Chito-oligosaccharide loaded with ACC (1-aminocyclopropane-1-carboxylate) deaminase-producing bacteria improved soil microecology and cotton growth in salinized lands

Abstract

Chitosan oligosaccharide, as a soil amendment, is widely used to cope with various environmental stresses. This study comparatively analyzed the effects of chitosan oligosaccharide-based polymer amendment (COS-PA), ACC deaminase-producing bacterial agent (A1), and their combination (COS loaded with ACC deaminase-producing strains, CAS) on cotton vegetative growth under salt stress. The results showed that the application of COS-PA, A1, and CAS significantly reduced soil pH and EC values compared with the control. Numerically, the CAS treatment showed the most significant effects among the treatments, with soil sucrase, urease, and alkaline phosphatase activities 27.84%, 111.4%, and 33.79% higher than those of the control, respectively. Additionally, this treatment also held significantly higher leaf peroxidase and catalase activities and cotton vegetative growth parameters. Microbial analysis showed that CAS mainly improved soil microecology by changing the relative abundance and composition of fungal (Basidiomycota) and bacterial (Atescibacteria and Actinobacteria) communities. In summary, under the specific conditions of this pot experiment, the CAS application was associated with improved soil enzyme activities and cotton growth parameters. Field validation is required. This study provides a reference for the development of new soil amendments and contributes to the sustainable utilization of salinized lands.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09495-6.

Highlights

Application of chito-oligosaccharide loaded with ACC deaminase-producing bacteria was associated with improved soil microecology.

Application of chito-oligosaccharide loaded with ACC deaminase-producing bacteria was correlated with enhanced cotton growth under saline-alkali stress in pot experiments.

Shifts in the soil fungal community were associated with changes in soil quality parameters.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09495-6.

Introduction

Soil salinization is a common soil degradation problem in arid, semi-arid, and coastal regions. Soil salinization weakens many ecological functions of soil, limiting the sustainable agricultural development [1]. According to reports, soil salinization not only disrupts soil water balance, leading to soil compaction [2], but also subjects plants to osmotic stress, ion toxicity, acid-base imbalance, and oxidative stress. These affect water and nutrient absorption of crop roots, inhibiting crop growth and yield formation [3, 4]. Therefore, remediating salinized lands has become an important task in sustainable agricultural development.

Traditionally, salinized soils are remediated using physical, chemical, or biological technologies. The use of each technology often disrupts the balance of the soil environment and has a limited effect on crop yield improvement. Rhizobacteria enhance soil fertility through nitrogen fixation and phosphorus and potassium solubilization [5]. They also regulate plant growth and development by producing plant hormones and enhance plant stress resistance by activating plant defense genes [6, 7]. In addition, some rhizobacteria can alleviate drought stress in plants by secreting extracellular polysaccharides, and help to resist salt stress by reducing osmotic potential. The bacteria known to enhance plant salt resistance are mainly Bacillus, Pseudomonas, Enterobacter, and Azotobacter [8, 9]. Although these bacteria have a positive effect on soil quality and crop yield, direct application of exogenous bacteria poses risks to soil ecological balance and may induce nutrient competition among species. Therefore, it is necessary to choose suitable carriers. Chitosan oligosaccharides (COS) are a candidate material for carriers. Studies have shown that COS carry a lot of free amino groups and are positively charged in mildly acidic environments. Chitosan oligosaccharides can instantly and strongly bind through electrostatic interactions with negatively charged groups (i.e., teichoic acids and lipopolysaccharides) on the surface of microbial cells [10]. The slow-release characteristics of COS provide a steady supply of nutrients for bacteria [11]. However, conventional carriers (i.e., diatomaceous earth, zeolite, and bentonite) cannot provide nutrients. They just act as adsorbents and encapsulation materials and can’t support microbial metabolism and survival. Additionally, materials like molasses and starch-based additives are quickly metabolized by microbes, leading to overgrowth of unwanted bacteria.

Chitosan oligosaccharide (COS) is a degradation product of chitosan. Because of characteristics such as low viscosity, high solubility, high biocompatibility, and easy absorption and utilization by organisms, COS and its derivatives are widely used in agricultural production for pest and disease control [11, 12]. Zhao et al. [12] found that the Cr absorption capacity of COS was as high as 148.1 mg/g, and COS application increased the biomass and heavy metal stress resistance of Cannabis sativa L. compared with the control [13]. Wang et al. [14] found that after applying chitosan hydrogel microspheres loaded with Bacillus subtilis for remediating soil heavy metal pollution, the biomass of Lactuca sativa L. var. ramosa Hort. was increased by 88%, the Cr concentration in plants was reduced by 17%, and the plant nutrient absorption and the abundance of bacteria with heavy metal stress resistance (Proteobacteria, Actinomycete, and Cerobacteria) were significantly increased. Microorganisms in salinized soil ecosystems use a series of biochemical processes and their metabolites to decompose the applied organic amendments, converting organic matter into humus and releasing essential nutrients. This can greatly enhance soil structure and nutrient levels [1]. Arshad et al. [15] found that the application of 1-aminocyclopropane-1-carboxylate (ACC)-deaminase producing rhizobacteria (EC 4.1.99.4) significantly alleviated the negative effects of external abiotic stress on the growth and yield of Pisum sativum L. It can be seen that rhizobacteria and COS have positive effects on farmland soil quality and crop growth under external stress.

Chitosan oligosaccharide can reduce the toxicity of harmful ions in the soil and promote crop growth under abiotic stress conditions. However, research on the effects of COS and COS loaded with rhizobacteria on the salinized soil microenvironment is limited [12]. Changes in microbial diversity, like an increase in the relative abundance of core bacteria, are one of the main ways soil microbes resist abiotic stress [16]. Therefore, this pot experiment was conducted. The objectives were to reveal the effect of COS, ACC deaminase-producing bacterial agent (A1), and the COS loaded with ACC deaminase-producing strains (CAS) on (1) soil physicochemical properties and cotton growth and physiological characteristics at the budding stage, (2) dominant soil bacterial and fungal community in salinized soil, and (3) salinized soil remediation. It was hypothesized that the application of COS and CAS might increase biomass accumulation during the budding stage and the diversity of key soil bacteria and fungi, thus improving cotton yield and salinized soil. This study will provide a reference for the development of new soil amendments and contribute to the sustainable utilization of salinized lands.

Methods

Experimental materials

The cotton variety “Xinlu Zao72” bred by Xinjiang Huiyuan Seed Industry Co., Ltd., China was used in this experiment. It has strong resistance to bacterial blight and verticillium wilt. The soil for the experiment was collected from Dafeng Town, Hutubi County, Xinjiang, China (44.18°N, 86.60°E). After removing residues such as stones and branches, the soil was thoroughly mixed with perlite and vermiculite in a ratio of 4: 1: 1. The mixture was used as the substrate for planting cotton. The organic carbon content of the soil mixture was 82.3 g/kg. The total nitrogen content was 3.11 g/kg. The total phosphorus content was 966 mg/kg. The total potassium content was 20.2 g/kg. The available phosphorus content was 18.58 mg/kg. The readily available potassium content was 627 mg/kg. The sodium content was 11.9 g/kg, the pH was 7.78, and the electrical conductivity (EC) was 308 µS/cm.

The 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase-producing bacterial agent (A1) was prepared by following procedures: Pseudomonas cedrina (DY1-3) and Rhizobium pusense (W9), with the ability to produce ACC deaminase, were extracted in the laboratory [17]. Then, ACC was hydrolyzed by ACC dehydrogenase into α-butanone acid and ammonia. The 2,4-dinitrophenylhydrazine reagent was used to stain α-ketobutyrate. The content of α-ketobutyrate in the bacterial solution was determined to know the ACC-deaminase activity of this strain [18]. After that, LB medium was prepared. Specifically, 10 g of peptone, 5 g of yeast extract, and 10 g of NaCl were dissolved in deionized water. After making the volume to 1 L using deionized water, the pH was adjusted to 7.0 with NaOH (1.5% agar was added to the preparation of solid media). After activating W9 and DY1-3, the strains were cultured in the LB medium at 33℃ at 180 rpm for 12 h. The fermentation broth (mixture of W9 and DY1-3 (1:1)) was adjusted to OD₆₀₀ = 0.8 with sterile water. Finally, A1 was obtained [19]. This study didn’t directly measure the pH and EC values of A1 before application. However, based on the optimal pH of 8.5 for ACC deaminase in Pseudomonas [20] and the broad pH-tolerance range of Rhizobium pusense (5.5-9.0), it’s assumed that A1 was neutral to slightly alkaline.

The chitosan oligosaccharide-based polymer amendment (COS-PA) was prepared by mixing COS (Miclin Biochemical Technology Co., Ltd., Shanghai, China) (Table 1) with manganese sulfate (MnSO₄), an amino acid complex, γ-polyglutamic acid (γ-PGA), polyaluminum chloride (PAC), magnesium sulfate (MgSO₄), and calcium hydrogen phosphate (CaHPO₄) (Figure S1). The detailed preparation procedure followed the protocol described by An et al. [21]. This experiment did not measure the pH of the COS-PA solution before dilution. However, our previous research [21] showed that the pH of the prepared COS-PA was 4.6–4.8. The EC value was not measured directly; instead, the impacts of the amendments on soil salinity were assessed indirectly by changes in soil EC after application (Table 2). The resulting COS-PA solution was then diluted to a final chitosan concentration of 100 mg/L before being applied to the root zone.

Table 1.

Physicochemical characteristics of the chito-oligosaccharide product

Parameter	Value	Unit
Moisture content	1.05	g/100 g
Monosaccharide content	Not detected	-
Convertible glucosamine content	5.93	g/100 g
Chito-oligosaccharide content	90.2	g/100 g
pH	4.86	-
Relative molecular mass	1643	Da

Table 2.

Soil pH and electrical conductivity (EC)

	pH	EC (µS/cm)
CK	8.03 ± 0.06 b	229.50 ± 22.54 c
S	8.69 ± 0.09 a	499.10 ± 24.58 a
T1	7.79 ± 0.30 b	241.67 ± 24.83 c
T2	7.82 ± 0.08 b	177.73 ± 19.89 c
T3	7.58 ± 0.27 b	237.33 ± 15.51 c
T1S	7.82 ± 0.08 b	367.80 ± 28.72 b
T2S	7.84 ± 0.27 b	346.17 ± 19.75 b
T3S	7.78 ± 0.07 b	341.93 ± 22.41 b

Experimental design

There were eight treatments in the experiment (Table 3), namely: CK (no salt stress & no A1 & COS-PA application), S (salt stress & no A1 and COS-PA application), T1 (no salt stress & COS-PA application), T2 (no salt stress & A1 application (W9: DY1-3 = 1: 1)), T3 (no salt stress & CAS (COS-PA: A1 = 1: 1) application); T1S (salt stress & COS-PA application), T2S (salt stress & A1 application), and T3S (salt stress & CAS application). To simulate the actual ion composition of salinized lands in Xinjiang, China, NaCl, Na₂SO₄, NaHCO₃, and Na₂CO₃ were mixed in a 12: 9: 8: 1 ratio (molar ratio) [22]. Then, the salts were dissolved in 2.5 L of distilled water and stirred to prepare the salt solution (pH: 8.77; EC: 71,600 µS/cm).

Table 3.

Experimental design

	Salt solution	Chitosan oligosaccharide-based polymer amendment (COS-PA)	ACC deaminase-producing bacteria agent (A1)
CK	-	-	-
S	+	-	-
T1	-	+	-
T2	-	-	+
T3	-	+	+
T1S	+	+	-
T2S	+	-	+
T3S	+	+	+

“+” indicates addition, and “-” indicates no addition

Forty pots (inner diameter: 19 cm; height: 24.7 cm) were used, and each pot was filled with 5 kg of the prepared substrate. Each treatment had five repetitions. Cotton seeds were sterilized in 9% hydrogen peroxide for 20 min, rinsed with pure water, and sown (5 seeds per pot) in vermiculite containing salt solution. The seeds were germinated at 25 °C until the cotyledons were fully expanded. After germination, the seedlings were thinned to four uniform plants per pot for further growth. On May 25, the first fertilization was carried out, and the COS-PA, A1, and CAS were applied to the substrate (400 mL of A1 per pot, 300 mL of COS-PA per pot, and 700 mL CAS (400 mL of A1 & 300 mL of COS-PA) per pot). Before application, the amendment liquid for each pot was made up to 700 mL with distilled water. Then, the liquids were added slowly and evenly to prevent any leaks. On June 10, only two plants were retained in each pot, and the second fertilization was carried out. Fertilization (0.077 g of urea, 0.128 g of diammonium, and 0.026 g of potassium sulfate per 1 kg of soil) was carried out every 7 days.

During the budding stage (60 days after transplantation), cotton plant samples and root-zone soil samples were randomly collected (three replicates per treatment). Pots were arranged in a completely randomized design. Following the principle of even distribution (avoiding the pot walls and densely rooted areas), three subsamples were randomly collected from each pot and mixed thoroughly. This mixture was used as the representative soil sample for the pot. For each treatment (five pots), subsamples were collected in the same way. Then, the five representative samples from the same treatment were mixed. Finally, the mixture was divided into three replicates for subsequent measurements. This sampling method could minimize human error caused by spatial heterogeneity. Visible plant residues and gravels were immediately removed from the collected soil samples. Then, the soil samples were divided into two parts after sieving (2 mm). One part was stored at -80 °C for subsequent analysis. The other part was dried to constant weight in an oven at 70 °C for 72 h, ground, and sieved (0.15 mm) for later use.

Determination of cotton growth indicators

Determination of plant height and fresh weight

Cotton plants were pulled out of the pots. The root systems were slowly rinsed with running water to remove the attached soil. After absorbing surface water using absorbent paper, the vertical height (plant height) was determined using a ruler with an accuracy of 0.1 cm. The fresh weight of the entire plant was determined using an electronic balance with an accuracy of 0.01 g.

Calculation of maximal leaf area

The fully expanded functional leaves at the top of the plants were selected to measure the main vein length (L) and the maximal leaf width (W) using a digital caliper. Then, the maximal leaf area (A) was calculated using the following formula [23]:

1

Determination of dry weight of the shoots and roots

The shoots and roots of the plants were separated and put into Kraft bags. The plant samples were dried in an oven at 105℃ for 30 min to inactivate the enzyme. Then, the samples were dried to constant weight at 80℃. The dry weight of each part was recorded.

Determination of main root length and stem diameter

The entire root and stem system was laid flat on a clean table. Then, a laser rangefinder was used to measure the maximum straight-line distance between the two ends of the main root and stem diameter.

Determination of cotton physiological parameters

Determination of relative chlorophyll content

The SPAD-502 chlorophyll meter (Minolta, Japan) was used to determine the SPAD value of the second functional leaves on the apex of the cotton plants, based on the light absorption difference between 650 and 940 nm. According to Jiang et al. [24], three points on each leaf that were not on the main vein were selected to determine the SPAD value, and the average was calculated as the relative chlorophyll content of the leaf.

Determination of chlorophyll content

The second functional leaf on the plant apex was used to determine chlorophyll content. Leaf samples (0.3 g) that were not on the main vein were collected, frozen in liquid nitrogen, sliced, and transferred into a centrifuge tube. After that, 20 mL of pre-cooled 95% ethanol was added, followed by extraction at room temperature for 24 h under light conditions until the leaves were completely decolorized. Then, the samples were centrifuged at 12,000 × g for 15 min at 4 °C. The supernatant was collected as the pigment extract. A UV-1200 spectrophotometer preheated for 30 min was used to determine the absorbance (OD value) of the extract at 664.2, 648.6, and 470 nm with an ethanol solution as a reference. The measurement was repeated three times for each sample. The contents of chlorophyll a (Chl a), chlorophyll b (Chl b), and total chlorophyll (Chl) were calculated according to the formulas of Schoefs [25].

2

3

where A664.2 and A648.6 represent the absorbance values at 664.2 and 648.6 nm, respectively.

The carotenoid content (Car) was calculated using the following formula [25]:

4

Antioxidant enzyme activity assay

Fresh leaves (0.1 g, frozen in liquid nitrogen) were homogenized with 1 mL of pre-cooled normal saline (9 g/L NaCl) in a tissue homogenizer at 15,000 rpm for 30 s under ice bath conditions. Then, the samples were centrifuged at 12,000 × g for 15 min at 4℃. The supernatant was collected as the crude enzyme extract. The activity of superoxide dismutase (SOD) was determined using the BC5165 kit produced by Solarbio Science & Technology Co., Ltd., Beijing, China, based on nitro-blue tetrazolium (NBT) photoreduction [26]. The amount of SOD that inhibited NBT photoreduction by 50% was defined as 1 U. The activity of peroxidase (POD) was determined using the oxidation of guaiacol [26]. According to the requirement of the BC0095 kit (Beijing Solarbio Science & Technology Co., Ltd., China), the absorbance change at 470 nm was determined. ΔOD470 increased by 0.01 per minute was defined as 1 U. To determine the activity of catalase (CAT), the decomposition rate of H₂O₂ at 240 nm was detected by UV spectrophotometry [26]. The decomposition of 1 µmol of H₂O₂ per minute was defined as 1 U according to the A007-1 kit (Beijing Solarbio Science & Technology Co., Ltd., China). All determinations were completed using a Synergy H1 enzyme-linked immunosorbent assay (BioTek, USA), with three replications. The malondialdehyde (MDA) content was determined using the thiobarbituric acid (TBA)-colorimetry [26]. The specific procedure followed the instructions of the Glutaraldehyde Assay Kit (Catalog No. BC0025, Solarbio Technology Co., Ltd., Beijing, China). Firstly, a crude enzyme extract was prepared. Then, 0.5 g of tissue sample was homogenized with the extract in an ice bath. After centrifugation, the supernatant was collected for later use. In the assay, a certain amount of crude enzyme extract was thoroughly mixed with an equal volume of 0.5% thiobarbituric acid (TBA) solution. The mixture was placed in a boiling water bath for 15 min to fully complete the color development. After the reaction, the test tube was immediately placed in an ice bath to stop the reaction. Once the sample was cooled to room temperature, centrifugation was conducted to collect the supernatant. A spectrophotometer or microplate reader was used to determine the absorbance of the reaction solution at 532 (OD532) and 600 nm (OD600), followed by the calculation of the MDA content.

Determination of soil enzyme activity

Soil enzyme activity was determined using kits produced by Sorbo Biotechnology Co., Ltd., Beijing, China. Soil alkaline phosphatase (EC 3.1.3.1) activity in soil samples was determined using the S-AKP/ALP activity detection kit, based on the reaction of the phenolic products generated by the enzymatic reaction with the chromogenic reagent under alkaline conditions. According to the instructions of the kit (Catalog No.: S-AKP/ALP), air-dried and sieved soil samples were weighed (0.1 g). Then, the buffer and substrate were added, followed by the incubation in the dark at 37 °C for 24 h in a constant-temperature incubator. After the reaction, the absorbance of the samples at 400–420 nm was quickly determined. One unit was defined as the amount of enzyme that released 1 nanomole of phenol per gram of soil per day at 37 °C.

Soil sucrase (EC 3.2.1.26) activity was determined using the S-SC activity detection kit, based on the 3,5-dinitrosalicylic acid (DNS)-based colorimetry. Specifically, a soil sample was weighed (0.03 g), and a sucrose solution prepared with acetate buffer was added. Then, the sample was incubated at 37 °C for 24 h. Sucrase catalyzes the hydrolysis of sucrose to produce reducing sugars such as glucose and fructose. After incubation, the DNS reagent was added, followed by a boiling water bath for color development. After cooling, the volume was made to 0.3 mL. After filtering with filter paper, the absorbance was determined at 510 nm. One unit was defined as the amount of enzyme that produced 1 mg of reducing sugar per gram of soil per day at 37 °C.

Soil urease (EC 3.5.1.5) activity was detected using the S-UE activity detection kit based on the indophenol blue-based colorimetry. Soil samples (0.1 g) were weighed, and citrate buffer and urea substrate were added, followed by incubation at 37 °C for 24 h. The ammonium nitrogen produced by urease hydrolysis of urea reacted with alkaline sodium hypochlorite and sodium phenol solution to generate blue indophenol dye. This colored compound has a characteristic absorption peak at 578 nm, and the color intensity is proportional to the ammonium nitrogen content [27]. One unit was defined as the amount of enzyme that produced 1 µg of NH_₃-N per gram of soil per day.

Determination of soil pH and EC

Soil pH was determined according to the methodology described by Nielsen et al. [28]. Specifically, 10.0 g of dried soil sample was mixed with 25 mL of deionized water (1: 2.5, w/v) in a 50 mL centrifuge tube, followed by shaking at 200 rpm at room temperature for 30 min. After 1 h, the supernatant was collected, and its pH was determined by a pH meter (METTLER TOLEDO FE28, Switzerland). The EC value (µS/cm) was determined using a conductivity meter (HANNA HI2300, Italy; µS/cm) at 25 °C.

High-throughput sequencing and quality control

Total DNA was extracted from the cotton root-zone soil samples using the MOBIO soil DNA extraction kit (Qiagen). DNA samples were sent to Beijing Baimaike Biotechnology Co., Ltd., China for PCR amplification and high-throughput sequencing. The 16 S rRNA sequence of root-zone bacteria was amplified using universal primer pairs 799 F/1115 and 338f/806r. The ITS rRNA sequence of root-zone fungi was amplified using universal primer pairs ITS1F/ITS2 and ITS1F/ITS4. High-throughput sequencing was performed using Illumina’s MiSeq sequencing platform.

After sequencing, the raw reads were filtered using Trimmomatic v0.33. Then, the primer sequence was identified and removed using the cutadapt 1.9.1 to obtain clean reads. Denoising was completed using the dada2 in QIIME2 2020.6. After merging paired-end sequencing data and removing chimeras, the final non-chimeric reads were obtained.

Statistical analysis

The experimental data were sorted and preliminarily calculated using Microsoft Excel 2016 (Microsoft Corp., USA). The bar charts were plotted in Origin 8.0. Statistical analysis was performed using SPSS 26 software (IBM Corp., USA). The data in the charts are mean ± standard deviation (mean ± SD, n = 3). The significance of differences between different treatments was tested by one-way analysis of variance (ANOVA). Before performing ANOVA, the homogeneity of variance was first tested. Based on this, when the data met the homogeneity of variance, one-way ANOVA combined with Duncan’s multiple range test was used to compare differences between treatments. For data that did not meet homogeneity of variance, Welch’s ANOVA combined with the Games-Howell post hoc test was used for statistical inference. The significance level was set at p < 0.05. The random forest classification method was used to classify microbial communities at the phylum level based on relative abundance, using randomForest in R 3.6.1 (https://www.rproject.org/). Then, key microbial communities were extracted based on mean decrease accuracy > 0.7. LEFSe (Linear discriminant analysis Effect Size) was performed at the phylum and class levels to determine the specific microbiota in response to different treatments. Mantel analysis was performed to explore the relationship between soil physicochemical properties, cotton growth, cotton physiological parameters, and key soil microbial communities under different treatments.

Results

Analysis of soil physicochemical properties

The soil pH of the COS treatments (COS-PA and CAS) was 2.62%-11.08% lower than that of CK, and the soil EC value was 22.56%-35.70% lower than that of CK (p < 0.05) (Table 2). The soil sucrose and urease activities of the S treatment were lower than those of CK. The soil sucrose activity of the T2 treatment and the soil sucrose and urease activities of the T3 treatment were 61.86%, 77.91%, and 54.89% higher than those of CK, respectively (p < 0.05). The soil alkaline phosphatase activity of the T2 and T3 treatments was 31.38% and 40.05% higher than that of CK, respectively (p < 0.05). The activities of soil urease and alkaline phosphatase of the T1S treatment and the activity of soil urease of the T2S treatment were 50.41%, 36.18%, and 103.4% higher than those of the S treatment, respectively (p < 0.05). The activities of soil sucrose, urease, and alkaline phosphatase of the T3S treatment were 27.84%, 111.4%, and 33.79% higher than those of the S treatment, respectively (p < 0.05) (Fig. 1).

Fig. 1.

Soil enzyme activities under different treatments. a Soil sucrase; b Soil urease; c Soil alkaline phosphatase. CK, S, T1, T2, T3, T1S, T2S, T3S. Treatment abbreviations are defined in Table 3. Different lowercase letters indicate significant differences between treatments (p < 0.05). Data represent the means of three replications. The same below

Analysis of cotton growth and development at the budding stage

The main root length of the T1, T2, and T3 treatments was 32.42%, 33.01%, and 37.66% higher than that of CK, respectively (p < 0.05), and the root dry weight was 38.66%, 36.01%, and 49.02% higher than that of CK, respectively (p < 0.05). The stem diameter, main root length, leaf area, shoot dry weight, and root dry weight of the T1S, T2S, and T3S treatments were numerically higher than those of the S treatment. Among them, the plant height, stem diameter, root length, leaf area, shoot dry weight, and root dry weight of the T3S treatment were 35.56%, 35.25%, 64.83%, 43.56%, 104.1%, and 116.3% higher than those of the S treatment, respectively (p < 0.05). Besides, there were differences in plant height, root length, shoot dry weight, and root dry weight between T3S and T1S treatments (p < 0.05) (Fig. 2).

Fig. 2.

Cotton growth under different treatments. a shoot height; b stem diameter; c main root length; d leaf area; e shoot dry weight; f, root dry weight. Different lowercase letters indicate significant differences between treatments (p < 0.05). Data represent the means of three replications

Analysis of cotton physiological characteristics at the budding stage

The Chl a content and SPAD value of cotton leaves of the T2 treatment were 32.14% and 10.47% higher than those of CK, respectively; The Chl a content and SPAD value of the T3 treatment were 61.71% and 27.12% higher than those of CK, respectively (p < 0.05). The Chl a content of the T1S, T2S, and T3S treatments was 26.20%, 66.01%, and 92.26% higher than that of the S treatment, respectively (p < 0.05), and the SPAD value was 36.56%, 29.70%, and 47.18% higher than that of the S treatment, respectively (p < 0.05). The Chl b content of the T2S and T3S treatments was 41.37% and 64.19% higher than that of the S treatment, respectively (p < 0.05), and the Car content was 48.07% and 54.06% higher than that of the S treatment, respectively (p < 0.05) (Fig. 3a). The leaf CAT activity of the T3S treatment was 126.1% higher than that of the S treatment (p < 0.05), and the MDA content of the T1S, T2S, and T3S treatments was 41.25%, 30.46%, and 29.11% lower than that of the S treatment (p < 0.05) (Fig. 3b).

Fig. 3.

Pigment content (a) and antioxidant enzyme activity (b) in cotton leaves under different treatments. Different lowercase letters indicate significant differences between treatments (p < 0.05). Data represent the means of three replications

Analysis of soil bacterial and fungal diversity

There were significant differences in fungal community composition among different treatments (Fig. 4a, b). The addition of COS and A1 did not affect bacterial diversity indices (ACE, Chao1, Simpson, and Shannon) (p > 0.05) (Fig. 4c). The fungal Chao1 and ACE indices of the T2 and T3 treatments were lower than those of CK (p < 0.05). The fungal Chao1 and ACE indices of the T1S, T2S, and T3S treatments were lower than those of the S treatment (p < 0.05), but there was no difference in Simpson and Shannon indices (p > 0.05) (Fig. 4d).

Fig. 4.

Soil microbial diversity under different treatments. Principal component analysis (PCA) based on Bray-Cutis distance metrics of soil bacterial (a) and fungal (b) communities. The 95% confidence ellipses are shown around the samples. Box plots of α-diversity of soil bacterial (c) and fungal (d) communities

Analysis of soil bacterial and fungal community composition

The analysis of the soil bacterial community (Fig. 5a) showed that Proteobacteria were the dominant phylum. The abundance of Bacteroidota of T1, T2, and T3 treatments was 4.11%, 4.97%, and 6.77% lower than that of CK, respectively. The relative abundance of Bacteroidota of T1S, T2S, and T3S treatments was 5.49%, 5.23%, and 5.08% lower than that of S treatment, respectively. The relative abundance of Proteobacteria of T1, T2, and T3 treatments was 3.41%, 7.31%, and 11.17% higher than that of CK, respectively. The relative abundance of Acidobacteriota of T1S, T2S, and T3S treatments was 4.24%, 0.67%, and 3.58% higher than that of S treatment, respectively. The relative abundance of Actinobacteriota of T2 and T3 treatments was 30.97% and 4.05% higher than that of CK, respectively. The relative abundance of Actinobacteriota of T1S, T2S, and T3S treatments was 10.44%, 12.16%, and 58.82% higher than that of S treatment, respectively.

Fig. 5.

Soil microbial community composition under different treatments. The relative abundance of 10 most abundant microbial phyla (a bacterial, b fungal) and classes (c bacterial, d fungal)

The relative abundance of the dominant fungal phylum Ascomycota of the T1 and T2 treatments was 27.41% and 3.34% higher than that of CK, respectively, while its abundance in the T3 treatment was 12.24% lower than that of CK. The relative abundance of Ascomycota of the T1S and T2S treatments was 3.82% and 5.26% higher than that of the S treatment, respectively, while its abundance in the T3S treatment was 13.03% lower. The relative abundance of Basidiomycota of the T1, T2, and T3 treatments was 9.96%, 3.56%, and 2.25% lower than that of CK, respectively, and its abundance in the T1S, T2S, and T3S treatments was 11.78%, 6.22%, and 10.79% lower than that of the S treatment, respectively. The relative abundance of Glomeromycota of the T3 treatment was 24.16% higher than that of CK. The relative abundance of Glomeromycota of the T3S treatment was 15.38% higher than that of the S treatment (Fig. 5b).

At the class level, the relative abundance of the dominant bacterial class Gammaproteobacteria of the T1, T2, and T3 treatments was 3.94%, 8.41%, and 3.45% higher than that of CK, respectively. The relative abundance of the bacterial class Bacteroides of the T1, T2, and T3 treatments was 3.91%, 4.72%, and 6.45% lower than that of CK, respectively. The relative abundance of Bacteroides of the T1S, T2S, and T3S treatments was 5.12%, 4.85%, and 4.88% lower than that of the S treatment, respectively. The relative abundance of the bacterial class Vicinamibacteria of the T1S, T2S, and T3S treatments was 3.98%, 0.68%, and 3.01% higher than that of the S treatment, respectively (Fig. 5c). The relative abundance of the dominant fungal class Glomeromycetes of the T3 treatment was 24.09% higher than that of CK, and its abundance in the T3S treatment was 15.48% higher than that of the S treatment. The relative abundance of the fungal class Sordariomycetes of the T1S, T2S, and T3S treatments was 6.28%, 3.96%, and 1.12% lower than that of the S treatment, respectively. The relative abundance of the fungal class Eurotiomycetes of the T1, T2, and T3 treatments was 6.13%, 1.64%, and 3.65% lower than that of CK, respectively (Fig. 5d).

Selection of key soil bacterial and fungal communities based on the random forest classification model

Key bacterial and fungal communities can be seen in Fig. 6a. Among them, bacterial phyla with a mean decrease accuracy greater than 0.7 were Firmicutes, Fibrobacterota, Bacteroidota, Elusimicrobiota, Cyanobacteria, Myxococcota, Planctomycetota, Patescibacteria, Acidobacteriota, Sumerlaeota, Spirochaetota, Bdellovibrionota, Actinobacteriota, and Dependentiae. Fungal phyla with a mean decrease accuracy greater than 0.7 were Basidiomycota, Ascomycota, Glomeromycota, unclassified Fungi, Chytridiomycota, Mortierellomycota, Rozellomycota, Mucoromycota, and Calcarisporiellomycota.

Fig. 6.

Random forest analysis intuitively displaying the importance of bacterial and fungal phyla (a bacterial, b fungal). Histogram of linear discriminant analysis (LDA) scores for discriminating microbial phyla and classes between different treatments (c bacterial, d fungal) (LDA score > 4.0)

The LEfSe analysis detected 12 key bacterial communities and 10 key fungal communities (Fig. 6b). Among them, Bacteroidota, Firmicutes, and Basidiomycota had higher abundance in the S treatment. Myxococcota and Ascomycota had higher abundance in the T1 treatment. Actinobacteriota had higher abundance in the T2 treatment. Patescibacteria and Glomeromycota had higher abundance in the T3 treatment (Fig. 6c). Acidobacteria and Ascomycota had higher abundance in the T1S treatment. Proteobacteria and Basidiomycota had higher abundance in the T2S treatment (Fig. 6d).

Integrated analysis of soil physicochemical properties, cotton growth and physiological characteristics, and soil bacterial communities

There was no significant difference in the effects of COS-PA, A1, and CAS on soil bacterial function under salt stress (Fig. 7a, b). The function Arbuscular Mycorrhizal of the T2 and T3 treatments was 9.89% and 24.97% higher than that of CK, respectively. The function Arbuscular Mycorrhizal of the T1S and T3S treatments was 17.53% and 23.42% higher than that of the S treatment, respectively. The function Plant Pathogen of the T1 and T3 treatments was 5.35% and 4.51% lower than that of CK, respectively. The function Plant Pathogen of the T1S and T3S treatments was 7.71% and 8.91% lower than that of the S treatment, respectively (Fig. 7b).

Fig. 7.

Annotation of bacterial metabolic pathways based on Picrust2 prediction (a) and fungal metabolic pathways based on FunGuild prediction (b). Relationships between the dominant soil microbial phyla, soil physicochemical properties, plant growth, and plant physiology, as revealed by the Mantel test (c) and Pearson’s correlations (d)

Mantel analysis showed a correlation between soil physicochemical properties and the abundance of Ascomycota, Glomeromycota, and Acidobacteriota in the soil under salt stress (Mantel’s, p < 0.01). There was a correlation between cotton growth characteristics and the abundance of Basidiomycota (p < 0.01), Actinobacteriota (p < 0.01), and Patescibacteria (p < 0.05) in the soil under salt stress. There was also a correlation between the physiological characteristics of cotton and the abundance of Patescibacteria (p < 0.05) and Actinobacteriota (p < 0.01) (Fig. 7c).

The abundance of soil bacteria Patescibacteria and Actinobacteriota was negatively correlated with soil EC (p < 0.05), but positively correlated with plant height, stem diameter, shoot dry weight, root dry weight, leaf area, and chlorophyll content (p < 0.05). The abundance of Acidobacteriota was positively correlated with the activities of soil urease and alkaline phosphatase (p < 0.05). The abundance of soil fungus Basidiomycota was positively correlated with soil pH and EC (p < 0.01), but negatively correlated with plant height, stem diameter, shoot dry weight, and root dry weight (p < 0.01). The abundance of soil fungus Ascomycota was negatively correlated with the activities of soil urease and alkaline phosphatase (p < 0.05). The abundance of soil fungus Glomeromycota was negatively correlated with soil pH (p < 0.05), and positively correlated with the activities of soil sucrase and urease (p < 0.01) (Fig. 7d).

Discussion

The application of COA-PA, A1, and CAS enhanced soil quality

Soil enzymes, as key catalysts for the formation and decomposition of soil organic matter and the decomposition of exogenous substances, have a significant impact on the cycling of carbon, nitrogen, and phosphorus in soil ecosystems. Soil enzyme activity has been used as an indicator of soil quality [29]. In this study, both COS-PA and A1 applications reduced the pH and EC of salinized soil and increased soil enzyme activity compared with the control. The CAS application (T3S treatment) triggered the greatest increase in soil enzyme activity. The decrease in soil pH was related to the acidity property of COS (Tables 1 and 2). Chitosan is a biodegradable and non-toxic substance that can improve soil quality, plant growth, and plant resilience. Through a specialized enzymatic process, chitosan can be degraded to produce COS, a low-molecular-weight oligosaccharide with high solubility and high bioavailability [30].

The ACC deaminase produced by plant growth-promoting rhizobacteria can increase plant biomass and tolerance to various abiotic stresses [31, 32]. In this study, the CAS application significantly reduced soil pH and EC, and significantly increased soil enzyme activity compared with the control (Fig. 1). This indicates that the application of ACC deaminase can enhance crop resistance to salt stress and soil quality. ACC deaminase is important for effective nodulation and is closely related to the production and cycling of carbon and nitrogen in the soil. Applying rhizobacteria that produce ACC deaminase can promote soil microorganisms to produce soil enzymes, enhancing soil quality [33]. Besides, Singh et al. [34] also found that inoculation with ACC deaminase-producing rhizobacteria could reduce Na⁺ accumulation and increase uptake of K⁺ in stems and roots of wheat (Tritium aestivum L.), resulting in a favorable K⁺/Na⁺ ratio and alleviating the toxic effects of salt stress. Besides, ACC deaminase-producing rhizobacteria can alleviate various biotic and abiotic stresses by reducing the ethylene level in stressed plants [34]. However, notably, the carrier COS-PA contains multiple components (γ-PGA, amino acids, and minerals). The specific active component responsible for the observed effects is yet to be determined. In addition, this study didn’t systematically characterize the physicochemical interactions between COS and the strains (i.e., zeta potential, Fourier Transform Infrared Spectroscopy, and encapsulation efficiency), so the loading mechanism is still unclear.

The application of COS-PA and CAS promoted cotton growth

Plants tolerate salt stress through enhancing osmotic regulation, synthesizing organic solutes, scavenging free radicals and toxic compounds, and increasing the activities of soil microorganisms [35]. The tolerance of plants to high salinity stress depends on their ability to remove salts from shoots or tolerate high concentrations of salt in leaves (tissue tolerance), which is closely related to plant biomass and physiological regulation [36]. Ma et al. [37] found that the combined application of COS and arbuscular mycorrhizal fungi enhanced the root vitality of tobacco, leaf chlorophyll content, antioxidant enzyme activity, disease resistance, caprylic acid content, and the abundance of Bacillus in rhizosphere soil compared with the control. This is consistent with the results of this study. In this study, the impacts of CAS on root length and dry weight were significant. Additionally, root length and root dry weight were negatively correlated with the fungus Basidiomycota, and positively correlated with the bacteria Patescibacteria and Actinobacteriota. These correlative findings suggest that CAS has a greater impact on the directly contacted root system than COS-PA and A1, and also directly promotes root growth and nutrient absorption. The positive correlation between root traits and the abundance of Patescibacteria and Actinobacteriota implies a potential link between these bacterial groups and cotton salt tolerance. The reason is that ACC deaminase-producing rhizobacteria and ACC deaminase can reduce ethylene levels in plants exposed to biotic and abiotic stresses to promote plant growth, particularly promoting the growth of plant roots under abiotic stress [38, 39]. A large number of studies have also confirmed that rhizobacteria that produce ACC deaminase have a positive impact on plant growth. For example, under salt stress, inoculation with ACC deaminase-producing Mesorhizobium ciceri G-55 could stimulate the nodulation of the legume chickpea (Cicer arietinum L.) and increase plant biomass compared with the control [40]. Overexpression of ACC deaminase in the rhizobium Sinorhizobium meliloti of alfalfa could alleviate the negative effects of copper on nodulation, root ethylene production, and shoot biomass accumulation, and enhance the expression of antioxidant genes [41]. In addition, in this study, the CAS treatment exhibited the highest numerical values for cotton biomass, POD, and CAT activities among the COS-PA, A1, and CAS treatments (Figs. 2e and f and 3b). It’s worth noting that the strain used in this study (A1) not only has ACC deaminase activity but also has potential growth-promoting traits like IAA production, siderophore secretion, phosphate solubilization, and nitrogen fixation [42]. A highly effective PGPR often possesses several of these traits at the same time, and they work together to help plants cope with abiotic stresses, usually in ways that can’t be completely separated functionally. Since this study focuses on the ACC deaminase-producing bacteria, the discussion mainly centers on ACC deaminase activity. However, this study acknowledges that the observed growth-promoting effects can’t be strictly distinguished as solely due to ACC deaminase versus other PGPR traits. Therefore, the growth-promoting effect of CAS is hard to attribute to a single mechanism and is more likely the result of multiple traits working together. This indicates that the combined application of COS-PA and A1, i.e., CAS, can better exert the antagonistic activity of rhizobacteria and alleviate the negative impact of salt stress on cotton growth and development. Studies have shown that the combined application of arbuscular mycorrhizal fungi and COS has a better effect on reducing ROS accumulation than single applications, significantly reducing damage to the plant cell membrane caused by external stress [37]. Under external stress, the plant cell membrane plays an important role in polar transport and signal transduction. Maintaining cell membrane integrity is of great significance to enhancing plant antioxidant properties and stress tolerance [43].

As a plant stimulant, COS can promote the synthesis of resistance-related enzymes and secondary metabolites in plants and enhance the ability to scavenge reactive oxygen species and resist abiotic stress [44]. In addition, COS influences the expression of root genes related to strainoctone metabolism and transport, and triggers the synthesis of secondary metabolites isoflavones and terpenes and the expression of lipid and mineral nutrient transporter proteins. This shows that COS promotes the formation of symbiosis-oriented metabolic and physiological environments, which supports the research hypothesis of this study, that COS can be used as a medium for loading rhizobacteria; It stimulates fungal symbiosis and microbial activity in salinized soils. When the rhizobacteria that produce ACC-deaminase cannot completely offset the negative effects of salt stress on cotton growth, using COS loaded with rhizobacteria that produce ACC-deaminase is necessary [45–47].

The application of COS-PA and CAS promoted cotton growth by regulating soil microbial communities

The composition of soil bacterial and fungal communities is impacted by multiple factors, among which exogenous substance application can directly alter soil pH [37, 48]. Bacteria and fungi have specific pH tolerance ranges. The pH range 4–7 is most suitable for bacterial survival, while soil fungi generally have a broader pH tolerance range [48]. In this study, the application of COS and A1 significantly impacted the composition of both soil bacterial and fungal communities (Fig. 5a, b). However, their impact on fungal community composition was greater, with a reduced soil fungal Chao1 index. This may be related to the exogenous amendment lowering soil pH (Table 1). This further indicates that soil fungi in this study are more sensitive to changes in soil pH. At the same time, COS has certain antimicrobial activity, particularly inhibiting some pathogenic fungi [49]. Notably, ACC deaminase-producing rhizobacteria may indirectly or directly inhibit the reproduction and survival of some fungi by competing with fungi for nutrients or niches [50].

Moreover, in this study, the relative abundances of the dominant bacteria Patescibacteria and Actinobacteriota were significantly positively correlated with cotton growth (biomass and chlorophyll synthesis) (Figs. 6 and 7c and d). The relative abundance of Patescibacteria and Actinobacteriota was positively correlated with cotton growth parameters, suggesting a potential association that warrants further functional investigation [51, 52]. It was also found that under salt stress, the application of exogenous chitosan oligosaccharides and ACC deaminase-producing rhizobacteria was associated with increased abundance of Actinobacteriota compared with the control (Fig. 5a). This suggests a possible association between the application of exogenous substances and increased Actinobacteriota abundance, which may be related to cotton growth promotion under the conditions tested. Yin et al. [53] found that Actinobacteriota was a species whose abundance was correlated with tea plant biomass. This is consistent with the results of this study. Therefore, it can be hypothesized that ACC deaminase-producing rhizobacteria and Actinobacteriota may act synergistically in promoting plant growth under salt stress, although further functional studies are needed to confirm this [53]. This synergy may occur because ACC deaminase-producing rhizobacteria can reduce ethylene levels by breaking down ACC, alleviating stress-induced inhibition on plants [54]. Additionally, COS may activate the plant’s defense response, inducing antioxidant enzyme activity, and provide a stable microenvironment for bacteria to enhance their ACC deaminase production capacity [55]. Previous studies have also shown that Actinobacteriota prefer nutrient-rich environments. Carbon-rich environments can stimulate their rapid growth, resulting in their enrichment in plants and rhizosphere soils. Chitosan oligosaccharide, a carbon source used by microorganisms, can enhance soil organic carbon availability. This further coincided with increased metabolic activity and population expansion of carbon-preference microorganisms such as actinomycetes, providing favorable conditions for the survival and reproduction of Actinobacteriota [30]. Based on these correlational findings, it is possible that COS contributes to cotton root development and photosynthesis, potentially through modifying soil nutrient availability and the microbial environment. However, direct mechanistic evidence is required to establish causation. In this study, both COS and ACC deaminase-producing rhizobacteria reduced the relative abundance of the soil fungus Basidiomycota compared with the control. This may be because (1) the smaller size of bacterial cells relative to fungal cells leads to a larger surface area-to-volume ratio, allowing bacteria to contact and absorb smaller dissolved organic compounds more quickly [56]. (2) The growth rate of bacteria is much faster than that of fungi, allowing them to occupy organic resources faster [56]. (3) The cell membrane structure of bacteria is simpler than that of fungi, which helps bacteria to directly absorb small compounds from soil solutions, leading to a decrease in the relative abundance of fungi [57]. Under salt stress, the application of COS-PA and ACC deaminase-producing rhizobacteria increased the relative abundance of the soil bacterium Acidobacteriota compared with the control, and there was a significant correlation between Acidobacteriota abundance and soil properties. The abundance of Acidobacteriota was positively correlated with soil enzyme activities, raising the hypothesis that COS may influence this group and that such influence might be associated with soil quality improvement [58]. At the same time, ACC deaminase-producing rhizobacteria can enhance soil microbial activity and plant root vitality by increasing soil nutrients, which in turn enhances soil enzyme activity under salt stress, thus alleviating abiotic stress in plants [59–61]. This is confirmed by the results of microbial function prediction in this study. To address the above limitations, future research will introduce metagenomics and metatranscriptomics techniques to directly verify the functions and activities of core microbes at the levels of gene composition and expression. This will fully clarify the molecular mechanisms by which amendments regulate salinized soil and enhance cotton’s salt tolerance and growth.

Limitations and prospects

Although this study obtained some new findings, there are some limitations. (1) The pot experiments were conducted under controlled greenhouse conditions, which cannot fully reflect the complex field environment, necessitating field validations. (2) Measurements were only taken during the budding stage. Therefore, the dynamics of CAS’s effects throughout the entire growing season are unknown. (3) The impacts on yield and fiber quality were not assessed. In the future, field trials will be carried out with multiple sampling points throughout the growing season (seedling, budding, flowering, boll-forming, and boll-opening stages) to verify the dynamic effects of COS and to optimize its concentration, application frequency, and timing. This will provide a foundation for further soil quality and crop improvements. In addition, the combined effects of COS with practices such as straw incorporation, drip irrigation, and fertilizer reduction are also worth further exploration.

Besides, this study still has some limitations in explaining the mechanisms. Mainly, this study didn’t set up a carrier-only control (COS-PA without bacteria) or a heat-killed bacterial control, and also didn’t directly track the strain (A1) after application to see whether it colonized and survived. So, it’s hard to precisely say whether the growth-promoting effects come from chitosan oligosaccharide or from ACC deaminase. Also, the microbial function predictions were just computational inferences and need to be verified with metatranscriptomics or metagenomics. Actually, the strain used (A1) has multiple PGPR traits, including IAA production, siderophore secretion, phosphate solubilization, and nitrogen fixation (Etesami et al., 2018). So, its growth-promoting effect is likely due to a combination of these traits and the carrier effect. Future research will involve creating ACC deaminase knockout mutants and using metagenomics and fluorescent tagging to separate the contributions of each mechanism.

Concluson

The application of chito-oligosaccharide and ACC deaminase-producing rhizobacteria could lower soil pH and increase soil enzyme activity, cotton root length, root biomass, and leaf POD and CAT activity compared with the control. These finally reduce physiological damage and growth inhibition caused by salt stress. The combined application of chito-oligosaccharide and ACC deaminase-producing rhizobacteria promoted cotton growth more significantly than single applications. The application of chito-oligosaccharide and ACC deaminase-producing rhizobacteria could induce changes in the rhizosphere microbiota to resist external stress. They could regulate the relative abundance of the soil bacterium Acidobacteriota and the soil fungi Basidiomycota, Ascomycota, and Glomeromycota, promoting cotton growth and enhancing soil quality. Compared with traditional inert carriers such as sodium alginate, the COS used in this study not only serves as a carrier for loading strains but also possesses growth-promoting effects. It can serve as a carbon source to promote rhizosphere bacterial colonization, and its degradation products can induce plant defense responses [44]. In contrast, inert carriers only play a physical encapsulation role and lack a growth-promoting effect. Therefore, CAS, through the carrier role and growth-promoting effect of chitosan oligosaccharide, protects the microbial strains in the agent while synergistically promoting cotton growth and soil improvement. This is its key advantage over traditional carriers. Notably, the strains used in this study not only have ACC deaminase activity but also possess multiple growth-promoting traits and nitrogen-fixing ability. Therefore, the effects of CAS result from the combined action of multiple growth-promoting mechanisms.

Authors’ contributions

Mengjie An: Conceptualization, Methodology, Formal analysis, Software, Visualization, Investigation, Funding acquisition, Writing-original draft, Writing-review & editing. Kaidi Ren and Yanfeng Jia: Formal analysis, Software, Visualization, Investigation, Methodology, Writing-original draft. Linlin Zhang: Visualization, Investigation, Methodology, Writing-original draft. Qianqian Wang: Visualization, Investigation, Methodology, Writing-original draft. Yongqi Zhu: Investigation, Methodology, Writing-review & editing, Supervision. Yonghong Fan: Resources, Funding acquisition, Supervision, Conceptualization, Writing - review & editing.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This work was supported by the National Key Research and Development Program of China (2024YFD1200200), the Key R&D Program Project of Xinjiang Uygur Autonomous Region of China (2022B02053-2), and the Natural Science Foundation of Xinjiang Uygur Autonomous Region (2023D01C16).

Data availability

The data generated in this study have been deposited and are available in the public database NCBI (the National Center for Biotechnology Information). All raw data from 16 S rRNA and ITS rRNA gene sequence analysis can be found in the Sequence Read Archive (SRA) (accession number: PRJNA1431344). Other data generated or analyzed in this study are included in the main text and the supplemental materials.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.