Influences of soil amendments on soil properties
The Tibetan hulless barley growth cycle was completed at approximately 135 days, during which physicochemical properties of soils from different amendment treatments varied significantly (Fig. 1 and Table S2). Random forest analysis identified that 5 soil physicochemical factors: pH, AP, TOC, C/N ratio, and Zn corresponded strongly to soil amendment treatment types (Fig. 1a). The TOC, C/N ratio, and pH were significantly higher in MIX and ORG groups than the DAP group; AP and Zn were significantly higher in the DAP group compared to the other two groups (Fig. 1b). The higher values of these physicochemical parameters most likely resulted from added soil amendments. For example, high organic matter content in the organic soil amendment resulted in higher amounts of TOC and C/N ratio while higher AP in the DAP group was due to the high P content of the corresponding inorganic fertilizer.
Data from the PCoA analysis (Fig. 1c) indicated that there were significant differences (P < 0.001) between the DAP, MIX, and ORG treated soil groups, three growth stages, and bulk and rhizosphere soils (P < 0.01). During the three growth stages, NO2− was significantly different in DAP, MIX, and ORG treated soils; TOC and levels of metal elements such as Fe, Al, and Zn changed significantly in DAP and MIX treated soil groups. In comparison, properties of the soil amended with organic fertilizer remained the most stable in terms of its parameters (Table S3). Additionally, values of the physicochemical parameters associated with nitrogen were significantly higher in organic amendments than that of bulk soil (Table S4). This is attributable to the rhizosphere microorganisms that benefit plant growth by participating in the nitrogen cycle through processes such as nitrogen fixation [40, 41].
Influences of different soil amendments on farmland soil bacteria and fungi communities
Soil amendments are known to influence soil microbial communities, microbial biomass, composition, and diversity [42, 43]. During the hulless barley growth stage, the rhizosphere microbial biomass in the DAP treated soil decreased: bacterial biomass decreased from 4.19 × 109 copies·g− 1 to 2.21 × 109 copies·g− 1 and fungal biomass decreased from 5.06 × 108 copies·g− 1 to 3.18 × 108 copies·g− 1 (decreases of 47.2 and 37.2%, respectively). In contrast, the rhizosphere microbial biomass in the MIX treated soil increased: bacterial biomass increased from 1.91 × 109 copies·g− 1 to 1.16 × 1010 copies·g− 1 and fungal biomass increased from 3.64 × 108 copies·g− 1 to 6.02 × 109 copies·g− 1 (increases of 507 and 65.4%, respectively) (Fig. 2a). The bacterial and fungal biomass in the ORG treated soil increased by 202% and 90.5%, respectively.
The microbial biomass of the DAP treated soil peaked during thee tillering stage and decreased following hulless barley growth. In contrast, the microbial biomass in ORG and MIX treated soils increased with crop growth. Inorganic fertilizers tend to be less stable in soils over time and are readily consumed, mobilized, and leached out. Organic soil amendments exhibit slow-release characteristics that last over a longer period of time [10, 12]. This long-term supply of organic carbon and nutrients optimally benefit and sustain microbial biomass growth and activities.
Densities of the rhizosphere bacterial and fungal biomasses were significantly higher in the treated soil compared to bulk soil. During the ripening stage, the difference between the bulk and rhizosphere bacteria biomass in MIX treated soil was approximately 9 × 109 copies·g− 1; other changes in microbial biomass did not exceed 3 × 109 copies·g− 1. Changes in fugal biomass were minimal (Fig. 2a). Through 16S rRNA and ITS gene sequencing of 108 soil samples, 6877 bacterial OTUs and 2597 fungal OTUs were analyzed. The data indicates that mixed organic and inorganic soil amendments increase the diversity of microbial communities in the rhizosphere (Fig. S2). The diversity of microbial communities, particularly for bacteria in bulk soil has a higher level of response to soil amendments compared to rhizosphere microbial communities. Lower diversity was observed for fungi. This response was similarly observed in other studies connecting microbial community responses to various soil amendments [44, 45]. The dominant bacteria phyla are Actinobacteriota, Bacteroidota, Firmicutes, and Proteobacteria; dominant fungi phyla are Ascomycota, Basidiomycota, and Mortierellomycota. At the genus level, the relative abundance of top 20 dominant bacteria are Arthrobacter, Sphingomonas, Chitinophagaceae, Thermomonas, Paeniglutamicibacter, and Rhodanobacter;dominant fungi genera are Gibberella, Didymella, Plectosphaerella, Naganishia, Cystofilobasidium, and Cladosporium (Fig. S3). The composition of bacterial and fungal communities exhibit significant differences across amendment/fertilizer treatments (Fig. 2b). The highest differences are seen between different soil amendment treatments and sampling locations in bulk and rhizosphere soils, most sensitively to early Tibetan hulless barley growth.
PCoA further identified the effect of different soil amendments on bacterial and fungal communities (Fig. 2c). Both bacteria and fungi were separated into different groups based on treatment, time, and site characteristics. PCoA indicate that fungi communities were significantly impacted by soil amendments and to a lesser degree rhizosphere bacterial communities across the stages. Based on Bary-Curtis distance, fungi communities are more sensitive than bacteria communities to rhizosphere-bulk soil samples and soil amendment treatments (Fig. 2d). Relative to DAP treated soils, rhizosphere bacteria communities are similar between ORG and MIX treated soils. The similarity stems from early stage of crop growth, where microorganisms originating from the bulk soil form a new rhizosphere microbiome [46–49]. The correlation between bacterial and fungal communities and environmental factors was analyzed (Fig. 2e), the composition of the bacteria community was correlated with SON and TP and the fungi community with TOC and trace metal elements (Mg, Cu, Al, and Mn). Both bacteria and fungi communities were possibly influenced by TOC, SON, TP, Zn, and Mg in soil. These environmental factors are consistent with the analysis of soil physicochemical factors (Fig. 1a). Nitrogen and phosphorus were possible limiting factors for bacterial growth in Tibetan hulless barley farmland soil. Fungi were possibly impacted by the presence of TOC and trace metal elements in fertilizer. It was reported that bacterial metabolism altered nitrogen-phosphorus synergy while percentage of fungi increase with higher of organic carbon content [50, 51].
Amendment microbial indicators in Tibetan Hulless Barley farmland soil
A total of 210 bacterial OTUs (1.76% of total soil community sequences) and 121 fungal OTUs (3.95% of total soil community sequences) were significantly altered (p < 0.05) by soil amendment treatments across growth stages. The primary bacteria phyla were Actinobacteriota, Chloroflexi, and Firmicutes. The primary fungi phyla were Ascomycota, Basidiomycota, and Mortierellomycota. For bulk soil, a total of 247 bacterial OTUs (3.54% of total soil community sequences) and 131 fungal OTUs (3.51% of total soil community sequences) were identified (p < 0.05) as indicators across fertilizer treatments. Several functional microbiotas responded strongly to soil amendment treatments (Fig. 3). In contrast, soil amendments have a diminished effect on abundant microbiota species due to their strong resistance functionalities [42, 52, 53]. The number of indicator species in rhizospheric soil was marginally lower than that of bulk soil due to plant growth and soil amendment impacts on the microbiome and rhizosphere. For example, rhizosphere bacteria genera Angustibacter degrade gelatin and aesculin, Flavobacterium degrades Casein and gelatin for auxin production and P-solubilization[54], and Gemmatimonas modulates C and N intakes depending on environmental stimuli[55]. Fungal genera identified as indicator species are represented by Saprotroph and other guilds. The effects of soil amendments on the rhizosphere and bulk soil bacteria and fungi vary. Rhizospheric soil amendment indicator species possess metabolic functions promoting plant growth while bulk soil indicators lack these functions.
A total of 44 bacterial OTUs (13.47% of total soil community sequences) and 41 fungal OTUs (13.87% of total soil community sequences) were identified as growth stages indicator species (relative abundance > 0.1%, p < 0.05). These values were lower than the number of indicator species based on soil amendment treatments (Fig. S4). A total of 25 bacterial OTUs (9.65% of total soil community sequences) and 14 fungal OTUs (5.03% of total soil community sequences) were identified in bulk soil. These values were lower than the abundance of indicator species present during rhizospheric growth as well (relative abundance > 0.1%, p < 0.05). Plant growth-promoting rhizobacteria (PGPR) is the indicator species that changes significantly across Tibetan hulless barley growth stages, allowing for better plant adaptation to environmental stressors; for example, the relative abundances of Arthrobacter, Devosia, Bacillus, and Oceanobacillus (Fig. S4). Previous studies on wheat or barley (Hordeum vulgare L.) include these genera as significant PGPR indicators[56–58]. These plant growth indicator species have high relative abundances (reddish in heatmap color, Fig. S4) and their metabolic functions play an important role in plant growth and soil nutrient cycling. The number of plant growth indicators was low, but the abundance was high.
Soil amendment effects on microbial co-occurrence patterns
Co-occurrence network analysis examines interactions between species. The microbial community of the MIX treated soil has the most complex microbial co-occurrence pattern with the highest node and edge counts as well as clustering coefficient; it has the lowest ratio of negative to positive connections (Figs. 4a-b and S5). This indicates that microbial cooperation is dominant, improving the metabolic efficiency of microorganisms and the utilization rate of nutrients by plant uptake as a result [59]. The percentages of bacterial and bacterial connections compared to total connections was > 80%, connections between bacteria and fungi > 10%, and between fungi and fungi between 1% -2%. The MIX treated soil had most bacterial-fungal connections, accounting for 16.4% of the total number of connections (Figs. 4b-d). Most nodes species are present in MIX treated soil bacteria phylum Actinobacteriota and fungus phylum Ascomycota are dominant (Figs. 4c-e). Microorganisms with top 20 connectivity were considered as key species. The MIX treated soil has the highest average number of important species connections, followed in descending order by the ORG treated soil with the DAP treated soil possessing the least connections (Figs. 4c and 4d). Key genera Mizugakiibacter, Granulicella, Planococcus, Mizugakiibacter, and Planococcus have inhibitory and resistance effects on plant diseases while Granulicella promotes plant growth as well [60, 61].
The co-occurrence patterns of different fertilizer treatments corresponding to the growth stages of Tibetan hulless barley show that the tight complexity of interactions gradually resulted in a decrease across growth stages (Fig. S6). The complexity of the network structure for fungal indicators increased from tillering to jointing growth stage before decreasing in the ripening growth stage. The connectivity of each network varied across the growth stages as well. Bacterial interaction structure varied while fungal interaction structure exhibited a slow-release response time. This was due to the bacteria having a shorter turnover time and responding more rapidly to environmental changes while fungi show resistance to environmental changes, resulting in a longer period of growth turnover time [19, 62].
The combined application of organic and inorganic soil amendments significantly increases the abundance of key species, major soil ecological groups in relation to key taxon, and improves interrelated interactions between microbes and their functionalities. These complex and tight interactions can enhance the utilization of resources and resistance to environmental stressors and disease, thus promoting the healthy growth of Tibetan hulless barley in an inarable plateau environment [17, 63–65].
Correlation between microbial communities and soil amendment treatments
Plant traits showed that there was no significant difference in plant heights across soil amendment treatments during the tillering stage; plant heights fromDAP soil amendments were significantly lower compared to the other soil amendment treatments during the jointing and ripening stages (Fig. 5). Thousand Grain Weight (TGW) among the three soil amendment treatments were ranked MIX > ORG > DAP; a similar trend was observed for number of plants (Fig. 5 and S7a). The estimated yield was highest in the ORG treatment with no significant difference between DAP and MIX treatments. The MIX treatment achieved the most optimal growth and crop yield (546, 973 kg/km2) (Fig. 5a and S7a).A total of 9 OTUs from rhizosphere microorganisms share a significant correlation with TGW and plant height (Fig. 5b; ρ > 0.5, p < 0.05). These are as follows: bacteria OTU216 (Devosia, 0.71%), OTU913 (Nocardioides, 0.04%), OTU3105 (Cereibacter, 0.03%), OTU4330 and OTU4827 (Brachybacterium, 0.06% and 0.03%), OTU4942 (Flavisolibacter, < 0.01%), OTU4686 (norank family Caloramatoraceae, < 0.01%), OTU6855 (norank family Vicinamibacteraceae, 0.01%), and fungi OTU2172 (Mortierella, < 0.01%) (Fig. 5b). The number of microorganisms involved in a positive correlation with plant height did not vary significantly across soil amendment treatments (3 from DAP, 2 from MIX, and 1 from DAP & ORG). Species correlation with Tibetan hulless barley yield, grain number, and plant number show that most plant growth indicator species share a significant positive correlation with plant height and enriched in the jointing stage (Fig. S7b). The soil amendment indicator and abundant species share a notable positive correlation with grain counts. These species mostly belonged to the MIX group and ORG group, with a small portion belonging to the DAP group (Fig. S7c). Compared with different growth stages, highly sensitive bacteria belonged to phylum Proteobacteria. Bacteria responsive to soil amendment treatments mostly belonged to phyla Actinobacteria and Ascomycota (Fig. 3).
The PLS-SEM model showed that the effect of soil amendment treatments on soil pH was most observable among environmental factors. The analysis interpretation (R2) of soil organic matter sources show that MIX treatment had the most significant positive effect on the increase of soil organic matter. The growth of bacteria and fungi was positively affected by soil organic matter. Soil properties (pH, inorganic matter, organic matter) had a stronger effect on the rhizosphere bacterial community than fungi community. It indicates that bacteria are more sensitive to fertilization than fungi during plant growth. The MIX treatment contributed the most to changes in soil organic concentration for the microbiome as well as the growth and yield of Tibetan hulless barley (Fig. 5c).
The health and quality of farmland is driven by soil physicochemical factors and the microbiome in relation to rhizosphere microorganisms. Through linking soil amendment treatments with soil physicochemical parameters, growth stages, and the microbiome, we can establish a relationship between the response model of the complex soil-microbe-crop system and Tibetan hulless barley. Specific microbial groups from the soil around the root zone of Tibetan hulless barley, more specifically unique bacteria including Arthrobacter, Microbacterium, Planomicrobium, Nocardioides, and so on were optimally enriched. Compared to fungi, bacteria are more diverse in their response. However, fungi can more robustly maintain ecological plateau farmland systems. Rhizosphere microorganisms form complex interactions and are subjected to a dynamic process impacted by host type, plant growth environment, and physiological status [41].
The relationship between Tibetan Hulless Barley rhizosphere microbiome and metabolic functions
The metagenomic data was effectively assembled in 18 relative high quality draft MAGs (Table S5). MAG coverage of the dominant and representative rhizosphere bacteria (Nocardioides, Rhodanobacter, Sphingomonas, Mycobacterium, Flexivirga, and so on) helps us to infer presence and absence of pathways and understand comprehensive metabolic functions of rhizosphere microorganisms. In carbon biodegradation pathway, chitin-degrading genes degrade inactive carbon in the soil, replenishing the supply of carbon sources under extreme environments and conditions [66]. Rhizosphere microbial functions during carbon fixation, such as glycerol utilization, pyruvate metabolism, formate metabolism, ethanol fermentation, and acetyl-CoA metabolism are all related to the rTCA cycle. This cycle is prevalent in anaerobic bacteria that thrive in the low-oxygen conditions of the Tibetan Plateau. A large variety of acetic acid fermentation genes were discovered as well. There is increasing evidence that acetic acid production plays an important role in the organic carbon cycle in extreme microaerobic or anaerobic habitats [67]. Nitrogen cycling driven by the rhizosphere microbiome is important for crop growth [45]. Assimilation and dissimilatory nitrate reduction and glutamate synthesis are both related to the conversion of ammonia into organic nitrogen and an ammonia transporter helps plant roots absorb ammonium nitrogen [68, 69]. In extreme environments, Tibetan hulless barley rhizosphere microorganisms can synthesize large amounts of amino acids for resistance against environmental stressors. As a result, a consistent input of ammonia is crucial for a thriving microbial community. Assimilatory sulfate reduction consumes sulfate in the environment to synthesize sulfur-containing amino acids. Cysteine can be degraded and catalyzed by cysteine desulfhydrase to generate hydrogen sulfide, pyruvate, and ammonia-all of which possess regulatory effects on plant growth [70, 71]. Phosphorus metabolism includes inorganic phosphorus, polyphosphoric hydrochloride, phosphate, and organic phosphorus metabolism. Energy is released, promoting plant resistance to harsh environmental stressors and inhibitors to growth [72]. Additionally, some functional genes in rhizospheric microorganisms such as heat shock protein genes, genes related to heavy metal resistance, and genes that resist oxidative stress help plants adapt to harsh environments and promote plant growth as well.
Fertilizers promote changes in microbial abundance and significantly impact various metabolic functions such as carbon, nitrogen, phosphorus, and sulfur cycles (Fig S8). Genes in higher abundance within the MIX treatment include fructose and mannose metabolism genes mtlK, pfp, and mtlA in the carbon degradation pathway; TCA cycle genes such as pyruvate metabolism genes fumA and fumB; glyoxydicarboxylic acid metabolism genes acnB and PCCA in the carbon fixation pathway; and acetate fermentation gene aldH. This indicates that the microbial community of the MIX treatment group has more carbon source pathways. Additionally, nitrate reduction genes associated with the nitrogen cycle, nrtC and narB as well as amino acid metabolism-related genes (arginine metabolism genes GDH2 and ureC, cysteine and methionine metabolism genes gshB, and valine, leucine, and isoleucine metabolism genes vdh and accD6) were more abundant in the MIX treatment group. Ammonia was likely converted into organic nitrogen, which is beneficial for crop production [73].
Sulfur-related sulfate-reducing genes cysA and ssuD are more abundant in the MIX treatment group and could fix arsenic, reducing its effectiveness in rice plants and providing an effective way to inhibit the infiltration of arsenic into the food chain [74]. The abundance of nucleotide metabolism-related gene xdhB, cofactor and vitamin metabolism-related genes cobN and cobIJ, terpenoid metabolism-related gene atuF, and lipid metabolism-related primary bile acid genes AMACR and mcr in the MIX group is higher compared to the other treatment groups. Vitamins and terpenoid metabolites play an important role in promoting the production of essential compounds in plants and bacteria, inducing resistance to pathogens and promoting plant growth as a result[75, 76]. Additionally, primary bile acids contribute to crop disease resistance [77]. Therefore, mixed fertilizer can be used optimally to promote Tibetan hulless barley resistance to negative plateau environmental stressors and inhibitors to plant health.