Analysis of Soil Nutrients and Microbial Community Characteristics in Rainfed Rice–Potato Cropping Systems

Abstract

Background: Rainfed rice–potato cropping systems represent an emerging agricultural pattern in Yunnan Province. This study investigates the dynamics of soil nutrient release and microbial community structure under rainfed rice–potato cropping systems. Methods: Four experimental treatments were established using two rice cultivation methods (flooded and rainfed cultivation) as the preceding crop, followed by two distinct potato cultivars: rainfed rice–potato Dianshu 23 (DR), rainfed rice–potato Dianshu 1418 (DY), flooded rice–potato Dianshu 23 (WR), and flooded rice–potato Dianshu 1418 (WY). Soil samples were collected before rice planting and at harvest, as well as before potato planting and at 40-, 80-, and 120-days post-planting. Soil nutrient release dynamics and microbial community composition were analyzed across all treatments. Results: Flooded rice cultivation as the preceding crop exhibited higher soil nutrient depletion compared to rainfed systems, accompanied by more pronounced increases in soil urease and invertase activities. Following potato establishment, rainfed rice–potato systems demonstrated an accelerated release of available nitrogen and potassium during the initial growth period relative to flooded rice–potato systems. At potato harvest, soil urease and invertase activities increased in rainfed rice–potato systems compared to pre-planting levels, while decreasing in flooded rice–potato systems. Proteobacteria constituted the dominant bacterial phylum across all treatments. Rainfed rice cultivation significantly enhanced Cyanobacteria relative abundance, whereas flooded rice cultivation promoted increased Thermodesulfobacteria abundance. Ascomycota dominated fungal communities, with flooded rice showing significantly greater reductions in Ascomycota relative abundance compared to rainfed systems. Rainfed rice–potato systems exhibited superior soil microbial community richness, diversity, and species abundance relative to flooded rice–potato systems. Bacterial genera associated with nitrogen metabolism showed higher relative abundance in rainfed rice–potato systems, as did pathogenic fungal genera. Conclusions: Soil nutrient characteristics and microbial community profiles in rainfed rice–potato cropping systems differ markedly from traditional flooded rice–potato rotation practices. These findings provide valuable insights for optimizing water and nutrient management strategies in rainfed rice–potato cropping systems.

1. Introduction

The potato (Solanum tuberosum L.), a member of the Solanaceae family, serves as both a staple food and vegetable crop, representing the fourth most important crop globally after rice, wheat, and maize [1,2]. The diverse altitudinal climate zones of Yunnan Province enable continuous potato production throughout the year. Rainfed rice cultivation represents an upland production system that relies predominantly on natural precipitation with minimal supplemental irrigation [3]. The widespread adoption of rainfed rice cultivation in Yunnan Province has led to the emergence of an innovative rainfed rice–potato multiple cropping system.

Continuous potato monoculture results in soil nutrient depletion and an increased incidence of soil-borne diseases, ultimately compromising both productivity and long-term sustainability [4]. Crop rotation enhances nutrient cycling through the accelerated decomposition of crop residues and improved soil microbial community structure [5], thereby increasing nutrient use efficiency [6]. Additionally, rotation reduces pathogenic microorganisms in the rhizosphere while promoting beneficial fungal populations [7]. Comprehensive five-year rotation studies demonstrate that potato-rice systems achieve optimal yield performance, with sustainable yield indices consistently exceeding 0.8 [8]. Further research indicates substantial potential for potato–rice rotations in terms of reducing phosphorus fertilizer requirements and minimizing environmental impacts [9].

Previous research has predominantly examined the interrelated effects of soil nutrient cycling and microbial community dynamics in conventional flooded–upland rotation systems. However, investigations of soil nutrient cycling and microbial community changes under the emerging rainfed rice–potato cropping pattern remain unreported. This investigation established rainfed rice–potato and flooded rice–potato cropping systems and employed high-throughput sequencing to examine soil nutrient dynamics and microbial community variations under contrasting rice cultivation methods, along with their subsequent effects on soil properties during the cultivation of different potato cultivars. The research aims to provide theoretical foundations for optimizing field management practices in rainfed rice–potato cropping systems.

2. Materials and Methods

2.1. Experimental Site Characteristics

The pot experiment was conducted from March 2024 to January 2025 in a greenhouse facility at the Teaching Experimental Farm of Yunnan Agricultural University. Cylindrical pots measuring 30 cm in height and 40 cm in diameter were used for cultivation. The experimental site is located at 102°42′ E, 25°22′ N, with an elevation of 1920 m. The region experiences a subtropical monsoon climate characterized by distinct dry and wet seasons. Rice was cultivated from March through September 2024, followed by potato cultivation from September 2024 through January 2025. The growth medium consisted of red soil with uniform pre-planting treatment applied across both upland and flooded rice cultivation systems. The experimental design employed fifteen pots randomly allocated into three biological replicates of five pots each. Initial soil chemical properties were characterized as follows: pH 6.07, organic matter 13.72 g kg⁻¹, alkali-hydrolyzable nitrogen 57.63 mg kg⁻¹, available phosphorus 3.28 mg kg⁻¹, and available potassium 133.88 mg kg⁻¹. Baseline soil enzyme activities were measured as urease 65.96 U g⁻¹, invertase 7.81 U g⁻¹, and catalase 15.47 U g⁻¹.

2.2. Experimental Materials

Potato cultivars Dianshu 23 and Dianshu 1418 were supplied by the Tuber Crop Research Institute of Yunnan Agricultural University. Rice variety Dianhe You 615 was provided by the Rice Research Institute of Yunnan Agricultural University.

2.3. Experimental Design

The experiment comprised nine treatments: pre-planting control soil under upland and flooded rice systems (CK), monoculture potato cultivars Dianshu 23 (R) and Dianshu 1418 (Y), upland rice followed by potato cultivar Dianshu 23 (DR) or Dianshu 1418 (DY), flooded rice followed by potato cultivar Dianshu 23 (WR) or Dianshu 1418 (WY), and aggregate upland rice–potato rotations (D) and flooded rice–potato rotations (W) representing combined data from both potato cultivars within each water management system. The experiment utilized pot cultivation with containers measuring 30 cm in height and 40 cm in diameter. These dimensions were selected to replicate field conditions, with the 30 cm depth providing adequate rooting volume equivalent to a standard tillage layer. Fertilizer quantities for individual pots were calculated proportionally based on the pot surface area, using field application rates as the reference standard. Prior to rice sowing, basal fertilizer was applied and thoroughly incorporated into the soil at the following rates: commercial organic fertilizer at 6000 kg ha⁻¹, compound fertilizer (N:P₂O₅:K₂O = 20:5:15) at 600 kg ha⁻¹, calcium fertilizer at 600 kg ha⁻¹, and magnesium fertilizer at 600 kg ha⁻¹. Rice was established through direct seeding in both upland and flooded systems, with 36 pots allocated to each treatment. Following sowing, soil moisture was maintained to promote germination. Seedlings were thinned 40 days post-emergence to retain four vigorous plants per pot. Flooded rice treatments received continuous irrigation to maintain saturated soil conditions throughout the growing season. Following rice harvest, 18 pots from each rice water management system were planted with potato cultivar Dianshu 23, and an additional 18 pots with cultivar Dianshu 1418. Prior to potato planting, basal fertilizer was applied and incorporated at the following rates: commercial organic fertilizer at 4500 kg ha⁻¹ and compound fertilizer (N:P₂O₅:K₂O = 20:5:15) at 1200 kg ha⁻¹. Each pot received one virus-free pre-basic seed tuber weighing approximately 50 g.

2.4. Experimental Methodology

2.4.1. Soil Sample Collection Protocol

On 16 March 2024, basal fertilizers were incorporated into the soil, baseline soil samples were collected to establish pre-planting controls (CK), and rice was sown. Rice was harvested on 2 September 2024, with post-harvest soil samples collected immediately following crop removal. On 4 September 2024, basal fertilizers were incorporated, pre-planting soil samples were collected, and potatoes were planted. Subsequently, soil samples were collected from the 15–20 cm depth at 40-day intervals throughout the potato growing season (40, 80, and 120 days after planting). These soil samples enabled the assessment of physicochemical properties and evaluation of how antecedent rice cultivation systems influenced bacterial and fungal community composition. For each treatment, soil samples were collected from fifteen randomly selected pots, immediately placed in sterile bags, and transported to the laboratory on ice to preserve microbial community integrity. Five samples were homogenized and subsequently divided into two equal portions using the quartering method to constitute one biological replicate. This procedure was repeated three times, yielding three independent biological replicates per treatment. One portion was stored at −80 °C for subsequent extraction of total soil DNA and measurement of soil enzyme activities, while the other portion was air-dried at room temperature under ventilated conditions for determination of soil physicochemical properties.

2.4.2. Analytical Procedures for Soil Properties

Physicochemical analysis utilized standard protocols following sample preparation through 2.000 mm and 0.149 mm sieves. Measurements included pH determination via digital pH meter, available nitrogen quantification through alkaline diffusion methodology, available phosphorus assessment using sodium bicarbonate extraction with molybdenum-antimony-scandium colorimetry, available potassium determination via ammonium acetate extraction and flame photometry, and organic matter content analysis through low-temperature potassium dichromate oxidation colorimetry.

Enzyme activity analysis employed samples dried at 37 °C and processed through 0.600 mm sieves. Urease (S-UE), sucrase (S-SC), and catalase (S-CAT) activities were quantified using standardized detection kits from Beijing Boxbio Science & Technology Co., Ltd. (Beijing, China).

2.4.3. Molecular Analysis Procedures

Each treatment consisted of fifteen pots arranged in a randomized design with three biological replicates of five pots each. Genomic DNA was extracted from soil samples, and DNA integrity was assessed by 1% agarose gel electrophoresis. DNA concentration and purity were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The bacterial V3-V4 hypervariable region of the 16S rRNA gene was amplified using universal primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). For fungal community analysis, the ITS2 region was amplified using primers ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′). PCR reactions were performed in 20 μL volumes containing 4 μL of 5 × TransStart FastPfu buffer, 2 μL of 2.5 mM dNTPs, 0.8 μL each of forward and reverse primers (5 μM), 0.4 μL of TransStart FastPfu DNA polymerase, and 10 ng of template DNA. PCR amplification was performed under the following thermal cycling conditions: initial denaturation at 95 °C for 3 min, followed by 35 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, with a final extension at 72 °C for 10 min. Reactions were held at 4 °C upon completion using an ABI GeneAmp^® 9700 thermal cycler (Thermo Fisher Scientific (China) Co., Ltd., Shanghai, China). PCR products were purified using the AxyPrep DNA Gel Recovery Kit (Axygen Biosciences, Union City, CA, USA) following gel excision and eluted with Tris-HCl buffer. Purified products were verified by 2% agarose gel electrophoresis and quantified using the QuantiFluor™-ST fluorometric quantification system (Promega, Madison, WI, USA). Sequencing libraries were prepared from purified amplicons using the NEXTflex Rapid DNA-Seq Kit (Illumina (China) Scientific Equipment Co., Ltd., Shanghai, China) and sequenced on an Illumina NovaSeq 6000 platform (Majorbio Bio-Pharm Technology Co., Ltd., Shanghai, China). Quality-filtered and merged sequences were processed through the DADA2 plugin within the QIIME2 pipeline for denoising, generating amplicon sequence variants (ASVs). Taxonomic classification of ASVs was performed using the naive Bayes classifier implemented in QIIME2 against the Silva 16S rRNA gene database (version 138) for bacterial sequences and the UNITE database (version 9.0) for fungal sequences. Raw sequence data have been deposited in the NCBI Sequence Read Archive under BioProject accession number PRJNA1346805.

2.5. Statistical Analysis Framework

Data compilation and organization were performed using Microsoft Excel 2021. Statistical comparisons between treatment groups were conducted using Student’s t-test in DPS (version 9.01) software. Microbial community data were analyzed using the Majorbio Cloud Platform (https://cloud.majorbio.com). Alpha diversity indices, including Chao1 and Shannon index, were calculated using mothur (version 1.30) software (http://www.mothur.org/wiki/Calculators) URL (accessed on 15 June 2025). Between-group differences in alpha diversity were assessed using the Wilcoxon rank-sum test. Beta diversity was assessed through principal coordinate analysis (PCoA) based on the Bray–Curtis dissimilarity matrix. Statistical significance of compositional differences between treatment groups was evaluated using permutational multivariate analysis of variance (PERMANOVA). Differentially abundant taxa from phylum to genus level were identified using linear discriminant analysis effect size (LEfSe) with thresholds of LDA score > 4 and p < 0.05. Co-occurrence network analysis was constructed based on Spearman correlation coefficients with selection criteria of |r| > 0.6 and p < 0.05. All statistical analyses and data visualizations were performed using R version 3.3.1 and Python version 2.7.

3. Results

3.1. Effects of Different Rice Cultivation Methods on Soil Nutrients and Enzyme Activity for Subsequent Potato Crops

As shown in

Table 1. Effects of different rice cultivation methods on soil nutrients for subsequent potato crops.

Sample Time (d)	Treatment	pH	SOM (g/kg)	AN (mg/kg)	AP (mg/kg)	AK (mg/kg)
Before rice planting	CK	6.07 ± 0.02	13.72 ± 1.18	57.63 ± 8.47	3.28 ± 0.41	133.88 ± 4.08
Rice harvest	D	6.02 ± 0.02	22.04 ± 0.61 **	61.83 ± 12.72	3.13 ± 0.30	111.97 ± 2.33 **
	W	6.24 ± 0.04 **	16.42 ± 0.51	52.27 ± 2.65	2.68 ± 0.30	71.19 ± 2.65
Before planting potatoes	D0	5.91 ± 0.02	22.32 ± 1.41	278.83 ± 5.35 *	5.74 ± 0.78	154.64 ± 3.95
	W0	6.09 ± 0.05 **	24.17 ± 0.97	258.30 ± 6.42	6.59 ± 0.59	176.94 ± 2.17 **
40	D1	5.25 ± 0.09	19.91 ± 2.03	65.57 ± 6.18	4.01 ± 0.38	159.69 ± 6.37 **
	W1	5.40 ± 0.07 **	19.03 ± 0.64	63.47 ± 4.55	4.41 ± 0.15 *	143.49 ± 3.49
80	D2	5.69 ± 0.03	21.01 ± 0.59 **	32.67 ± 2.15	5.19 ± 0.34 *	122.93 ± 3.42 *
	W2	5.70 ± 0.02	18.47 ± 0.75	30.92 ± 1.50	4.37 ± 0.54	112.58 ± 9.24
Potato harvest	D3	5.78 ± 0.21	25.60 ± 0.72	33.25 ± 4.44	3.86 ± 0.48	117.89 ± 8.82
	W3	5.76 ± 0.09	25.12 ± 0.71	29.40 ± 5.05	5.06 ± 0.21 **	120.05 ± 6.07

Note: Data are presented as mean ± standard deviation. Statistical significance levels are denoted as * p ≤ 0.05 and ** p ≤ 0.01. Within each sampling period, asterisks indicate significant differences between treatments. This notation system applies throughout.

, rice cultivation induced notable changes in soil chemical properties. Soil pH increased under flooded rice management, while both cultivation systems enhanced soil organic matter content, with upland rice demonstrating substantially greater accumulation than flooded rice. However, post-harvest soil concentrations of alkali-hydrolyzable nitrogen, available phosphorus, and available potassium were consistently lower in flooded rice systems compared to upland rice systems.

Following potato establishment, soil pH and organic matter (SOM) exhibited a biphasic pattern characterized by initial decline followed by recovery during crop development. Available phosphorus (AP) demonstrated a consistent downward trend, whereas available nitrogen (AN) and potassium (AK) displayed progressive depletion throughout the growing period. Soil pH reached minimum values at 40 days post-planting, with rainfed treatments (D) exhibiting lower values than waterlogged treatments (W) initially, before converging during subsequent growth phases. Soil organic matter concentrations remained consistently elevated in rainfed systems compared to waterlogged systems, while available phosphorus exhibited the inverse relationship. Available nitrogen and potassium showed substantial early to mid-season decline, with rainfed treatments experiencing less severe depletion than waterlogged treatments. Both nutrients subsequently stabilized at approximately 30 and 120 mg/kg, respectively. Comparison of soil nutrients between rice harvest and potato harvest revealed decreased pH and available nitrogen, while soil organic matter, available phosphorus, and available potassium increased to varying degrees. Waterlogged treatments demonstrated particularly notable improvements in available phosphorus.

These findings indicate that potato cultivation accelerated soil nutrient mobilization throughout crop development. Rainfed rice–potato systems demonstrated enhanced release rates of available nitrogen and potassium during early to mid-season growth compared to waterlogged rice–potato systems. Furthermore, the cropping sequence exhibited nutrient conservation properties that support subsequent crop establishment and development.

Table 2. Effects of different rice cultivation methods on enzyme activity for subsequent potato crops.

Sample Time (d)	Treatment	S-UE (U/g)	S-SC (U/g)	S-CAT (U/g)
Before rice planting	CK	65.96 ± 18.71	7.81 ± 0.37	15.47 ± 0.45
Before planting potatoes	D0	128.69 ± 6.90	35.49 ± 0.18	17.31 ± 1.05
	W0	134.98 ± 18.02	51.04 ± 4.31 *	17.66 ± 0.45
40	D1	97.15 ± 42.23	36.62 ± 14.58 *	15.87 ± 0.77
	W1	87.59 ± 15.37	20.29 ± 4.63	16.51 ± 0.75
80	D2	146.29 ± 14.51	25.36 ± 6.66 **	17.28 ± 0.83
	W2	136.54 ± 8.64	10.90 ± 2.28	17.38 ± 0.80
Potato harvest	D3	152.07 ± 16.69 **	53.50 ± 35.77	17.97 ± 0.50
	W3	112.92 ± 20.65	20.98 ± 6.61	17.86 ± 0.57

Note: Statistical significance levels are denoted as * p ≤ 0.05 and ** p ≤ 0.01. Within each sampling period, asterisks indicate significant differences between treatments.

demonstrates that rice cultivation substantially enhanced soil enzyme activities, including urease (S-UE), sucrase (S-SC), and catalase (S-CAT), across both cultivation methods. Waterlogged rice systems exhibited greater enzyme activity enhancement compared to rainfed systems. These results suggest that waterlogged rice cultivation provides superior conditions for enhancing soil urease and sucrase activities, potentially creating more favorable soil conditions for subsequent potato establishment.

Following potato establishment, soil urease activity exhibited contrasting patterns between treatments: rainfed systems (D) demonstrated overall increases while waterlogged systems (W) showed declining activity. Both treatments initially experienced substantial decreases during early growth phases, followed by marked increases during mid to late-season development. Soil sucrase activity displayed distinct temporal patterns: rainfed systems maintained stable levels during early to mid-season growth before increasing substantially in late season, whereas waterlogged systems exhibited consistent decline throughout the growing period. Soil catalase activity demonstrated similar biphasic patterns in both treatments, characterized by initial decline followed by recovery, with comparable activity levels between systems.

Comparative analysis between pre-planting and harvest stages revealed that soil urease and sucrase activities increased substantially in rainfed systems while decreasing markedly in waterlogged systems, resulting in significantly higher enzyme activities in rainfed treatments at harvest. Soil catalase activity showed modest increases in both systems with similar final levels. These findings demonstrate that potato cultivation under rainfed rice–potato systems enhances soil urease and sucrase activities throughout the growing season, while waterlogged rice–potato systems experience declining enzyme activities. Consequently, rainfed rice–potato cropping systems create more favorable soil biochemical conditions for subsequent crop production.

3.2. Effects of Different Rice Cultivation Methods on Soil Microbial Community Diversity in Subsequent Potato Crops

3.2.1. Soil Microbial Alpha-Diversity Analysis in Rainfed and Waterlogged Rice–Potato Cropping Systems

Illumina high-throughput sequencing technology was employed to assess the alpha-diversity parameters of soil bacterial and fungal community structures.

Table 3. Effects of different rice cultivation systems on soil bacterial and fungal α diversity in subsequent potato crops of different varieties.

Type	Sample Time (d)	Treatment	Sobs Value	Shannon Index	Chao Index	Coverage
Bacteria	Before rice planting	CK	1766.00 ± 69.20	6.49 ± 0.11	1776.92 ± 73.37	99.89%
	Before planting potatoes	D0	1581.67 ± 39.55 *	5.83 ± 0.26	1640.12 ± 83.18 *	99.70%
		W0	1425.00 ± 50.48	6.07 ± 0.19	1437.22 ± 52.51	99.89%
	40	DR1	1369.67 ± 300.85 a	5.66 ± 0.44 a	1398.96 ± 314.65 a	99.83%
		WR1	1387.67 ± 166.16 a	6.01 ± 0.19 a	1397.57 ± 172.78 a	99.90%
		DY1	1553.00 ± 172.27 a	6.13 ± 0.19 a	1568.74 ± 172.65 a	99.87%
		WY1	1254.00 ± 292.80 a	5.68 ± 0.36 a	1274.11 ± 303.34 a	99.86%
	80	DR2	1565.33 ± 76.27 a **	6.09 ± 0.22 a	1591.99 ± 75.61 a **	99.83%
		WR2	1254.00 ± 21.93 ab	5.83 ± 0.07 a	1259.72 ± 22.50 b	99.93%
		DY2	1418.67 ± 171.2 ab	5.94 ± 0.40 a	1431.34 ± 168.88 ab	99.89%
		WY2	1093.67 ± 272.89 b	5.63 ± 0.62 a	1102.77 ± 279.01 b	99.92%
	Potato harvest	DR3	1224.67 ± 196.17 a	5.84 ± 0.37 a	1235.14 ± 202.87 a	99.91%
		WR3	1062.00 ± 177.39 a	5.77 ± 0.41 a	1063.68 ± 177.45 a	99.97%
		DY3	1341.00 ± 156.14 a	6.07 ± 0.21 a	1354.65 ± 168.70 a	99.88%
		WY3	1105.67 ± 165.08 a	5.74 ± 0.41 a	1115.33 ± 162.32 a	99.91%
Fungi	Before rice planting	CK	479.00 ± 35.00	4.70 ± 0.33	479.43 ± 34.62	99.99%
	Before planting potatoes	D0	438.33 ± 148.19	4.03 ± 0.69	445.47 ± 144.87	99.95%
		W0	345.33 ± 32.72	4.43 ± 0.39	345.44 ± 32.84	100.00%
	40	DR1	464.33 ± 59.34 a	4.08 ± 0.60 ab	469.19 ± 61.29 a	99.96%
		WR1	370.00 ± 115.53 ab	4.68 ± 0.32 a	370.25 ± 115.90 ab	100.00%
		DY1	335.67 ± 36.86 ab	4.06 ± 0.90 ab	335.98 ± 37.03 ab	99.99%
		WY1	240.00 ± 94.25 b	3.31 ± 0.39 b	242.59 ± 94.17 b	99.98%
	80	DR2	488.00 ± 85.56 a *	3.98 ± 0.53 ab	492.96 ± 87.57 a *	99.95%
		WR2	344.67 ± 12.66 b	3.64 ± 0.32 b	347.97 ± 12.73 b	99.98%
		DY2	465.33 ± 20.40 a *	4.46 ± 0.24 a	469.55 ± 22.77 a *	99.97%
		WY2	350.67 ± 46.92 b	4.02 ± 0.38 ab	352.02 ± 46.32 b	99.98%
	Potato harvest	DR3	407.67 ± 29.02 a	3.99 ± 0.05 a	420.33 ± 40.13 a	99.93%
		WR3	343.67 ± 106.52 a	3.41 ± 0.69 a	347.42 ± 106.86 a	99.97%
		DY3	436.67 ± 39.43 a	4.18 ± 0.29 a	438.77 ± 40.68 a	99.97%
		WY3	309.67 ± 71.04 a	3.66 ± 0.37 a	315.08 ± 74.77 a	99.96%

Note: For multi-group comparisons, the Kruskal–Wallis rank-sum test was employed with false discovery rate (FDR) correction for multiple testing. Post hoc pairwise comparisons were conducted using the Tukey–Kramer test at a 95% confidence level. Within each sampling period, treatments sharing different lowercase letters differ significantly (p < 0.05). Two-group comparisons were analyzed using the two-tailed Student’s t-test. Statistical significance is indicated by asterisks: * p ≤ 0.05 and ** p ≤ 0.01. Within each variety and sampling period, asterisks denote significant differences between treatments.

demonstrates that all Coverage indices exceeded 99%, confirming adequate sequencing depth to accurately represent soil bacterial and fungal community compositions.

Bacterial community analysis revealed that rice cultivation reduced all diversity indices, including Sobs values, Shannon index, and Chao index. Waterlogged rice systems exhibited more pronounced reductions in Sobs values and Chao index compared to rainfed systems, whereas Shannon index displayed the inverse relationship. Following potato establishment, both Sobs values and Chao index demonstrated progressive decline throughout crop development. Shannon index diverged between treatments, with rainfed systems showing moderate increases while waterlogged systems continued declining. At harvest, rainfed systems consistently demonstrated superior diversity indices compared to waterlogged systems across all measured parameters.

Fungal community analysis revealed that Sobs values and Chao index followed patterns consistent with bacterial communities, while Shannon index showed similar trends with the exception of the rainfed rice–potato treatment. These findings demonstrate that rainfed rice cultivation maintained soil microbial community stability more effectively than waterlogged cultivation, exhibiting reduced rates of diversity loss and more stable community structure dynamics. During the potato growing season, both microbial community richness and the number of observed taxa exhibited progressive decline. However, treatment-specific responses varied considerably. Within the upland rice rotation system, microbial community diversity increased in the DY treatment while decreasing in the DR treatment. In contrast, the flooded rice rotation system showed consistent diversity reduction across all potato phases.

3.2.2. Soil Microbial Beta-Diversity Analysis in Rainfed and Waterlogged Rice–Potato Cropping Systems

Principal coordinate analysis (PCoA) of microbial community composition at the amplicon sequence variant (ASV) level revealed distinct clustering patterns for bacterial communities. Figure 1c demonstrates clear spatial separation between the control (CK), rainfed rice–potato systems (D treatment), and waterlogged rice–potato systems (W treatment). Rainfed systems clustered on the right side of the PC1 axis, while waterlogged systems positioned on the left side, both distributed along the PC2 axis.

Figure 1a demonstrates that different potato varieties within each treatment system clustered closely or overlapped during crop development. This pattern indicates that the rice cultivation method represents the primary driver of bacterial community structure variation, while potato variety effects demonstrate common crop characteristics across treatments.

Fungal community analysis presented in Figure 1d revealed comparable spatial separation between control, rainfed, and waterlogged treatment systems. Waterlogged systems occupied the upper and middle regions of PC1 axis while positioning predominantly on the right side of the PC2 axis. Conversely, rainfed systems clustered in the lower PC1 region. Throughout potato development, both waterlogged and rainfed systems demonstrated temporal progression from left to right along the PC2 axis.

Figure 1b analysis reveals that rainfed system samples (D0, DR1, DY1) exhibited clustering overlap, while waterlogged system samples (WR3, WY3) showed similar convergence patterns. These findings confirm that the rice cultivation environment represents the dominant factor influencing fungal community structure differentiation. Additionally, the analysis demonstrates significant correlations with potato variety selection and developmental timing, indicating that both agricultural management decisions and temporal factors contribute to microbial community dynamics in these cropping systems.

3.2.3. Similarity Analysis of Soil Microbial Communities in Dryland/Irrigated Rice–Potato Double Cropping Systems

Cross-treatment analysis identified 158 shared bacterial amplicon sequence variants (ASVs) and 50 shared fungal ASVs across all experimental conditions.

Bacterial community analysis presented in Figure 2a demonstrates that initial rainfed treatment (D0) showed a 10.01% reduction compared to control, while initial waterlogged treatment (W0) exhibited a substantially greater 43.44% decline. Throughout potato development, all treatments experienced continued bacterial ASV decline. Rainfed systems showed reductions of 50.93% (DR3) and 35.08% (DY3) relative to baseline, while waterlogged systems demonstrated decreases of 38.95% (WR3) and 23.69% (WY3) from their respective starting points. These findings indicate that rainfed rice–potato systems experienced more pronounced bacterial diversity loss during potato cultivation compared to waterlogged rice–potato systems.

Fungal community analysis in Figure 2b reveals that rainfed treatment (D0) decreased by 36.85% compared to control, while waterlogged treatment (W0) showed a more substantial 56.77% reduction. During potato development, most treatments demonstrated continued fungal ASV decline, with a notable exception of the waterlogged rice–potato treatment WR3, which increased by 17.05%. Rainfed systems showed decreases of 25.87% (DR3) and 7.89% (DY3), while waterlogged system WY3 declined by 13.82%.

These findings demonstrate that flooded rice cultivation resulted in substantially greater reductions in both bacterial and fungal ASV richness compared to upland rice systems. During the subsequent potato cropping phase, bacterial ASV richness declined across all rotation systems, with the magnitude of reduction following the order DR3 > WR3 > DY3 > WY3, indicating that upland rice–potato rotations (DR3, DY3) generally exhibited more pronounced bacterial community contraction than flooded rice–potato systems (WR3, WY3). The waterlogged rice–potato system WR3 represents the only treatment showing increased fungal diversity, while rainfed system DR3 exhibits the most pronounced fungal diversity decline at 25.87%.

3.2.4. Soil Microbial Community Composition Analysis at Phylum Level

Bacterial Community Structure

Figure 3a demonstrates that dominant bacterial phyla (relative abundance > 1%) across all treatments comprised Pseudomonadota (21.76–42.63%), Chloroflexota (13.84–28.20%), Actinomycetota (6.40–14.23%), Acidobacteriota (3.70–15.82%), and Bacteroidota (1.53–9.70%).

Rice cultivation resulted in an increased relative abundance of Pseudomonadota and Bacteroidota compared to control conditions, while Chloroflexota, Actinomycetota, and Acidobacteriota exhibited decreased abundance. Comparative analysis between cultivation methods reveals that rainfed systems maintained similar Pseudomonadota abundance to waterlogged systems. However, rainfed cultivation demonstrated reduced Chloroflexota and Acidobacteriota abundance, while exhibiting enhanced Actinomycetota and Bacteroidota populations.

Throughout potato development, both rainfed and waterlogged systems demonstrated increasing Pseudomonadota and Bacteroidota abundance. Chloroflexota and Acidobacteriota exhibited divergent patterns: rainfed systems showed increasing trends while waterlogged systems declined. Actinomycetota displayed the inverse relationship. Thermodesulfobacteria abundance declined consistently across treatments, with waterlogged systems maintaining substantially higher levels than rainfed systems. Cyanobacteria demonstrated biphasic dynamics, initially declining before recovering, with rainfed systems achieving markedly higher abundance than waterlogged systems.

At harvest, rainfed systems exhibited higher Actinomycetota abundance compared to waterlogged systems, while Chloroflexota, Acidobacteriota, and Bacteroidota abundance remained lower in rainfed treatments.

Fungal Community Structure

Figure 3b analysis identifies the predominant fungal phyla (relative abundance > 1%) as Ascomycota (38.12–76.14%), Basidiomycota (4.96–17.46%), Mortierellomycota (3.56–10.09%), and Chytridiomycota (1.03–14.72%). Ascomycota maintained dominant status across all developmental stages.

Rice cultivation produced contrasting effects on fungal communities. Ascomycota and Mortierellomycota abundance declined compared to control conditions, with waterlogged systems showing greater Ascomycota reduction and rainfed systems experiencing more substantial Mortierellomycota decline. Basidiomycota abundance decreased under rainfed cultivation but increased under waterlogged conditions, while Chytridiomycota responded inversely.

During potato cultivation, Ascomycota abundance increased across both systems, with waterlogged treatments demonstrating greater enhancement. Basidiomycota and Mortierellomycota abundance increased in rainfed systems while declining in waterlogged systems. Chytridiomycota abundance decreased substantially in both treatment types.

Final harvest analysis revealed that waterlogged systems maintained higher Ascomycota and Basidiomycota abundance compared to rainfed systems, while rainfed systems demonstrated superior Mortierellomycota and Chytridiomycota abundance.

3.2.5. Soil Microbial Community Analysis at Genus Level

Bacterial Community Composition

Figure 4a demonstrates that predominant bacterial genera comprised Sphingomonas (2.79–7.95%), norank_c__AD3 (1.82–10.87%), Ammoniphilus (0.59–7.80%), norank_f__Chitinophagaceae (1.56–4.71%), and HSB_OF53-F07 (1.29–5.78%).

Rice cultivation produced differential effects on bacterial genera relative to control conditions. Sphingomonas, Ammoniphilus, and norank_f__Chitinophagaceae demonstrated increased abundance, with rainfed systems (D0) exhibiting substantially greater enhancement of norank_f__Chitinophagaceae compared to waterlogged systems (W0). Conversely, norank_c__AD3 and HSB_OF53-F07 abundance declined significantly, with rainfed systems showing more pronounced reductions.

Throughout potato development, rainfed systems demonstrated biphasic patterns for Sphingomonas and norank_c__AD3 abundance, characterized by initial increases followed by decline, resulting in net positive changes. Waterlogged systems exhibited consistent declining trends for these genera. Ammoniphilus abundance increased in rainfed systems while decreasing in waterlogged systems. Both Rhodanobacter and Nitrosospira demonstrated consistent abundance increases across all treatments.

At harvest, rainfed systems maintained superior abundance levels for Sphingomonas, Ammoniphilus, norank_c__AD3, Rhodanobacter, and Nitrosospira compared to waterlogged systems, indicating enhanced bacterial diversity under rainfed cultivation conditions.

Fungal Community Composition

Figure 4b analysis reveals that dominant fungal genera included Cladosporium (2.56–14.49%), Albifimbria (0.58–12.02%), Mortierella (1.70–8.42%), and Fusarium (2.17–6.06%).

Rice cultivation effects varied among fungal genera. Cladosporium and Albifimbria abundance increased compared to control conditions, with rainfed systems demonstrating greater enhancement than waterlogged systems. Mortierella responded differently between cultivation methods, decreasing under rainfed conditions while increasing under waterlogged cultivation. Fusarium abundance declined in both systems, with waterlogged treatments showing more substantial reductions.

During potato cultivation, Cladosporium and Albifimbria abundance increased in rainfed systems while declining in waterlogged systems. Mortierella exhibited similar divergent patterns, increasing under rainfed cultivation and decreasing under waterlogged conditions. Both genera demonstrated biphasic temporal dynamics throughout the growing season. Fusarium abundance increased consistently across both cultivation systems.

Late-season fungal dynamics revealed notable patterns. Botrytis and Plectosphaerella demonstrated substantial abundance increases during late-season sampling periods (80 and 120 days post-planting). Cladosporium and Albifimbria abundance increased dramatically at 80 days post-planting across treatments.

Final harvest analysis confirmed that rainfed systems supported the higher abundance of Cladosporium, Albifimbria, and Fusarium compared to waterlogged systems, while waterlogged systems maintained superior Mortierella populations. These results indicate that cultivation method significantly influences fungal community composition and temporal dynamics throughout the cropping cycle.

3.2.6. Analysis of Differences in Soil Microbial Community Species Composition

Linear discriminant analysis (LDA) was conducted to identify taxonomically distinct microbial species across hierarchical levels from phylum to genus for bacterial and fungal communities in treatment groups D0, W0, D3, and W3, employing an LDA effect size threshold of 4.0.

Bacterial Community Differentiation

Bacterial community analysis revealed differential species enrichment patterns across treatment groups (Figure 5a,c), with D0, W0, D3, and W3 treatments exhibiting 10, 13, 3, and 9 significantly enriched taxonomic groups, respectively. Taxonomic overlap analysis identified two shared bacterial groups between D3 and D0 treatments (c_Cyanobacteriia and p_Cyanobacteriota), while W3 and W0 treatments shared five bacterial groups (o_Geobacterales, f_Geobacteraceae, c_Desulfuromonadia, p_Thermodesulfobacteriota, and f_Rhodocyclaceae).

These findings demonstrate that waterlogged rice cultivation produced marginally higher bacterial species differentiation compared to rainfed cultivation systems. Both cultivation methods generated substantial taxonomic variation, indicating that hydrological conditions exert pronounced influence on bacterial community structure. During the potato cultivation phase, waterlogged rice–potato systems exhibited substantially greater bacterial species enrichment and inter-treatment differentiation compared to rainfed rice–potato systems, while maintaining stable dominant species composition.

Fungal Community Differentiation

Fungal community analysis revealed distinct enrichment patterns, with treatments (Figure 5b,d) D0, W0, D3, and W3 demonstrating 4, 9, 11, and 2 significantly differentiated taxonomic groups, respectively. Comparative analysis identified two shared fungal taxa between D3 and D0 treatments (o_Chaetothyriales and f_Herpotrichiellaceae), whereas W3 and W0 treatments exhibited no overlapping species meeting the established LDA threshold criteria.

The analysis indicates that waterlogged rice cultivation generated substantially higher fungal species differentiation compared to rainfed systems during the initial cultivation phase. However, the pattern reversed during potato cultivation, with rainfed rice–potato systems demonstrating markedly greater fungal species enrichment and differentiation than waterlogged rice–potato systems. This temporal shift suggests that fungal communities respond dynamically to changing cultivation conditions, while maintaining relatively stable dominant species populations throughout the cropping sequence.

3.3. Correlation Analysis of Soil Physicochemical Properties and Microbial Communities Across Rice Cultivation Systems and Potato Varieties

Spearman correlation analysis was conducted separately for rainfed (D) and waterlogged (W) treatments to examine relationships between the twenty most abundant soil bacterial and fungal genera and environmental variables, with subsequent average clustering analysis.

3.3.1. Rainfed System Bacterial Correlations

Bacterial community analysis for rainfed systems (Figure 6a) revealed distinct correlation patterns with soil parameters. Pseudarthrobacter (Actinomycetota) and Lysobacter (Pseudomonadota) demonstrated significant positive correlations with available phosphorus, whereas Bradyrhizobium (Pseudomonadota) and Gemmatimonas (Gemmatimonadota) exhibited significant negative correlations with this nutrient parameter.

Additional correlations included positive relationships between Devosia (Pseudomonadota) and soil pH, while norank_o_Chloroplast (Cyanobacteriota) showed significant to highly significant positive correlations with pH, soil organic matter, urease activity, and catalase activity, coupled with significant negative correlation with available potassium. These findings indicate that rainfed cultivation systems exhibit strong bacterial community responses to key soil parameters, including pH, soil organic matter, available phosphorus, available potassium, and urease activity.

3.3.2. Waterlogged System Bacterial Correlations

Waterlogged system (Figure 6b) analysis demonstrated that major bacterial phyla, including Pseudomonadota, Bacillota, and Bacteroidota, predominantly exhibited negative correlations with environmental parameters. Notably, Bradyrhizobium and Nitrosospira (Pseudomonadota) and Ammoniphilus (Bacillota) achieved statistically significant correlation coefficients.

Contrasting patterns emerged with positive correlations, including highly significant relationships between Dechloromonas (Pseudomonadota) and both available nitrogen and available potassium, as well as a significant positive correlation between Pseudarthrobacter (Actinomycetota) and soil organic matter. The analysis demonstrates that waterlogged cultivation systems show pronounced bacterial community sensitivity to pH, available nitrogen, and available potassium parameters.

The comparative analysis reveals fundamental differences in microbial–environmental interactions between cultivation systems, with rainfed systems demonstrating greater sensitivity to phosphorus availability and enzyme activities, while waterlogged systems exhibit stronger responses to nitrogen and potassium dynamics. These distinct correlation patterns reflect the underlying differences in soil biochemical processes between the two cultivation approaches.

3.3.3. Rainfed System Fungal Correlations

Fungal community correlation analysis for rainfed systems (Figure 7a) revealed that Ascomycota predominantly exhibited negative correlations with available nitrogen and available potassium while demonstrating positive correlations with urease and catalase activities. Several genera within this phylum achieved statistical significance, including Poaceascoma, Fusidium, Botrytis, and Plectosphaerella. Additionally, Acremonium and Cladosporium demonstrated highly significant positive correlations with available phosphorus.

Within Mortierellomycota, the genus Mortierella exhibited significant negative correlations with available nitrogen, phosphorus, and potassium. These results demonstrate that rainfed cultivation systems show pronounced fungal community sensitivity to nutrient availability and soil enzyme activities.

3.3.4. Waterlogged System Fungal Correlations

Waterlogged system (Figure 7b) analysis revealed that unclassified fungal taxa, Mortierellomycota, and Basidiomycota predominantly exhibited positive correlations with environmental parameters. Notably, Rhizoctonia within Basidiomycota demonstrated significant to highly significant positive correlations with pH, soil organic matter, and available phosphorus.

Conversely, Ascomycota demonstrated predominantly negative correlations with available nitrogen, phosphorus, potassium, and sucrase activity. Multiple genera within this phylum, including Botrytis, Plectosphaerella, Gibellulopsis, and Verticillium, achieved statistically significant negative correlation coefficients. Furthermore, Spizellomyces within Chytridiomycota exhibited significant to highly significant negative correlations with pH, available phosphorus, and catalase activity.

The comprehensive analysis indicates that waterlogged cultivation systems demonstrate strong fungal community responsiveness to pH, soil organic matter, available phosphorus, available potassium, and urease activity. The contrasting correlation patterns between rainfed and waterlogged systems reflect fundamental differences in fungal community–environment interactions, with each cultivation method creating distinct selective pressures that shape microbial community structure and function.

4. Discussion

4.1. Effects of Rice Cultivation Systems on Soil Physicochemical Properties in Rice–Potato Cropping Sequences

Paddy soils constitute distinctive agroecological systems characterized by their capacity to regulate redox-driven processes that govern soil organic matter dynamics and nutrient cycling [10,11]. Rice–upland crop rotations provide multiple agronomic benefits, including reduced pest and disease pressure, enhanced land use efficiency, and improved grain productivity [12,13]. Our experimental findings demonstrate that both cultivation methods enhanced soil nutrient mobilization, creating favorable conditions for subsequent potato development.

Previous research demonstrates the long-term benefits of crop rotation systems. Seven-year continuous rice–rapeseed rotations improved soil quality through increased organic matter content and enhanced sucrase and urease activities while moderating the decline rates of soil pH, available nitrogen, and available phosphorus [14]. Our results align with these findings, showing that subsequent potato cultivation decelerated the decline rates of soil pH, available nitrogen, phosphorus, and potassium while promoting soil organic matter accumulation. These observations corroborate similar results reported by Meng et al. [15].

During the rice cultivation phase, flooded rice systems exhibited lower concentrations of alkali-hydrolyzable nitrogen (AN), available phosphorus (AP), and available potassium (AK) compared to upland rice systems. However, soil enzyme activities, including urease, invertase, and catalase, demonstrated significantly higher levels under flooded conditions. During the subsequent potato cropping phase, nutrient depletion occurred more rapidly in the flooded rice–potato rotation than in the upland rice–potato rotation throughout early and mid-growth stages. Conversely, soil enzyme activities remained higher in the upland rice–potato system compared to the flooded rice–potato rotation. Biomass production patterns corroborated these nutrient dynamics. Aboveground biomass under flooded rice cultivation reached 1.71-fold that of upland rice, demonstrating superior growth performance under saturated conditions and correspondingly greater nutrient uptake. In the potato cropping phase, the flooded rice–potato rotation generated 0.91-fold the aboveground biomass, 1.27-fold the fresh tuber weight, and 1.14-fold the total biomass (combining aboveground and tuber components) relative to the upland rice–potato rotation. Although flooded rice cultivation resulted in greater nutrient depletion than upland rice during the initial phase, the subsequent potato crop in the flooded rice–potato rotation system exhibited enhanced overall biomass production compared to the upland rice–potato system. This differential performance may reflect the legacy effects of soil conditions established during the antecedent flooded rice phase.

Harvest-time soil analysis confirmed the sustainability benefits of both cropping systems. Compared to pre-planting conditions, soil pH, organic matter, available phosphorus, and available potassium increased across all treatments. Soil urease and sucrase activities increased in rainfed rice–potato systems while decreasing in waterlogged rice–potato systems relative to baseline conditions. These results demonstrate that both rice cultivation systems can effectively maintain soil fertility when integrated with potato production, providing viable pathways for sustainable agricultural intensification.

4.2. Microbial Community Dynamics in Rice–Potato Cropping Systems

Previous studies have demonstrated that prolonged monoculture significantly alters soil microbial community structure, with continuous cropping reducing fungal diversity and richness [16]. However, under crop rotation systems, bacterial richness and diversity decreased during the crop maturation phase. Our experimental results confirm this pattern, showing decreased microbial ASV numbers following both rice cultivation methods, with continued decline throughout subsequent potato development.

Multiple studies support the resource competition theory, which predicts maximum microbial diversity under moderate resource limitation and reduced diversity when environmental resources become severely constrained [17,18]. Soil moisture represents a critical factor in microbial community development. Irrigation creates anaerobic environments that promote rapid anaerobic microbial proliferation while altering existing community structures. Under aerobic conditions, organisms such as methane-oxidizing bacteria become dominant [19,20].

The biochemical basis for these patterns involves enzymatic adaptations. Aerobic microorganisms have evolved protective enzymes including peroxidase and superoxide dismutase to neutralize oxygen-containing compounds. Anaerobic microorganisms lack these metabolic capabilities and experience rapid mortality under aerobic conditions [21].

Our results demonstrate that waterlogged rice cultivation produced greater microbial ASV reductions compared to rainfed cultivation. Waterlogged rice–potato systems undergo environmental transitions from anaerobic to aerobic conditions, resulting in substantial anaerobic microbial mortality and corresponding rapid ASV decline. In contrast, rainfed rice–potato systems maintain consistent aerobic conditions with only crop species changes, producing more moderate ASV reductions.

Linear discriminant analysis with a threshold of 4.0 identified treatment-specific microbial enrichment patterns. Bacterial community analysis revealed two shared species between treatments D3 and D0, while treatments W3 and W0 shared five species. Fungal community analysis identified two shared species between treatments D3 and D0. These results indicate that during potato cultivation, rainfed systems demonstrate accelerated bacterial community turnover but reduced fungal community changes relative to waterlogged systems.

The findings demonstrate that cultivation method significantly influences microbial community stability and composition throughout the cropping sequence, with important implications for sustainable agricultural management.

4.2.1. Bacterial Community Analysis at Phylum Level

Phylum-level bacterial analysis reveals that Actinobacteria demonstrates superior enzymatic capabilities for degrading both aliphatic and aromatic compounds [22]. Within soil ecosystems, Acidobacteria contributes significantly to nutrient cycling and organic matter decomposition [23,24], with previous research establishing their roles in plant polymer degradation and the cycling of carbon, nitrogen, sulfur, and trace metals [25].

Our results demonstrate contrasting abundance patterns between cultivation systems. Actinobacteria abundance exceeded that of waterlogged systems under rainfed cultivation, while Acidobacteria exhibited the inverse relationship. These phyla likely exhibit competitive dynamics, with Actinobacteria demonstrating competitive advantages under rainfed conditions while Acidobacteria predominates in waterlogged systems.

Proteobacteria, characterized as fast-growing eutrophic bacterial groups, showed similar enhancement patterns across both cultivation methods. Previous research demonstrates that combined organic–inorganic fertilization stimulates rhizosphere Pseudomonas growth, which contributes significantly to carbon fixation and degradation processes [26]. Our fertilization approach aligns with these practices, and the similar Proteobacteria responses across treatments suggest strong environmental adaptability.

Thermodesulfobacteria abundance patterns reflect cultivation-specific environmental conditions. Research indicates positive correlations between Thermodesulfobacteria and the removal efficiency of COD, NH4⁺-N, and total phosphorus in wastewater treatment systems [27]. Waterlogged treatments supported dominant Thermodesulfobacteria populations, while rainfed treatments showed minimal abundance, indicating aquatic environment preferences and the beneficial effects of preceding rice flooding conditions.

Cyanobacteria represent the sole phylum capable of oxygenic photosynthesis, with eukaryotic phototrophy widely attributed to cyanobacterial endosymbiosis [28]. Rainfed systems demonstrated significantly higher Cyanobacteria abundance compared to waterlogged systems. These patterns suggest that optimal moisture conditions in rainfed systems promote enhanced Cyanobacteria proliferation.

4.2.2. Fungal Community Analysis at Phylum Level

Fungal decomposition processes involve temporal specialization between major phyla. Ascomycota primarily degrades labile straw components during early decomposition phases, while Basidiomycota specializes in refractory organic matter breakdown during later stages [29,30]. Rainfed systems supported significantly higher Ascomycota abundance compared to waterlogged systems, while Basidiomycota showed inverse patterns. These results suggest that anaerobic irrigation environments accelerate labile residue degradation, reducing Ascomycota community proportions.

4.2.3. Genus-Level Community Analysis

Bacterial genus analysis revealed an elevated abundance of nitrogen metabolism-associated taxa in rainfed systems, including Sphingomonas, Ammoniphilus, Rhodanobacter, and Nitrosospira. Sphingomonas provides multiple ecosystem services, including biodegradation, phytoremediation support, and plant stress tolerance enhancement [31]. Novel Rhodanobacter isolates from tomato rhizospheres demonstrate plant growth promotion capabilities [32]. The higher abundance of nitrogen-metabolizing genera in rainfed systems aligns with elevated soil available nitrogen content, indicating environmental suitability for these functional groups.

Fungal genus analysis revealed differential pathogen distributions between cultivation systems. Rainfed systems supported a higher abundance of Cladosporium, Albifimbria, and Fusarium, while Mortierella predominated in waterlogged systems. Several taxa represent agricultural concerns: Cladosporium species contribute to pear rot disease [33], Albifimbria verrucaria causes leaf spot disease [34], and Fusarium triggers potato wilt and dry rot diseases [35,36]. Conversely, Mortierella alpina represents a beneficial oleaginous fungus producing polyunsaturated fatty acids for industrial food applications [37] and bioactive compounds with pharmaceutical potential [38].

Late-season fungal dynamics revealed substantial increases in Botrytis, Plectosphaerella, Cladosporium, and Albifimbria abundance at 80 days post-planting, suggesting potential late blight disease pressure. These findings indicate that cultivation method selection influences both beneficial and pathogenic microbial populations, with important implications for integrated crop management strategies.

4.3. Environmental–Microbial Correlations at the Genus Level in Rice–Potato Cropping Systems

4.3.1. Bacterial Community Correlations

Cross-treatment analysis identified three bacterial genera—Pseudarthrobacter, Bradyrhizobium, and unclassified Roseiflexaceae—that demonstrated significant correlations with environmental parameters across both cultivation systems. These genera represent functionally important soil microorganisms with diverse metabolic capabilities.

Pseudarthrobacter exhibits multifunctional properties including degradation capabilities [39] and nitrogen fixation with phosphorus solubilization functions [40]. The genus exhibited differential functional responses between systems, showing a significant positive correlation with available phosphorus in rainfed treatments and with soil organic matter in waterlogged treatments. These patterns suggest that Pseudarthrobacter emphasizes degradation functions under rainfed conditions while prioritizing phosphorus solubilization activities in waterlogged environments.

Bradyrhizobium provides critical soil ecosystem services including nitrogen fixation, photosynthesis, denitrification, and aromatic compound degradation [41,42,43]. The genus demonstrated a significant negative correlation with available phosphorus in rainfed systems and with pH and catalase activity in waterlogged systems. Roseiflexaceae contributes substantially to ammonium nitrogen and chemical oxygen demand removal processes [44], showing consistent positive correlations with urease activity across both cultivation methods.

Previous research indicates that waterlogged anaerobic conditions promote soil denitrification and ammonium accumulation [45], with rainfed systems potentially exhibiting similar effects. During potato development, rainfed systems experienced declining available phosphorus alongside enhanced urease activity, creating favorable conditions for Bradyrhizobium and Roseiflexaceae proliferation. Waterlogged systems demonstrated inverse patterns.

Treatment-specific correlations revealed additional functional relationships. Lysobacter in rainfed systems showed a significant positive correlation with available phosphorus. This genus provides broad-spectrum antimicrobial activity and serves as an effective biocontrol agent against plant pathogenic fungi [46]. In waterlogged systems, Dechloromonas demonstrated highly significant positive correlations with available nitrogen and potassium. This genus achieves simultaneous nitrogen and phosphorus removal through denitrifying phosphorus removal pathways, reducing carbon source requirements and aeration energy demands [47].

Analysis reveals that key environmental drivers of dominant bacterial genera include pH, soil organic matter, available phosphorus, available potassium, and urease activity in rainfed systems, compared to pH, soil organic matter, available nitrogen, and available potassium in waterlogged systems. The differential nitrogen release patterns and urease activity levels between systems create distinct selective pressures that favor different microbial functional groups.

4.3.2. Fungal Community Correlations

Ascomycota genera dominated both cultivation systems, comprising 13 of the top 20 most abundant taxa in each treatment, consistent with phylum-level abundance patterns. Three genera—Botrytis, Plectosphaerella, and Saitozyma—demonstrated significant correlations with environmental factors across both systems.

Botrytis and Plectosphaerella exhibited highly significant negative correlations with available nitrogen and potassium in both treatments, indicating accelerated growth during nutrient depletion phases. Saitozyma demonstrated a significant positive correlation with available potassium across systems. Research indicates that Saitozyma podzolica demonstrates plant growth promotion properties and pathogen growth inhibition capabilities [48].

These three genera showed significant correlations with urease activity specifically in rainfed systems, corresponding to the increasing urease activity trends observed throughout potato development. Rainfed conditions proved more favorable for beneficial Saitozyma growth compared to waterlogged systems.

Mortierella, representing Mortierellomycota, comprises common soil fungi with demonstrated capabilities for degrading phenylurea herbicides including isoproturon [49] and diuron [50]. The genus exhibited significant negative correlations with available nitrogen, phosphorus, and potassium in rainfed systems, indicating enhanced growth during nutrient consumption periods.

Environmental factors demonstrating strong correlations with dominant fungal genera included available nitrogen, phosphorus, potassium, urease, and catalase activities in rainfed systems, compared to pH, soil organic matter, available nitrogen, phosphorus, potassium, and urease activity in waterlogged systems. Waterlogged systems showed predominantly negative correlations between dominant genera and soil nutrients, with accelerated growth during nutrient depletion. Rainfed systems demonstrated negative nutrient correlations coupled with positive enzyme activity correlations, indicating that nutrient consumption combined with enhanced enzyme activity optimally supports dominant microbial community development.

These contrasting patterns reflect the fundamental differences between rice cultivation environments, which create distinct fungal community structures during subsequent potato cultivation and generate differential soil nutrient release and enzyme activity patterns throughout the cropping sequence.

5. Conclusions

Waterlogged rice cultivation as the preceding crop generated enhanced soil nutrient mobilization and elevated urease, sucrase, and catalase activities compared to rainfed rice systems. During subsequent potato cultivation, rainfed rice–potato systems demonstrated accelerated nutrient release dynamics and enhanced urease and catalase activity levels throughout crop development.

Rainfed rice–potato systems consistently maintained superior microbial community richness, diversity, and species abundance relative to waterlogged rice–potato systems. Notably, nitrogen metabolism-associated bacterial genera, including Sphingomonas, Rhodanobacter, and Nitrosospira, achieved higher relative abundance in rainfed systems compared to waterlogged systems. Conversely, pathogenic fungal genera also demonstrated elevated abundance in rainfed rice–potato systems relative to waterlogged systems.

These findings indicate that rainfed rice–potato cropping systems require integrated fertilization strategies combining organic and inorganic inputs, coupled with enhanced soil-borne disease management protocols. The higher pathogenic fungal loads observed in rainfed systems necessitate proactive disease prevention measures to ensure sustainable productivity while capitalizing on the superior nutrient cycling and microbial diversity benefits these systems provide.