Elevation Explains Moss Rhizoid Bacterial Variation but Not Endophytic Composition

Bacteria residing at the rhizoid-soil interface of alpine mosses on Segrila Mountain shift significantly across elevational gradients and soil properties, whereas endophytic bacterial communities located inside moss tissues show no such environmental dependency. In an original research study published in Frontiers in Microbiology within the Terrestrial Microbiology section, a research collaboration led by Ruihong Wang, Yuquan Wei, and colleagues investigated microbial communities associated with dominant alpine mosses in southeastern Tibet on the Tibetan Plateau. The investigation directly contrasts bacteria inhabiting the external rhizoid-sphere—the physical interface where moss rhizoids contact the soil—with endophytic bacteria living within moss tissues, resolving previously poorly characterized ecological distinctions across high-altitude alpine habitats.

To evaluate these microbial assemblages across environmental transitions, the authors conducted field sampling across four elevation zones ranging from 3,700 to 4,300 meters on Segrila Mountain. The research team, representing Huazhong Agricultural University in Wuhan, China Agricultural University in Beijing, and Xizang Agricultural and Animal Husbandry University in Nyingchi, gathered a total of 36 rhizoid-sphere samples alongside 12 endophytic samples. These collections represented three dominant alpine moss species: Actinothuidium hookeri, Ptilium crista-castrensis, and Dicranum fuscescens. By gathering samples across this elevation range, the researchers designed their sampling framework to examine how habitat, elevation, soil properties, and bacterial community attributes were related to moss traits.

The team deployed high-throughput 16S rRNA gene amplicon sequencing to characterize the taxonomic diversity and composition of both the rhizoid-sphere and endophytic bacterial communities. Beyond basic taxonomic profiling, the researchers integrated a multi-tiered quantitative analytical approach combining community composition analysis, SPIEC-EASI co-occurrence networks, random-forest machine-learning models, and structural path modeling. This comprehensive analytical pipeline allowed the investigators to separate the specific contributions of external abiotic factors, host taxonomic identity, and community-level interactions in shaping both external and internal bacterial assemblages across the alpine gradient.

Statistical community analysis revealed that elevation and host species identity significantly explained variation in rhizoid-sphere bacterial composition, while neither factor significantly explained endophytic bacterial composition. Elevation uniquely explained 25.9 percent of the taxonomic variation in rhizoid-sphere bacterial communities, an association backed by high statistical significance with a p-value of 0.0001. Host moss species identity accounted for an additional 7.9 percent of the variation in rhizoid-sphere community composition, which was also statistically significant with a p-value of 0.0093. In marked contrast, neither elevation nor moss species significantly explained endophytic bacterial composition, with significance testing for both variables yielding p-values exceeding 0.05.

A corresponding pattern of environmental sensitivity emerged when the investigators modeled the relationship between local edaphic factors and bacterial community composition. Soil properties significantly explained community variation in the rhizoid-sphere, generating an adjusted R-squared value of 0.199 and a statistically significant p-value of 0.001. Conversely, the statistical model evaluating the relationship between soil properties and endophytic community variation was statistically nonsignificant, producing an adjusted R-squared of only 0.029 and a p-value of 0.239. These findings show that soil properties significantly explained bacterial community variation at the moss rhizoid-soil interface, whereas the endophytic model was statistically nonsignificant.

Co-occurrence network analysis conducted via SPIEC-EASI highlighted pronounced structural differences between the internal and external bacterial communities. The endophytic bacterial network demonstrated substantially greater network clustering and modularity than the rhizoid-sphere network, exhibiting a clustering coefficient of 0.665 compared to 0.305 for the rhizoid-sphere, and a modularity score of 0.84 compared to 0.79 for the rhizoid-sphere. In addition, the endophytic network displayed a substantially larger proportion of positive partial correlations at 94.71 percent, compared with 83.48 percent positive partial correlations observed within the rhizoid-sphere network. These network metrics indicate that endophytic bacteria maintain more cohesive, modular, and positively correlated associations inside moss tissues.

To determine how bacterial community attributes correlate with host plant characteristics, the researchers applied random-forest modeling to predict functional moss traits from the microbial and environmental data. Using leave-one-out cross-validation, the random-forest models proved most effective at predicting elemental nutrient concentrations within moss tissues. Specifically, the models most strongly predicted moss tissue potassium concentrations with a leave-one-out cross-validation R-squared of 0.58. Moss tissue nitrogen concentrations were likewise strongly predicted by the random-forest models, achieving a leave-one-out cross-validation R-squared of 0.48, establishing a measurable predictive link between community attributes and moss nutritional chemistry.

Structural path modeling provided further resolution into direct relationships between microbial diversity and specific moss trait dimensions. In the path models, rhizoid-sphere bacterial Shannon diversity was positively associated with moss morphological traits, demonstrating a standardized path coefficient beta of 0.500 and a p-value of 0.0005. Simultaneously, rhizoid-sphere Shannon diversity was negatively associated with moss nutrient traits, yielding a standardized path coefficient beta of negative 0.263 and a p-value of 0.0036. The researchers concluded that these distinct findings position the rhizoid-sphere as a pivotal ecological interface linking elevational gradients, soil conditions, bacterial community variation, and moss functional traits in alpine ecosystems.

Subject of Research: Bacterial community variation across alpine moss rhizoid and endophytic habitats along an elevational gradient

Article Title: Distinct associations of rhizoid-sphere and endophytic bacteria with dominant moss traits along an elevational gradient on Segrila Mountain

Article References: Wang, R., Liu, W., Wang, L., Dong, M., Lei, Y., Ma, H., Song, Y., Ma, H., Hu, R., Jiang, Y., & Wei, Y. (2026). Distinct associations of rhizoid-sphere and endophytic bacteria with dominant moss traits along an elevational gradient on Segrila Mountain. Frontiers in Microbiology, 17, Article 2019778. https://doi.org/10.3389/fmicb.2026.2019778

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.2019778

Keywords: alpine mosses, rhizoid-sphere bacteria, endophytic bacteria, Segrila Mountain, southeastern Tibet, Tibetan Plateau, 16S rRNA gene amplicon sequencing, Actinothuidium hookeri, Ptilium crista-castrensis, Dicranum fuscescens, elevational gradient, SPIEC-EASI networks, random-forest models, path modeling, Shannon diversity, moss tissue nitrogen, moss tissue potassium, rhizoid-soil interface