Community differentiation and assembly of soil nosZ-type denitrifiers across dominant plant communities in an alpine wetland of Qinghai Lake

High on the northeastern Qinghai–Tibet Plateau, where the alpine wetlands of Qinghai Lake sprawl across a basin perched more than 3,000 meters above sea level, a quiet microbial drama unfolds beneath the roots of sedges and grasses. It involves one of the most consequential gases in the climate system: nitrous oxide, a long-lived greenhouse gas whose rising atmospheric burden contributes both to climate forcing and to the depletion of stratospheric ozone. The only known enzymatic biological sink for this gas is the reduction of nitrous oxide to benign dinitrogen, a reaction catalyzed by nitrous oxide reductase and encoded by the nosZ gene. A new study of the wetlands ringing Qinghai Lake now shows that the communities of nosZ-harboring microorganisms in these soils are far from uniform: they differ sharply, in both diversity and taxonomic makeup, depending on which dominant plant community holds sway above ground.

The stakes are considerable. During denitrification, microbes sequentially reduce nitrate through nitrite and nitric oxide to nitrous oxide, and then ideally onward to dinitrogen. Whether nitrous oxide accumulates as an intermediate or is consumed at the terminal step depends on the microorganisms carrying nosZ. Crucially, this gene is not confined to canonical denitrifiers; it also occurs among phylogenetically diverse microorganisms that lack a complete denitrification pathway, and experimental evidence shows that such non-denitrifying nitrous oxide reducers can measurably lower soil nitrous oxide production. Alpine wetlands, with their strong hydrological and redox gradients and tightly coupled microbial nitrogen transformations, are exactly the kind of environmentally heterogeneous ecosystems where the organization of this functional guild matters most for the balance between nitrous oxide generation and consumption.

Researchers from Qinghai Normal University and affiliated field stations focused on four naturally occurring dominant plant communities in the wetlands surrounding Qinghai Lake, each named for its prevailing species: Blysmus sinocompressus, Carex regescens, Elymus nutans, and Leymus secalinus. In August 2020, during the peak growing season, the team sampled ten spatially separated sites distributed around the lake, establishing five one-square-meter quadrats at each site and collecting surface soil from zero to ten centimeters depth along S-shaped transects. The five subsamples within each quadrat were homogenized into one composite sample, yielding fifty independently collected and sequenced quadrat-level soil samples, with the site serving as the higher-level spatial replicate for inference about plant-community type. Alongside the molecular work, the researchers measured aboveground and belowground vegetation biomass and a suite of soil properties, including pH, moisture, temperature, total carbon, total nitrogen, ammonium nitrogen, nitrate nitrogen, and the soil carbon-to-nitrogen ratio.

The nosZ gene was amplified with the primer pair nosZ-F and 1622R, and the products were sequenced on an Illumina HiSeq 2500 platform. After merging, quality filtering, and chimera screening, clean sequences were clustered into operational taxonomic units at 97 percent sequence similarity. Rarefaction curves based on ACE and Chao1 estimates approached saturation once sequencing depth reached roughly 28,000 reads, indicating that the effort adequately captured richness patterns. The results were striking. A total of 40,225 nosZ OTUs were recovered, yet only 859 were shared among all four plant communities. Soils beneath Leymus secalinus contained by far the largest pool of community-specific OTUs—16,644—compared with 6,787 under Elymus nutans, 2,816 under Carex regescens, and 2,411 under Blysmus sinocompressus. Leymus soils also showed significantly higher ACE and Chao1 richness than the other three community types, while Carex soils registered the lowest values.

Community composition told an equally clear story. Ordination analyses revealed pronounced differentiation among the four plant communities: in a non-metric multidimensional scaling configuration with an acceptable stress value of 0.1147, samples from different vegetation types occupied distinct regions of ordination space, with Blysmus and Leymus samples particularly well separated. Multivariate tests confirmed significant compositional differences, with permutational analysis of variance explaining 28 percent of the variation and analysis of similarities yielding a statistic of 0.52, both highly significant. At the phylum level, Proteobacteria dominated the nosZ-bearing communities in all four vegetation types, although its relative contribution declined toward Leymus soils, where unclassified taxa expanded. At the genus level, Bradyrhizobium was the predominant component overall, with Mesorhizobium, Herbaspirillum, Azospirillum, Paracoccus, Polaromonas, Shinella, and unclassified taxa contributing appreciably and varying in relative representation among communities. This dominance of Proteobacteria and Bradyrhizobium is consistent with earlier nosZ-based surveys in which Rhizobiaceae members and other Proteobacteria figured as major components of nitrous oxide-reducing communities.

One of the most intriguing findings was the clear decoupling between the environmental variables associated with community diversity and those tied to taxonomic structure. Belowground biomass made the largest contribution to variation in diversity, at 21.1 percent, followed by aboveground biomass at 9.7 percent, soil temperature at 9.2 percent, total carbon at 8.7 percent, and total nitrogen at 7.4 percent. Compositional variation, by contrast, was most strongly associated with the soil carbon-to-nitrogen ratio, which accounted for 22.5 percent of the total contribution, followed by pH at 16.5 percent and total carbon at 13.4 percent, with soil moisture and nitrate nitrogen showing intermediate contributions. Mantel analyses at both the phylum and genus levels confirmed significant relationships between community dissimilarity and subsets of vegetation and soil properties, although the strength and significance of these relationships differed between taxonomic resolutions. The researchers interpret the strong association of diversity with belowground biomass as reflecting the integrative influence of the rhizosphere, where roots simultaneously modify carbon availability, oxygen status, and nutrient supply, rather than a direct effect of root mass itself.

To move beyond descriptive comparisons, the team applied phylogeny-based null models that partition compositional turnover into ecological processes. The beta-nearest taxon index, or betaNTI, was consistently positive across all four plant communities and substantially exceeded the conventional threshold of plus two, indicating that phylogenetic turnover deviated strongly from null expectations and, within this framework, was predominantly associated with variable selection—heterogeneous environmental filtering imposing different fitness constraints on microbial lineages across contrasting habitats. Elymus and Leymus soils generally exhibited higher betaNTI values than Blysmus and Carex soils. When the relative contributions of assembly processes were partitioned, variable selection was prominent in Carex and Elymus communities, whereas Blysmus soils showed comparatively similar contributions of variable selection and dispersal limitation, and Leymus soils registered a larger fraction categorized as undominated drift. Homogeneous selection and homogenizing dispersal contributed only minimally across all four vegetation types.

These assembly patterns carry ecological weight. Environmental heterogeneity can strengthen deterministic sorting because spatially contrasting habitats impose different constraints on microbial lineages, and soil pH provides a well-established example, with increasingly restrictive pH conditions shifting bacterial assembly toward stronger deterministic selection. The prominence of the carbon-to-nitrogen ratio, pH, and total carbon for taxonomic composition suggests stronger niche-based sorting of community membership along soil chemical gradients, and prior work has shown that pH can influence denitrification at multiple biological levels, including gene expression, enzyme maturation, and the kinetics of nitrous oxide reduction. At the same time, the authors caution that because vegetation and soil properties covaried across sites in this observational design, the observed differentiation should be read as plant-community-associated rather than as a direct causal effect of plant identity, and the contribution values are best regarded as measures of statistical importance rather than experimentally demonstrated causal effects.

Several limitations frame the conclusions. Sampling occurred during a single growing-season period, and multiple quadrats were nested within a limited number of sites, so seasonal dynamics and site-level spatial dependence cannot be fully resolved. The conventional Proteobacteria-oriented nosZ primers may underrepresent diversity in non-Proteobacterial groups such as Firmicutes and Bacteroidetes, meaning the pronounced Proteobacteria dominance reflects the primer-defined fraction of the community rather than a complete inventory of all nitrous oxide reducers. Moreover, DNA-based amplicon sequencing documents the occurrence and relative representation of targeted lineages but does not quantify absolute gene abundance, transcription, enzyme activity, or realized nitrous oxide fluxes. The authors therefore call for future work combining broader or clade-resolved nosZ assays with quantitative PCR, transcript-level measurements, direct gas-flux determinations, repeated seasonal sampling, and greater site-level replication. Even so, the study demonstrates that the dominant plant communities of the Qinghai Lake alpine wetland are associated with coordinated differentiation in belowground environments, primer-defined nosZ community composition, and inferred assembly processes—evidence that the vegetation mosaic above ground helps shape the microbial guild that decides the fate of nitrous oxide beneath it.

Subject of Research: Soil nosZ-type denitrifier community differentiation and assembly across dominant plant communities in the Qinghai Lake alpine wetland

Article Title: Community differentiation and assembly of soil nosZ-type denitrifiers across dominant plant communities in an alpine wetland of Qinghai Lake

Article References: Zhou, S., Zhang, N., Tang, J., & Chen, K. (2026). Community differentiation and assembly of soil nosZ-type denitrifiers across dominant plant communities in an alpine wetland of Qinghai Lake. Frontiers in Microbiology, 17, Article 1972695. https://doi.org/10.3389/fmicb.2026.1972695

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1972695

Keywords: alpine wetland, Qinghai Lake, nosZ, denitrifiers, nitrous oxide, community assembly, variable selection, Bradyrhizobium, Proteobacteria, null model, Qinghai–Tibet Plateau, soil microbiome