Reducing Interdisciplinary Roadblocks Through Multi‐Network Collaboration on Plant–Microbial Interactions

Résumé fourni par la source

The world beneath our feet is a final frontier in terrestrial ecology, where plant roots, fungal hyphae, and bacteria interact with soil, supporting plant growth. Research has focused on the connections between plant genotypes and phenotypes above and below ground, and to a lesser extent the microbial symbionts influencing plant phenotypes, growth, survival, and ecological strategies (Classen et al. 2015). These interactions shape ecological processes such as nutrient cycling and productivity (Phillips et al. 2013). Despite their importance, we lack comprehensive knowledge about the variability of plant–microbial interactions globally and their responses to global change (Segal and Kivlin 2025). Historically, research communities on plants and microbes have been siloed, hindering advancements in understanding these interactions. For example, plant physiologists rarely interact with microbial systematists and biogeochemists (but see Cheaib et al. 2025). Overcoming these challenges requires interdisciplinary and international collaborations. To this end, we founded the Multi-Institute Collaborative Research on BElowground plant–microbial interactions Network of Networks: MICROBENet^Net, a network fostering international collaboration and data harmonization focused on plant and fungal interactions. Our initial aim is to predict plant–fungal traits across environments and understand their ecosystem consequences. We begin by focusing on plant–fungal interactions, expecting to explore other microbes (e.g., bacteria) as our network expands. In surveying over 60 national and international plant and fungal ecologists and evolutionary biologists across career stages, we identified eight key challenges and roadblocks that currently hinder collaboration: (1) structural challenges to harmonize data among network hubs, (2) lack of disciplinary overlap and a common lexicon among fields and siloed research hubs (e.g., plant ecology and fungal ecology but not plant–fungal ecology), (3) lack of overlap in scientific conferences inhibiting inter-network formation, (4) sparse interdisciplinary mentoring, (5) geographic/cultural barriers as well as (6) time zone barriers to international collaboration, and limited funding sources for (7) sustaining and (8) growing international multi-network collaborations preclude multi-network longevity and expansion (Fig. 1). Scientific language barriers provide a crosscutting example of these roadblocks. Individual plant, fungal, and ecosystem research networks have units and terminology that are not shared with, or have completely different meanings to, researchers in other networks. This disconnect arises because measured traits and functions can differ in what they represent to the community of scientists (i.e., the spatial and temporal scale at which they are measured and the importance of internal dynamics compared to the more coarse or granular values expected by other disciplines). For example, it is common for researchers to look for functionally important patterns between plant and fungal diversity (Van Nuland et al. 2024), even though these diversities are measured differently and represent communities at very different spatial and temporal scales. To remove these scientific roadblocks, we integrate scientists from various fields at in-person colloquiums to set goals for convergent research and train the next generation of scientists to lead international collaborations. Additional MICROBENet^Net activities include research exchanges among siloed research disciplines and geographic locations. We also built a vibrant online community on Slack for more frequent dialog and will host regular online meetings to discuss crosscutting plant–microbial interaction research. Our network integrates previously discipline-specific working groups across four hubs (Figs 1, 2). Hub 1: Trait ecology and evolution connects fungal species identification and global mapping networks (e.g., MaarjAM, MycoPortal, UNITE) with trait-based datasets focused on both fungi and plants (e.g., FRED, MacroMycoFunc). It also includes networks examining ecosystem modules such as plant trait–soil interactions (e.g., New Roots for Restoration Biology Integration Institute) and trait evolution (e.g., TeaMoPTE), as well as phylogenetic frameworks (e.g., PyPHLAWD, FePhyFoPhum). Collectively, these networks contribute to spatial and trait databases (e.g., Fun^Fun) that link fungal traits and distributions to plant hosts globally. These founding networks represent only an initial subset. Many additional plant–microbe databases provide complementary expertise, including FUNGuild (Nguyen et al. 2016), the TRY plant trait database (Kattge et al. 2011), Open Traits (https://opentraits.org/), FungalTraits (Põlme et al. 2020), Not a Trait Database (https://github.com/akoontz11/natdb), the 1,000 Fungal Genomes Project (https://1000.fungalgenomes.org/), MycoFlor, and plant–bacterial/archaeal databases. Hub 2: Plant and microbial interaction outcomes expands this foundation by integrating networks that define which plant and fungal taxa interact (e.g., GlobalFungi (https://globalfungi.com/), GloBI (https://www.globalbioticinteractions.org/), MaarjAM (https://maarjam.ut.ee/)) and the ecological consequences of those interactions (e.g., MycoDB -Chaudhary et al. 2016). Hub 3: Biogeography brings together plant and fungal trait and interaction data to map and model how these relationships vary across space and time. It incorporates large-scale distribution datasets such as GBIF (https://www.gbif.org/), SoilBON (https://geobon.org/bons/thematic-bon/soil-bon/), RefSoil+, and GlobalFungi (https://globalfungi.com/). Hub 4: Earth system modeling uses this integrated framework to identify key parameters and processes for inclusion in predictive models. Several ecosystem models already include microbial components (e.g., FUN-CORPSE (https://github.com/bsulman/FUN-CORPSE), MIMICS (Wieder et al. 2014), CESM (https://www.cesm.ucar.edu)) and will serve as platforms for translating linked empirical data into forecasts of plant–microbial dynamics and functional consequences. We held an initial colloquium in March 2025 to bring together scientists and integrate datasets among research hubs within our network (Fig. 2). Over 30 established researchers and early career colleagues convened for 3 days to create convergent research goals. We used the following framework to advance interdisciplinary, international research: We kicked off our first colloquium discussing how to create a new research network and cohort that works collaboratively across topics and countries. We worked collectively and in small groups to discuss how we approach mentoring, being a mentee, and working in diverse collaborative groups. Overall, the goal was to build trust among team members in order to have in-depth, respectful, and productive collaborations. A number of themes emerged: These steps ensure that collaborative groups are able to focus on the possibilities, have fun, and remain open minded. We hold “update” calls for all active collaborative groups every 3–4 months at multiple rotating times, circulating the agenda and minutes via Slack or WhatsApp depending on regional access. Over 2 days, we discussed a number of exciting frontiers, challenges and barriers, as well as solutions. In sum, a defining feature of MICROBENet^Net is its emphasis on a positive, forward-looking mindset. Rather than dwelling on limitations, the network chooses to focus on what is possible—empowering participants to build on existing resources, strengthen complementary networks, and pursue shared scientific goals. As the community grows, it aims to foster open communication, interdisciplinary training of early career researchers with a shared vocabulary, and the integration of diverse datasets. While barriers such as communication challenges and resource limitations remain, the network is committed to inclusivity and long-term collabora

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.