Dutch elm disease resistance tracks with a different seed mycobiome in Ulmus minor
Seed biology is doing more of the work than tree pathologists once assumed. In Ulmus minor, resistance to Dutch elm disease appears to align with a distinct fungal signature inside the samara, the winged fruit that carries the next generation of trees into the landscape.
The preprint examined whether mother trees with contrasting levels of resistance to Dutch elm disease also differ in the fungal communities associated with their seeds. Rather than treating seed fungi as incidental passengers, the study framed them as part of an early-life microbial filter that could shape seedling performance from the moment dispersal begins. The design focused on nine U. minor genotypes spanning resistant, intermediate, and susceptible groups, and the investigators used high-throughput sequencing of the ITS1 region to map the seed-associated mycobiome.
The result was not a uniform fungal background shared across all trees. It was a community that shifted with host resistance status.
Richer fungal communities in susceptible trees, leaner assemblages in resistant ones
Across the samaras analyzed, the team identified 522 fungal amplicon sequence variants, including 74 members of a core samara mycobiome. The dominant fungal classes were Dothideomycetes and Tremellomycetes, but the overall picture changed once the communities were separated by resistance group. Susceptible genotypes carried richer fungal communities, while resistant genotypes supported less diverse but more distinct assemblages.
That pattern matters because microbial richness does not always mean microbial stability or benefit. In a seed environment, a more crowded fungal community may reflect a looser ecological filter, a greater opportunity for opportunistic taxa to persist, or a host phenotype that allows broader colonization. By contrast, a narrower but more sharply differentiated assemblage in resistant trees suggests stronger selection during seed development or seed maturation, with the host exerting a tighter ecological sieve on which fungi are allowed to establish.
A one-sentence summary captures the central signal: resistance was associated with less diverse, more distinctive seed fungal communities.
Candidate taxa separated resistant from intermediate and susceptible genotypes
Differential abundance analysis identified 21 genera associated with host resistance level. Among the clearest contrasts, Protofenestella was enriched in resistant genotypes, while Taphrina was enriched in intermediate and susceptible genotypes. Those shifts do not prove causality, but they do provide candidate organisms for functional follow-up and raise the possibility that specific fungal taxa accompany, or perhaps participate in, resistance-linked host states.
Because the data come from seed tissue rather than mature wood or infected vascular tissue, the signal is especially interesting. DED resistance is usually discussed in terms of the tree’s response to pathogen invasion later in life, when the fungus driving the disease is already interacting with xylem defenses and vascular physiology. Here, the microbial difference appears much earlier, in reproductive material that influences the microbial starting point for the next generation. If seed-associated fungi are transmitted, filtered, or shaped in concert with host resistance traits, the mycobiome could become part of the inheritance environment even before a seedling faces any pathogen pressure.
Network structure suggested stronger ecological compartmentalization in resistant trees
The community analysis went beyond taxonomy and looked at how fungi were organized within each resistance group. Co-occurrence network analysis showed contrasting patterns of community structure, with resistant genotypes forming more modular and less connected networks than susceptible genotypes. In ecological terms, a modular network implies more compartmentalized associations among fungal taxa, rather than a single tightly interwoven community.
That pattern is biologically plausible. A more modular seed mycobiome may limit the spread of any one taxon through the community and reduce the chance that a dominant colonizer reshapes the entire assemblage. Less connected networks can also reflect stronger host control, because host-mediated filtering may prevent broad, dense associations from accumulating. In practical terms, the resistant genotype appears to host a fungal community that is not merely smaller, but structurally different in the way its members coexist.
The difference is not cosmetic. It points to distinct assembly rules operating inside the reproductive tissues of resistant and susceptible elms.
Why a seed mycobiome could matter for tree resistance
Forest trees live on a different timescale than annual crops or many human clinical models. A seed fungal community can influence germination, seedling vigor, early stress tolerance, and later exposure to pathogens long before the tree has had a chance to express its full adult phenotype. For a disease such as DED, where long-lived host resistance is central to conservation and breeding strategies, even subtle microbial differences at the seed stage may have outsized importance.
The study does not claim that the mycobiome determines resistance on its own. It does show that resistance status and seed fungal assembly travel together in a consistent pattern across the genotypes examined. That makes the samara mycobiome a credible candidate trait for future work, especially if breeders want to understand whether fungal associates can serve as markers of resistant germplasm or as functional contributors to early development.
What remains unresolved is the direction of effect. Resistant trees may actively shape a more selective mycobiome, the mycobiome may help reinforce traits associated with resistance, or both may reflect deeper host genetic differences that influence multiple biological layers at once. The study also leaves open whether these fungal differences persist after dispersal or translate into measurable effects on seedling health and eventual disease susceptibility.
Functional follow-up will decide how much of this signal is biology and how much is association
The preprint ends with a restrained but important message: the samara mycobiome offers candidate fungal taxa for future validation and may have applications in elm conservation and mycobiome-informed breeding. That is a credible next step, but it is still a next step. Sequencing can reveal who is present and how communities differ; it cannot yet show whether those fungi alter host resistance, simply accompany it, or arise because resistant and susceptible genotypes create different seed environments from the start.
The most useful next experiments will need to test function directly. Candidate genera such as Protofenestella and Taphrina will need to be assessed in controlled settings, alongside seedling emergence, colonization dynamics, and downstream disease outcomes. Until then, the main advance is conceptual: resistance to Dutch elm disease is not confined to bark, xylem, or canopy symptoms. It may be written, in part, into the fungal community that rides inside the seed.