Effects of Different Land Use Types on Soil Quality and Microbial Diversity in Paddy Soil

Abstract

This study investigated the effects of three land use patterns—rice (Oryza sativa L.)–rapeseed (Brassica napus L.) rotation (Rapeseed), rice–shrimp (Procambarus clarkii G.) rotation (Shrimp), and the conversion of paddy fields to forestland (Forestland)—on aggregate structure, nutrient content, and microbial diversity in rice soils in Chuandian Town, Jingzhou District, Jianghan Plain, central China. The results revealed that the Shrimp treatment significantly increased soil organic matter (SOM), available nitrogen (AN), and available phosphorus (AP) content in the surface soil (0–10 cm) while reducing soil bulk density and improving pore structure. Forestland exhibited higher aggregate stability in deeper soil layers (20–40 cm), particularly in the 0.053–0.25 mm size fraction. Microbial diversity analysis showed that bacterial richness (Chao1 index) and diversity (Shannon index) were significantly higher in the Shrimp and Rapeseed treatments compared to those in the Forestland treatment, with Proteobacteria and Chloroflexi being the dominant bacterial phyla. Fungal communities were dominated by Ascomycota, withfForestland showing greater fungal richness in deeper soil. Soil depth significantly influenced aggregates, nutrients, and microbial diversity, with surface soil exhibiting higher values for these parameters than deeper layers. Redundancy analysis indicated that SOM, AP, and pH were the key drivers of bacterial community variation, while fungal communities were more influenced by nitrogen and porosity. Path analysis further demonstrated that land use patterns indirectly affected microbial diversity via altering aggregate structure and nutrient availability. Overall, the Shrimp treatment outperformed others in improving soil structure and nutrient supply, whereas the Forestland treatment was more conducive to promoting aggregate stability in deeper soil. Land use patterns indirectly regulated microbial communities through modifying soil aggregate structure and nutrient status, thereby influencing soil ecosystem health and stability. This study provides a theoretical basis for the sustainable management of rice soils, suggesting the optimization of rotation patterns in agricultural production to synergistically enhance soil physical, chemical, and biological properties.

1. Introduction

Soil is a heterogeneous and complex entity that is influenced by multiple factors, including natural factors such as topography and parent material, as well as anthropogenic factors such as tillage and fertilization. Tillage and fertilization practices often lead to the uneven distribution of soil nutrients, exerting profound impacts on soil ecosystems [1]. As the foundation for rice growth, soil physicochemical properties—such as the soil structure, nutrient content, water status, pH, and redox potential—directly affect rice root development, nutrient uptake, and water use efficiency [2]. Soil aggregates play a critical role in physical processes such as erosion, compaction, and crusting, serving as the physical basis for soil fertility. In addition, soil aggregates contribute to the protection and sustained supply of essential nutrients such as carbon, nitrogen, and phosphorus [3,4]. These aggregates significantly influence soil physicochemical properties, making their study essential. As fundamental components of soil structure, soil aggregates markedly affect soil porosity, temperature, and fertility, thus serving as key indicators for evaluating soil texture [5,6,7]. Changes in land use patterns influence the composition, stability, and phosphorus stoichiometry of rice soil aggregates, with aggregates of different sizes exhibiting varying capacities to support soil functions and supply phosphorus [6].

Microorganisms play an indispensable role in soil formation through facilitating organic matter decomposition, improving soil structure, promoting plant growth, and remediating contaminated or degraded soils [6,8]. In soil ecosystems, microorganisms exhibit high biological activity and rapid responsiveness, participating in various biochemical processes. During the decomposition of organic matter, soil microorganisms convert organic matter into inorganic nutrients (nitrogen, phosphorus, and potassium) that plants can readily absorb, thereby enhancing nutrient cycling and utilization efficiency [9]. The distribution of soil microorganisms exhibits spatial heterogeneity, with microbial community structure and abundance varying across soil depths and types, which adds complexity and diversity to their distribution and functions [10,11,12]. Cao et al. [13] found that rice field rotations were more conducive to soil organic carbon accumulation and aggregate stability than forestland treatment. Kang et al. [14] investigated the impact of changes in rice soil use patterns on soil aggregates, revealing that long-term paddy–upland rotations increased the percentage of large aggregates (>0.25 mm) and aggregate stability while reducing aggregate-associated carbon storage and soil mineralization. Liu et al. [15] explored the effects of land use conversion on soil fertility, demonstrating significant increases in phosphorus content across different aggregate sizes in rice soils, with higher carbon, nitrogen, and phosphorus contents in larger aggregates compared to smaller ones. Li et al. [16] examined the responses of soil properties to land use changes and found that paddy fields exhibited the highest pH, soil organic matter (SOM), and total nitrogen (TN) values, while forestlands also had high SOM values; in contrast, drylands exhibited the lowest TN, SOM, and pH values. Luo et al. [17] studied microbial changes under paddy field conversion and reported higher biological indices, total anion concentrations, and prokaryotic community α-diversity in rice soils compared to drylands, although fungal α-diversity was lower in rice soils. Zhou et al. [18] investigated microbial community characteristics following the conversion of drylands to paddy fields, noting higher relative abundances of Acidobacteria in drylands, while Bacteroidetes, Chloroflexi, and Nitrospirae were more abundant in paddy fields; conversely, the relative abundances of Acidobacteria, Bacilli, and Planctomycetes significantly decreased after conversion.

Despite these findings, previous studies have not sufficiently explored the coupled relationships among soil structure, nutrient dynamics, and microbial diversity. Most research has focused on separately describing changes in soil nutrient content and microbial diversity, leaving the complex interactions, feedback loops, and co-evolutionary patterns under different rice soil use patterns poorly understood. Building on this foundation, this study was conducted in central China—a typical paddy–upland rotation planting area—to investigate the effects of different land use patterns on rice soil aggregate structure and phosphorus distribution. This study aimed to provide a theoretical reference for understanding rice soil aggregate structure, nutrient supplementation, and phosphorus dynamics. Specifically, this work addresses the following questions: (1) how do paddy–upland rotations alter soil nutrient forms and availability, thereby affecting soil fungal abundance and community structure? (2) How does the soil depth under different land use patterns influence rice soil microbial communities, and what factors collectively drive microbial community succession? This study aimed to elucidate the responses of soil microbial community succession to changes in soil physicochemical properties, chemical characteristics, and internal biological structures under different land use patterns in rice soils.

2. Materials and Methods

2.1. Description of Experimental Site

Soil samples were collected from three different land use types: rice–rapeseed rotation (Rapeseed), rice–shrimp rotation (Shrimp), and the conversion of paddy fields to forestland (Forestland), in Hubei Province, central China (N 30°33′24″, E 112°04′56″, altitude 52 m). The dominant soil type was paddy soil, which had been converted from conventional rice fields since 2011 [2]. The distribution of different rice field utilization patterns is presented in Figure 1.

2.2. Experimental Design

Field experiments were conducted in December, 2024, with soil samples collected from four soil layers (0–10, 10–20, 20–30, and 30–40 cm) under rice–rapeseed rotation, rice–shrimp rotation, and paddy field converted to forestland soils in the main rice planting area of Lijiachang Village, Chuandian Town. Each treatment was replicated four times. The samples were transported back to the laboratory in ice bags, and plant and animal debris were removed from the soil samples. The soil was then mixed thoroughly, with one part used to measure the physicochemical properties of the soil and the remaining part stored at −80 °C for microbial diversity analysis.

2.3. Sample Measurements

The soil pH was measured using a pH meter with a water-to-soil ratio of 1:2.5. The SOM content was determined according to the potassium dichromate external heating method. The available nitrogen (AN), available phosphorus (AP), available potassium (AK), and cation exchange capacity (CEC) were determined using flow analysis, molybdenum–antimony colorimetry, flame spectrophotometry [19], and ammonium acetate exchange methods, respectively. The soil bulk density was obtained following the ring knife method. The soil porosity was calculated based on the bulk density as follows: porosity (%) = (1 − bulk density/density) × 100, where the soil density was 2.65 g cm⁻³. Soil total DNA was extracted from soil samples using the Omega Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA). The extracted DNA was analyzed to determine the fragment size, concentration, purity, and integrity using 1.2% agarose gel electrophoresis (5 V/cm, 20 min) and the NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) [20] and stored at −80 °C for subsequent analysis. Polymerase chain reaction (PCR) amplification was performed in a 25 μL reaction system using high-fidelity DNA polymerase. The bacterial 16S rDNA V4–V5 region was amplified with primers 515F (5′-GTGYCAGCMGCCGCGGTAA-3′) and 806R (5′-GGACTACNVGGGTWTCTAAT-3′) [21], while the fungal ITS region was amplified with primers ITS5 (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and ITS2 (5′-GCTGCGTTCTTCATCGATGC-3′) [5]. The PCR protocol consisted of initial denaturation at 98 °C for 2 min, followed by 27 cycles (for bacterial 16S) or 35 cycles (for fungal ITS) of denaturation at 98 °C for 15 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, with final extension at 72 °C for 5 min and holding at 10 °C [22,23]. The PCR products were purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA), quantified with the Quantus Fluorometer (Promega, Madison, WI, USA), and subjected to high-throughput sequencing on the Illumina NovaSeq platform. Microbial diversity analysis was conducted by Beijing Biomarker Biotechnology Co., Ltd. (Beijing, China).

2.4. Statistical Analysis

After initial quality control of the high-throughput sequencing data, demultiplexing was performed using index and barcode information for precise sample differentiation, followed by barcode sequence removal. Interactive visualization of the community taxonomic composition was conducted using Krona software (version v 2.6) [22,23,24]. QIIME2 (version 2020.6) was employed to calculate Chao1 and species richness indices based on non-rarefied amplicon sequence variant (ASV) data to assess the microbial community richness in soil samples [13,25]. Furthermore, Shannon and Simpson diversity indices [26,27,28] were computed to evaluate species diversity within the communities. The linear discriminant analysis effect size (LEfSe) algorithm was applied to identify significantly affected taxa among different treatments [22]. All statistical analyses were performed in SPSS 18.0 (SPSS Inc., Chicago, IL, USA). Statistical significance was defined at p < 0.05, with p < 0.01 considered highly significant. Data visualization and graphical editing were conducted using Origin 2024 and Adobe Illustrator 2024 [29].

3. Results

3.1. Soil Physical Properties

3.1.1. Soil Porosity

Table 1. Soil structure of paddy soil under different land use types.

Soil Layer	Treatment	Bulk Density (cm3)	Porosity (%)	Capillary Pore (%)	Aeration Pore Space (%)	Water Content (%)
0–10 cm	Rapeseed	1.59 ± 0.09 b	39.8 ± 3.72 b	60.2 ± 3.94 ab	3.92 ± 0.37 a	32.7 ± 0.01 ab
	Shrimp	1.31 ± 0.02 c	50.3 ± 0.98 a	48.2 ± 1.69 c	4.95 ± 0.08 a	58.6 ± 0.01 a
	Forestland	1.36 ± 0.03 bc	48.5 ± 1.37 a	47.1 ± 1.21 c	4.82 ± 0.13 a	21.4 ± 0.01 b
10–20 cm	Rapeseed	1.79 ± 0.01 a	32.2 ± 0.65 bc	69.3 ± 1.11 a	3.18 ± 0.05 bc	20.0 ± 0.04 b
	Shrimp	1.70 ± 0.01 ab	35.8 ± 0.48 bc	64.2 ± 2.75 a	3.52 ± 0.04 ab	34.8 ± 0.01 a
	Forestland	1.81 ± 0.13 a	31.4 ± 5.18 c	64.2 ± 4.35 a	3.10 ± 0.52 c	21.5 ± 0.02 b
20–30 cm	Rapeseed	1.73 ± 0.00 ab	34.7 ± 0.27 bc	68.5 ± 0.84 a	3.42 ± 0.02 ab	24.3 ± 0.01 ab
	Shrimp	1.42 ± 0.12 bc	46.1 ± 4.59 a	54.0 ± 3.16 bc	4.56 ± 0.46 a	29.3 ± 0.01 ab
	Forestland	1.66 ± 0.01 b	37.3 ± 0.51 b	55.0 ± 3.48 b	3.70 ± 0.05 ab	18.0 ± 0.02 b
30–40 cm	Rapeseed	1.71 ± 0.03 ab	35.3 ± 1.14 bc	68.5 ± 1.62 a	3.49 ± 0.11 ab	24.2 ± 0.03 ab
	Shrimp	1.46 ± 0.01 b	44.8 ± 0.42 a	53.4 ± 3.78 bc	4.45 ± 0.04 a	26.0 ± 0.02 ab
	Forestland	1.62 ± 0.00 b	38.5 ± 0.36 b	53.3 ± 1.28 bc	3.83 ± 0.03 a	17.7 ± 0.03 b

Note: Different lowercase letters for the same item indicate significant differences at p < 0.05.

presents the impact of three different rice field land use types on soil physical properties at various soil depths (0–40 cm). The bulk density of Rapeseed soil samples was higher in all soil layers, with the maximum bulk density observed at the 0–10 cm layer, reaching 1.59 g cm⁻³ Compared to Rapeseed, Shrimp showed the greatest reduction in bulk density at the 0–10 cm, 20–30 cm, and 30–40 cm layers, with decreases of 17.6%, 17.9%, and 14.6%, respectively. The reduction at the 10–20 cm layer was relatively smaller at 5.0%. Forestland showed a 14.4% decrease in bulk density at the 0–10 cm layer compared to Rapeseed, but a slight increase of 1.11% at the 10–20 cm layer, with other layers showing varying degrees of reduction. Both Rapeseed and Forestland exhibited higher soil densities, with the soil density of Forestland being slightly lower than that of Rapeseed. At the 0–10 cm layer, Shrimp showed the most significant decrease in porosity, increasing by 26.38% compared to Rapeseed, while Forestland saw a decrease of about 21.86%. At the 10–20 cm layer, the porosity of Shrimp decreased by approximately 11.18% compared to that of Rapeseed, while the porosity of Forestland increased by about 2.48%. At the 20–30 cm layer, Shrimp showed a more significant reduction in porosity compared to Rapeseed (down by about 32.85%), while Forestland’s impact was relatively smaller (a decrease of about 7.49%). Shrimp showed significantly reduced porosity at the 20–30 cm soil layer, while Forestland had a lesser effect. At the 30–40 cm layer, the porosity of Shrimp soil decreased by about 26.91% compared to that of Rapeseed, and the porosity of Forestland soil exhibited a decrease of about 9.07%. Across the 0–40 cm soil depth, Rapeseed exhibited the highest capillary porosity, while Shrimp and Forestland had relatively similar proportions. At the 0–10 cm layer, Shrimp showed a more significant advantage. The capillary porosity of Rapeseed at the 0–10 cm layer was 60.2 ± 3.94%, which was the highest among all land use types. The capillary porosity of Shrimp was the highest at 64.2 ± 1.69%, and Forestland had the lowest capillary porosity at 47.1 ± 1.21%. At the 20–40 cm soil depth, Rapeseed maintained the highest capillary porosity, with the capillary porosity of Forestland slightly exceeding that of Shrimp. Shrimp exhibited the highest aeration porosity in the 0–30 cm soil layers, while Rapeseed had the lowest aeration porosity. Shrimp soil also contained higher water content in all soil layers, especially in the 30–40 cm layer. The water content of Forestland was generally close to that of Rapeseed, particularly at the 10–20 cm layer, with only a 1.5% difference from that of Rapeseed, but still lower than that of Shrimp. Rapeseed consistently showed the lowest water content compared to the other two land use types, with the greatest difference observed at the 30–40 cm layer. Capillary pores (<0.05 mm) retain water, which is essential for plant uptake and nutrient availability, while aeration pores (>0.05 mm) facilitate oxygen supply, which is critical for root respiration and microbial activity. Our results emphasize their importance in maintaining soil health, noting that the higher aeration porosity of Shrimp (e.g., 4.95% at 0–10 cm) supports aerobic microbial processes, while Rapeseed’s higher capillary porosity (60.2% at 0–10 cm) enhances water retention; these findings are consistent with those reported by Xu et al. [30].

3.1.2. Soil Aggregates

Figure 2 depicts the relative content of aggregates in different soil layers (0–40 cm) under three land use types. The 2 mm aggregate content in Rapeseed (98.1%) soil was significantly higher than that in Shrimp (94.3%) and Forestland (96.3%) soils. The difference between Rapeseed and Forestland was small, with values of 96.6% and 96.3%, respectively, while the 2 mm aggregate content of Shrimp was slightly lower (94.5%). Although the 2 mm aggregate content in the 10–20 cm soil layer was slightly lower under Shrimp, it remained relatively stable. The 2 mm aggregate contents in the 20–30 cm and 30–40 cm layers of Shrimp soil were 95.6% and 95.6%, respectively, both of which were higher compared to that of Rapeseed in the same layers (92.9% and 89.7%, respectively). Rapeseed soil exhibited the highest relative content of aggregates in the surface layer (0–10 cm), reaching 98.1%, with the relative content gradually decreasing with soil depth. The aggregate content in Shrimp showed little variation between different soil layers, maintaining a value of over 94.0% overall. The relative content of 0.25–2 mm aggregates in Shrimp rotation soil also increased, rising from 3.78% to 7.04%. The relative content of 0.25–2 mm aggregates in Forestland soil was higher than that in both Rapeseed and Shrimp soil at all levels, showing that forestland promoted stronger soil aggregate formation. The relative content of 0.25–2 mm aggregates under the Shrimp rotation was generally lower than that in Forestland. In the 0–10 cm layer, Forestland soil’s 0.25–0.053 mm aggregate content (5.23%) was much higher than that of Rapeseed soil (0.35%). In the 10–20, 20–30, and 30–40 cm layers, the 0.053–0.25 mm aggregate content of Forestland was consistently higher than that of Rapeseed. In all soil layers (0–40 cm), the relative content of 0.053–0.25 mm aggregates in Forestland soil was higher than that in Shrimp soil, especially in the 30–40 cm layer, where the aggregate content of Forestland (9.25%) was 1.06% higher than that of Shrimp (8.19%). In the 20–30 cm layer, the <0.053 mm aggregate content of Shrimp was 3.4% higher than that of Rapeseed, while in the 30–40 cm layer, Shrimp’s <0.053 mm aggregate content was 8.19% compared to Rapeseed’s 2.25%. Shrimp outperformed Rapeseed in all soil layers, particularly in deeper soil layers, where the aggregate content of Shrimp was significantly higher than that of Rapeseed.

3.2. Soil Chemical Properties

Table 2. Impacts of different land use practices on soil nutrients in rice paddy soils.

Soil Layer	Treatment	pH	SOM (g/kg)	TN (g/kg)	AN (mg/kg)	AP (mg/kg)	AK (mg/kg)	CEC (cmol/kg)
0–10 cm	Rapeseed	6.21 ± 0.01 cd	30.9 ± 0.60 a	2.94 ± 0.27 a	212.1 ± 2.39 ab	4.98 ± 0.88 b	67.1 ± 3.24 c	17.8 ± 1.57 bc
	Shrimp	6.51 ± 0.01 c	35.16 ± 0.9 a	3.31 ± 0.58 a	421.4 ± 1.63 a	6.34 ± 0.24 a	69.3 ± 2.53 c	16.7 ± 2.92 bc
	Forestland	5.98 ± 0.02 d	27.1 ± 2.36 ab	2.81 ± 0.11 a	132.3 ± 1.27 c	4.35 ± 0.00 b	78.5 ± 14.14 bc	11.2 ± 1.25 c
10–20 cm	Rapeseed	7.25 ± 0.05 a	29.2 ± 1.40 ab	2.74 ± 0.19 a	194.1 ± 1.25 b	2.98 ± 0.13 c	86.1 ± 5.12 ab	10.8 ± 2.80 c
	Shrimp	7.06 ± 0.02 b	27.8 ± 0.82 ab	2.47 ± 0.29 ab	325.3 ± 0.82 a	5.89 ± 0.11 a	69.3 ± 0.44 c	6.75 ± 2.42 c
	Forestland	6.8 ± 0.05 c	23.1 ± 3.90 c	2.72 ± 0.23 a	105.9 ± 2.06 c	2.71 ± 0.01 c	53.3 ± 0.55 d	14.2 ± 0.85 c
20–30 cm	Rapeseed	7.74 ± 0.00 a	25.2 ± 2.41 bc	1.65 ± 0.20 b	188.02 ± 1.73 b	2.52 ± 0.00 c	65.0 ± 3.56 c	21.9 ± 5.35 b
	Shrimp	6.87 ± 0.01 b	26.7 ± 1.75 bc	2.58 ± 0.06 a	304.2 ± 1.74 a	2.70 ± 0.04 c	71.3 ± 5.83 bc	38.2 ± 2.47 a
	Forestland	6.5 ± 0.01 c	21.9 ± 0.47 cd	1.96 ± 0.04 b	94.2 ± 0.83 c	3.39 ± 0.03 c	102.1 ± 1.77 a	45.0 ± 0.58 a
30–40 cm	Rapeseed	6.43 ± 0.02 cd	21.6 ± 0.84 cd	0.77 ± 0.24 c	165.7 ± 1.96 b	2.86 ± 0.01 c	91.2 ± 0.77 a	46.0 ± 0.31 a
	Shrimp	7.01 ± 0.03 b	23.0 ± 0.91 c	0.61 ± 0.07 c	156.4 ± 2.69 b	2.57 ± 0.02 c	81.3 ± 1.09 ab	44.1 ± 0.82 a
	Forestland	6.93 ± 0.03 b	19.0 ± 0.97 d	0.42 ± 0.25 c	82.1 ± 1.29 c	2.71 ± 0.01 c	83.3 ± 1.12 ab	30.4 ± 8.70 b

Note: Different lowercase letters for the same item indicate significant differences at p < 0.05. SOM: soil organic matter; TN: total nitrogen; AN: available nitrogen; AP: available phosphorus; AK: available potassium; CEC: cation exchange capacity.

shows the effects of different land use types (Rapeseed, Shrimp, and Forestland) on soil nutrients at different soil depths (0–10, 10–20, 20–30, and 30–40 cm) in paddy soils. The TN content in the 0–10 cm layer of Shrimp was 17.7% higher than that of Forestland. The AN content in the 0–40 cm soil layer of Shrimp was 37.0% and 45.4% higher than that of Rapeseed and Forestland, respectively. Shrimp also exhibited higher AP content (6.34 mg·kg⁻¹ in the 0–10 cm layer. The SOM content in the 0–10 cm layer of Shrimp was 26.9% and 45.7% higher than that of Rapeseed and Forestland, respectively. Forestland’s soil pH was lower in the 0–10 cm and 10–20 cm layers, indicating slightly acidic soil conditions. The AN content in Forestland was significantly lower, with average contents of 119.1 mg·kg⁻¹ in the 0–10 cm and 10–20 cm layers. The SOM content in the 0–10 cm layer of Forestland was 22.7% lower than that in Shrimp. In the 0–10 cm layer, the TN content in Shrimp soil was significantly higher than that in Forestland soil, with an increase of 17.8%, and it was also higher than that in Rapeseed soil, with an increase of 12.6%. In the 0–10 cm layer, the AN content of Shrimp was significantly higher than that of Forestland, representing an increase of 218.5%, and it was higher than that of Rapeseed, with an increase of 98.7%. In the 10–20 cm layer, the AN content of Shrimp was also significantly elevated compared to that of Forestland and Rapeseed, with increases of 207.6% and 67.5%, respectively. In the 0–10 cm layer, the SOM content of Shrimp was significantly higher than that of Forestland, with an increase of 29.7%, and it was higher than that of Rapeseed, with an increase of 13.7%, although the difference between treatments was not significant. In the 10–20 cm layer, the AK content of Forestland was lower compared to that of Shrimp and Rapeseed, with reductions of 23.1% and 38.0%, respectively. The CEC of Shrimp was higher compared to that of Rapeseed and Forestland in the 0–40 cm soil layer.

3.3. Microbial Composition and Diversity

3.3.1. Alpha Diversity of Soil Bacteria and Fungi

As shown in Figure S1, with the increase in sequencing reads, the number of features gradually increased. The dilution curves for bacteria and fungi exhibited typical saturation curves. This indicates that while increasing the sampling size enabled the detection of more microbial species, after reaching a certain number, the increase in the number of species became smaller and eventually stabilized. This finding suggests that the sequencing data utilized in this study are reasonable and effective.

Different land use practices significantly affected the richness and diversity of bacterial and fungal communities in paddy soils (

Table 3. Impact of different land use practices on soil nutrients in rice paddy soils.

Microbial Group	Soil Layer	Treatment	Richness Index		Diversity Index		Coverage
			Chao 1	Species	Simpson	Shannon
Bacteria	0–10 cm	Rapeseed	3470 a	3470 a	0.9983 a	10.51 a	0.9988 b
		Shrimp	3297 a	3293 a	0.9985 a	10.51 a	0.9987 b
		Forestland	2926 a	2926 a	0.9966 b	9.70 b	0.9996 a
	10–20 cm	Rapeseed	2791 a	2789 a	0.9972 a	9.92 a	0.9989 b
		Shrimp	2883 a	2882 a	0.9977 a	10.06 a	0.9988 b
		Forestland	3011 a	3011 a	0.9968 a	9.78 a	0.9996 a
	20–30 cm	Rapeseed	2484 a	2484 a	0.9960 b	9.51 b	0.9994 a
		Shrimp	3353 a	3350 a	0.9984 a	10.49 a	0.9987 b
		Forestland	2847 a	2847 a	0.9969 ab	9.74 b	0.9995 a
	30–40 cm	Rapeseed	2576 a	2575 a	0.9945 a	9.95 a	0.9996 a
		Shrimp	2509 a	2505 a	0.9974 a	9.91 a	0.9988 b
		Forestland	2767 a	2767 a	0.9966 a	9.62 a	0.9995 a
Fungi	0–10 cm	Rapeseed	226 b	226 b	0.9464 a	6.15 a	1.000 a
		Shrimp	251 b	251 b	0.8631 a	5.17 a	1.000 a
		Forestland	451 a	450 a	0.9412 a	5.69 a	0.999 a
	10–20 cm	Rapeseed	272 b	271 b	0.9790 a	6.85 a	1.000 a
		Shrimp	259 b	259 b	0.9038 a	5.96 a	1.000 a
		Forestland	424 a	424 a	0.9503 a	6.13 a	1.000 a
	20–30 cm	Rapeseed	352 b	352 b	0.9744 a	7.10 a	1.000 a
		Shrimp	249 b	249 b	0.9594 a	6.32 a	1.000 a
		Forestland	530 a	529 a	0.9462 a	6.02 a	0.999 b
	30–40 cm	Rapeseed	269 a	269 a	0.9828 a	6.92 a	1.000 a
		Shrimp	404 a	403 a	0.8506 a	6.01 a	1.000 a
		Forestland	423 a	422 a	0.9481 a	5.89 a	0.999 a

Note: Different lowercase letters for the same item indicate significant differences at p < 0.05.

). Regarding bacterial community, the Rapeseed group showed the highest Chao index in the 0–10 cm soil layer, both at 3470, while the Forestland group had the lowest richness index in this layer, at 2926. In the 10–40 cm soil layers, Forestland had the highest Chao 1 index and Species index. According to the Simpson diversity index, the Rapeseed group exhibited higher diversity in the 0–10 cm (0.9983) and 10–20 cm (0.9972) soil layers, while the Forestland group showed lower diversity, especially in the 0–10 cm layer (0.9966). The Shannon diversity index indicated that the Rapeseed and Shrimp groups had higher values in the 0–10 cm and 10–20 cm soil layers. The Forestland group maintained a relatively high coverage level (0.9996); although the coverage in the 30–40 cm soil layer (0.999) was slightly lower compared to that in other layers, it still approached 1.0. Regarding fungal community, the fungal diversity index for all three treatments increased with soil depth. The Chao 1 index and Species index for all three treatments reached their maximum values in the 30–40 cm soil layer. There was a negative correlation between fungal diversity and soil depth, while fungal richness showed a positive correlation with soil depth.

3.3.2. Soil Bacterial and Fungal Composition

At the bacterial phylum level, Chloroflexi exhibited relatively high abundance across all soil layers, indicating its dominance in the study area (Figure 3a). The abundance of Chloroflexi in the Rapeseed 3 and Shrimp 4 soil layers was 23.03% and 21.02%, respectively, with an average abundance of 17.47%, the highest among all bacterial phyla. The abundance of Proteobacteria was also significant, especially in the Shrimp (1, 2, 3) and Forestland (1, 2, 3) soil layers, with an average abundance of 15.30%, ranking second among all bacterial phyla. The abundance of Acidobacteriota was relatively stable across all soil layers, with an average abundance of 11.76%, ranking third. The abundance of this phylum was more balanced, particularly in the Shrimp (1, 2, 3) soil layers. Actinobacteriota had a high abundance of 25.84% in the Forestland 1 soil layer, with an average abundance of 8.97%, ranking fourth. The abundance of Verrucomicrobiota and Planctomycetota was relatively low, with an average abundance of 6.12%, and their distribution across different soil layers was relatively uniform. Among fungi, Ascomycota and Basidiomycota were the dominant fungal phyla (Figure 3b). Ascomycota exhibited significant abundance in all soil layers, especially in the Shrimp (1, 3) and Forestland (1, 3) soil layers, where the abundance reached 68.79% and 71.17%, respectively, showing the highest abundance across all soil layers. Basidiomycota also showed significant abundance in all soil layers, particularly in the Rapeseed (3, 4) and Shrimp (2, 4) soil layers, where the abundance was 19.53%, 22.17%, and 22.45%, 28.32%, respectively, with an average abundance of 22.78%. Mortierellomycota showed adaptability to specific soil environments. The abundance of Mortierellomycota was higher in the Rapeseed (1, 2) and Shrimp (1, 2) soil layers at 10.36% and 15.78%, respectively, while it was lower in other soil layers, particularly in the Forestland (3, 4) soil layers, where the abundance dropped to 3.43%. The average abundance of Mortierellomycota was 6.70%.

As shown in Figure 4, the abundance of the bacterial (Figure 4a) and fungal (Figure 4b) communities was closely related to environmental treatments. Different soil types and treatments significantly affected the composition of microbial communities. Among all bacterial phyla, Proteobacteria and Desulfobacterota exhibited significant linear discriminant analysis (LDA) scores, especially in the Forestland 1 and Rapeseed 1 treatment groups, indicating that these two bacterial phyla had a strong growth advantage under specific soil or treatment conditions. Spirochaetota and Nitrospirota had higher LDA scores in the Rapeseed 3 and Shrimp 2 treatment groups, while Acidobacteriota and Bacteroidota showed lower LDA scores.

Among fungal phyla, Mortierellomycota and Rozellomycota exhibited significantly higher LDA scores in the Forestland 1 and Shrimp 3 treatment groups, while Mucoromycota and Kickxellomycota showed lower abundances across different treatments. Glomeromycota had lower abundance in the Rapeseed 2 and Forestland 4 treatment groups.

Figure 5 shows the co-occurrence networks of bacterial and fungal communities in soil under different land use types. The results show that the interactions between bacterial ecological networks across different treatments were mainly characterized by synergistic relationships, with positive correlation coefficients above 50% (

Table 4. Molecular ecological network parameters of soil microbial communities under different rice field land use practices.

Network Graph Parameters	Bacteria	Fungi
Edge count	11	15
Node count	9	17
Positive correlation ratio (%)	57.14	60
Average degree	1.56	1.76
Average weighted degree	5.74	14.19
Graph density	0.194	0.061
Connected components	3	1
Average clustering coefficient	0.333	0.35

). Among bacteria, Micrarchaeota and Actinobacteriota had the highest centrality in the network. Planctomycetota and Desulfobacterota showed a positive correlation, while Spirochaetota and unclassified_Bacteria showed a negative correlation in abundance. A competitive relationship was identified between Bacteroidota and Spirochaetota. The positive correlation between Planctomycetota and Desulfobacterota indicates their synergistic growth under similar resource demands. Among fungi, Basidiomycota, Mortierellomycota, and Ascomycota formed key groups in the molecular ecological network. These key groups occupied central positions in the microbial community network, connecting the most nodes and therefore playing a crucial role in maintaining the stability of the microbial community.

Microbial community composition at the genus level under different rice soil land use types is presented in Figure 6. At the bacterial genus level, as shown in Figure 6a, the abundance of Isosphaeraceae (maximum variation: 6.35%) showed significant changes under different soil management conditions. Specifically, the abundance of Isosphaeraceae ranged from 0.15% (Rapeseed 3) to 6.50% (Forestland 3). Anaerolineaceae (maximum variation: 5.57%) ranged in abundance from 2.11% (Forestland 4) to 7.73% (Rapeseed 3). Vicinamibacterales (maximum variation: 4.04%) showed relatively small changes in abundance, fluctuating smoothly between 0.73% (Forestland 4) and 3.77% (Rapeseed 1). SC_I_84 (maximum variation: 3.50%) exhibited a smaller range of variation. At the fungal genus level, as shown in Figure 6b, Fusarium had the largest variation in abundance (16.24%). The variation in abundance of Fusarium between Forestland 2 (16.22%) and Shrimp 1 (1.53%) indicated a particularly strong response to soil conditions. Mortierella showed a maximum variation of 14.08%, with its abundance ranging from 0.67% (Forestland 3) to 15.18% (Rapeseed 2) and higher abundance detected in soils rich in SOM. Didymella exhibited a variation of 11.73%, with abundance approaching zero under Forestland 3 and Forestland 4 conditions and reaching its peak at 12.28% under Rapeseed 3 conditions. Hygrocybe and Saitozyma showed variation ranges of 6.54% and 5.54%, respectively, with higher abundance under Forestland 2 and Forestland 4 conditions. Cladosporium and Penicillium showed smaller variations, with changes of 3.94% and 2.97%, respectively.

3.4. Beta Diversity of Soil Bacteria and Fungi

In the bacterial community, as shown in Figure 7a, the x-axis (PC1) explained 8.70% of the variance, while the y-axis (PC2) explained 5.39%, with a total contribution rate of 14.09%. There was a clear separation between the Rapeseed, Shrimp, and Forestland groups, with lower diversity observed in the Rapeseed group. The Shrimp group exhibited richer bacterial community diversity, while the Forestland group had moderate bacterial community diversity. In the fungal community, as shown in Figure 7b, the PC1 axis explained 6.81% of the variance and the PC2 axis explained 3.84% of the variance, with a total contribution rate of 10.65%. Significant variation in fungal community structure was observed between the Rapeseed and Forestland groups, while the fungal community structure of the Shrimp group was more consistent. The confidence ellipse for the Forestland group did not overlap with the other two groups, demonstrating that the fungal community structure under paddy field soil converted to forestland differed significantly from the other two land use types. Although the fungal communities of the Rapeseed and Shrimp groups exhibited similarities, the variation in the Rapeseed group was larger, while the Shrimp group was more concentrated.

In the bacterial community, the number of shared operational taxonomic units (OTUs) across all samples was 89 (Figure 8a). Forestland 2 had the highest number of unique OTUs, accounting for 26.2% of the total OTUs in the Forestland group. The average number of unique OTUs in the Shrimp group (3792) was lower than that in the Rapeseed group but exhibited less variation. The number of unique OTUs in the Shrimp group showed greater variability, with a high value of 4627 in Shrimp 3, which may have been related to the wetland environment, while the low value of 2543 in Shrimp 4 may reflect local environmental differences. Regarding fungal community, Forestland 4 had the highest number of unique OTUs, accounting for 29.1% of the total OTUs in the Forestland group (Figure 8b). The average number of unique OTUs in the Forestland group was the highest, indicating that paddy field soil converted to forestland had the strongest shaping effect on the fungal community. The average number of unique OTUs in the Rapeseed and Shrimp groups was similar. The number of shared OTUs was very low at only seven, indicating that the bacterial communities under the three land use types were highly differentiated.

To perform a more comprehensive comparison of the species composition differences at the genus level under different treatments, a heatmap (Figure 9) was designed based on cluster analysis of the relative abundances of ASVs. The depth of the soil layer significantly affected the distribution and relative abundance of dominant bacteria and fungi in paddy soil. In general, bacterial abundance tends to be higher in shallow soil layers and lower in deeper soil layers. In the shallow soil layers in the present study, certain bacterial genera, such as Nitrobacter and Bacillus, exhibited higher abundance (Figure 9a). In the deeper soil layers, Clostridium and Firmicutes showed stronger adaptability. Among fungi, as shown in Figure 9b, a gradual decrease in the abundance of certain fungal genera was observed from Forestland 1 to Forestland 4 (from shallow to deep soil layers). From Shrimp 1 to Shrimp 4, the fungal abundance varied with soil depth. Similarly, from Rapeseed 1 to Rapeseed 4, the fungal abundance declined as the soil depth increased.

3.5. Correlation Between Soil Physicochemical Properties and Dominant Microbial Species

Redundancy analysis (RDA) was performed utilizing all physical, chemical, and biological data from the 0–40 cm soil layer to examine the correlations between soil properties and dominant microbial species. For bacteria, the first principal component axis (PCoA Axis 1) explained 59.7% of the community variation, and the second principal component axis (PCoA Axis 2) explained 5.7% of the variation (Figure 10a). This indicates that the soil microbial community structure was significantly driven by environmental gradients. Proteobacteria and Firmicutes showed significant positive correlations with SOM, AK, AP, CEC, and pH. In contrast, Acidobacteriota and Chloroflexi had weaker correlations with these environmental factors. Additionally, Verrucomicrobiota exhibited a strong positive correlation with AN, while Planctomycetota and Clenarchaeota showed stronger correlations with TN, reflecting the role of nitrogen cycling in driving the distribution of these bacterial groups. For fungi (Figure 10b), PCoA Axis 1 explained 37.4% of the community variation, while PCoA Axis 2 explained 16.1% of the community variation, which suggests that the soil fungal community structure was also significantly driven by environmental gradients. Ascomycota and Monoblepharomycota exhibited significant positive correlations with SOM, TN, AN, and pH (Figure 2). In contrast, the correlations between these environmental factors and Basidiomycota, Glomeromycota, and Kickxellomycota were weaker. Mucoromycota and Calcarisporiellomycota exhibited strong positive correlations with AN, while Rozellomycota and Aphelidiomycota showed higher correlations with TN and pH, reflecting the driving influence of nitrogen and pH on the distribution of these fungal communities. Notably, Entorrhizomycota and Chytridiomycota were positively correlated with soil AP and AK, while they were negatively correlated with soil CEC.

Partial least squares path modeling (PLS-PM) analysis revealed that land use types significantly influenced soil bacterial and fungal community diversity through influencing the soil aggregate structure and soil nutrients (Figure 11). The model showed good fit (χ² = 18.456, df = 15, p = 0.238; CFI = 0.985; TLI = 0.972; RMSEA = 0.038), confirming that the path model effectively explains the relationships between variables. Path analysis demonstrated that land use types indirectly drove changes in soil microbial diversity by impacting the soil aggregate structure and soil nutrient status. In Forestland soil, the reduced soil bulk density and increased SOM significantly enhanced bacterial community diversity, which was consistent with the distribution of Proteobacteria and Acidobacteriota along the SOM gradient revealed by the PCoA. Fungal community diversity was significantly enhanced in Rapeseed soil due to the increased content of >2 mm aggregates and AN, which was related to the increased abundance of Mucoromycota. The Shrimp treatment promoted fungal community diversity through increasing soil porosity and AN content, which may have been associated with the growth of Entorrhizomycota. Elevated porosity and SOM content in the surface soil (0–10 cm) supported the diversity of bacterial and fungal communities, while anaerobic groups dominated in deeper soil layers (30–40 cm). This result is consistent with the trend described earlier in this article whereby bacterial richness is negatively correlated with soil depth, while fungal richness is positively correlated with soil depth (Section 3.3.1).

4. Discussion

4.1. Impacts of Land Use Types on Soil Aggregate Structure and Physical Properties

Soil aggregates, as fundamental units of soil structure, directly influence the physical properties, nutrient retention, and erosion resistance of soil. This study identified significant variations in soil aggregate structure and physical properties across land use types. Rapeseed exhibited the highest proportion of >2 mm aggregates (98.1%) in topsoil (0–10 cm), significantly exceeding the proportions of Shrimp (94.3%) and Forestland (96.3%) (Figure 3). This was likely due to Rapeseed root exudates (e.g., polysaccharides and organic acids) and crop residue inputs, which promoted particle binding via cementation [2], as well as intensive tillage, thereby promoting large aggregate formation. Lower tillage frequency in the Rapeseed treatments minimized soil disturbance, favoring macroaggregate formation and stability; this finding is consistent the study by with Xu et al. [14,30], which reported increased > 0.25 mm aggregates under long-term rice–wheat rotation. Conversely, Forestland showed greater aggregate stability in deeper soil (20–40 cm), particularly for 0.053–0.25 mm fractions (Figure 3), likely because deep-rooted vegetation (trees) and litter inputs enhanced the organic matter content and microaggregate formation [11,31]. Long-term vegetation cover in the Forestland treatments reduced erosion and mechanical disturbance, further stabilizing aggregates [13]. The Shrimp treatment showed uniform aggregate distribution, with significantly higher 0.053–0.25 mm and <0.053 mm fractions in deeper soil (30–40 cm) compared to Rapeseed treatment (Figure 3). This may result from shrimp activities such as burrowing and excretion, which improve soil pore structure and particle contact, thereby facilitating aggregate formation [32]. Changes in soil bulk density and porosity supported these findings. Shrimp had a significantly lower bulk density, with reductions of 17.6% and 17.9% at 0–10 cm and 20–30 cm, respectively (Table 1). This suggests that water management and bioturbation enhanced soil looseness, increasing porosity by up to 50.3% at 0–10 cm, in alignment with previous reports that shrimp culture improves soil structure [32]. Forestland showed slightly elevated porosity at 10–20 cm (31.4%), reflecting stable structure formation during long-term restoration. In contrast, Rapeseed’s higher bulk density (1.59 g·cm⁻³ at 0–10 cm) and lower porosity (39.8%) likely resulted from tillage-induced compaction [30]. It is worth noting that Rapeseed’s bulk density (1.59 g·cm⁻³ at 0–10 cm) exceeds the typical range for paddy soils (1.2–1.4 g·cm⁻³) reported by Xu et al. [30], indicating moderate compaction likely due to tillage. These physical changes affect the distribution of soil water and oxygen, profoundly impacting nutrient cycling and microbial communities.

4.2. Driving Mechanisms of Land Use Types in Relation to Soil Nutrient Distribution and Availability

Soil nutrient content is a critical indicator of soil quality, with land use significantly influencing nutrient accumulation and availability. In this study, the Shrimp rotation treatments exhibited the highest levels of soil SOM, AN, and AP in topsoil (0–10 cm), likely resulting from the accumulation of shrimp excrement and residual feed, enriching organic matter. This effect was further enhanced by high moisture levels, which promote decomposition, as noted by Ma et al. [33]. Shrimp surpassed Rapeseed and Forestland by 26.9% and 45.7%, respectively, in terms of SOM, and by 98.7% and 218.5%, respectively, in terms of nitrogen. Shrimp also had significantly higher phosphorus at 6.34 mg·kg⁻¹ (Table 2). This is likely because shrimp excrement and residual feed increase the SOM content, while high moisture promotes SOM decomposition and nutrient release [34]. The higher pH of the Shrimp treatment (6.51 at 0–10 cm), compared to that of the Forestland treatment (5.98), creates a more alkaline environment that can enhance nutrient availability, such as phosphorus solubility [35]. In contrast, Forestland’s lower pH (5.98 at 0–10 cm) compared to Shrimp (6.51) and Rapeseed (6.21) may be attributed to organic acid accumulation from litter decomposition and reduced liming, as supported by Li et al. [16]. Notably, Rapeseed also showed elevated pH (6.21 at 0–10 cm, up to 7.74 in deeper soil) (Table 2), possibly due to the regulation of acidity by lime application or root exudates [16,36]. However, its low AP (4.12 mg·kg⁻¹ at 0–10 cm) may reflect enhanced phosphorus fixation, particularly in central China soils with high iron–aluminum oxide content [35]. We suggest that Rapeseed’s low AP content (4.98 mg·kg⁻¹ at 0–10 cm), which is likely due to phosphorus fixation, may necessitate the application of P fertilization or organic amendments (e.g., manure) to enhance availability, aligning with the findings of Biswas et al. [35]. Forestland had lower nutrient levels, especially SOM and nitrogen in deeper soil (Table 2), likely due to slower SOM decomposition and long-term nutrient uptake by vegetation [37]. Notably, the AP content of Forestland soil was higher (132.5 mg·kg⁻¹ at 0–10 cm), which is possibly linked to potassium enrichment by tree roots and enhanced clay mineral adsorption [16]. Soil depth significantly influenced nutrient distribution. Topsoil (0–10 cm) contained higher SOM, nitrogen, and phosphorus levels than deeper soil (20–40 cm), reflecting greater organic inputs and microbial activity [9,37]. For example, the SOM content of Shrimp reached 32.6 g·kg⁻¹ at 0–10 cm, but it dropped to only 19.8 g·kg⁻¹ at 30–40 cm (Table 2). These vertical differences likely shape microbial community structure and function.

4.3. Regulation of Microbial Community Diversity and Function by Land Use and Soil Depth

Microbial community structure and function were highly sensitive to land use. Alpha diversity analysis revealed that the Rapeseed and Shrimp treatments had significantly higher bacterial richness (Chao1) and diversity (Shannon) compared to the Forestland treatment, particularly in topsoil (0–10 cm) (Table 3). For example, Shrimp’s Chao1 value reached 3470 at 0–10 cm, compared to 2926 for Forestland. This is in alignment with nutrient distribution, indicating that nutrient availability (SOM and AN) drove bacterial diversity [38]. High SOM and nutrient contents in topsoil provide carbon and energy, enhancing diversity [37,39]. Conversely, Forestland exhibited higher fungal richness in deeper soil (30–40 cm) (Table 3), likely due to litter inputs and anaerobic conditions that favor fungi such as Ascomycota [40,41]. At the phylum level, Proteobacteria and Chloroflexi dominated the bacterial communities (Figure 5a). Proteobacteria showed high abundance in Shrimp and Forestland (mean 15.30%), positively correlating with SOM, AP, and pH (Figure 10a), underscoring the importance of these phyla in SOM decomposition and nutrient cycling [39]. Chloroflexi were most abundant in Rapeseed and Shrimp (mean 17.47%), likely due to their ability to adapt to high-moisture, low-oxygen environments [41]. The enrichment of Actinobacteriota in Forestland (up to 25.84%) reflects their ability to decompose lignin and cellulose in low-nutrient settings [32]. Fungal communities were dominated by Ascomycota and Basidiomycota (Figure 5b). Ascomycota predominated in Shrimp and Forestland (up to 71.17%), indicating their adaptation to high moisture and SOM [24]. Basidiomycota were more abundant in Rapeseed and Shrimp (mean 22.78%), which is likely linked to higher organic residue inputs. At the genus level, Fusarium abundance varied significantly in Shrimp (up to 16.24%), possibly due to moisture fluctuations caused by shrimp activity, as this fungus thrives in wet environments [24]. The high Fusarium abundance (16.24% variation) in the Shrimp treatment, likely linked to moisture fluctuations caused by shrimp activity, could pose a risk to rice health due to its pathogenic potential (e.g., root rot), as reported by Jia et al. [24]. We recommend future monitoring of Fusarium levels and plant disease incidence to assess potential long-term implications. Forestland showed higher Mortierella abundance, which is likely associated with its decomposition of complex SOM-like litter [41]. Beta diversity analysis highlighted microbial community differences (Figure 9). Bacterial communities in Rapeseed were separated from those of Shrimp and Forestland (PC1: 8.70% variance), while fungal communities in Forestland showed no overlap with the others (PC1: 6.81% variance), which is suggestive of stronger fungal community shaping in Forestland soils driven by limited nutrients and high litter input [40,42]. Co-occurrence network analysis (Figure 7) indicated that bacterial communities were dominated by synergistic interactions (positive correlations > 50%), such as Planctomycetota and Desulfobacterota, reflecting cooperative growth under similar resource demands. Fungal communities relied on Ascomycota and Basidiomycota as keystone taxa, stabilizing community structure [31,38].

4.4. Coupling Mechanisms of Soil Physicochemical Properties and Microbial Communities and Their Ecological Implications

RDA and path analysis were conducted to elucidate soil property–microbial interactions. The results showed that bacterial distribution was driven by SOM, AP, and pH (Figure 10a), with Proteobacteria and Firmicutes showing strong positive correlations, highlighting their role in SOM decomposition and nutrient cycling [39]. Fungal communities were more influenced by nitrogen (AN) and porosity (Figure 10b), with Ascomycota and Monoblepharomycota positively correlated with SOM and nitrogen, reflecting their contributions to nitrogen cycling [41]. Notably, some fungi (Entorrhizomycota) were positively correlated with AP and potassium but negatively correlated with CEC, possibly due to high potassium enrichment and low CEC in Forestland [34,37]. Path analysis (Figure 11 showed that land use indirectly drove microbial diversity by altering soil aggregate structure and nutrient availability. Forestland soil exhibited increased bacterial diversity due to a reduced bulk density and increased SOM, which is consistent with the distribution of Proteobacteria and Acidobacteriota along the SOM gradient [33]. Rapeseed soil showed enhanced fungal diversity because of its higher > 2 mm aggregate and nitrogen contents, which correlated with elevated Mortierella abundance [24,43]. Shrimp boosted fungal diversity via increased soil porosity and nitrogen content, which may have been associated with Entorrhizomycota growth [44]. Soil depth was critical: high porosity and SOM in topsoil (0–10 cm) supported bacterial and fungal diversity, while anaerobic taxa (Chloroflexi and Basidiomycota) dominated deeper soil (30–40 cm) (Table 3), in alignment with depth-driven microbial distribution [17]. These findings highlight coupled physical–chemical–biological processes and demonstrate that land use shapes microbial structure and function through altering soil aggregates and nutrient contents, thereby affecting ecosystem health and stability. The Shrimp treatment enhanced nutrient availability and porosity, boosting bacterial decomposition and nutrient cycling. Under the Forestland treatment, long-term carbon sequestration and erosion resistance were promoted by stable deep aggregates and fungal enrichment [11,45]. However, the low AP content of Rapeseed may have limited microbial activity, particularly in phosphorus-limited rice soils [46,47].

5. Conclusions

This study elucidated land use impacts on rice soil ecosystems. The findings of this study offer practical insights, showing that rice–shrimp rotation enhances soil quality and agricultural productivity. The rapeseed rotation treatments optimized soil structure and nutrient availability, while the conversion of paddy fields to forestland enhanced deep soil aggregate stability and long-term carbon sequestration, which is ideal for ecological restoration. Rapeseed requires optimized phosphorus management to reduce fixation and improve nutrient efficiency [35]. Soil depth significantly influenced microbial distribution, with active bacteria dominating the topsoil and stress-tolerant fungi prevailing in deeper layers, reflecting environmental filtering and biotic interactions [4,48,49]. Future research should integrate long-term experiments to explore microbial functional genes and nutrient cycling for sustainable soil management. The conversion of paddy fields to forestland requires topsoil nutrient supplementation to sustain microbial activity; and rice–rapeseed rotation needs optimized phosphorus management to mitigate fixation. Paddy–upland rotation regulates nutrient availability (nitrogen and phosphorus) and soil properties (pH and porosity), directly selecting fungal taxa adapted to distinct niches, consequently altering community structure and function. Soil depth, through gradients in properties (nutrients and oxygen) and aggregates, shapes microbial vertical distribution alongside land use. The topsoil hosts active bacteria, while deeper soil favors stress-tolerant fungi, reflecting combined environmental and biotic effects.