixel-level prediction comes with two types of statistical uncertainty: the coefficient of variation and an estimate of model extrapolation. The coefficient of variation measures the relative dispersion of predicted values around the prediction mean, and is calculated using the standard deviation divided by the prediction mean across all bootstrapped runs. Areas with higher uncertainty indicate a wider confidence interval around model predictions. Extrapolation is estimated using a principle component approach to determine the degree to which pixels are geographically and environmentally represented in the training data. Areas with high extrapolation indicate pixels that are environmentally unique and/or far from nearby sampled locations.

See more details on the technical approach and cross-validation procedures in the published article here [Link to Nature Article] and additional code resources here [Link to GitHub].

About

These data products were developed jointly by the following contributors (in order of research article authorship):

Michael E. Van Nuland

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA)

Colin Averill

(Institute of Integrative Biology, Department of Environmental Systems Science, ETH Zürich, Zürich, Switzerland; Funga Public Benefit Corporation, Austin TX USA)

Justin D. Stewart

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA; Amsterdam Institute for Life and Environment (A-LIFE), Section Ecology & Evolution, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands)

Oleh Prylutskyi

(Falz-Fein Biosphere Reserve "Askania Nova", Kherson Oblast, Ukraine)

Adriana Corrales

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA)

Laura G. van Galen

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA; Institute of Integrative Biology, Department of Environmental Systems Science, ETH Zürich, Zürich, Switzerland)

Bethan F. Manley

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA)

Clara Qin

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA)

Thomas Lauber

(Institute of Integrative Biology, Department of Environmental Systems Science, ETH Zürich, Zürich, Switzerland)

Vladimir Mikryukov

(Mycology and Microbiology Center, University of Tartu, Tartu, Estonia)

Olesia Dulia

(Mycology and Microbiology Center, University of Tartu, Tartu, Estonia)

Giuliana Furci

(Fungi Foundation, Brooklyn, NY, USA)

César Marín

(Amsterdam Institute for Life and Environment (A-LIFE), Section Ecology & Evolution, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands; Centro de Investigación e Innovación para el Cambio Climático (CiiCC), Universidad Santo Tomás, Valdivia, Chile)

Merlin Sheldrake

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA; Amsterdam Institute for Life and Environment (A-LIFE), Section Ecology & Evolution, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands)

James Weedon

(Amsterdam Institute for Life and Environment (A-LIFE), Section Systems Ecology, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands)

Kabir G. Peay

(Department of Earth System Science, Stanford University, Stanford, CA, USA; Department of Biology, Stanford University, Stanford, CA, USA)

Charlie K. Cornwallis

(Department of Biology, Lund University, Lund, Sweden)

Tomáš Větrovský

(Laboratory of Environmental Microbiology, Institute of Microbiology of the Czech Academy of Sciences, Czech Republic)

Petr Kohout

(Laboratory of Environmental Microbiology, Institute of Microbiology of the Czech Academy of Sciences, Czech Republic)

Petr Baldrian

(Laboratory of Environmental Microbiology, Institute of Microbiology of the Czech Academy of Sciences, Czech Republic)

Leho Tedersoo

(Mycology and Microbiology Center, University of Tartu, Tartu, Estonia; College of Science, King Saud University, Riyadh, Saudi Arabia)

Stuart A. West

(Department of Biology, Oxford University, Oxford, United Kingdom)

Thomas W. Crowther

(Institute of Integrative Biology, Department of Environmental Systems Science, ETH Zürich, Zürich, Switzerland)

E. Toby Kiers

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA; Amsterdam Institute for Life and Environment (A-LIFE), Section Ecology & Evolution, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands)

SPUN Mapping Consortium

Johan van den Hoogen

(Society for the Protection of Underground Networks (SPUN), Dover, DE, USA; Institute of Integrative Biology, Department of Environmental Systems Science, ETH Zürich, Zürich, Switzerland

The development of this project was funded, in part, through the following:

• SPUN: Jeremy and Hannelore Grantham Environmental Trust, Paul Allen Family Foundation, the Schmidt Family Foundation, Quadrature Climate Foundation, and the Bezos Earth Fund.
• Kiers: NWO-VICI (202.012), NWO-Spinoza (SPI.2023.2) and HFSP (RGP 0029).
• Averill: Ambizione grant no. PZ00P3_17990 from the Swiss National Science Foundation.
• Stewart: NOW-Gravity Grant Microp (024.004.014).
• Crowther Lab: DOB Ecology and the Bernina Foundation
• GlobalFungi: Czech Science Foundation (21-17749S), MEYS (LC23152, LM2023055)
• Tedersoo: Estonian Science Foundation (PRG632)
• Peay: CIFAR program Fungal Kingdom: Threats and Opportunities, US NSF (DEB-1845544) and DOE (DE-SC0023661)
• Marín: ANID – Chile projects SIA No. SA77210019 (2021), Fondecyt Regular Project No. 1240186 (2024).