Diversity Map Test
Mycorrhizal fungi are tiny organisms that live in the soil and form partnerships with plant roots, helping plants get nutrients. They play a big role in keeping Earth’s ecosystems healthy, but we don’t fully understand where different types of these fungi are found around the world. This knowledge gap makes it hard to track and protect these important underground networks.
article link
This study employs machine learning on a global dataset of 25,000 geolocated soil and root samples with over 2.8 billion fungal sequences to produce high-resolution (1 km²) spatial models of mycorrhizal fungal richness.
It integrates comprehensive environmental covariates and assesses model uncertainty and extrapolation, making it the most detailed map of mycorrhizal diversity to date.
The analysis reveals that mycorrhizal diversity hotspots, especially arbuscular mycorrhizal (AM) fungi, are poorly protected globally, highlighting gaps in conservation efforts.
Why is it important to collect and analyze these data in this way?
Mycorrhizal fungi are critical to ecosystem functions, such as carbon cycling and plant nutrient acquisition, yet remain underrepresented in global biodiversity studies. High-resolution data is necessary to identify underground biodiversity hotspots and inform precise conservation and land management strategies. Without accurate baselines, it is impossible to measure the impacts of biodiversity loss or develop effective conservation strategies.
How do we measure biodiversity?
The study uses Hill numbers (q=0) to estimate effective mycorrhizal species richness from OTUs (Operational Taxonomic Units) or Virtual Taxa (VTs) identified in DNA sequencing datasets. Diversity is measured through species accumulation curves, which standardize comparisons across sampling methods.
What are the most important predictors of mycorrhizal diversity, and what is the importance of this finding?
Key predictors include temperature, evapotranspiration, soil organic carbon (SOC), aboveground plant biomass, and anthropogenic land cover. For AM fungi, higher temperatures and human-modified landscapes increase richness, reflecting the role of ruderal species and facultative AM hosts. For ectomycorrhizal (EcM) fungi, SOC and host plant distribution are critical, linking fungal diversity with carbon dynamics and forest ecosystems.
These findings underline the coupling of fungal and plant biodiversity and highlight their sensitivity to environmental and climate changes.
How do we make predictions?
The study uses Random Forest machine-learning models trained on geolocated fungal richness data and environmental predictors. Ensemble modeling techniques and spatial eigenvector mapping reduce bias and improve the robustness of predictions.
How do we measure uncertainty, and what is its importance?
Uncertainty is quantified as the coefficient of variation from bootstrapped model predictions. High extrapolation values indicate regions poorly represented in the training data, requiring cautious interpretation. Understanding uncertainty is essential for refining predictions and directing future sampling efforts.
How protected is mycorrhizal diversity, and how do we measure this?
Only 5% of AM fungal hotspots and 15% of EcM hotspots overlap with protected areas. Protection is evaluated by intersecting predicted richness hotspots with the World Database of Protected Areas.
How will these data drive field research priorities?
The maps highlight under-sampled and high-uncertainty regions, such as tropical forests and deserts, guiding future sampling campaigns. They also identify ecologically critical and vulnerable regions where targeted studies can improve conservation strategies.
How might these data influence conservation and restoration decision-making?
By pinpointing poorly protected hotspots, the data can inform policies to expand protected areas and integrate microbial biodiversity into conservation plans. Restoration efforts can prioritize areas with rich fungal diversity to enhance ecosystem functionality.
What are the biases and limitations of the current state of the work?
Geographic biases in the training dataset lead to high uncertainty in regions like tropical forests and deserts. The reliance on molecular sequencing methods introduces biases from primer selection and barcode region variability. Predictions at 1 km² resolution might oversimplify fine-scale diversity patterns within each pixel.
What is our roadmap, and what is it driving towards?
The roadmap aims to improve data representation across under-sampled regions and develop tools for rapid fungal monitoring using new sequencing technologies. Future goals include ground-truthing predictions, integrating socio-economic factors into land management, and expanding analyses to other fungal symbioses like ericoid and orchid mycorrhizas. Ultimately, the work drives toward a global conservation framework that includes underground biodiversity as a key component of ecosystem health.
Hotspots overview
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed vestibulum mi quis purus vehicula rutrum. Aenean maximus convallis egestas. Donec ligula sapien, mattis non felis id, mattis rutrum mauris. Mauris augue risus, ultrices non dolor in, tincidunt sagittis ligula. Sed eu efficitur erat. Quisque eget quam ac orci dapibus ullamcorper. Curabitur fermentum eros ut lobortis maximus. Phasellus volutpat mi nec suscipit sagittis.
.svg)
