Eight Crops, One Salty Field: How Plant Identity Rebuilds the Rhizosphere Microbiome

Saline-alkaline farmland is one of the harshest stages in modern agriculture, and a new field experiment from northern China shows that the crop a farmer chooses may quietly rebuild the entire underground world beneath it. Reporting in Frontiers in Microbiology, a research team led by Xiaoling Zhang, Hanghang Hou and Baili Feng of Northwest A&F University grew eight crop species side by side in the same salt-stressed soil and found that crop identity alone explained 78.1 percent of the variation in bacterial communities and 83.5 percent of the variation in fungal communities in the rhizosphere, the narrow zone of soil hugging the living root. The finding is striking because every plot shared the same preceding crop, the same tillage, the same fertilization and the same irrigation, meaning the dramatic underground differences were generated by the plants themselves.

The scale of the problem the study addresses is enormous. Salinized soils cover approximately 1.38 billion hectares worldwide, and the affected area continues to expand each year according to the Food and Agriculture Organization of the United Nations. Salt constrains crop growth through osmotic stress, sodium and chloride toxicity, and the deterioration of soil aggregation, aeration and water permeability. Saline-alkaline stress can also disrupt microbial nitrogen and phosphorus cycling, impairing nitrification, promoting denitrification and suppressing phosphorus desorption, mineralization and solubilization, which reinforces declines in soil fertility and crop performance. Crop-based rehabilitation is attractive precisely because it can combine productive land use with gradual soil improvement, but it has remained unclear how different crops shape rhizosphere conditions when grown under the same field background.

To answer that question, the team conducted a common-field comparison at a saline-alkaline experimental station on the northern Loess Plateau in Inner Mongolia, at 1,026 meters above sea level in a temperate continental semi-arid climate with mean annual precipitation of just 265.34 millimeters and annual evaporation of 1,800 to 2,200 millimeters. The sandy loam soil had an initial pH of 8.82, an electrical conductivity of 3.21 dS per meter in a 1:5 soil-to-water extract, and a total salt content of 1.03 percent, classifying it as salt-affected alkaline soil under FAO diagnostic standards. The eight crops, alfalfa, broomcorn millet, Jerusalem artichoke, oat, quinoa, rapeseed, sugar beet and sunflower, were arranged in a randomized complete block design with three biological replicates per treatment, giving 24 plots of 30 square meters each, separated by bare buffers and isolation strips to prevent cross-contamination.

The technical rigor of the experiment is what makes its results so compelling. All plots received identical basal fertilization of 160 kilograms per hectare of urea, 320 kilograms per hectare of superphosphate and 130 kilograms per hectare of potassium sulfate, incorporated evenly by plowing. Rhizosphere sampling was aligned with crop phenology rather than calendar date, with six crops sampled on 21 July 2025 and two on 6 August 2025, all at the full flowering stage except sugar beet, which was sampled at the rapid taproot enlargement stage. This phenology-based approach was adopted to reduce developmental-stage heterogeneity, because root-derived inputs, nutrient demand and rhizosphere microbial processes can vary markedly with plant developmental stage. Five plants from the centre of each plot were pooled to form one biological replicate, and approximately 500 grams of pooled rhizosphere soil was collected per plot.

The results revealed that crop identity generated pronounced differences in rhizosphere nutrient status, salinity, moisture and enzyme activity. Alfalfa had the highest total nitrogen, organic carbon and total phosphorus contents, with values of 0.64, 9.98 and 0.82 grams per kilogram respectively, while rapeseed had the highest total potassium content at 21.26 grams per kilogram. Rapeseed also showed the highest ammonium nitrogen content at 6.73 milligrams per kilogram, quinoa the highest nitrate nitrogen at 22.17 milligrams per kilogram, and oat the highest alkali-hydrolysable nitrogen at 56.49 milligrams per kilogram. Salinity-related traits showed similarly distinct patterns: rapeseed had the highest water-soluble sodium at 1,493.67 milligrams per kilogram, the highest total soluble salts at 1.65 percent and the highest moisture content at 13.54 percent, while oat had the lowest water-soluble sodium at 352.83 milligrams per kilogram and Jerusalem artichoke the lowest total salt content at 0.28 percent. Urease activity was highest in oat and dehydrogenase activity highest in sugar beet, indicating that crop identity altered not only which nutrients were present but also the biochemical activity of the rhizosphere.

Plant ion allocation patterns were equally revealing. Quinoa combined high potassium concentrations with low sodium-to-potassium ratios, a characteristic consistent with the maintenance of potassium-dependent metabolic and osmotic functions under saline conditions. Sunflower showed a different allocation strategy, with high root sodium, restricted sodium translocation and strong potassium translocation to aboveground tissues, indicating a predominantly root-associated sodium distribution while maintaining potassium supply to shoots. Sugar beet showed high aboveground sodium and the highest sodium translocation factor, indicating preferential sodium distribution to aboveground tissues. These patterns describe crop-specific tissue ion partitioning rather than whole-plant salt tolerance, highlighting substantial variation in how different crops distribute sodium and potassium between roots and shoots under identical saline-alkaline field conditions.

Microbial community analysis using bacterial 16S rRNA gene and fungal ITS amplicon sequencing on an Illumina MiSeq platform confirmed the strong effect of crop identity on rhizosphere diversity. All six alpha-diversity indices, including Chao1 richness, Shannon diversity and Faith’s phylogenetic diversity for both bacteria and fungi, remained significant after Benjamini-Hochberg false discovery rate correction. Jerusalem artichoke had significantly higher bacterial Chao1 richness and Shannon diversity than rapeseed, and bacterial phylogenetic diversity was higher in both Jerusalem artichoke and sunflower than in rapeseed. For fungi, Shannon diversity was significantly higher in broomcorn millet and sunflower than in sugar beet. PERMANOVA based on Bray-Curtis distances showed that crop identity explained 78.1 percent of bacterial community variation and 83.5 percent of fungal community variation, with the higher value for fungi suggesting that fungal community composition was slightly more strongly structured by crop identity than bacterial community composition. Bacterial communities contained a larger shared operational taxonomic unit core, with 1,188 OTUs shared across all crop rhizospheres, while fungal communities had only 87 shared OTUs and greater crop-specific differentiation.

Redundancy analysis revealed distinct environmental associations underlying bacterial and fungal community differentiation. Bacterial community structure was significantly associated with a broad set of rhizosphere variables, including total nitrogen, organic carbon, total phosphorus, available nitrogen, pH, urease activity, total salts and soil moisture, indicating that bacterial differentiation was linked to combined gradients in nutrient status, salinity and alkalinity. In contrast, fungal community variation was more strongly associated with soil moisture and nutrient-related variables, particularly total nitrogen, organic carbon and total potassium. Environment-responsive OTUs provided taxon-level evidence for these associations, with the top 10 bacterial OTUs showing correlation values ranging from 0.806 to 0.879 and the top 10 fungal OTUs showing even stronger correlations from 0.933 to 0.967. These responsive taxa included bacterial genera such as Skermanella, Erythrobacter and Echinicola, and fungal genera such as Botryotrichum, Mrakia, Psathyrella and Mortierella, indicating that crop identity mainly altered the relative representation of particular microbial groups rather than completely replacing the resident community.

Co-occurrence network analysis added another dimension to the findings, showing that crop identity influenced not only which taxa were present but also how microbial associations were organized. Sunflower and rapeseed supported the most densely connected networks, with sunflower having 4,247 edges, an average degree of 12.909 and a network density of 0.020, while Jerusalem artichoke and oat showed higher overall network complexity, with values of 0.568 and 0.563 respectively. Bacterial nodes dominated the networks across all crop rhizospheres, accounting for 79.39 to 85.14 percent of total nodes, with Proteobacteria and Actinobacteriota central to network organization. Keystone analysis identified connector nodes in all crop networks, whereas module hubs were detected only in Jerusalem artichoke, oat and sunflower, suggesting that different crops supported different patterns of microbial ecological organization even within the same field soil.

The most integrative result came from multiple factor analysis, which organized 44 variables into six quantitative blocks covering soil nutrients, salinity and moisture, enzyme activities, plant nutrient and ion traits, microbial diversity and community structure, and network and inferred functional characteristics. The first four retained dimensions explained 73.6 percent of total variation, and Ward.D2 hierarchical clustering separated the eight crops into two stable groups confirmed by 1,000 bootstrap iterations. Group I, comprising Jerusalem artichoke, oat, alfalfa, broomcorn millet and quinoa, was generally associated with lower rhizosphere salinity and relatively stronger nutrient- and enzyme-related characteristics, while Group II, comprising sunflower, rapeseed and sugar beet, was characterized by higher rhizosphere salinity and greater contributions from sodium allocation traits. Dehydrogenase activity and urease showed the highest weighted contributions across the retained dimensions at 8.6 and 7.4 percent respectively, followed by soil total salts, moisture content and water-soluble sodium. The authors emphasize that these two groups are best interpreted as broad multivariate crop-associated profiles rather than discrete physiological strategies, and that linking these profiles with biomass, yield and longer-term field performance in future experiments will clarify their relevance to crop selection and saline-alkaline land management. The study provides an integrated framework for identifying informative rhizosphere and plant traits for future crop evaluation in a world where salinized farmland continues to expand.

Subject of Research: Microbiology

Article Title: Crop-associated differentiation of rhizosphere microbial communities and soil functional traits in saline-alkaline farmland

Article References: Zhang, X., Yuan, Y., Shi, Y., Chen, S., Gao, Z., Yang, Q., Hou, H., & Feng, B. (2026). Crop-associated differentiation of rhizosphere microbial communities and soil functional traits in saline-alkaline farmland. Frontiers in Microbiology, 17, Article 1952914. https://doi.org/10.3389/fmicb.2026.1952914

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1952914

Keywords: rhizosphere microbiome, saline-alkaline soil, crop identity, plant ion allocation, soil enzymes, microbial communities, Frontiers in Microbiology, Crop-associated, differentiation, rhizosphere, microbial, communities