Synergistic Effects of Deep Rotary Tillage and Microbial Decomposition Agents on Straw Decomposition, Soil Nutrient Dynamics, and Microbial Communities in Rice Systems

Abstract

This study evaluated the synergistic effects of microbial decomposition agents and deep rotary tillage on rice straw decomposition, soil nutrient dynamics, and microbial communities in paddy fields of southern China. A two-factor randomized block experiment was conducted, with straw decomposition dynamics modeled using a modified Olson decay model, and microbial communities were assessed via high-throughput sequencing and network analysis. The combined treatment significantly increased the decomposition rate constant, reduced the time for 50% decomposition to 81 days, and enhanced soil nutrient availability, especially total nitrogen, phosphorus, and potassium. Microbial richness, diversity, and network complexity were also improved. Structural equation modeling indicated that nutrient availability, rather than microbial α-diversity, was the main driver of decomposition processes. These findings suggest that integrating microbial agents with deep tillage offers an effective strategy for optimizing straw return, improving soil fertility, and enhancing microbial functional resilience in rice systems.

1. Introduction

With the growing consensus on sustainable agricultural development, straw return has been increasingly adopted—particularly in rice cultivation regions of China—as an important practice to enhance soil fertility and promote nutrient cycling [1,2]. Rice straw is rich in essential nutrients such as nitrogen, phosphorus, and potassium. Its efficient decomposition and nutrient release are crucial for improving soil physicochemical properties and supporting the growth of subsequent crops [3]. However, the slow decomposition rate and low nutrient release efficiency of straw significantly limit the practical effectiveness of straw return techniques [2,4]. In humid regions of southern China, the practice of direct straw incorporation can lead to adverse effects on crop seedling growth and an increased risk of pest and disease outbreaks. This phenomenon is influenced by several factors, including pathogen dynamics, soil health, and environmental conditions [5,6].

While globally there is a growing consensus that reduced- or no-tillage practices can improve soil structure and biodiversity—especially through enhanced soil aggregate stability, moisture retention, and increased soil fauna [7,8]—their effects are context-dependent. In regions with compacted soils and abundant rainfall, such as southern China’s rice-growing areas, these practices may result in slow straw decomposition and hinder seedling establishment. Therefore, site-specific and adaptive tillage strategies are necessary. Deep rotary tillage, under such regional conditions, can improve soil aeration, break compacted layers, and enhance straw–soil–microbe contact, thereby accelerating straw decomposition and nutrient cycling. In field studies conducted in Henan Province, China, deep tillage was shown to disrupt compacted layers, improve soil structure, and enhance oxygen availability, which promoted root growth and stimulated aerobic microbial activity essential for organic matter decomposition [9]. Incorporating straw during deep tillage enhances microbial abundance, enzymatic activities, and colonization on straw surfaces, accelerating fragmentation and nutrient release [10]. Moreover, effective straw burial through rotary tillage facilitates the integration of organic matter into the soil, supporting microbial diversity and metabolic functions over time [11].

The introduction of microbial decomposition agents composed of efficient microbial communities significantly shortens the decomposition cycle of agricultural residues, promoting rapid nutrient recycling and enhancing soil health [12]. These agents typically include diverse bacteria and fungi, such as Bacillus subtilis, Trichoderma viride, and Aspergillus niger, which synergistically enhance organic matter degradation and nutrient release [13,14]. Additionally, microbial agents offer environmental benefits by mitigating pollution from agricultural waste burning and improving soil fertility without adverse effects [15,16].

Although numerous studies have investigated the effects of straw incorporation on soil microbial communities and nutrient cycling, many have focused on surface or conventional tillage systems with limited mechanistic analysis. For instance, long-term straw return with potassium fertilization significantly improved rice and oilseed rape yields by 23.0% and 20.5%, respectively, while enhancing the abundance of copiotrophic bacteria such as Firmicutes and Proteobacteria [17]. Inversion tillage combined with straw return, as studied in northeastern China, also enhanced soil multi-nutrient cycling and reshaped microbial network structures by increasing modularity and path length [18]. Additionally, straw incorporation across a latitudinal gradient significantly altered microbial co-occurrence patterns and increased stochasticity in community assembly [19]. Despite these findings, few studies have explored how deep rotary tillage combined with microbial decomposition agents influences straw decomposition and microbial communities in humid, intensively cultivated rice-growing regions, such as those found in southern China, Southeast Asia, and parts of South America and Africa. To address this region-specific knowledge gap, the present study aimed to determine (i) the decomposition dynamics of rice straw and residue decay modeling using a modified Olson index model; (ii) changes in soil microbial alpha diversity, community structure, and microbial interaction patterns under different straw return approaches; and (iii) the relationships among soil physicochemical properties, microbial community composition, and straw decomposition processes.

The aim of this research is to model the microbial and environmental mechanisms hypothesized to regulate straw decomposition and nutrient release under deep rotary tillage combined with decomposition agents. This work provides a theoretical basis for optimizing straw resource utilization, enhancing soil health, and promoting sustainable agricultural development.

The aim of this research is to model putative microbial and environmental regulatory mechanisms governing straw decomposition and nutrient release under deep rotary tillage combined with decomposition agents.

2. Materials and Methods

2.1. Experimental Site

The field experiment was conducted from 2022 to 2023 at the Yunyuan Experimental Base and the Life Sciences Building of Hunan Agricultural University, Changsha, Hunan Province, China (112°59′ E, 28°12′ N). The experimental site is located in a subtropical monsoon climate zone, with an annual frost-free period of 255–293 days, an average annual temperature of 16.8–17.3 °C, an annual sunshine duration of 1550–1750 h, and annual precipitation ranging from 1358.6 to 1552.5 mm.

The basic physicochemical properties of the 0–15 cm soil layer were determined prior to the experiment. The results were as follows: bulk density 1.21 g·cm⁻³, organic matter 20.32 g·kg⁻¹, total nitrogen (TN) 1.13 g·kg⁻¹, total phosphorus (TP) 1.07 g·kg⁻¹, total potassium (TK) 13.83 g·kg⁻¹, alkali-hydrolyzable nitrogen (AN) 64.85 mg·kg⁻¹, available phosphorus (AP) 7.33 mg·kg⁻¹, available potassium (AK) 139.13 mg·kg⁻¹, and pH 6.81. The measurements were conducted following standard procedures described by Bao (2000) [20].

The experimental field was under a tobacco–rice rotation system. The preceding late rice cultivar was ‘Liangyou 336’, sown in mid-May 2022 and harvested around October 2022. The rice straw burial decomposition experiment commenced in December 2022. Meteorological data, including daily temperature and precipitation, were recorded during the experimental period from 1 December 2022 to 30 April 2023, as shown in Figure 1. These data were obtained from the China Meteorological Data Service Center (http://data.cma.cn, accessed on 28 June 2024), which provides official weather records for the region near the experimental site.

2.2. Experimental Design

A two-factor randomized block design was employed to evaluate the effects of microbial decomposition agent application (M) and tillage method (D) on straw decomposition and soil properties. Factor levels were as follows: M0: no microbial agent applied; M1: microbial decomposition agent applied; D0: conventional rotary tillage; D1: deep rotary tillage. Thus, four treatments were established: M0D0 (no microbial agent + conventional tillage), M1D0 (microbial agent + conventional tillage), M0D1 (no microbial agent + deep tillage), and M1D1 (microbial agent + deep tillage), each replicated three times. The area of each plot was 50 m² (4 m × 15 m).

The microbial decomposition agent used in this study was the Straw-Specific Decomposition Agent (product information available at http://www.wobaosw.com/shows/10/35.html, accessed on 28 June 2024), provided by Henan Wobao Biotechnology Co., Ltd. (Hebi, China). The product was manufactured on 15 October 2022 and stored in a cool, dry environment at room temperature (15–25 °C). It was applied within 12 months after production. The formulation consisted of a mixed microbial community with a viable count ≥ 10¹⁰ CFU g⁻¹, including Bacillus spp., filamentous fungi, and lactic acid bacteria. Specific proportions of these microbial components were not disclosed by the manufacturer. The powdered agent was evenly applied to the rice straw prior to mechanical incorporation into the soil through rotary tillage.

Conventional rotary tillage was performed using a 1GZ-230B rotary tiller (Hunan Nongyou Agricultural Machinery Group Co., Ltd., Changsha, China) at a depth of 20–25 cm, while deep rotary tillage was carried out with a 1S-100 deep tillage machine equipped with a spiral auger blade (Hunan Tianye Modern Intelligent Equipment Co., Ltd., Changsha, China) at a depth of 30–35 cm.

For the straw decomposition assay, rice straw was collected from the experimental field, oven-dried at 40 °C, and cut into 2–3 cm segments. Approximately 9 g (oven-dried basis) of straw was placed into nylon mesh bags (10 cm × 15 cm, 0.15 mm mesh size). The application rate of the decomposition agent was calculated based on the local straw yield (400 kg·ha⁻¹) at 2 kg ha⁻¹. The decomposition agent, straw, and an equal mass of field soil were thoroughly mixed, packed into the mesh bags, and sealed.

The nylon bags were evenly buried at a depth of 10–15 cm in each plot following a five-point sampling layout. To allow for destructive sampling at eight time points (0, 10, 20, 30, 60, 90, 120, and 150 days after burial) and to prevent sample loss due to animal interference or environmental disturbance, ten nylon bags were placed in each of the three replicate plots per treatment, resulting in 30 bags per treatment and 120 bags in total across all four treatments. The burial date was 1 December 2022. The strings of the bags were left exposed above the soil surface to facilitate periodic retrieval. No crops were grown during the experimental period.

2.3. Sample Collection and Determination

Rice straw residues were retrieved at 0, 10, 20, 30, 60, 90, 120, and 150 days after burial. The collected straw samples were gently washed with distilled water, oven-dried at 60 °C to a constant weight, and the decomposition rate was determined using the gravimetric weight loss method.

After 150 days of straw return, soil samples were collected from the 0–20 cm layer using a soil auger following the five-point sampling method, with five replicates per treatment. The collected soil samples were thoroughly homogenized and divided into two subsamples. One subsample was air-dried at room temperature, passed through a 2 mm sieve after removal of visible stones and debris, and used for soil nutrient analysis. The other subsample was placed into sterile tubes, transported on ice to the laboratory, and immediately stored at −80 °C for subsequent metagenomic sequencing.

Soil physicochemical properties were determined following standard procedures. Soil pH was measured using a pH meter with a soil-to-water ratio of 1:2.5 (w/v) [20]. Soil organic matter (SOM) content was determined using the potassium dichromate oxidation method with external heating. Total nitrogen (TN) was measured using the Kjeldahl digestion method [21]. Total phosphorus (TP) was determined by the colorimetric molybdenum antimony blue method after acid digestion [21]. Total potassium (TK) was measured by flame photometry following acid digestion [20]. Available phosphorus (AP) was extracted with 0.5 mol·L⁻¹ sodium bicarbonate (NaHCO₃) solution and measured using the colorimetric molybdenum antimony blue method. Available potassium (AK) was extracted with 1 mol·L⁻¹ ammonium acetate (NH₄OAc) solution and measured by flame photometry. Alkali-hydrolyzable nitrogen (AN) was determined by the alkali hydrolysis diffusion method [20].

2.4. Soil Microbial DNA Extraction and Sequencing

Soil DNA was extracted using the E.Z.N.A.^® Soil DNA Kit (Omega Bio-Tek, Norcross, GA, USA) following the manufacturer’s instructions. DNA quality was assessed by agarose gel electrophoresis (2% agarose), and DNA concentration was quantified using a UV spectrophotometer.

The bacterial 16S rRNA gene V3–V4 regions were amplified using the primer pair 341F (5′-CCTACGGGNGGCWGCAG-3′) and 805R (5′-GACTACHVGGGTATCTAATCC-3′) [22]. Fungal internal transcribed spacer 2 (ITS2) regions were amplified using the primers ITS1FI2 (5′-GTGARTCATCGAATCTTTG-3′) and ITS2 (5′-TCCTCCGCTTATTGATATGC-3′) [23].

PCR products were verified via 2% agarose gel electrophoresis and purified using the AxyPrep PCR Cleanup Kit (Axygen Biosciences, Union City, CA, USA). The purified amplicons were quantified using a Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) on a Qubit 2.0 fluorometer, and only libraries with concentrations >2 nM were used for sequencing.

Qualified libraries were diluted to the appropriate concentration, pooled in equal molar ratios according to the target sequencing output, and denatured with NaOH to produce single-stranded DNA templates. Sequencing was performed on an Illumina NovaSeq platform (Illumina Inc., San Diego, CA, USA) with a paired-end 2 × 250 bp strategy by LC-Bio Technology Co., Ltd. (Hangzhou, China).

Raw sequencing reads were assigned to each sample based on unique barcodes and trimmed to remove barcode and primer sequences. Paired-end reads were merged using FLASH (v1.2.11). Quality filtering of the raw reads was conducted using fqtrim (v0.94) under specific parameters to obtain high-quality clean reads. Chimeric sequences were detected and removed using Vsearch (v2.3.4). Denoising and feature table generation were performed using DADA2.

Alpha diversity metrics, including observed OTU richness, the Shannon diversity index, and the Chao1 richness estimator, were calculated after rarefying the feature table to an equal sequencing depth of 30,000 reads per sample to ensure comparability across samples. Rarefaction and diversity calculations were performed using the QIIME2 (version 2022.2; QIIME 2 Development Team, Northern Arizona University, Flagstaff, AZ, USA) [24] pipeline. Principal component analysis (PCA) was performed based on normalized taxonomic abundance data to explore the overall variation in microbial community composition across samples. Taxonomic classification of features was assigned based on the SILVA database (release 132; Max Planck Institute for Marine Microbiology and Jacobs University, Bremen, Germany) [25] for 16S rRNA genes and the UNITE database (release 8.3, 2021-07-01; UNITE Community, Department of Forest Mycology and Plant Pathology, Swedish University of Agricultural Sciences, Uppsala, Sweden) [26] for ITS sequences. Feature abundances were normalized by the relative abundance within each sample. Amplicon sequence variants (ASVs) generated using the DADA2 pipeline were used for co-occurrence network analysis.

2.5. Data Analysis

Data processing was conducted using Microsoft Excel 2015 (Microsoft Corp., Redmond, WA, USA). Straw decomposition rate and cumulative decomposition rate were calculated following the methods described by He et al. (2022) [27], using the following formulas:

$$\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{w} \mathrm{d}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}(\mathrm{g}·d^{-1})={\displaystyle \frac{M_n-M_m}{\mathrm{n}-\mathrm{m}}}$$

$$\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{w} \mathrm{c}\mathrm{u}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{d}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\left(\mathrm{\%}\right)={\displaystyle \frac{M_0-M_t}{M_0}}\times 100\mathrm{\%}$$

where $M_n$ is the straw mass (g) on day n, $M_m$ is the straw mass (g) on day m, $M_0$ is the initial straw mass (g), and $M_t$ is the straw mass (g) after t days of decomposition.

To model the straw decomposition dynamics, a modified Olson exponential decay model was fitted independently to each treatment group’s data to estimate treatment-specific decomposition rates, based on the method of Olson (1963) [28]. The model is expressed as follows:

$$y=a\times e^{-kt}={\displaystyle \frac{M_t}{M_0}}$$

where $y$ is the remaining mass fraction, $k$ is the decomposition rate constant (d⁻¹), $a$ is the adjustment coefficient representing the initial residue proportion, and $t$ is the time (days). Non-linear regression was performed using the least squares optimization method in SPSS 19.0 software (IBM Corp., Armonk, NY, USA), and model fit was evaluated using the root mean square error (RMSE). The models were fitted based on the mean values of each treatment across three replicates, rather than on individual plots, to minimize within-treatment variability.

Based on the fitted $k$ values, the times required for 50% and 95% decomposition ($T_{500}$ and $T_{95}$, respectively) were calculated using the following equations:

$$T_{50}={\displaystyle \frac{\ln 0.5-\ln a}{(-\mathrm{k})}}$$

$$T_{95}={\displaystyle \frac{\ln 0.05-\ln a}{(-\mathrm{k})}}$$

Statistical analyses were performed using SPSS 19.0 software. A two-way analysis of variance (ANOVA) was conducted to evaluate the main effects of microbial agents, deep tillage, and their interaction on soil nutrient parameters. Where applicable, block effects were included as a random factor in the ANOVA model to account for spatial variability across replicates. Significant differences among treatment means were determined using Duncan’s multiple range test at the p < 0.05 level.

Bar plots, Venn diagrams, and redundancy analysis (RDA) were performed and visualized using OriginPro software (Version 2024, OriginLab Corporation, Northampton, MA, USA). For the Venn diagram analysis, OTUs clustered at 97% similarity were used to illustrate the overlap and uniqueness of microbial taxa among treatments. Redundancy analysis (RDA) was performed in Canoco 5.0 (Microcomputer Power, Ithaca, NY, USA) [29] to explore the effects of microbial agent application (M) and deep tillage (D) (both coded as binary variables: 0 = absence; 1 = presence) on soil physicochemical properties, including total nitrogen (TN), total phosphorus (TP), total potassium (TK), available nitrogen (AN), available phosphorus (AP), available potassium (AK), soil organic matter (SOM), and pH. Boxplots, principal component analysis (PCA), and correlation heatmaps were generated using the Chiplot online platform (https://www.chiplot.online/, accessed on 28 June 2024). Principal component analysis (PCA) was used to visualize the overall variation in bacterial community composition at the phylum level across different treatments. Correlation heatmaps were generated based on Pearson correlation analyses to explore the associations between bacterial and fungal community composition and soil physicochemical properties, including nutrient factors (TN, TP, TK, AN, AP, and AK), soil organic matter (SOM), and pH.

Co-occurrence network analysis was performed to explore microbial interactions within soil bacterial and fungal communities. Amplicon sequence variants (ASVs) with a detection frequency in at least 20% of samples and ranking among the top 100 in relative abundance were selected for analysis. Pairwise Spearman correlation coefficients between ASVs were computed using the psych package in R software (version 4.3.2, R Core Team, Vienna, Austria). Correlations with |r| > 0.6 and p < 0.01 were retained to construct the adjacency matrices. The resulting co-occurrence networks were visualized using Chiplot (https://www.chiplot.online/, accessed on 28 June 2024). In the networks, nodes represent ASVs, and edges represent statistically significant correlations between them.

Structural equation modeling (SEM) was carried out using AMOS software (Version 29.0, IBM Corp., Armonk, NY, USA) to explore the direct and indirect relationships among straw decomposition rate, soil nutrient variables, and microbial alpha diversity indices. Given the limited sample size (n = 20), a simplified path analysis using only observed variables was employed rather than a full latent-variable SEM.

Model fit was evaluated using chi-square (χ²), root mean square error of approximation (RMSEA), and the goodness-of-fit index (GFI). Path coefficients were estimated using maximum likelihood estimation (MLE), and standardized regression weights were used to indicate the strength and significance of the relationships. All interpretations were made with caution due to the sample size limitations.

3. Results

3.1. Straw Decomposition Dynamics

The dynamic changes in straw decomposition rate and cumulative decomposition rate over a 150-day period are presented in Figure 2. The straw decomposition rate (Figure 2A) exhibited a rapid increase during the initial 10 days across all treatments, reaching a peak at day 10, followed by a decline at day 20 and a slight rebound at day 30. After 60 days, the decomposition rate declined significantly and stabilized at a relatively low level. Although M1D1 and M0D1 showed marginally higher rates at certain time points, the differences were not consistent throughout the decomposition period. By day 150, cumulative decomposition rates differed among treatments, with M1D1 exhibiting the highest value, followed by M0D1, M1D0, and M0D0 (Figure 2B). Two-way ANOVA indicated that at 150 days, both microbial agent application (F = 8.23, p = 0.007) and deep tillage (F = 10.65, p = 0.003) significantly increased the cumulative straw decomposition rate, while their interaction effect was not significant (F = 2.13, p > 0.05).

The straw decomposition process was further described using a first-order kinetic model (

Table 1. Rice straw residue rate index model.

Treatment	Fitted Equation	RMSE	k	T50 (d)	T95 (d)
M0D0	y = 0.9447e−0.0058t	0.0385	0.0058	109	506
M0D1	y = 0.96e−0.0067t	0.0378	0.0067	97	438
M1D0	y = 0.9885e−0.0077t	0.0204	0.0077	88	387
M1D1	y = 0.9665e−0.0081t	0.0284	0.0081	81	364

y: straw residue rate; t: decomposition time (days); RMSE: root mean square error, calculated from residuals between observed and predicted values under each treatment; k: decomposition rate constant; T₅₀ and T₉₅: time required for 50% and 95% straw decomposition, respectively (days).

). All treatments exhibited high model fit. The decomposition rate constant (k) was highest in the M1D1 treatment (k = 0.0081 d⁻¹), followed by M1D0 (0.0077 d⁻¹), M0D1 (0.0067 d⁻¹), and M0D0 (0.0058 d⁻¹) (Table 1). Correspondingly, the time required for 50% decomposition (T₅₀) decreased from 109 days in the control to 81 days in M1D1. Similarly, the time required for 95% decomposition (T₉₅) was shortest under M1D1 (364 days), compared to 506, 438, and 387 days for M0D0, M0D1, and M1D0, respectively, indicating accelerated decomposition under combined microbial agent application and deep tillage (Table 1).

Overall, these results demonstrate that both microbial agent application and deep tillage independently promoted straw degradation, with the combined treatment (M1D1) showing the most pronounced effect.

3.2. Soil Fertility Under Different Treatments

Promoting straw returning significantly improved soil pH, SOM, TN, TP, TK, AN, AP, and AK contents, with the M1D1 treatment showing the highest overall values (

Table 2. Effects of microbial decomposition agent (M), tillage method (D), and their interaction (M × D) on soil physicochemical properties.

Treatment	pH	SOM(g∙kg−1)	TN(g∙kg−1)	TP(g∙kg−1)	TK(g∙kg−1)	AN (mg∙kg−1)	AP (mg∙kg−1)	AK (mg∙kg−1)
M0D0	5.90 ± 0.02 ab	29.60 ± 0.35 b	1.22 ± 0.01 d	1.40 ± 0.01 c	4.40 ± 0.30 c	134.32 ± 2.81 c	9.14 ± 0.23 c	228.65 ± 7.29 c
M0D1	5.86 ± 0.05 b	30.33 ± 0.45 ab	1.25 ± 0.01 c	1.42 ± 0.01 bc	4.97 ± 0.08 b	137.41 ± 1.34 c	9.76 ± 0.11 b	234.52 ± 8.79 bc
M1D0	6.06 ± 0.13 a	30.88 ± 0.67 ab	1.40 ± 0.01 b	1.44 ± 0.01 b	5.38 ± 0.07 a	144.65 ± 1.87 b	9.90 ± 0.06 b	247.80 ± 2.69 ab
M1D1	5.97 ± 0.04 ab	31.43 ± 0.88 a	1.37 ± 0.01 a	1.52 ± 0.03 a	5.58 ± 0.08 a	150.39 ± 0.94 a	10.28 ± 0.08 a	255.65 ± 6.74 a
M	*	*	**	**	**	ns	**	**
D	ns	ns	ns	**	*	ns	**	ns
M × D	ns	ns	**	*	ns	ns	ns	ns

Different lowercase letters indicate significant differences among treatments at the p < 0.05 level according to multiple comparisons. ** and * indicate significant differences at the 0.01 and 0.05 levels, respectively, based on two-way ANOVA; “ns” indicates no significant difference.

). Compared to M0D0, the M1D0 treatment increased pH to 6.06 ± 0.13, which alleviated soil acidity. It also improved SOM to 30.88 ± 0.67 g·kg⁻¹, TN to 1.40 ± 0.01 g·kg⁻¹, and TP to 1.44 ± 0.01 g·kg⁻¹. M1D1 further enhanced TN to 1.37 ± 0.01 g·kg⁻¹ and TP to 1.52 ± 0.03 g·kg⁻¹, and significantly elevated TK to 5.58 ± 0.08 g·kg⁻¹ (p < 0.05) (Table 2). In terms of available nutrients, AN, AP, and AK under M1D1 increased to 150.39 ± 0.94 mg·kg⁻¹, 10.28 ± 0.08 mg·kg⁻¹, and 255.65 ± 6.74 mg·kg⁻¹, respectively, exhibiting a significant advantage (p < 0.05) (Table 2). Analysis of variance indicated that the microbial agent exerted a significant main effect on most indicators, while deep tillage treatment notably affected TP, TK, and AP, and a significant interaction between the microbial agent and deep tillage was observed for TN and TP. Overall, the application of a microbial agent played a dominant role in improving soil physicochemical properties, while deep tillage exerted a regulatory effect on certain nutrient parameters.

3.3. Soil Microbial Community Diversity Indices Under Different Treatments

The diversity indices of both bacterial and fungal communities varied significantly across treatments involving microbial agent application and deep tillage (Figure 3). For bacterial communities, the combined treatment of microbial agent application with deep tillage (M1D1) exhibited the highest OTU richness, with an average of 2613.75 ± 217.58, compared to the lowest observed in the control without microbial agent and without deep tillage (M0D0) at 1680.75 ± 342.88 (Figure 3A). The bacterial Shannon diversity index remained high across treatments, all above 9.30 ± 0.60, with M1D1 reaching the maximum at 10.53 ± 0.09, indicating enhanced bacterial diversity under combined management (Figure 3B). The Chao1 richness estimator followed a similar trend, with the highest value observed in M1D1 and the lowest in M0D0 (Figure 3C). For fungal communities, the highest OTU richness was also recorded under M1D1, while the lowest was found in M0D0 (Figure 3A). The Shannon index for fungi peaked at 6.54 ± 0.15 in M1D1, whereas M0D0 recorded the lowest at 5.43 ± 1.33 (Figure 3B). Fungal Chao1 richness estimates were highest in M1D1 and lowest in M0D0 (Figure 3C). Overall, the application of microbial agents, particularly when combined with deep tillage, significantly improved microbial richness (p < 0.05), diversity, and community evenness.

The relative abundance profiles of bacterial and fungal communities at the phylum level varied significantly among treatments (p < 0.05) (Figure 4A,B). For bacteria, Proteobacteria and Acidobacteriota were the dominant phyla across all groups, followed by Chloroflexi and Actinobacteriota. Compared to the control (M0D0), the relative abundance of Chloroflexi increased markedly in the M1D1 treatment, whereas Actinobacteriota decreased slightly. Additionally, the application of microbial agents (M1) appeared to enrich minor phyla such as Gemmatimonadota and Methylomirabilota, particularly under deep tillage (D1) conditions. In fungal communities, Ascomycota overwhelmingly dominated across all treatments, with Basidiomycota and Mortierellomycota also present at lower abundances (Figure 4B). Notably, the M1D1 treatment promoted the proliferation of Glomeromycota and Mucoromycota, suggesting that the combination of microbial agent application and deep tillage positively influenced specific beneficial fungal groups.

Principal component analysis (PCA) illustrated distinct clustering patterns among treatments for both bacterial and fungal communities (Figure 4C,D). For bacteria, the first two principal coordinates (PC1 and PC2) explained 66.12% and 15.44% of the total variation, respectively. Samples from the M1D1 group formed a separate cluster, clearly distinct from the other treatments, indicating substantial shifts in community structure due to the combined effects of M and D. Similarly, for fungi, PC1 and PC2 accounted for 67.45% and 22.8% of the variation, respectively. Samples treated with M0D1 (deep tillage alone) exhibited the most pronounced separation from M0D0 along the PC1 axis, which explained 67.45% of the variance. In contrast, the M1D1 treatment clustered more closely with M1D0 and M0D0, suggesting that microbial agent application had a relatively limited additional effect on fungal community composition when combined with deep tillage.

Venn diagram analyses revealed notable differences in the numbers of unique and shared operational taxonomic units (OTUs) among treatments (Figure 4E,F). For bacteria (Figure 4E), 532 OTUs were shared among all four groups. The M1D1 treatment harbored the highest number of unique OTUs, indicating that combined microbial agent application and deep tillage enhanced bacterial richness. For fungal communities (Figure 4F), 177 OTUs were common to all treatments, and M1D1 exhibited the highest number of unique fungal OTUs. These results further supported that the combination of microbial agent addition and deep tillage promoted the diversification of both bacterial and fungal communities.

3.4. Effects of Different Treatments on Soil Microbial Co-Occurrence Network Structure

Based on R analysis, ASVs with a detection rate of at least 20% and among the top 100 in abundance were selected. Spearman correlation coefficients were calculated, and correlations with p < 0.01 were used to construct the bacterial and fungal co-occurrence networks (Figure 5). Microbial agent application, especially when combined with deep rotary tillage (M1D1), substantially increased network complexity, as reflected by the higher numbers of nodes and links. For bacteria, the number of nodes and edges increased from 969 and 68,927 in M0D0 to 1738 and 157,350 in M1D1, respectively. In fungal networks, edge numbers increased from 42,240 in M0D0 to 50,485 in M1D1. These increases suggest that microbial inoculants enhanced microbial richness and intensified potential microbial interactions. Notably, the proportion of positive correlations decreased in both bacterial and fungal networks under M1 treatments. For bacteria, positive correlations declined from 63.64% in M0D1 to 53.66% in M1D0, while for fungi, they dropped from 97.55% in M0D0 to 88.57% in M1D1. This trend may reflect a shift from highly cooperative interactions toward more competitive or functionally differentiated relationships within microbial communities.

At the phylum level, Proteobacteria and Acidobacteriota were dominant bacterial taxa across treatments, with their node counts further increasing in M1D1, highlighting their central ecological roles under microbial amendment. Firmicutes, associated with organic matter degradation, also markedly increased in M1D1, while Chloroflexi showed a slight decline, potentially due to competition with introduced microbes. For fungi, Ascomycota remained the dominant group across all treatments, maintaining a strong network presence even under microbial amendment. Basidiomycota showed a relatively stable presence, whereas Zygomycota, often considered fast-growing opportunists, showed a notable increase in M1D1, suggesting enhanced niche expansion or colonization potential in response to microbial agents.

Overall, these results indicate that microbial agent application, particularly under deep tillage conditions, significantly enhanced the structural complexity of both bacterial and fungal networks, promoted the activity of key phyla involved in organic matter turnover and nutrient cycling, and reshaped the ecological relationships among microbial taxa within the community.

3.5. Mechanistic Relationships Among Straw Decomposition, Soil Nutrients, and Microbial Diversity

The correlation analysis revealed that soil physicochemical properties were significantly associated with both bacterial and fungal community structures (p < 0.05) (Figure 6A). Pearson’s correlation coefficients showed that total nitrogen (TN), total phosphorus (TP), total potassium (TK), available nitrogen (AN), available phosphorus (AP), and available potassium (AK) were positively correlated with microbial communities, whereas soil pH exhibited a negative correlation, particularly with fungal communities. Mantel tests further confirmed these associations, with TK and AN displaying the strongest correlations (p < 0.01) for both bacteria and fungi. Notably, bacterial communities showed broader and stronger positive correlations with soil nutrient variables compared to fungal communities.

Redundancy analysis (RDA) indicated that environmental factors explained a substantial proportion of the variation in microbial community composition (Figure 6B). The first two RDA axes explained 49.73% and 48.29% of the total variance, respectively. Among the soil factors, TP, AP, SOM (soil organic matter), TK, and AN were strongly aligned with microbial community shifts along the RDA1 axis. The application of deep tillage and microbial agents promoted increased nutrient availability, indicating that both treatments effectively modified the soil environment to facilitate microbial community restructuring.

Path analysis was performed to investigate the putative causal pathways linking straw decomposition rate, soil nutrients, and microbial α-diversity (Figure 6C). The model revealed that straw decomposition rate had a significant positive effect on TP, TN, AN, AP, and AK. Regarding microbial communities, AP and AK were significantly associated with higher bacterial α-diversity. In contrast, AN, AP, and AK had strong positive effects on fungal α-diversity, revealing nutrient-specific microbial responses. All direct paths from straw decomposition to microbial α-diversity (bacterial or fungal) were statistically non-significant (p > 0.05), indicating that microbial diversity shifts are primarily driven by changes in soil nutrients rather than decomposition per se. Similarly, neither bacterial nor fungal α-diversity had a significant direct feedback effect on straw decomposition rate. Overall, the model supports a nutrient-mediated mechanism in which straw decomposition indirectly influences microbial community composition by altering soil nutrient availability. The model fit was acceptable (χ² = 37.694, df = 14, RMSEA = 0.072, GFI = 0.928), although the significant chi-square value may reflect sample size sensitivity. Given the relatively small sample size (n = 20), more weight was placed on RMSEA and GFI, which indicated a reasonable fit and supported the plausibility of the proposed pathways.

Overall, these results demonstrate that the improvements in soil nutrient status induced by microbial agent application and deep tillage were key drivers of changes in microbial community composition and function, ultimately promoting straw decomposition.

4. Discussion

This study systematically investigated the synergistic effects of microbial agent application and deep tillage on straw decomposition, soil fertility improvement, and soil microbial community structure enhancement. By integrating kinetic modeling, soil property analysis, and microbial community profiling, we demonstrated that this combined management approach effectively accelerates straw degradation and optimizes soil ecological function.

The kinetic analysis of straw decomposition in this study demonstrated that the combined application of microbial agents and deep rotary tillage (M1D1) significantly increased the decomposition rate constant and notably shortened the times required for 50% and 95% straw decomposition compared to the control. These results indicate that integrating biological and physical interventions can effectively overcome the slow degradation typically observed under conventional straw return practices [27,30]. Microbial agents such as Bacillus spp. enhance straw degradation via nutrient activation and enzymatic activity, thereby promoting the decomposition of organic carbon [31]. Our results suggest that the synergistic effect between enhanced microbial activity and improved soil aeration under deep tillage conditions further optimizes the decomposition environment. Specifically, deep tillage disrupts compacted soil layers, increasing oxygen availability and promoting microbial proliferation [32], while microbial agents directly contribute degradative enzymes necessary for efficient lignocellulose breakdown [33]. This dual mechanism may account for the superior enhancement observed.

Secondly, the combined treatment significantly improved soil fertility, particularly TN, TP, and available nutrient levels (AN, AP, and AK). Compared to M0D0, TN and TP increased by 12.3% and 8.6%, respectively, under M1D1. Previous work showed that microbial agents, such as Bacillus megaterium and Bacillus mucilaginous, can enhance nutrient cycling by improving phosphorus and potassium availability and promoting plant growth [34]. Here, the addition of deep tillage enhanced nutrient redistribution [35], explaining the broader improvements in nutrient availability observed. This synergistic effect addresses a major limitation of conventional straw incorporation, where early-stage nitrogen immobilization and nutrient lockup frequently constrain crop growth [3,36].

In this study, soil physicochemical properties were identified as the primary drivers of microbial community structure and straw decomposition rate. The correlation analysis and Mantel tests confirmed that nutrient-related parameters, particularly total potassium (TK), available nitrogen (AN), and available phosphorus (AP), exhibited strong positive associations with both bacterial and fungal communities. These findings are consistent with previous reports indicating that enhanced nutrient availability can directly stimulate microbial proliferation and functional diversity in agricultural soils [37,38]. Soil pH demonstrated only weak correlations with microbial community composition, particularly in fungal assemblages [39,40]. The redundancy analysis (RDA) further supported this observation, as pH contributed minimally to the explained variance in microbial structure compared to nutrient-related factors. This result suggests that, under the experimental conditions of this study, nutrient status played a more decisive role than pH in shaping microbial communities, likely due to the relatively narrow pH range in the soils examined [41,42].

Structural equation modeling (SEM) analysis revealed a nutrient-mediated regulatory pathway linking straw decomposition rate, soil nutrient status, and microbial α-diversity. Specifically, straw decomposition significantly increased soil total phosphorus (TP), total nitrogen (TN), and available nutrient levels (AN, AP, and AK), suggesting strong feedback between residue turnover and nutrient release. In turn, available phosphorus (AP) and potassium (AK) exhibited significant positive effects on bacterial α-diversity, while available nitrogen (AN), AP, and AK strongly promoted fungal α-diversity. Notably, straw decomposition did not directly influence microbial α-diversity, and microbial α-diversity had no significant feedback effect on decomposition rate. These findings indicate that microbial diversity shifts are primarily driven by nutrient enrichment resulting from straw decomposition, rather than by decomposition itself. This pattern highlights the importance of nutrient availability—especially phosphorus and potassium—in shaping soil microbial communities during organic residue turnover. It also aligns with previous research suggesting that microbial functional responses are often constrained by nutrient limitations, particularly nitrogen and phosphorus availability, which are critical for sustaining microbial growth [43,44].

Microbial community analysis revealed that the combined treatment (M1D1) markedly enhanced bacterial and fungal richness, diversity, and co-occurrence network complexity, suggesting improved microbial ecological stability and functional resilience. These enhancements are likely attributable to the synergistic effects of microbial agent application and deep rotary tillage. This interpretation aligns with previous findings showing that deep vertical rotary tillage can improve soil nutrient availability and alter physicochemical conditions, thereby promoting microbial diversity and network interactions [45,46].

Specifically, beneficial microbial groups such as Gemmatimonadota and Glomeromycota were significantly enriched, supporting improved organic matter turnover and nutrient cycling capacities. Gemmatimonadota were enriched under M1D1 treatment, suggesting their potential involvement in the decomposition of complex organic compounds and carbon cycling, as reported in diverse soil environments [47,48]. The enrichment of Glomeromycota under the M1D1 treatment (Figure 4B) suggests a potential stimulation of arbuscular mycorrhizal associations. As arbuscular mycorrhizal fungi, Glomeromycota play a well-established role in improving soil phosphorus availability and facilitating plant–microbe symbiosis, thereby enhancing soil fertility and promoting sustainable plant growth [48,49]. The combined activity of these microbial groups improves soil structure, nutrient availability, and ecosystem resilience, although their function can be modulated by environmental factors such as soil type, climate, and management practices [50,51].

Recent studies have emphasized the pivotal role of microbial functional diversity in regulating lignocellulose degradation during straw decomposition. For instance, Anna M. Alessi et al. [52] employed multi-omics techniques to reveal that microbial enrichment during wheat straw degradation involved a succession towards Bacteroidetes- and Proteobacteria-dominated communities, with key genera such as Asticcacaulis, Leadbetterella, and Truepera contributing significantly to lignocellulolytic enzyme production. Consistent with these findings, our study observed increased relative abundances of Proteobacteria under the M1D1 treatment, suggesting that microbial amendment combined with deep tillage not only restructures bacterial communities but may also enrich lignocellulolytic taxa. Similarly, Qi Shen et al. [53] demonstrated that straw addition in an ectopic fermentation system enhanced the abundance of Sphingobacterium, Pseudomonas, and Bacillus, and boosted the expression of carbohydrate-active enzymes (CAZymes), leading to significantly greater cellulose, hemicellulose, and lignin degradation. These findings support our observation that microbial agents and tillage practices not only restructure microbial communities but also likely enrich taxa with potent lignocellulolytic functions, thereby accelerating straw turnover and nutrient cycling.

In summary, our findings suggest that the combination of microbial agent application and deep tillage enhanced straw decomposition rates and improved soil nutrient status, alongside observable shifts in microbial community structure. The results further indicate that nutrient availability—rather than microbial diversity itself—may play a more direct role in facilitating straw decomposition under the studied conditions. These observations provide insights into how integrated field practices may influence soil processes, though further long-term studies are needed to confirm broader implications for soil health.

5. Conclusions

This study investigated the effects of microbial decomposition agents and deep rotary tillage on rice straw decomposition, soil nutrient dynamics, and microbial community characteristics in paddy fields in southern China. The combined treatment (M1D1) was associated with increased straw decomposition rates; elevated levels of total nitrogen, available phosphorus, and available potassium; and shifts in microbial community structure, including higher bacterial and fungal diversity and co-occurrence network complexity. Path analysis further suggested that the influence of straw decomposition on microbial α-diversity was indirect and mediated through changes in soil nutrient availability, particularly AP, AK, and AN. Microbial α-diversity itself did not show significant direct feedback on decomposition rates. These findings indicate that management practices integrating microbial agents and tillage may enhance soil nutrient status and shape microbial communities through nutrient-mediated pathways. Further validation across different soil types and cropping systems is needed to generalize these observations.