Atractylodes macrocephala Root Rot Affects Microbial Communities in Various Root-Associated Niches

Abstract

Atractylodes macrocephala, a perennial herb widely used in traditional Chinese medicine, is highly prone to root rot, which significantly reduces its yield and quality. This study compared the physicochemical properties of soil from healthy and diseased A. macrocephala plants and analyzed the microbial diversity in the endophytic, rhizosphere, and root zone soils. The results showed that the diseased plants had higher levels of available potassium and electrical conductivity in the rhizosphere, both positively correlated with the severity of root rot, while soil pH was negatively correlated. The diversity and richness of endophytic bacterial and fungal communities were significantly reduced in diseased plants. Additionally, root rot led to major changes in the rhizosphere microbial community, with an increased abundance of Proteobacteria and Ascomycota, and a decrease in Firmicutes, Bacteroidetes, Actinobacteria, and Basidiomycota. Fusarium oxysporum, Fusarium solani, and Fusarium fujikuroi were identified as key pathogens associated with root rot. This study enhances our understanding of the microbial interactions in soils affected by root rot, offering a foundation for developing soil improvement and biological control strategies to mitigate this disease in A. macrocephala cultivation.

1. Introduction

Atractylodes macrocephala, a perennial herb from the Asteraceae family, is extensively cultivated in East Asia, especially in China [1]. Its rhizome, known as “Baizhu” in Chinese, is commonly used in traditional Chinese medicine. The use of A. macrocephala dates back to the Shennong’s Classic of Materia Medica (Shen-Nong-Ben-Cao-Jing) from the Eastern Han Dynasty. For millennia, this plant has been employed to address various health issues, including spleen hypofunction, poor appetite, bloating, diarrhea, dizziness, and heart palpitations [2,3]. Modern pharmacological studies have revealed that A. macrocephala offers a range of benefits, including enhanced gastrointestinal function and gonad hormone regulation, as well as antitumor, anti-inflammatory, anti-aging, anti-oxidative, and anti-osteoporotic effects. Further, the antibacterial, antitocolytic, and neuroprotective properties of A. macrocephala have also been documented [4]. Today, A. macrocephala is commonly prescribed as a Chinese medicinal herb in clinical settings, and there is a high demand for its use in pharmaceutical preparations. Wild A. macrocephala is found in at least 14 provinces across China, particularly in the Zhejiang, Jiangsu, Jiangxi, Hunan, and Anhui provinces [5]. However, as this plant is primarily cultivated through artificial means, its long-term cultivation often leads to reduced genetic diversity, which in turn decreases disease resistance and quality stability and increases its susceptibility to various diseases. Among these potential diseases, root rot is especially detrimental.

A. macrocephala root rot is a prevalent issue that inflicts considerable damage to plants, leading to substantial root decay and seedling mortality. This disease has thus severely impacted the agricultural industry in China. Interestingly, A. macrocephala root rot has often been linked to the practice of continuous cropping [6]. Studies show that after several cycles of continuous cropping, the incidence rate of root rot may escalate to 60% or higher [7]. Changes in soil physicochemical properties and the disruption of soil nutrient balance are typically considered the culprits behind diseases related to continuous cropping. Additionally, root rot is predominantly caused by fungal pathogens, such as Fusarium oxysporum, Rhizoctonia solani, and Ceratobasidium sp. [8]. However, the etiology of A. macrocephala root rot remains unclear, and the diversity of pathogenic microorganisms presents considerable challenges for its management.

Plants serve as hosts for a diverse range of microorganisms within their rhizospheres, phyllospheres, and endospheres, forming intricate ecosystems known as plant microbiomes or phytobiomes. These communities encompass bacteria, archaea, fungi, actinomycetes, and protists [9,10,11]. These microorganisms can exist within plant tissues as endophytes, on plant surfaces as epiphytes, or in the soil as rhizosphere inhabitants [12]. Notably, these plant-associated microbes play a pivotal role in promoting plant growth and maintaining plant health through functions such as nitrogen fixation, plant hormone production, drought tolerance improvement, and pathogen suppression [13,14,15,16]. Endophytic communities are dynamic and adapt in response to plant diseases. The rhizosphere, a critical and intricately structured zone surrounding plant roots, is a key area of focus in plant microbiome research [17,18]. Rhizosphere microorganisms mediate key processes such as soil decomposition, mineralization, and aggregate formation, which in turn bolster plant resilience [19]. The composition of plant microbiomes is intricately shaped by a multitude of biotic and abiotic factors, including the genotype of the host plant, the stage of plant development, prevailing climatic conditions, soil types, and the array of field management practices employed [20,21]. Plant diseases frequently arise due to pathogen infections, which disrupt the composition of the plant microbiome [22]. Consequently, elucidating the rhizosphere microbiome’s response to disease is essential for uncovering the mechanisms underpinning microbiome-mediated plant health enhancement and for crafting effective disease mitigation strategies. Several recent studies have examined the influence of indigenous and rhizosphere microorganisms on the incidence of root rot in plants. Extensive studies in a variety of tuberous and rhizomatous plants, such as potatoes [23], Pinellia ternate [24], Aconitum [25], and turmeric [26], have consistently pointed to pathogenic fungi as the primary agents causing root rot. Concurrently, emerging evidence has also demonstrated shifts within the plant’s microbial communities [27] and defensive mechanisms [28,29] subsequent to root rot episodes. An analysis of the microbiome of rhizosphere soils [30] revealed a correlation between the severity of root rot in sugarcane and the microbial community composition of the rhizosphere. The richness of fungal communities in rhizosphere soil samples from mildly diseased plants was found to be significantly lower than that in rhizosphere soil samples from healthy plants or severely diseased plants, indicating a decrease in potential pathogenic fungi. Notably, increased disease severity has been linked to reductions in root dimensions and fungal diversity, alongside an increase in operational taxonomic unit (OTU) abundance in affected fields. Additionally, research shows that the incidence of potato common scab is correlated with soil–root microbiome alterations, highlighting the pivotal role of the microbiome in plant health and disease dynamics [23]. This underscores the urgent need for microbiome-focused plant disease research. In this context, the lack of research on the specific role of the A. macrocephala microbiome in the severity of root rot warrants attention.

In this study, we systematically collected A. macrocephala plant samples and corresponding soil samples from plants at varying stages of root rot. We evaluated the physicochemical properties of both the rhizosphere and root zone soils and utilized high-throughput sequencing technology to analyze the microbial community composition and diversity of endophytic, rhizosphere, and root zone microbes. The objectives of our study are to (1) delineate endophytic, rhizosphere, and soil microbial communities across healthy to disease-affected A. macrocephala plants and to determine their respective functions; (2) identify potential pathogens associated with root rot; and (3) establish correlations among environmental factors, the plant microbiota, and the occurrence of A. macrocephala root rot, with the aim of developing potential risk mitigation strategies. Through these objectives, we aim to gain a deeper understanding of the impact of root rot on the health of A. macrocephala plants and explore strategies to mitigate the disease by modulating physicochemical soil properties and microbial communities. We believe that these findings will provide a scientific basis for the development of new disease management methods.

2. Materials and Methods

2.1. Sample Collection

Soil and plant samples were collected from a test field in Hangzhou City, Zhejiang Province (119°57′ E, 30°03′ N), China. Meteorological data at the experimental sites were recorded from January through November, covering the typical growing season for A. macrocephala. Temperature and precipitation patterns during the 2020 and 2021 growing seasons are presented in Figure S1, showing similar trends in daily temperatures and fluctuations in precipitation for both years. After 2 years of monoculture, A. macrocephala plants in the field exhibited symptoms of root rot disease, leading to clusters of dead plants randomly distributed throughout the field. To simulate the local natural conditions, we did not exert excessive control over environmental variables, allowing the plants to grow under natural rainfall, temperature, and humidity conditions. The plants were divided into three pathological groups: (i) healthy plants with no symptoms of disease on the leaves or stem base; (ii) moderately diseased plants with root rot, showing weak or yellow leaves and typical disease-associated scab on the stem base; and (iii) severely diseased plants with root rot (Figure S2). Samples of healthy, moderately diseased, and severely diseased plant roots (for endosphere evaluation), rhizosphere soil, and bulk soil were collected at random and marked as AE_H, RS_H, RZ_H, AE_M; RS_M, RZ_M; and AE_S, RS_S, RZ_S, respectively. The rhizosphere (representing the soil still attached to the roots after gentle shaking) and bulk soils (the soil shaken off roots) [31] were collected separately, mixed well, and packed into Ziplock bags for analyses of soil physicochemical properties. Simultaneously, about 10 g of fresh soil was added to sterilized tubes and stored in liquid nitrogen for the analysis of microbial abundance and diversity.

2.2. Soil Physicochemical Properties

The collected soil samples were air-dried on kraft paper and sifted through a 1 mm screen. The soil that passed through the screen was considered dry soil. The physicochemical properties of the soil were assessed using the outlined protocol. Specifically, water was incorporated into the soil samples at a ratio of 2.5:1. Subsequently, the mixtures were agitated at a speed of 200 revolutions per minute (rpm) for 10 min using the TS-2 shaker (Kylin-Bell, Haimen, Jiangsu, China). Following agitation, the solutions were left undisturbed for 30 min. Then, soil pH was assessed using the FE28 Standard pH meter sourced from Mettler-Toledo (Columbus, OH, USA). The potassium dichromate external heating method was employed to measure the organic matter content of the soil. Meanwhile, the alkali-hydrolyzable nitrogen content was determined using the diffusion–absorption method [32]. The available phosphorus content was determined using the hydrochloric acid–sulfuric acid double acid extraction colorimetric method. Furthermore, the available potassium content was determined using the ammonium acetate extraction flame photometric method [15]. All samples were analyzed using three biological replicates and two technical replicates.

2.3. DNA Extraction, PCR Amplification, and Illumina NovaSeq Sequencing

The FastDNA^® Spin Kit from MP Biomedicals (Irvine, CA, USA) was employed to extract the total microbial genomic DNA from soil and plant samples, adhering strictly to the manufacturer’s guidelines. Then, 1.0% agarose gel electrophoresis and the NanoDrop^® ND-2000 spectrophotometer (Thermo Scientific Inc., Waltham, MA, USA) were utilized to assess the quality and concentration of the extracted DNA. The DNA samples were stored at −80 °C for subsequent use.

A two-step PCR process was employed to generate bacterial 16S rRNA amplicon libraries. Individual DNA samples obtained from all specimens were amplified using a thermocycler PCR system (GeneAmp 9700, ABI, Foster City, CA, USA). Drawing from Bulgarelli’s optimization experiments with 16S rRNA primer pairs, we selected primer 799F (5′-AACMGGATTAGATACCCKG-3′), which exhibits three mismatches with the poplar chloroplast 16S rRNA, and primer 1392R (5′-ACGGGCGGTGTGTRC-3′) for our study. The first round of PCR amplification was conducted using these selected primers. Each PCR was replicated thrice using a 20 µL mixture consisting of 4 µL of 5× FastPfu Buffer, 2 µL of 2.5 mM dNTPs, 0.8 µL of each primer (5 µM), 0.4 µL of FastPfu Polymerase, 0.2 µL of bovine serum albumin (BSA), and 10 ng of template DNA. The cycling conditions included an initial denaturation step at 94 °C for 3 min, followed by 27 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 45 s. Finally, the terminal extension phase was conducted at 72 °C for 10 min. To eliminate residual primers and primer dimers, the obtained PCR products were separated on a 2% agarose gel (100 V for 30 min). To isolate the target products (which had an amplicon length of 593 bp), mitochondrial by-products (1000 bp) were removed. The DNA was then extracted and purified from the gel slices using an AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer’s instructions. After purification, a second round of PCR amplification was performed using the primers 799F (5′-AACMGGATTAGATACCCKG-3′) and 1193R (5′-ACGTCATCCCCACCTTCC-3′) to shorten the amplicon length to 394 bp for sequencing purposes. The reaction conditions and steps were identical to those used in the first round, except that the number of PCR cycles was reduced to 13. Simultaneously, fungal ITS amplicon libraries were constructed using the specific primers ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′). The reaction conditions, purification steps, and quantification procedures used for the second round of PCR were identical to those employed in the first round. The purified amplicons were pooled at equimolar concentrations and subjected to paired-end sequencing on an Illumina NovaSeq PE250 platform (Illumina, San Diego, CA, USA) based on standard protocols by Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). The raw sequencing data were submitted to the NCBI Sequence Read Archive (SRA) database under the Accession Number PRJNA1110161.

2.4. Data Processing

Raw FASTQ files were de-multiplexed using a custom Perl script. Subsequently, quality filtration was performed using Fastp version 0.19.6 [33], and the retained sequences were merged with FLASH version 1.2.7 [34] based on the following criteria: (i) Reads were truncated at any position with an average quality score below 20 over a sliding window of 50 base pairs. Truncated reads shorter than 50 base pairs or containing ambiguous characters were discarded. (ii) Only overlapping sequences longer than 10 base pairs were assembled, with consideration given to their overlapping regions. The maximum mismatch ratio allowed in the overlap region was 0.2. Reads that could not be successfully assembled based on these criteria were discarded. (iii) Samples were distinguished based on barcode and primer sequences, and the sequence direction was adjusted accordingly. An exact match was required for barcode matching. However, up to two nucleotide mismatches were allowed for primer matching. After optimization, the sequences were clustered into OTUs based on a sequence similarity threshold of 97% using UPARSE version 7.1 [35,36]. For each OTU, the most abundant sequence was selected as the representative sequence. To mitigate the influence of sequencing depth on alpha and beta diversity evaluations, the number of 16S rRNA gene sequences obtained from each sample was standardized to 20,000 to ensure an average coverage of 99.09% across all samples.

Each representative OTU sequence was taxonomically classified using RDP Classifier version 2.2 [37], against the 16S rRNA gene database (Silva v138) and the UNITE version 8 ITS database (specific to fungi). The confidence threshold was 0.7.

2.5. Statistical Analysis

The bioinformatic analysis of the soil/gut microbiota was conducted using the Majorbio Cloud platform (https://cloud.majorbio.com, accessed on 5 March 2024). Rarefaction curves and various alpha diversity indices—including observed OTUs, Chao1 richness, the Shannon index, Ace, and the Simpson and Sobs index, as well as beta diversity metrics using Bray–Curtis dissimilarity—were computed based on the OTU data using the Mothur software version 1.30.1. Beta diversity metrics were analyzed using the Vegan package (version 2.5-3) and the results were assessed using the ANOSIM test [38]. BugBase (https://bugbase.cs.umn.edu/index.html, accessed on 5 March 2024) was used to predict bacterial phenotypes [39], including biofilm formation, pathogenicity, mobile elements, and oxygen utilization. FUNGuild [40] was used to categorize fungal communities into pathotrophs, symbiotrophs, and saprotrophs based on their nutritional modes. By integrating the bacterial and fungal classifications obtained through these functions, the taxonomic composition of both bacteria and fungi was predicted. The influence of soil physicochemical factors on microbial community structure was assessed via redundancy analysis (RDA) and ordinal regression. Statistical analyses were conducted using SPSS v21 software (IBM, Armonk, NY, USA), with a two-way ANOVA to assess the effects of disease status and soil type on microbial diversity. A significance level of p < 0.05 was applied. Tukey’s HSD test was used for post hoc comparisons to identify significant differences between groups, with normality and homoscedasticity assumptions checked beforehand. Additionally, the Kruskal–Wallis H test and Welch’s test, with false discovery rate (FDR) correction for multiple comparisons, were employed to analyze microbial abundance differences in bacteria and fungi across root-associated niches (p < 0.05 significance level).

2.6. Isolation and Identification of A. macrocephala Root Rot Pathogens

Diseased root samples were inoculated onto potato dextrose agar (PDA) plates and incubated in the dark at 25 °C for 7 days to isolate potential pathogenic fungi. DNA isolation was performed using a modified version of the cetyltrimethylammonium bromide (CTAB) method, as described by Doyle in 1987 [41]. The ITS sequences were referenced against NCBI’s ITS (fungi) sequence databases using nucleotide BLAST (https://blast.ncbi.nlm.nih.gov/, accessed on 15 October 2023) to ascertain their approximate phylogenetic affiliations. Phylogenetic relationships were visualized using MEGA (version 7.0) software, which generated a phylogenetic tree based on the neighbor joining method.

3. Results

3.1. Comparison of Soil Physicochemical Properties

Table 1. Physiochemical characteristics of the rhizosphere and non-rhizosphere soil of healthy and diseased Atractylodes macrocephala.

	pH	EC (US/CM)	OM (g/kg−1)	AN (mg/kg−1)	AP (mg/kg−1)	AK (mg/kg−1)
RS_H	7.29 ± 0.38 a	65.77 ± 8.56 b	28.52 ± 0.25 a	178.70 ± 2.81 a	66.53 ± 3.12 a	59.80 ± 10.23 b
RS_M	6.93 ± 0.14 ab	82.77 ± 15.56 b	28.17 ± 1.15 a	174.10 ± 5.30 ab	67.54 ± 8.43 a	66.40 ± 2.97 b
RS_S	6.64 ± 0.30 b	109.7 ± 16.89 a	28.92 ± 0.22 a	171.70 ± 5.46 b	70.50 ± 3.29 a	151.30 ± 44.29 a
RZ_H	7.07 ± 0.15 a	53.40 ± 3.10 b	27.35 ± 3.54 a	172.70 ± 11.83 a	65.16 ± 17.57 a	54.00 ± 8.60 b
RZ_M	6.71 ± 0.13 b	73.03 ± 12.34 a	27.36 ± 1.04 a	169.40 ± 10.84 a	64.86 ± 10.06 a	59.40 ± 4.61 ab
RZ_S	6.50 ± 0.22 b	69.57 ± 4.65 a	26.17 ± 0.74 a	168.00 ± 3.13 a	62.86 ± 3.17 a	68.60 ± 9.76 a
Soil	ns	**	ns	ns	ns	**
Diseased	**	**	ns	ns	ns	**
Soil × Diseased	ns	ns	ns	ns	ns	**

Note: EC, OM, AN, AP, and AK represent electric conductivity, organic matter, available nitrogen, available phosphorus, and available potassium, respectively; H = healthy; M = moderately diseased S = severely diseased; RS = rhizosphere; RZ = non-rhizosphere. Different letters in columns indicate significant differences (ns p > 0.05, ** p < 0.01, n = 5).

presents a comparative analysis of the physicochemical properties of rhizosphere (RS) and bulk (RZ) soil samples from healthy (H) and diseased (M, S) A. macrocephala plants. The parameters assessed included soil pH, organic matter content, electrical conductivity, alkali-hydrolyzable nitrogen content, available phosphorus content, and available potassium content. Organic matter content, a critical indicator of soil fertility, showed no significant variation across the different groups (26.17 to 28.92 g/kg). Soil pH, a key determinant of microbial activity and nutrient availability, was found to differ significantly between the different groups. The pH of soil samples from diseased plants was significantly lower than that of soil samples from healthy plants, with the rhizosphere and bulk soils of severely diseased plants (RS_S and RZ_S) showing the most significant decrease (p < 0.05). Electrical conductivity, a measure of soil salinity, also varied significantly among the samples. The highest electrical conductivity was observed in the rhizosphere of severely diseased plants (RS_S; 109.70 US/cm). Meanwhile, the highest available potassium content was detected in the rhizosphere soil of severely diseased plants (RZ_S), and the difference (compared to the rhizosphere [RS_H] and bulk [RZ_H] soils of healthy plants) was statistically significant (p < 0.05). The other physicochemical properties, including alkali-hydrolyzable nitrogen and available phosphorus content, were comparable among the different groups of samples.

3.2. Comparison of Root-Associated Microbial Community Diversity Between Healthy and Diseased Plants

Alpha and beta diversity analyses were performed to assess the richness and diversity of microbial communities associated with healthy and diseased plants. At a threshold of 97%, the rarefaction curves of both prokaryotic and fungal communities approached their asymptotes. This indicated that the sequencing depth for each sample was adequate and that the sequencing library captured the majority of prokaryotic and fungal communities present in the samples.

3.2.1. Alpha Diversity Indices

The Shannon and Simpson indices were used to assess community diversity, while the Ace, Chao, and Sobs indices represented community richness. A notable decline in the diversity and richness of endophytic bacterial communities was observed upon the emergence of disease (p < 0.05). Specifically, the Shannon index fell from 4.7919 to 2.2712, the Simpson index rose from 0.0238 to 0.2079, Chao1 index decreased from 558.5100 to 259.5000, and the Sobs index plummeted from 398.3300 to 124.6700. Meanwhile, the variations in rhizosphere microorganisms were not substantial, with the Shannon index fluctuating between 6.3846 and 6.1666 and the Ace index ranging from 3337.3000 to 2951.8000. Moreover, the bacterial populations in the root zone also remained largely unaltered (

Table 2. Bacterial community diversity indices.

	Shannon	Simpson	Sobs	Ace	Chao1
AE_H	4.79 ± 0.03 a	0.0238 ± 0.0026 b	398.33 ± 11.01 a	559.44 ± 30.79 a	558.51 ± 35.49 a
AE_M	4.53 ± 0.24 a	0.0348 ± 0.0087 b	372.00 ± 49.27 a	508.04 ± 83.28 a	478.90 ± 80.72 a
AE_S	2.27 ± 0.17 b	0.2079 ± 0.0663 a	124.67 ± 17.79 b	470.76 ± 136.16 a	259.50 ± 46.07 b
RS_H	6.38 ± 0.12 a	0.0037 ± 0.0011 b	1101.20 ± 49.43 a	3337.30 ± 173.49 a	2233.30 ± 46.19 a
RS_M	6.23 ± 0.19 a	0.0061 ± 0.0029 ab	1068.00 ± 80.60 a	3181.20 ± 351.33 a	2201.30 ± 171.68 a
RS_S	6.17 ± 0.22 a	0.0070 ± 0.0023 a	1025.40 ± 100.50 a	2951.80 ± 567.09 a	2062.30 ± 309.10 a
RZ_H	6.21 ± 0.12 a	0.0055 ± 0.0021 a	1030.20 ± 26.78 b	2942.70 ± 198.89 a	2083.10 ± 122.19 a
RZ_M	6.20 ± 0.15 a	0.0061 ± 0.0025 a	1050.60 ± 54.53 ab	3210.10 ± 455.79 a	2182.60 ± 320.21 a
RZ_S	6.35 ± 0.16 a	0.0046 ± 0.0017 a	1114.60 ± 59.94 a	3334.50 ± 296.48 a	2271.00 ± 149.93 a
Bacteria	**	**	**	*	*
Diseased	*	**	**	ns	ns
Bacteria × Diseased	**	**	**	ns	ns

Note: The diversity indexes of endophytic rhizosphere bacteria and non-rhizosphere bacteria were tested by LSD in Two-Way ANOVA at the genus level. H = healthy; M = moderately diseased; S = severely diseased; AE = A. macrocephala endophyte (M ± SD, n = 3); RS = rhizosphere; RZ = non-rhizosphere (M ± SD, n = 5). Different letters in columns indicate significant differences (ns = non-significant, * = p < 0.05, ** = p < 0.01).

). Notably, the trends in endophytic fungal community richness and diversity mirrored those for endophytic bacteria, showing a significant decline with disease development (p < 0.05). Meanwhile, the changes in rhizosphere fungal communities were subtle, but the trend was opposite to that of the bacterial communities (

Table 3. Fungal community diversity indices.

	Shannon	Simpson	Sobs	Ace	Chao1
AE_H	2.66 ± 0.25 a	0.1960 ± 0.0142 b	707.33 ± 176.16 a	790.08 ± 184.31 a	774.95 ± 172.12 a
AE_M	2.51 ± 0.02 a	0.2742 ± 0.0264 a	734.67 ± 173.94 a	833.59 ± 191.20 a	814.62 ± 181.33 a
AE_S	2.66 ± 0.30 a	0.1369 ± 0.0339 c	492.33 ± 242.00 a	599.36 ± 275.70 a	584.97 ± 261.50 a
RS_H	3.71 ± 0.26 a	0.0648 ± 0.0260 a	806.80 ± 128.88 a	1019.70 ± 187.86 a	995.27 ± 185.19 a
RS_M	3.79 ± 0.36 a	0.0667 ± 0.0335 a	930.80 ± 103.65 a	1167.10 ± 133.73 a	1143.70 ± 117.02 a
RS_S	3.92 ± 0.30 a	0.0490 ± 0.0177 a	842.00 ± 138.97 a	1046.50 ± 146.52 a	1017.70 ± 135.23 a
RZ_H	3.91 ± 0.07 a	0.0508 ± 0.0052 a	832.20 ± 55.22 a	1006.60 ± 105.17 a	974.81 ± 96.03 a
RZ_M	3.85 ± 0.11 a	0.0558 ± 0.0114 a	916.80 ± 84.07 a	1122.30 ± 107.63 a	1109.60 ± 110.24 a
RZ_S	3.75 ± 0.19 a	0.0616 ± 0.0148 a	869.80 ± 132.08 a	1114.70 ± 174.26 a	1063.80 ± 162.75 a
Fungi	*	**	*	*	*
Diseased	ns	**	ns	ns	ns
Fungi × Diseased	ns	**	ns	ns	ns

Note: The diversity indexes of endophytic rhizosphere fungi and non-rhizosphere fungi were tested by LSD in Two-Way ANOVA at the genus level. H = healthy; M = moderately diseased; S = severely diseased; AE = A. macrocephala endophyte (M ± SD, n = 3); RS = rhizosphere; RZ = non-rhizosphere (M ± SD, n = 5). Different letters in columns indicate significant differences (ns = non-significant, * = p < 0.05, ** = p < 0.01).

).

3.2.2. Beta Diversity Indices

To measure the similarity between the rhizosphere and endophytic communities in the different groups, we performed principal coordinate analysis (PCoA) using the weighted Bray–Curtis algorithm. The samples showing severe root rot could be clearly differentiated from the other samples along the first axis, as indicated by the ANOSIM test (p < 0.05). Notably, the AE-H group was significantly separated from AE-M group. The endophytic microbiome of the AE-S samples was predominantly found in the positive region of PCoA1, while the bacteria from AE-H and AE-M samples were primarily located in the negative region of PCoA1 (Figure 1A). A similar pattern was observed for the endogenous fungi (Figure 1B).

In contrast to the endophytic microbial communities, in the rhizosphere microbial communities, only the fungi displayed a significant change in beta diversity with disease progression. Based on the PCoA results, we found that the fungal communities were distinctly clustered into two groups. Specifically, the rhizosphere soil fungi were divided into two clusters, with RS-S forming one group and RS-H and RS-M forming the other. (Figure 1B). The fungal microbiome in the rhizosphere soil of the RS-S group was mainly located toward the positive region of PCoA1, while the RS-H and RS-M groups were predominantly detected in the negative region of PCoA1 (Figure 1B). As for the microorganisms in the root zone, there were no significant differences in bacteria or fungi among the different groups (Figure 1A,B).

3.3. Composition of Root-Associated Microbial Communities in Healthy and Diseased Plants

A total of 2,965,146 optimized fungal sequences were acquired from 39 samples. When clustered based on a 97% similarity threshold, 2084 OTUs were identified from root samples, 3248 from rhizosphere soil samples, and 3206 from root zone soil samples. Additionally, a total of 2,158,322 optimized bacterial sequences were also obtained. Based on a 97% similarity threshold, a total of 985, 10,070, and 9770 OTUs were identified in the root and stem samples, rhizosphere soil, and root zone soil, respectively. Figure 2 shows the comparative analysis of OTUs of the plant (AE), rhizosphere soil (RS), and bulk soil (RZ) samples from healthy (H) and diseased (M, S) plants. Notably, the soil bacterial communities exhibited greater richness than the root endophytic bacterial communities. Interestingly, both the endophyte and soil samples from the healthy plants had a higher number of bacterial OTUs (79 and 479, respectively) than the samples from diseased plants (21 and 378, respectively) (Figure 2A and Figure S3). Regarding fungi, the rhizosphere soil of diseased plants contained a higher number of fungal OTUs (221) than that of healthy plants (148) (Figure S4). In contrast, endophytic fungi showed a similar pattern in terms of the number of OTUs to the endophytic bacteria (Figure 2B).

Proteobacteria (41.76–90.92%) were the most abundant phylum across all samples, followed by Firmicutes (6.54–26.14%), Bacteroidetes (1.35–20.0%), and Actinobacteria (0.83–7.40%). As disease severity increased, the relative abundance of Proteobacteria increased significantly, whereas the relative abundance of Firmicutes, Bacteroidetes, and Actinobacteria decreased. The bacterial communities in the rhizosphere and root zone soil samples exhibited higher richness than those in the plant samples. In addition, Proteobacteria, Firmicutes, Bacteroidota, Actinobacteriota (16.63–21.86%), unclassified_k_norank_d_Bacteria (3.32–5.66%), Myxococcota (3.89–6.58%), Nitrospirota (2.97–4.03%), and Chloroflexota (2.76–3.58%) also showed a high abundance in the soil samples. The abundance of Firmicutes decreased significantly upon the occurrence of root rot (Figure 2A(a)). Notably, three phyla were identified within the plant samples, while four phyla were identified in the rhizosphere and root zone soil samples. Ascomycota was the most dominant phylum of endophytic fungi, with its relative abundance ranging from 82.04% to 84.56%. The other dominant fungal phyla included Basidiomycota (6.87–16.30%) and unclassified_k_Fungi (1.36–7.93%). With disease development, the relative abundance of Basidiomycota increased while that of unclassified_k_Fungi decreased. Apart from Ascomycota, Basidiomycota, and unclassified_k_Fungi, Mortierellomycota also exhibited a notable abundance in the rhizosphere and root zone soils. The fungal community structure within the rhizosphere soil contrasted with the fungal community structure in the plants. Notably, the abundance of Ascomycota increased significantly with disease occurrence, while that of Basidiomycota decreased (Figure 2B(a)).

To gain a deeper understanding of species composition, we conducted a genus-level community structure analysis. The predominant endophytic bacterial groups included Klebsiella (11.07–35.47%), Enterobacter (6.84–25.24%), norank_D_bacteria (0.08–1.18%), norank_F_Muribaculaceae (0.29–9.82%), and Burkholderia-Cavallonia-Papbirkholderia (1.61–11.69%). The abundance of Klebsiella and Enterobacter declined with disease progression. The abundance of bacteria in the rhizosphere and root zone soils was greater than that within the plants. Klebsiella, the dominant endophytic bacterial genus, was detected at very low levels in the rhizosphere and root zone soils. Conversely, Nitrospira (2.97–4.55%) and norank_o_Acidobacteriales (4.42–7.16%) were specific to these soil samples (Figure 2A(b)). Meanwhile, the analysis of fungal communities revealed that the most abundant endophytic fungi were Paraphoma (15.65–58.81%), Fusarium (1.11–6.08%), and Neocosmospora (2.14–20.50%). In the rhizosphere and root zone soils, the dominant fungal genera were Fusarium (9.12–13.49%), Neocosmospora (4.43–10.95%), Chaetomium (6.73–13.48%), and Trichocladium (5.95–12.10%). Notably, a significant increase in Neocosmospora was observed in the rhizosphere soil following disease occurrence, whereas Trichocladium demonstrated a significant decline (Figure 2B(b)).

3.4. Differences in Microbial Communities Between Healthy and Diseased Plants

The abundance values for the various microbial communities were compared among various samples using the Kruskal–Wallis rank sum test. To correct for multiple comparisons, the false discovery rate (fdr) method was employed. Subsequently, in order to specifically identify the species exhibiting significant variations in abundance, we utilized the Tukey–Kramer post hoc test. This approach allowed us to compare three distinct groups: the healthy group (H), the group with moderate root rot (M), and the group with severe root rot (S). Among the various endophytic bacteria, Klebsiella, Enterobacter, and Pseudomonas showed a significant increase in their relative abundance (p < 0.01) after root rot development, while norank_F_Muribaculaceae showed a significant decrease (p < 0.05). Among the rhizosphere bacteria, Pseudomonas and Sphingomonas increased significantly in abundance after moderate (p < 0.01) and severe (p < 0.01) root rot development, whereas Nitrospira and norank__f__Gemmatimonadaceae showed the opposite trend (p < 0.01). Notably, the relative abundance of Bryobacter exhibited a negative correlation with the severity of root rot disease (p < 0.01) (Figure 3A).

Among the fungi, endophytic Paraphoma showed a significant reduction in abundance (p < 0.001) after disease development, while endophytic Fusarium (p < 0.01) and Neocosmospora showed a significant increase (p < 0.05). Meanwhile, in the rhizosphere, the relative abundance of Chaetomium increased significantly under conditions of moderate disease but showed a downward trend in casaes of severe disease (p < 0.01). Additionally, the relative abundances of Trichocladium, Ascobolus, Mortierella, Gibberella, and Trichoderma all showed different degrees of down-regulation (p < 0.01) after disease occurrence. Notably, several species of Fusarium were isolated from diseased rhizomes, suggesting that fungi from this genus may be responsible for root rot in plants (Figure 3B). Following the RDA of the top eight dominant bacterial genera and rhizosphere characteristics, we identified significant differences in the rhizosphere bacterial community structure between healthy and root rot-diseased A. macrocephala plants (Figure 4A). Among the physicochemical properties of soil, pH had the highest impact on the bacterial community structure, followed by AK, OM, EC, AP, and AN. At the genus level, Lactobacillus emerged as a notable contributor to the bacterial community. Similarly, the RDA of the top eight dominant fungal genera and rhizosphere characteristics revealed significant variations in the rhizosphere microbial structure between healthy and diseased plants (Figure 4B). Among the physicochemical properties of the soil, AK had the highest impact on the fungal community structure, followed by EC, pH, AN, OM, and AP. Thus, available potassium and electric conductivity emerged as the primary environmental factors affecting the fungal community structure of the soil (Figure 4B). Notably, unclassified_f_Ceratobasidiaceae, Chaetomium, and Neocosmospora were identified as the key contributors to the rhizosphere fungal communities.

Additionally, soil pH showed a significant positive correlation with Fusarium, whereas Neocosmospora was positively correlated with electrical conductivity, available potassium, and organic matter. In contrast, Chaetomium exhibited a significant negative correlation with the organic matter content of the soil. These results provide valuable insights into the interactions between rhizosphere microbial communities and soil environmental factors in A. macrocephala.

3.5. Effects of Root Rot Disease on the Functional Profiles of Root-Associated Microbiomes

The BugBase database was utilized for the phenotypic prediction of the bacterial microbiome of the rhizosphere soil. The results revealed significant differences in anaerobic bacteria, facultatively anaerobic bacteria, Gram-positive bacteria, Gram-negative bacteria, and bacteria showing biofilm formation, presumptive pathogenicity, mobile genetic elements, and stress tolerance between healthy and diseased plants. Simultaneously, facultatively anaerobic microbes and microbial groups exhibiting stress tolerance, containing mobile elements, and showing biofilm formation properties were relatively abundant within the plants. Their relative abundances, ranging from 30.12% to 86.85%, 24.20% to 86.31%, 29.65% to 87.93%, and 24.30% to 86.31%, respectively, tended to increase with the onset of disease. Intriguingly, aerobic bacteria were primarily concentrated in the rhizosphere and root zone soils. The relative abundance of potentially pathogenic bacteria gradually increased from the rhizosphere soil to the tuber interior (3.59% to 89.96%) and increased even further as tuber decay progressed, showing an overall upward trend (29.46% to 89.96%) (Figure S6).

The functional prediction of the fungal microbiome was conducted utilizing the Fungi Functional Guild (FUNGuild) database. FUNGuild categorizes fungi into specific nutritional modes based on their roles as pathotrophs, saprotrophs, and symbiotrophs, uncovering their roles as plant pathogens. In this study, the fungal endophytic communities largely belonged to the fungal parasite–plant pathogen–plant saprotroph, undefined saprotroph, and plant pathogen groups. Following disease occurrence, the fungal parasite–plant pathogen–plant saprotroph functional groups exhibited a significant decrease, while the plant pathogen group showed a notable increase. In the rhizosphere and root zone soils, the fungal parasite–plant pathogen–plant saprotroph fungal functional group also demonstrated a marked reduction (Figure S5). It is important to note that the results derived from both the BugBase and FUNGuild databases represent predictions rather than direct experimental observations. Further empirical validation is needed to confirm the predicted phenotypes and interactions.

3.6. Isolation and Identification of Potential Fungal Pathogens

Samples from diseased plants yielded three fungal strains, which were designated as F3, AM9-71 and AM9-72, respectively. After cultivation on PDA for 7 days, F3 produced abundant colonies with a white filamentous mycelium that appeared opaque when viewed from underneath. Additionally, it produced a significant number of sickle-shaped conidia. Meanwhile, AM9-72 produced a mycelium with purplish-red pigmentation, along with numerous large, sickle-shaped conidia. Furthermore, AM9-71 produced a large number of white aerated mycelia with an abundance of large sickle-shaped conidia (Figure 5). A phylogenetic analysis based on the ITS and EF-1α regions demonstrated that the F3 strain clustered together with multiple strains of Fusarium solani. Additionally, the AM9-72 strain could be grouped with several strains of Fusarium oxysporum but exhibited a relatively greater genetic distance from other Fusarium species. Overall, strain F3 was identified as F. solani, strain AM9-72 as F. oxysporum, and strain AM9-71 as F. fujikuroi (Figure S7).

4. Discussion

A. macrocephala root rot is a serious soil-borne disease. Its onset is influenced by the growth and pathogenicity of causative microorganisms, as well as environmental factors, soil physicochemical properties, and the community composition of rhizosphere microbes [8,42]. To understand how soil physicochemical properties affect root rot severity in A. macrocephala, we analyzed soil pH, organic matter content, electrical conductivity, nutrient availability, and microbial diversity in the different root-associated niches of diseased and healthy plants. Our results revealed lower pH levels in the rhizosphere soil of diseased plants (Table 1). pH is a crucial factor influencing soil microbial diversity and community structure [43]. pH levels typically show a significant drop in the rhizosphere soil of diseased plants, including wheat, cruciferous plants, tobacco, strawberry, and ginseng [44]. This suggests that disease occurrence can alter the rhizosphere soil environment, with decreased pH being a significant indicator of these alterations [45]. Soil pH significantly affects the dynamics of microbial communities, and acidic soil conditions favor fungal growth over bacterial growth [46,47]. A previous study found that as soil pH increases, the ratio of fungal to bacterial biomass declines, although the biomass index for fungi shows a slight increase [48]. The equilibrium of soil micro-ecology and microbial diversity is essential for soil health and disease suppression in plants [17,49]. Additionally, ambient pH levels are required by pathogens to colonize, invade, and cause host mortality. These levels also influence the production of pathogenicity factors and modulate the expression of genes related to virulence and survival in pathogens [50]. Consequently, a decrease in pH or soil acidification can disrupt microbial diversity and the soil ecosystem, leading to an imbalance in the soil’s micro-ecology and an increase in soil-borne diseases in cultivated land.

In the present study, a significant increase in available potassium content was observed in the rhizosphere soil of diseased plants, while the alkali-hydrolyzable nitrogen and available phosphorus content did not show any discernible changes (Table 1). Rhizosphere soil plays a crucial role in plant growth and nutrient uptake. Root exudates significantly influence the properties, microbial communities, and functions of soil [51]. The release of organic acids (e.g., oxalate, citrate, malate, oxalate, succinate, and fumarate) by plant roots serves as a response mechanism to different conditions of stress [52,53]. The available potassium content in the rhizosphere soil increases due to these organic acids, thereby enhancing the rate of soil potassium utilization [54]. In plants damaged due to diseases caused by continuous cropping, soil acidification is often accompanied by an increase in the available potassium content of the rhizosphere soil [55,56,57]. This change is likely a direct consequence of the structural alterations in soil microbial communities caused by continuous cropping. The interplay among these microorganisms can activate specific potassium-solubilizing microbes, thereby enhancing potassium availability in the soil [58].

The plant microbiome, including endophytes and external epiphytic or rhizosphere microorganisms, plays a crucial role in mediating interactions between the host plant and its environment [12]. Previous studies have demonstrated that the diversity of rhizosphere microorganisms exceeds that of endophytes, and there are often significant differences between these microbial communities [59,60,61,62,63,64]. In the present study, we compared the microbial community composition (OTUs) of bulk soil, rhizosphere, and endophytes and observed that the microbial diversity in both the rhizosphere and bulk soils (Figures S3 and S4) was greater than the microbial diversity of endophytes (Figure 1B and Figure 2B). The alpha diversity analysis yielded similar results, confirming this observation (Table 2 and Table 3).

In contrast to healthy plants, diseased plants generally exhibit a lower species richness in the endosphere and a simpler network morphology [65]. Likewise, the abundance, biomass, and diversity of soil microbial communities serve as critical indicators of soil quality [66]. An increase in the microbial abundance of soil has been linked to the improved suppression of various soilborne pathogens, including Sclerotium rolfsii, which causes tomato southern blight and F. oxysporum, which causes banana wilt [67,68]. In the present study, we observed that endophytes were more susceptible to the effects of root rot disease than the broader soil microbial community. Through PCoA, a clear separation between the three groups AE-H, AE-M, and AE-S was observed for both bacteria and fungi (Figure 1A and Figure 2A). Notably, a marked decrease in the abundance and diversity of endophytic bacteria was detected during the progression of root rot. In comparison, the diversity of soil microorganisms was not significantly altered following the onset of root rot. Nevertheless, the diversity of the rhizosphere bacterial community exhibited a decreasing trend, whereas that of the rhizosphere fungal community showed an increasing trend (Table 2 and Table 3). This pattern of alterations in the rhizosphere microbiome is consistent with findings from previous studies on root rot in avocado and sugarcane. The findings indicate that the rhizosphere microbial characteristics of root rot may be similar across different plants [30,69]. Additionally, pathogen-infected plants may modify the composition of their root exudates, providing more diverse carbon sources for rhizosphere microorganisms. This could potentially contribute to the slight increase in microbial diversity observed in the rhizosphere of infected plants [20].

The microbial community of the plant rhizosphere, much like the gut microbiome of animals, plays a crucial role in host health [70]. Dysbiosis in this niche can increase the risk of plant diseases, and changes in microbial community structure are observed when plants are invaded by pathogenic bacteria [71]. In our study, we identified Proteobacteria, Firmicutes, Bacteroides, and Actinobacteria as the dominant phyla in the plant rhizosphere, with Proteobacteria being the most prevalent, consistent with previous findings (Figure 2A(a)). Bacteria from the Proteobacteria phylum play a key role in phylogenetic systems, ecology, pathology, and energy metabolism. [72,73]. Our analysis revealed a significant increase in the relative abundance of Proteobacteria in the endophytic bacterial community of A. macrocephala following the onset of root rot disease (Figure 2A(a)). We hypothesized that this increase may be a response to the disrupted physiological balance in this plant due to the disease. At the genus level, a notable increase in the relative abundance of Klebsiella and Enterobacter was detected in the endophytic bacterial community (Figure 3A). Previous studies have demonstrated that Klebsiella and Enterobacter can promote plant growth and inhibit various fungal pathogens [74], triggering the plant’s internal defense mechanisms to resist pathogenic invasions. Thus, the shift in the endophytic bacterial community observed in this study likely represented the plant’s physiological response to biotic stress. Many Pseudomonas species closely interact with plants, promoting plant health by suppressing phytopathogens and enhancing disease resistance and growth. These bacteria can inhabit both the endosphere [75] and rhizosphere [76] of plants. Additionally, Pseudomonas spp. produce secondary metabolites with antimicrobial properties [77]. Interestingly, a significant increase in the relative abundance of Pseudomonas was detected in the endosphere, while the opposite trend was observed in the soil (Figure 3A). This suggested that the plant recruited more Pseudomonas from the soil to bolster its defense against pathogens. Meanwhile, Sphingomonas—which was found in both the endophytic and soil bacterial communities in this study—showed a significant increase in the soil after disease onset. Notably Sphingomonas enhances plant tolerance to abiotic stresses, aids in bioremediation, and promotes the degradation of environmental pollutants. Sphingomonas species produce phytohormones such as gibberellins, salicylic acid, and indole-3-acetic acid to help plants cope with various stresses [78]. Thus, the increase in its relative abundance in the soil (Figure 3A) potentially reflected a defensive response by the plant against pathogenic bacteria.

Several fungi belonging to the phylum Ascomycota are recognized as plant pathogens responsible for soilborne diseases across various agricultural systems. Furthermore, they are known producers of toxins [79,80]. These fungi typically colonize plant roots, causing damage to the root surface and thereby facilitating infection by other pathogenic organisms. In our study, we observed a clear dominance of Ascomycetes within both the rhizomes and the rhizosphere/bulk soils of A. macrocephala. Specifically, the relative abundance of Ascomycetes was markedly higher in soils affected by root rot than in healthy soils (Figure 2B(a)), a pattern similar to that noted previously with tobacco bacterial wilt [81]. Hence, it is reasonable to speculate that the increased presence of Ascomycetes in the rhizosphere can create conditions favorable for the initiation and spread of root rot. In contrast, the trend of the relative abundance of Basidiomycetes was opposite to that of Ascomycetes (Figure 2B(a)). As crucial decomposers in the carbon cycle, Basidiomycetes can break down organic materials from plant debris, such as cellulose, lignocellulose, and lignin, through enzyme secretion [82]. Therefore, a higher relative abundance of Basidiomycetes in healthy soils not only enhances the decomposition of plant residues while promoting carbon cycling but also positively influences the rhizosphere microecosystem.

At the genus level, we found that Fusarium displayed a high relative abundance in both the rhizosphere and bulk soils of A. macrocephala (Figure 2B(b)). Furthermore, a significant increase in the abundance of Fusarium within the endosphere was observed in plants affected by root rot (Figure 3B). Prior studies have shown that the F. oxysporum species complex (FOSC) possesses a broad host range and can cause vascular wilt [83]. Thus, we speculated that the increased relative abundance of Fusarium within the rhizome of A. macrocephala may be associated with the development of root rot disease. Subsequently, we successfully isolated Fusarium oxysporum, Fusarium fujikuroi, and Fusarium solani from diseased plant samples (Figure 5) [84]. Interestingly, Neocosmospora, which has previously been shown to be a part of the FOSC, also demonstrated a positive correlation with disease severity post-infection, in both the endospheric plant and soil communities. Additionally, we found that the relative abundance of Fusarium was negatively correlated with disease severity in both the inter-root and root zone soils (Figure 3B). Notably, this pattern has previously been documented in sugarcane root rot [30]. One possible explanation is that the invasion of Fusarium into the plant may reduce its abundance within the soil and increase its accumulation within the plant, further intensifying the severity of root rot disease. The changes in microbial diversity not only affect soil health and the ability to suppress diseases but also have a direct link to plant physiological responses. For instance, shifts in the rhizosphere microbial community can activate plant defense mechanisms, enhancing resistance to pathogens [85]. In this study, we observed a significant decrease in the abundance and diversity of endophytic bacteria following the onset of root rot disease, which may reflect the plant’s physiological response to biotic stress. Notably, we found an increase in the relative abundance of Pseudomonas in the plant endophytic community, which may be related to the plant’s enhanced defense mechanisms, as Pseudomonas species are known to suppress plant pathogens and enhance plant disease resistance [86].

To investigate the intricate interactions between microorganisms and the soil environment, and to identify the key factors driving changes in microbial communities, the RDA method was employed. Accordingly, the relationship between the microbial communities and environmental factors was examined. Soil pH is known to be a key determinant of soil microbial diversity and community structure [43]. Decreased pH is closely linked to the occurrence of plant diseases and shows a significant negative correlation with soil microbial diversity [87]. In this study, soil pH was found to be negatively correlated with the occurrence of root rot in A. macrocephala, and the relative abundance of Fusarium significantly declined as the pH of the rhizosphere soil decreased (Table 1) (Figure 3B). This may be because Fusarium was the primary pathogen responsible for root rot in A. macrocephala. Additionally, available potassium and electric conductivity were identified as the critical factors influencing the shifts in the fungal community structure associated with root rot (Figure 4B). This suggests that these parameters should be the focal points of future studies on soil improvement for A. macrocephala cultivation. While our study provides valuable insights into the relationship between microbial diversity and root rot in A. macrocephala, several key limitations should be acknowledged. First, the sampling was conducted at specific time points, which may not capture the temporal dynamics of microbial communities. Second, the reliance on high-throughput sequencing methods, while robust, may not fully account for the functional capabilities of the microbial taxa identified. Additionally, environmental factors outside our control, such as varying weather conditions, may influence plant-microbe interactions and confound our results. Future research should address these limitations by incorporating more frequent sampling and exploring functional analyses to deepen our understanding of these complex interactions.

5. Conclusions

Soil properties and microbial communities are closely associated with plant health. In this study, we observed that A. macrocephala exhibited different soil physicochemical properties and microbial community characteristics under different severities of root rot disease despite being cultivated in the same field through uniform management strategies. Specifically, the pH of the rhizosphere decreased significantly following disease onset, while the content of available potassium increased. Concurrently, the plant-associated microbial communities underwent significant alterations, with the endophytic microbial community exhibiting more pronounced changes than the soil microbial community, particularly in the rhizosphere. Notably, after disease onset, the diversity of endophytic microbes declined, and the community composition changed markedly. Specifically, substantial invasion by potential pathogens (mainly fungi such as Fusarium and Neocosmospora) was observed. Meanwhile, the relative abundance of bacteria with biocontrol potential, such as Proteobacteria, increased significantly. Moreover, key microbial taxa and core module members showed altered abundance levels in the rhizospheres of diseased plants, potentially reducing the soil’s capacity to suppress pathogens and diseases. In response to pathogen attacks, host plants actively reshape their microbiomes by recruiting potential antagonists (e.g., Pseudomonas, Sphingomonas, Chaetomium, and Mortierella) into the rhizosphere. However, when pathogen invasion becomes widespread and root rot severity increases, these microbial antagonists are unable to save the plant, leading to a subsequent decline in their populations. Changes in microbial composition also induce functional shifts within the plant microbiome, with some pathogenic fungi increasing in abundance and potentially contributing to disease processes. Overall, the present study demonstrates the relationships among soil physicochemical properties, plant microbial communities, and the occurrence and severity of root rot in A. macrocephala. Importantly, it provides a theoretical foundation for the microecological management of root rot in this plant.