Trichoderma harzianum DQ002 Enhances Oriental Melon Resistance Against Fusarium oxysporum f.sp. melonis by Regulating Soil Microbial Communities in the Rhizosphere

Abstract

Continuous planting results in a higher occurrence rate of oriental melon Fusarium wilt caused by Fusarium oxysporum f. sp. melonis (FOM), and treatment with Trichoderma can considerably alleviate the incidence of disease. However, the tripartite interaction mechanisms among T. harzianum–melon–rhizosphere microorganisms remain poorly understood in current research. Pot experiments elucidate the growth-promoting, antagonistic, and rhizosphere-regulating effects of T. harzianum on oriental melon. The experiment consisted of two treatments: (1) water control (CK), and (2) T. harzianum inoculation (MM) with three repetitions per treatment. Illumina high-throughput sequencing was employed to analyze the microbial community and associated metabolic pathways. Additionally, a comprehensive correlation analysis clarified how T. harzianum-modulated physiological factors regulate soil microbial communities to enhance melon resistance to FOM. T. harzianum inoculation significantly promoted plant growth, decreased the incidence rate of Fusarium wilt by 41.85%, and increased rhizosphere nitrate-N, pH, EC, and soil enzyme activity (e.g., sucrose and alkaline phosphatase). Notably, T. harzianum inoculation altered the rhizosphere microbial community’s relative abundance and structure, with the most striking changes in the fungal community. Principal coordinate analysis showed this fungal restructuring accounted for 44.9% of total community variation (37% from PCo1, 7.9% from PCo2). Soil-borne pathogens (e.g., Fusarium, Verticillium, Phytophthora) decreased in relative abundance with the inoculation of T. harzianum. Meanwhile, the microbial community shifted from a “fungal-dominated” to “bacterial-dominated” state: fungal proportion decreased by 9.47% (from 23.95% in CK to 14.48% in MM), while bacterial proportion increased by 9.47% (from 76.05% in CK to 85.52% in MM). Microbial abundance shifts primarily impacted amino acid and cofactor biosynthesis metabolic pathways. The application of T. harzianum modified the soil environment, restructuring microbial communities through these changes, which in turn regulated microbial metabolic pathways, creating a soil environment conducive to melon growth and thereby enhancing oriental melon resistance to FOM, while mitigating the obstacles of continuous cropping.

1. Introduction

The oriental melon, a significant vegetable crop, is cultivated and globally accepted by customers for its nutritional values [1]. In China, the scale of melon cultivation is very considerable. As of 2022, China’s melon cultivation area was about 57 million mu, accounting for 48–50% of the global total area, with an annual output of 137.7 million tons, close to 60% of the world’s total output (FAOSTAT database, www.fao.org/faostat, accessed on 1 March 2025). China has long been the world’s largest producer and consumer of melons. The trend of highly intensive and long-term monocropping of melon crops has become increasingly prominent. This continuous cropping pattern refers to the continuous cultivation of the same or related crops on the same plot for many years, which induces the occurrence of continuous cropping obstacles in the soil, including nutrient imbalances [2], the accumulation of harmful substances [3], and an increase in pathogenic microorganisms [4], such as Fusarium oxysporum f sp. Melonis (FOM). Fusarium wilt caused by FOM poses a major threat to melon production in China, with yield reductions of 30–50% in some severely affected areas, severely impacting economic benefits [5]. Continuous cropping obstacles in the soil have been a bottleneck in practical production, in which a severe soil-borne disease can seriously impact crop cultivation and yield. Conventional control measures predominantly involve crop rotation [6], grafting [7], and the use of chemical pesticides [8]. However, the incidence and severity of wilt have also been increasing because it is extremely difficult to eradicate FOM from the soil. Additionally, land resources and costs are limited. Moreover, long-term use of chemicals has polluted the environment and caused resistance in pathogenic bacteria [9]. Consequently, there is a growing need to explore alternative, environmentally sustainable, and effective methods for disease control, and addressing plant wilt has become an urgent matter.

Biological control agents, such as Trichoderma spp., are commonly used in agricultural production due to their environmentally friendly nature and effectiveness in controlling various pathogens [10,11]. In particular, T. harzianum is renowned for its antagonistic properties against a range of pathogenic microorganisms, making it a focal point of research for combating melon wilt [12]. As of 2024, there are more than 80 kinds of T. harzianum biopesticide and soil conditioner products registered in China, covering a variety of types such as single-component preparations and compound microbial products, which can meet different application methods and market needs (China Pesticide Industry Association, 2024 Report, https://www.ccpia.org.cn/, accessed on 1 March 2025). In recent years, most studies have demonstrated the efficacy of T. harzianum in enhancing plant resistance to soil-borne disease, such as stimulating plant root growth [13], triggering the plant immune response [14], and directly suppressing the proliferation of pathogenic microorganisms [5], and the use of T. harzianum has been an effective means of mitigating soil-borne disease. However, the mechanism of improving the resistance of melon to FOM via treatment with T. harzianum was not clear.

Soil enzymes and microbial communities play crucial roles in ecosystems and interact to influence the soil environment. Long-term continuous cropping leads to changes in soil enzyme activity and soil physicochemical properties [15], which in turn alter soil microbial composition and reduce the diversity of soil microbial communities [16]. This shift creates conditions where certain microbial species, such as pathogenic microorganisms, may come to dominate [17], while the populations of beneficial microorganisms (e.g., facultative and antagonistic bacteria) are drastically reduced [18], and the natural presence of antagonisms in the soil is weakened [19]. For instance, as beneficial microbial populations dwindle, their ability to outcompete pathogens for specific resources (a key aspect of competitive exclusion) is impaired, and their capacity to occupy overlapping niches that would otherwise restrict pathogenic expansion is diminished. This dual breakdown not only weakens the inhibition of pathogens by beneficial microorganisms but also reduces the activity of soil defense enzymes [20], ultimately making it easier for pathogenic microorganisms to spread and infest crops. It also exacerbates the deterioration of soil physicochemical properties, including soil acidification, salinization, and reduction of organic matter [9]. Previous studies have found that T. harzianum application promoted the recruitment of soil-beneficial microorganisms (Trichoderma sp. t102, Aspergillus sp. t35, Actinomycetes) in cabbage [21] and sweet sorghum [14], but such effects are inherently driven by its strain-specific biocontrol traits. Distinct T. harzianum strains differ in shaping microbial assemblages: high chitinase-secreting strains (e.g., T22 [22]) preferentially boost Actinomycetes to enhance nutrient-related enzyme activities, while rhizosphere-competent strains (e.g., alkaline-tolerant T34 [23]) more effectively recruit Aspergillus sp. t35, promoting organic matter and nitrogen accumulation [24]. Strains with extensive niche overlap with beneficial taxa induce stronger microbial recombination, and strain-dependent recruitment and soil regulation determine the formation of a healthy soil microbiome, prompting the construction of an antagonistic network and demonstrating superior pathogen inhibition.

Numerous reports have shown that T. harzianum may effectively prevent the infection of pathogens. Nevertheless, there are fewer studies on the mechanism of T. harzianum DQ002 to Fusarium wilt in oriental melons, and the mechanism of action of T. harzianum is still not fully understood. This study aims to clarify the antagonistic mechanism of T. harzianum DQ002 against FOM under continuous cropping conditions by investigating the changes in enzyme activities, physicochemical properties, and microbial community structure abundance in oriental melon rhizosphere soil and their correlation among these indices under the application of T. harzianum DQ002.

2. Materials and Methods

2.1. Material

The “L-2” oriental melon cultivar is a self-pollinating line susceptible to Fusarium oxysporum f. sp. melonis. Trichoderma harzianum DQ002 (Serial number: KX63 2592.1) was isolated from soil and provided by the College of Horticulture and Landscape Architecture, Heilongjiang Bayi Agricultural University. The soil was collected from 8 years of continuous cultivation of melons in the greenhouse in Daqing (N 46°35′21.408″ and E 125°10′25.319″), containing 1.58 mg/kg of ammoniacal nitrogen, 6.72 mg/kg of nitrate nitrogen, 75.34 mg/kg of available phosphorus, and 140.8 mg/kg of available potassium. The pH of the soil was 7.65, and the electrical conductivity (EC) of the soil was 233 µS/cm.

2.2. Experiment Design

Pot experiments were conducted at the greenhouse from May to September 2023, at Heilongjiang Bayi Agricultural University, Daqing, China (N 46°35′21.408″ and E 125°10′25.319″). The experiment consisted of two treatments: (1) water control (CK); and (2) T. harzianum inoculation, a concentration of 10⁸ spores/g soil (MM). Three replicates were carried out for treatments, and each replicate consisted of 50 pots in a north–south arrangement, with 10 pots in each column. After sampling every 10 days, pot positions were changed back and forth to reduce positional effects caused by uneven light and temperature.

Oriental melon seeds, with the same fullness or without breakage, were selected for germination at 28 °C and sown in seedling trays with sand. At the three-true-leaf stage, the seedlings were transplanted into a pot containing 5 kg of mono-cropping melon soil (20 × 20 × 30 cm), and the pots were transferred to the greenhouse under a 16/8 h light/dark regime with a day temperature of 28 °C and a night temperature of 18 °C. When melon seedlings grew to the four-leaf stage, T. harzianum spore suspension with a concentration of 1 × 10⁹ spore/mL was used for root irrigation treatment, with 500 mL of root irrigation per plant; that is, the soil corresponding to MM treatment in the experiment had a spore concentration of 10⁸ spore/g soil (using solid-state fermented spore powder (containing 2 billion spores per gram), first dissolved in water and stirred uniformly, and then counted by a hemocyte counter to ensure accurate concentration and uniform dispersion). The control group (CK) was treated with 500 mL of water. The suspension of T. harzianum spores was inoculated every 20 days. Three inoculations of T. harzianum were made throughout the experimental period (0, 20, and 40d).

2.3. Plant and Soil Sample

Soil samples were collected at 10, 20, 30, 40, 50, and 60d (being sampled firstly at 20 and 40d, followed by inoculation with T. harzianum) after inoculation with T. harzianum or water. The choice of sample schedule (at 10, 20, 30, 40, 50, and 60d) and decision to avoid sampling after inoculation (0, 20, 40d) is generally intended to capture the temporal dynamics of microbial interactions and soil responses after T. harzianum application, rather than the immediate, transient effects of the inoculum itself.

At the same time, the plant height, stem thickness, root length, and leaf area of the collected melon samples were measured. To collect soil samples, the entire root system of melon seedlings was dug up, soil on the surface of the root system being removed, and the collected soil around the root system was sieved (20 mesh) and stored at −80 °C.

The Fusarium wilt disease rate and disease severity were investigated at 30, 50, and 60d after treatments. According to the wilting conditions of the plants, the severity of the disease was classified into five grades [24]:

Grade 0: healthy plant, no symptoms (leaves fully turgid, stems upright).

Grade 1: wilting is limited to ≤20% of the plant, starting from the bottom leaves. Wilting is defined as leaves losing turgor, drooping, and failing to recover after rehydration (confirmed by comparison with healthy leaves of the same age). Disease progression begins from the lowermost leaf, and the affected area is calculated as the percentage of wilting leaves out of the total leaves (for example, 3 wilting leaves out of 15 total leaves equals 20%).

Grade 2: the wilting range involves 21–40% of the plant, spreading upward from the bottom. At this time, the lower leaves may be affected by yellowing in addition to wilting, but the upper 60% of the leaves remain upright.

Grade 3: The wilting range reaches 50% or more of the plant, most leaves from the bottom to the middle layer are wilted and may shrink, and only the top 1–2 layers of leaves remain upright.

Grade 4: the whole plant wilts and dies.

Observer calibration was performed before formal scoring to ensure consistency: three investigators independently scored plants with known disease conditions, and formal investigations were carried out after good agreement; blinded scoring was used (the treatment group was not informed) to reduce subjective bias.

The calculation formula is as follows:

Disease rate = (number of wilted plants)/(total number of plants investigated) × 100%

Disease severities = (∑(Number of diseased plants at each level × Corresponding level value))/(Highest level value × Total number of plants surveyed) × 100%

2.4. Soil Enzymatic Activity and Physical and Chemical Properties

The activities of soil dehydrogenase, sucrase, urease, alkaline phosphatase, and polyphenol oxidase were measured by the TTC (2,3,5-triphenyl tetrazolium chloride) reduction method, the 3,5-dinitrosalicylic acid colorimetric method, the phenol-hypochlorite colorimetric method, the Sodium benzene phosphate colorimetric method, and pyrogallol colorimetry, respectively. The unit of enzyme activity is expressed as mg g⁻¹ h⁻¹. The soil organic matter content and available phosphorus content were determined using the externally heated potassium dichromate volumetric method and the extraction–molybdenum antimony anti-colorimetric method, respectively. The contents of nitrate and ammonium in soil were determined by UV spectrophotometry and phenol-hypochlorite colorimetry, respectively. Soil pH (soil/water = 1:2.5, w/v) was measured using a pH meter, and soil electrical conductivity (soil/water = 1:2.5, w/v) was determined using a conductivity meter.

2.5. Soil DNA Extraction

Total genomic DNA of 0.5 g soil samples was extracted using the OMEGA Soil DNA Kit (M5636-02) (Omega Bio-Tek, Norcross, GA, USA), following the manufacturer’s instructions, and stored at −20 °C before further analysis. The quantity and quality of extracted DNAs were measured using a NanoDrop NC2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis, respectively.

2.6. 16S rRNA and ITS Gene Amplicon Sequencing

PCR amplification was conducted on the V4V5 region of the bacterial 16S rRNA gene using forward primer 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and reverse primer 907R (5′-CCGTCAATTCMTTTRAGTTT-3′) [25]. Similarly, amplification of the ITS1 region of fungi was performed using forward primer ITS5F (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and reverse primer ITS1R (5′-GCTGCGTTCTTCATCGATGC-3′) [26]. Sample-specific 7 bp barcodes were incorporated into the primers for multiplex sequencing. PCR amplicons were purified with Vazyme VAHTSTM DNA Clean Beads (Vazyme, Nanjing, China) and quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, USA). After the individual quantification step, amplicons were pooled in equal amounts, and pair-end 2 × 250 bp sequencing was performed using the Illumina NovaSeq platform with NovaSeq 6000 SP Reagent Kit (500 cycles) at Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China).

2.7. Bioinformatics and Statistical Analysis

Sequence data analyses were primarily conducted using QIIME2 and R packages (v3.2.0). Taxonomic assignment of ASVs was performed against the SILVA database for microbial identification. ASV-level alpha diversity indices—including Chao1, Observed species, Shannon, and Simpson—were calculated from the ASV table in QIIME2. Ranked abundance curves at the ASV level were generated to compare the richness and evenness of ASVs across samples. Beta diversity analysis examined microbial community structural variations across samples using Jaccard [27], Bray–Curtis [28], and UniFrac distance metrics [29,30], and results were visualized via PCoA, NMDS, and UPGMA hierarchical clustering [31]. Principal component analysis (PCA) was also conducted based on the genus-level compositional profiles [31]. Differences in microbiota structure among groups were tested for significance using PERMANOVA [32], ANOSIM [33,34], and Permdisp [32] in QIIME2. Taxonomic compositions and abundances were visualized with MEGAN [35] and GraPhlAn [36]. A Venn diagram (using R package “VennDiagram”) visualized shared and unique ASVs among samples/groups, based on ASV occurrence regardless of relative abundance [37]. ASV-level taxonomic abundances were statistically compared among samples/groups using MetaGenomeSeq and displayed as Manhattan plots [38]. OPLS-DA (a supervised model implemented via R package “muma”) was used to reveal microbiota variations among groups [39]. Nested stratified five-fold cross-validation was used for automated hyperparameter optimization and sample prediction. Co-occurrence network analysis was performed via SparCC with a pseudocount of 10⁻⁶. The correlation coefficient cutoff (0.70) was determined using random matrix theory via the R package RMThreshold. The network was constructed with nodes as ASVs and edges as correlations, visualized using R packages igraph and ggraph. Microbial functions were predicted by PICRUSt2 [40] against MetaCyc and KEGG databases.

3. Results

3.1. The Effect of T. harzianum on the Growth of Melons and the Disease Incidence of Melon Fusarium Wilt

During the growth period of melons, in comparison with CK, the height of melon plants in MM treatment was significantly higher than that of CK at 10 and 20d (Figure S1b), and there was a remarkable increase in stem diameter at 20 and 30d (Figure S1a), and a noticeable increase in root length and both fresh and dry weights at 10, 20, and 30d (Figure 1a,b,d) (p < 0.05). Additionally, the incidence and severity of melon Fusarium wilt significantly decreased in MM treatment at 30, 40, 50, and 60d, compared to CK (p < 0.05). Especially at 60d, the incidence of melon Fusarium wilt in CK reached 90.57%, while being 48.72% in MM, and the severity of melon Fusarium wilt in CK amounted to 65.91%, while being decreased to 20.45% in MM (Figure 1c,d).

3.2. The Effect of T. harzianum on the Soil Microenvironment

According to these findings, soil dehydrogenase activity in MM consistently exceeded that of CK, reaching a peak at 30d, with a noticeable difference between MM and CK at 10d (Figure 2a). Moreover, the effect of T. harzianum application on soil sucrase activity was similar to that of soil dehydrogenase activity (Figure 2b) (p < 0.05). Compared to CK, MM significantly promoted soil urease activity at 10, 30, 50, and 60d (Figure 2c), the soil alkaline phosphatase activity in MM was substantially higher than that of CK at 20d (Figure 2d), and the activity of soil polyphenol oxidase was also enhanced by the MM, showing a significant increase at 20 and 50d, but decreasing at 60d (Figure 2e) (p < 0.05). In addition, the physicochemical properties of different treated soils had significant changes in each period. Specifically, the soil consistently maintained higher EC and pH values in MM than in CK (Figure 2f and Figure S2d) (p < 0.05). Compared with CK, the available phosphorus content in MM soil decreased significantly at 10, 20, 40, and 50d (Figure 3a), and the soil nitrogen content increased significantly at 10, 30, 50, and 60d, but decreased significantly at 40d (Figure 3b), and the content of soil ammonium nitrogen was significantly lower than CK at 20, 40, and 60d (Figure S2a), but the content of soil nitrate nitrogen was significantly higher than CK at 10, 30, and 60d (Figure S2b) (p < 0.05). The soil organic matter content also decreased significantly at 20, 30, 40, 50, and 60 d (Figure S2c) (p < 0.05). The phosphorus and nitrogen contents in melon plants of the MM group were higher than those of the CK group in all periods (Figure 3c,d) (p < 0.05).

3.3. Microbial Community Diversity and Structure in the Melon Rhizosphere Soil

3.3.1. Alpha-Diversity of the Soil Microbial Community

As shown in

Table 1. Alpha diversity of bacterial communities (at the cutoff level of 5%).

Sample	Chao1	Observed	Pielou	Shannon	Simpson	Coverage (%)
CK-10	1569.2 ± 26.6 b	1439.6 ± 11.5 b	0.9 ± 0.0 a	9.5 ± 0.0 b	0.9 ± 0.0 a	98.6
MM-10	1724.2 ± 18.2 a	1584.3 ± 5.5 a	0.9 ± 0.0 a	9.6 ± 0.0 a	0.9 ± 0.0 a	98.5
CK-20	1603.6 ± 4.1 b	1512.8 ± 16.2 a	0.9 ± 0.0 a	9.6 ± 0.0 b	0.9 ± 0.0 a	98.1
MM-20	1638.5 ± 0.3 a	1534.8 ± 12.0 a	0.9 ± 0.0 a	9.7 ± 0.0 a	0.9 ± 0.0 a	98.4
CK-30	1737.9 ± 3.1 b	1643.1 ± 14.5 b	0.9 ± 0.0 a	9.7 ± 0.0 a	0.9 ± 0.0 a	98.4
MM-30	1881.5 ± 5.5 a	1692.2 ± 5.4 a	0.9 ± 0.0 b	9.7 ± 0.0 a	0.9 ± 0.0 a	98.1
CK-40	2028.4 ± 13.4 a	1761.6 ± 8.2 a	0.9 ± 0.0 a	9.8 ± 0.0 a	0.9 ± 0.0 a	98.0
MM-40	1927.4 ± 19.1 b	1755.9 ± 12.5 a	0.9 ± 0.0 a	9.8 ± 0.0 a	0.9 ± 0.0 a	98.2
CK-50	1752.0 ± 11.9 a	1659.2 ± 11.6 a	0.9 ± 0.0 a	10.0 ± 0.0 a	0.9 ± 0.0 a	98.6
MM-50	1551.6 ± 17.8 b	1427.7 ± 8.3 b	0.9 ± 0.0 a	9.8 ± 0.0 b	0.9 ± 0.0 a	99.0
CK-60	1756.3 ± 15.9 a	1680.0 ± 9.9 a	0.9 ± 0.0 a	10.0 ± 0.0 a	0.9 ± 0.0 a	98.5
MM-60	1736.2 ± 7.3 a	1660.6 ± 14.6 a	0.9 ± 0.0 a	9.9 ± 0.0 a	0.9 ± 0.0 a	98.6

Values (mean ± SD) with different letters above the bar are significantly different at p < 0.05 (Welch’s t test; Tukey’s Honestly Significant Difference Test), n = 3. Different lowercase letters in the chart indicate significant differences between groups, and the same letter indicates no significant differences. CK: water; MM: inoculated with T. harzianum DQ002.

, the Chao1 index and Observed species index of the bacterial community in MM were notably higher than those of CK at 10, 20, and 30d, while being remarkably lower than that of CK at 50d (p < 0.05). In MM treatment, the Shannon diversity index of the bacterial community displayed a significant increase at 10 and 20d, following a notable decrease at 50d, and the Pielou index of the bacteria was significantly decreased at 30d after MM treatment compared to CK (p < 0.05). As shown in

Table 2. Alpha diversity in fungal communities (at the cutoff level of 5%).

Sample	Chao1	Observed	Pielou	Shannon	Simpson	Coverage (%)
CK-10	357.2 ± 9.2 a	356.6 ± 9.3 a	0.6 ± 0.0 a	5.6 ± 0.2 a	0.9 ± 0.0 a	99.9
MM-10	232.1 ± 9.2 b	224.7 ± 5.4 b	0.2 ± 0.0 b	1.7 ± 0.1 b	0.3 ± 0.0 b	99.9
CK-20	378.1 ± 7.2 a	377.0 ± 7.4 a	0.7 ± 0.0 a	6.0 ± 0.1 a	0.9 ± 0.0 a	99.9
MM-20	250.1 ± 8.7 b	260.5 ± 14.4 b	0.3 ± 0.0 b	2.9 ± 0.2 b	0.5 ± 0.0 b	99.9
CK-30	361.0 ± 10.1 a	373.2 ± 5.6 a	0.5 ± 0.0 a	5.5 ± 0.2 a	0.9 ± 0.0 a	99.9
MM-30	238.1 ± 3.2 b	237.5 ± 3.2 b a	0.4 ± 0.0 b	3.0 ± 0.1 b	0.6 ± 0.0 b	99.9
CK-40	289.9 ± 4.1 a	288.4 ± 3.3 a	0.6 ± 0.0 a	4.9 ± 0.1 a	0.9 ± 0.0 a	99.9
MM-40	233.8 ± 5.5 b	233.0 ± 5.8 b	0.4 ± 0.0 b	3.2 ± 0.1 b	0.6 ± 0.0 b	99.9
CK-50	402.5 ± 5.9 a	391.2 ± 7.3 a	0.6 ± 0.0 a	6.0 ± 0.1 a	0.9 ± 0.0 a	99.9
MM-50	260.9 ± 7.5 b	254.5 ± 6.5 b	0.4 ± 0.0 b	3.2 ± 0.0 b	0.6 ± 0.0 b	99.9
CK-60	372.1 ± 11.9 a	376.7 ± 0.1 a	0.6 ± 0.0 a	5.4 ± 0.0 a	0.9 ± 0.0 a	99.9
MM-60	249.9 ± 2.9 b	245.8 ± 5.4 b	0.3 ± 0.0 b	3.0 ± 0.1 b	0.6 ± 0.0 b	99.9

Values (mean ± SD) with different letters above the bar are significantly different at p < 0.05 (Welch’s t test; Tukey’s Honestly Significant Difference Test), n = 3. Different lowercase letters in the chart indicate significant differences between groups, and the same letter indicates no significant differences. CK: water; MM: inoculated with T. harzianum DQ002.

, Chao1, Observed species, Shannon, Pielou, and Simpson diversity index of the fungal community markedly decreased in MM, compared to CK at six periods (p < 0.05).

Furthermore, the changes in microbial community diversity promoted the restructuring of the modular structure of the microbial interspecies co-occurrence network. As shown in Table S1, the CK network comprises 405 nodes and 289 edges, with a modularity index of 0.987 and an average clustering coefficient of 0.846. The network contains 23.95% fungi and 76.05% bacteria. The MM network included 297 nodes and 213 edges, a 0.98 modularity index and 0.899 average clustering coefficient, and contained 14.48% fungi and 85.52% bacteria. Rhizosphere microbial interkingdom co-occurrence association networks indicated that there was a decrease in fungal taxa after inoculation with T. harzianum, compared with the CK network. The bacterial taxa in the MM treatment network increased, and the number of connections between nodes rose, and the interaction between microorganisms was more frequent and closer (Figure 4).

3.3.2. Relative Abundance of Soil Microbial Communities

Following the application of T. harzianum, the composition of the soil community changed. The MM group had lower relative abundances of bacterial and fungal communities than those of CK at different taxonomic levels (Table S2). At different periods, the bacterial community was dominated by Proteobacteria, Actinobacteriota, Acidobacteriota, Chloroflexi, Gemmatimonadota, and Bacteroidota (Figure S3a), while the fungal community was mainly composed of Ascomycota, Basidiomycota, Mortierellomycota, and Rozellomycota (Figure S3b). At the dominant bacterial phylum level (Figure S5a), the relative abundance of Proteobacteria increased significantly in MM except at 60 days, and the relative abundance of Acidobacteriota was markedly decreased in comparison with CK (p < 0.05), while the relative abundance of Actinobacteria and Gemmatimonadota showed dynamic fluctuations. At the dominant fungal phylum level (Figure S5b), the relative abundance of Ascomycota in MM treatment increased dramatically compared to CK, while the relative abundance of Basidiomycota, Mortierellomycota, and Rozellomycota was all observably reduced (p < 0.05). Additionally, we found that Vicinamibacteraceae, Lysobacter, RB41, JG30-KF-CM45, and S0134_terrestrial_group were prevalent in the bacterial genus (Figure S3c), and Trichoderma displayed the highest relative abundance in MM treatment amongst fungal genera (Figure S3d). In the dominating bacterial genus (Figure 5a), compared with CK, in the MM treatment, the relative abundances of Vicinamibacteraceae and RB41 decreased significantly in each period, while Lysobacter showed a remarkable increase in relative abundance at 10, 40, 50, and 60d (p < 0.05). The relative abundance of S0134_terrestrial_group was significantly higher than that of CK at 20 and 60d (p < 0.05). In addition, the relative abundance of Trichoderma signally increased in the dominant fungal genus (Figure 5b), while the relative abundances of Mycochlamys, Botryotrichum, and Fusarium reduced significantly (p < 0.05).

3.4. Correlation Between Soil Microbial Community and Environmental Factors

Various environmental factors typically regulate the changes in microbial community structure. Principal Coordinate Analysis (PCoA) revealed that T. harzianum treatment altered the functional structure of the soil microbial community, particularly the fungal community, which corresponded to the observed Fusarium wilt severities. According to PCoA, the second principal component (PCo2) of the rhizosphere bacterial communities between the MM treatment and CK differentiated the two groups at different growth stages with a variance contribution value of 6%, whereas the first principal component (PCo1) of rhizosphere bacterial communities between the MM treatment and CK was not significant (Figure 6c). Conversely, the principal component plot of the fungal community showed clear separation (Figure 6d), showing that the inoculation with T. harzianum induced a significant restructuring of the rhizosphere soil fungal microbial community in melons, accounting for 44.9% of the variation (PCo1 37% and PCo2 7.9%).

Correlation analysis was performed to clarify the relationship between microbial communities and various environmental factors. The beta diversity of bacterial community was extremely negatively correlated with Fusarium wilt severity (Figure 6b). In contrast, the alpha diversity of bacterial and fungal communities was not correlated with the disease severity (Figure 6a). In addition, there were different intensities of correlation between various levels of microbial communities and environmental factors. The bacterial phylum, bacterial genus, and fungal genus all showed significant negative correlations with Fusarium wilt severity (Figure S5a,b). Of these, the Trichoderma genus showed a substantial negative correlation with disease severity (Figure S5c). Concretely speaking, amongst the bacterial and fungal phyla, it was observed that disease severity exhibited positive correlations with Acidobacteriota and planctomycetota, while displaying negative correlations with Proteobacteria, Aphelidiomycota, and Kickxellomycota (Figure S6a,b). In bacterial and fungal genera, it was found that disease severity revealed a positive correlation with Vicnamibacteraceae, RB41, Plectosphaerella, and Fusarium, as well as displaying a negative correlation with Lysobacter, Mycochlamys, and Trichoderma (Figure S6c,d). The correlation analysis among environmental factors indicated that disease severity showed a significant negative correlation with sucrase activity, EC, and nitrate nitrogen, and a highly significant positive correlation with soil polyphenol oxidase activity and soil organic matter (Figure 6e). The results implied that the content of soil organic matter decreased with increasing Trichoderma abundance, and that soil EC and sucrase activity increased, which promoted changes in the abundance of bacteria and fungi. The relative abundance of Plectosphaerella, Fusarium, Vicinamibacteraceae, and RB41 was inhibited, and the relative abundance of Lysobacter and Mycochlamys increased, further reducing the disease severity (Figure 6f).

3.5. Functional Predictions of Soil Microbial Communities in Rhizosphere Soil

The functional unit compositions of the bacterial and fungal colonies were predicted using the sequencing results of the 16S rRNA gene, and subsequent statistical analysis identified the presence of seven metabolic pathways in bacteria and five metabolic pathways in fungi (Figure S7a,b). Biosynthetic pathways that were significantly expressed in bacteria and fungi, namely Amino Acid Biosynthesis, Cofactor, Prosthetic Group, Electron Carrier, Vitamin Biosynthesis and Nucleoside and Nucleotide Biosynthesis were found. These pathways were substantially clustered within the overall Biosynthesis pathway category.

Further analysis of the differential metabolic pathways (Figures S8 and S9) in bacteria and fungi was conducted at each period. We found that differential metabolic pathways were significantly enriched in bacteria and fungi at six periods after T. harzianum inoculation (Figure S10a,b). The differential metabolic pathways of bacteria were significantly enriched in Glycine betaine degradation I, Enterobacterial common antigen biosynthesis, Superpathway of taurine degradation, Superpathway of L-tryptophan biosynthesis, Superpathway of lipopolysaccharide biosynthesis, and L-glutamate degradation V. The differential metabolic pathways of fungi were significantly enriched in Octane oxidation, Adenosine ribonucleotides de novo biosynthesis, Pentose phosphate pathway, Superpathway phosphatidate biosynthesis (yeast), CDP-diacylglycerol biosynthesis I, NAD/NADH phosphorylation and dephosphorylation. These metabolic pathways may be involved in regulating plant growth, immune function and information transmission, and may be significantly related to plant resistance. Among them, differential metabolic pathways in bacteria are associated with the relative abundance of RB41, Lysobacter, and Vicinabacteraceae, while the metabolic pathways in fungi are associated with the relative abundance of Trichoderma, Fusarium, Plectosphaerella, Mycochlamys, and Verticillium. Furthermore, Trichoderma was able to colonize the rhizosphere soil of melon in MM and reach its maximum relative abundance there, participating in the metabolic pathway of individual fungi (Figure S10).

4. Discussion

T. harzianum, a filamentous fungus, is widely used not only as a biocontrol agent against phytopathogenic fungi but also as a growth promoter for plants [41]. T. harzianum has been reported to promote the growth and improve the quality of plants, such as cucumber [24], tomato [42], pepper [43], strawberry [44], and wheat [45], and so on. In particular, Trichoderma can produce phytoalexin, stimulate the division of apical meristem cells, and increase the number of lateral roots and root hairs. For example, anthranilic acid (2-AA) secreted by Trichoderma strains interacts with plant auxin signaling and transport networks to promote lateral root formation [46]. The present study demonstrated that T. harzianum promoted the growth of melon plants (Figure 1). The mechanism of promoting growth was not well clarified, which may be related to plant metabolites, such as the hormone signal transduction pathway, which need to be further studied in the future.

Research has shown that with continuous melon farming, soil secondary salinization and acidification intensified, and that soil enzyme activities and physicochemical properties weakened [47]. The material foundation and source of energy for soil enzymes is provided by soil nutrients. According to an increasing amount of research, applying Trichoderma spp., including T. viride T23 [48], T. harzianum LTR-2 [49], and T. harzianum MC2 [50], can improve the nutrient content of the soil, alleviate continuous cropping disorders, and encourage the activities of soil enzymes. For example, T. asperellum effectively improved soil nutrients, enhanced enzyme activities, and successfully regulated the soil ion balance in the rhizosphere of maize seedlings [51]. In this study, the activities of soil dehydrogenase, sucrase, and alkaline phosphatase increased significantly after T. harzianum inoculation (Figure 2), which participated in the transformation cycle, the oxidative decomposition of organic materials, and the efficient use of carbon, nitrogen, and phosphorus. Applying T. harzianum accelerated the decomposition and utilization of organic matter in the soil, improved soil enzyme activity, thus promoting the increase of available carbon, nitrogen, and phosphorus content in the soil (Figure 3a,b), and some inorganic ions were released, which significantly increased the soil EC value (Figure 2f). Many parties worked together to activate soil nutrients and improve the melon plants’ absorption rates of nitrogen and phosphorus (Figure 3c,d). Meanwhile, the application of T. harzianum also induced systemic resistance in plants [12], enhancing their resistance to diseases. It is speculated that T. harzianum itself can directly secrete functional enzymes to participate in soil material transformation, which is the most direct way to affect soil enzyme activity. By secreting volatile organic compounds (VOCs) or antibacterial substances, it can inhibit pathogenic bacteria while promoting the reproduction of beneficial microorganisms such as actinomycetes and Bacillus, reshape the soil microbial community structure and community interactions, indirectly drive enzyme activity changes, establish a symbiotic relationship with plant roots, and induce enzyme secretion through “plant–microbial signals”. The mode of action of microorganisms influenced by T. harzianum is unclear, such as secreted enzyme-inducing substances, their synthesis pathways, or the mechanism of action of related signal molecules. Furthermore, soil pH and organic matter content can significantly influence soil nutrient availability and microbial activity [48]. The study has indicated that Trichoderma can secrete organic acids, decreasing soil pH, facilitating the transformation of soil organic residues, and increasing soil organic matter content [52]. This study found that the soil pH value of the MM group was significantly higher than that of CK (Figure S2d), while the soil organic matter content was lower than that of CK (Figure S2c), which was inconsistent with previous research findings. The observed changes may be presumed to be in the flourishing period of melon growth at the time of sampling. The short-term effect of MM treatment may mask the long-term trend that plants have a surge in nutrient demand. MM treatment may accelerate the mineralization of organic matter (converted into inorganic nitrogen and phosphorus) by promoting microorganisms to meet plant demand, resulting in a temporary decrease in organic matter; during the mineralization process, microorganisms metabolize to produce alkaline substances (such as ammonification releasing NH₃), which increases soil pH in the short term; while CK is still in the “accumulation state” due to slow mineralization of organic matter, so it shows a trend of low pH and high organic matter. In addition, this trend change may also be that MM treatment enriches specific functional microbial groups (in contrast to previous studies), that is, T. harzianum DQ002 preferentially recruits alkali-producing microorganisms (such as actinomycetes and pseudomonas) in MM treatment, and their metabolic process releases ammonia (NH₃) or carbonate, directly increasing pH. At the same time, these microorganisms efficiently decompose organic matter (such as cellulose and lignin as carbon sources), resulting in a decrease in organic matter content. Subsequent validation should be combined with metatranscriptomic and metabolomic analyses.

Soil microbial diversity plays a critical role in sustaining ecosystem equilibrium and rehabilitating ecological functions in soils [53]. A high diversity of soil microorganisms contributes to the presence of various antagonistic species, thereby enhancing plant disease resistance [54]. Abundant soil microbial diversity is correlated with increases in the species and quantity of antagonistic bacteria, including Bacillus [55], Pseudomonas [56], Actinomyces, and Burkholderia [57]. These bacteria can utilize resource competition or produce inhibitory substances, which directly or indirectly regulate the growth and reproduction of pathogenic bacteria, such as Fusarium oxysporum [58]. Furthermore, high microbial diversity offered opportunities for microbial interactions and contributed to the stability of symbiotic relationships [59]. For instance, rhizobia and phosphorus-dissolving bacteria [60] effectively collaborate with plant roots to facilitate plant growth and nutrient uptake. This study showed a significant increase in bacterial diversity and a significant decrease in fungal diversity in MM (Table 1). Additionally, the soil microbial community shifted from fungal to bacterial dominance, which was consistent with previous studies [59]. The strong antagonistic effect of T. harzianum on other pathogenic fungi could rely on the secondary metabolites secreted by T. harzianum to reduce the spore germination of soil pathogenic fungi [61]. This phenomenon can attract bacteria to proliferate in melon root soil, thereby promoting an increase in bacterial diversity. Further investigation may focus on the specific pathways of the symbiotic relationship between Trichoderma and beneficial bacteria, including material exchange.

Soil microorganisms and the soil environment are the key factors that can induce plant disease resistance. The diversity, structure, and abundance of soil microbial communities are related to the soil environment, and there exists a remarkably close interaction between them [62]. Previous studies have demonstrated a substantial relationship between plant disease severity, the soil environment, and the relative abundance of soil microbial communities [63]. For example, the endophytic bacterium Lysobacter firmicutimachus was linked to rice root rot [64]. This study found that there were significant negative correlations between disease severity and soil sucrase activity, EC value and nitrate nitrogen content, and a positive correlation with soil organic matter (Figure 6e). Additionally, disease severity exhibited a positive correlation with the relative abundance of RB41, Plectosphaerella, Fusarium and Vicinamibacterace, while showing a negative correlation with the relative abundance of Lysohacter (Figure 6f), aligning with findings from prior studies. The MM treatment resulted in a significant decrease in the relative abundance of Acidobacteria within bacterial communities. The reduction in relative abundance suggested alterations in the recycling pathways of related substances in the soil due to the influence of T. harzianum. Vicinamibacteraceae, as acidophilic bacteria, live in acidic environments and usually participate in cycling processes of carbon, nitrogen, and other substances, and promote plant nutrient uptake in the soil ecosystem. The relative abundance of acidophilic bacteria in soil was significantly negatively correlated with soil pH [65]. In this study, the results demonstrated that inoculation with T. harzianum increased the soil pH, creating an alkaline soil environment, while the relative abundance of acidic bacilli decreased, with a positive correlation observed between acidic bacilli and disease severity. Nevertheless, the study indicated a negative correlation between acidic bacilli and Fusarium oxysporum [4], which was contrary to previous results. Based on our findings, it is speculated that this phenomenon may be related to the alternative mechanism of the microbial interaction network. Acidic bacilli may produce growth factors (such as the vitamin B group) required by Fusarium oxysporum by decomposing complex organic matter, while metabolites of Fusarium oxysporum (such as certain organic acids) can reduce local microenvironmental pH and feed back to acidic bacilli that prefer acidity, forming a “mutualistic cycle”. It may also be related to special metabolites secreted by T. harzianum. In addition, there is a potential limitation in this study: that is, the sampling sites in this study are concentrated in soil microdomains (such as specific areas in the rhizosphere), while previous studies considered bulk soils, which may lead to a synchronous increase in the abundance of the two in the local environment (such as a patch that simultaneously enriches the complex polysaccharides preferred by acid bacilli and host root exudates of Fusarium oxysporum), forming a “false positive correlation”. Moreover, small sample sizes may amplify random fluctuations, such as accidental collection of Fusarium oxysporum-infected soil (when plant root exudates surge and attract acid bacilli and pathogens), and insufficient repetition fails to offset this possibility. Future investigation may focus on the changes in various soil chemicals after T. harzianum treatment, including the ion concentrations, the trace element levels, and the identification of potential metabolites in soil, and the influence of T. harzianum on acid bacilli and Fusarium oxysporum, and the mechanism of antimicrobial substances, simultaneously expanding spatiotemporal sampling to distinguish contingent associations from stabilization mechanisms.

Previous studies have found RB41 as a predominant bacterium involved in the degradation of sulfadiazine [21], soil heavy metals and other elements [8], and in salt stress mitigation [66], suggesting that RB41 is crucial for the degradation of detrimental substances in soil and for improving plant stress resistance. This study indicated that MM significantly reduced the relative abundance of RB41, possibly due to the overlapping ecological niches of T. harzianum and RB41 or T. harzianum suppressing RB41 by recruiting beneficial bacteria from other genera.

Proteobacteria exhibited a variety of metabolic roles, with certain species involved in nitrogen transformation and the sulfur cycle [67]. For example, Sphingomonadaceae can degrade aromatic compounds in soil, improve the activity of crop antioxidant enzymes, and exhibit resistance to a variety of pathogenic bacteria in soil, and Sphingomonadaceae can maintain soil nitrogen balance and play a crucial role in plant growth [68]. In a study of drought tolerance in maize, Sphingomonas sp. Hbc-6 proved able to regulate maize physiological metabolism while recruiting beneficial rhizosphere bacteria, promoting plant growth and enhancing drought tolerance [69]. In this study, MM treatment significantly increased the relative richness of Sphingomonadaceae in melon rhizosphere soil, facilitating the development and reproduction of Sphingomonadaceae (Figure S11). Sphingomonadaceae can interact with plant roots and may stimulate the production of plant growth hormones, such as salicylic acid, indoleacetic acid, and abscisic acid to promote growth. In Arabidopsis thaliana, the same results have been observed [70], and indoleacetic acid production effectively inhibited the reproduction of pathogenic microorganisms and mitigated biological stress. In this study, Vicinamibacteraceae, Sphingomonadaceae, and RB41 as predominant bacterial families were found, suggesting that there was cooperation and competition within the soil ecosystem. This study demonstrated that T. harzianum inhibited the relative abundance of Vicinamibacteraceae and RB41, while enhancing the relative abundance of Sphingomonadaceae. The signal molecules generated by T. harzianum may positively influence the regulation, mutual recognition, and aggregation of Sphingomonadaceae, or activate certain metabolic pathways. However, the specific mechanism requires additional investigation.

Functional prediction analysis of microbial metabolic pathways found that soil microorganisms significantly changed in multiple metabolic pathways after T. harzianum treatment, which was in agreement with the changes in microbial species composition observed earlier. Amino acids are crucial in metabolic pathways, and participate in various biosynthetic activities, including protein synthesis, energy metabolism, and the creation of precursor synthesis of numerous secondary metabolites [71]. It has been found that amino acids are essential for plants to withstand drought stress, hypothermia, and disease, and act as antioxidants [72,73]. Furthermore, many cofactors and prosthetic groups were required for enzymatic reactions and played key roles in cellular respiratory processes, such as glycolysis, the tricarboxylic acid cycle, and oxidative phosphorylation [74]. Nucleotides and their derivatives were involved in the regulation of gene expression and an active nucleotide biosynthetic pathway, which ensures that cells synthesize enough cAMP and other signaling molecules to enable plants to initiate a series of defense responses [75]. This study revealed that the application of T. harzianum greatly enhanced amino acid biosynthesis, cofactors, cogroups, electron carriers, and vitamin biosynthesis, and nucleoside and nucleotide biosynthesis inside bacterial and fungal metabolic pathways. The metabolites of these metabolic pathways provided the protein precursors for the melon rhizosphere’s microbial growth and reproduction, promoted enzymatic reaction, ensured genetic material synthesis, provided energy, and regulated metabolism, and jointly enhanced the stress resistance of melon plants. In addition, after T. harzianum addition, the root microbial community was screened and reconstructed (Figure 6c,d). The alterations in microbial community composition and density may induce the modifications of root metabolic pathways, resulting in stronger and more effective collaboration among various microorganisms. RB41 participates in the glycolysis pathway, TCA cycle, and nitrogen metabolism pathway, generating substantial energy for various life activities of cells [76]. Lysobacter has the ability to synthesize a variety of secondary metabolites with antibacterial activity, including lysozyme and antibiotics. Simultaneously, Lysobacter participated in the iron metabolism pathway, facilitated intracellular electron transfer, regulated enzyme activity, and contributed to other critical physiological processes [77]. Vicinabacteraceae was involved in the metabolism of fatty acids and amino acids, gradually promoting the conversion of fatty acids to acetyl-CoA through the β-oxidation pathway, which subsequently enters the TCA cycle, while also contributing to the synthesis and catabolism of amino acids [61]. Furthermore, several members of the microbial flora have been found to be capable of synthesizing vitamins and participating in the synthesis and transformation of cofactors, electron carriers, vitamins, and other substances, along with the coenzyme synthesis of various metabolic enzymes [78].

The metabolites of these metabolic pathways initiated a series of signal transduction and stimulated plant immune responses [75]. In this study, the results indicated that RB41, Lysobacter, and Vicinabacteraceae were significantly clustered in the bacterial metabolic pathway (Figure S10a), whereas Trichoderma, Fusarium, Plectosphaerella, Mycochlamys, and Verticillium exhibited significant enrichment in fungal metabolic pathways (Figure S10b). Meanwhile, the relative abundance of disease-associated Fusarium, Plectosphaerella, and Verticillium under T. harzianum treatment were discovered (Figure 5). It was speculated that T. harzianum inoculation may enhance the complex signaling interactions between melon roots and microorganisms, thereby triggering a series of defense responses, including involvement in activating the expression of defense-related genes and stimulating the production of antimicrobial metabolites in melon roots, such as phenols and flavonoids [5]. Subsequently, we will further investigate the effect of the interaction between T. harzianum and rhizosphere microorganisms on the signaling and communication mechanisms of melon roots, or the upstream and downstream regulation of melon defense-related genes, and the mechanism of association with genes of other metabolic pathways.

5. Conclusions

In conclusion, the application of T. harzianum modified the physical and chemical characteristics of soil and enzyme activities, altered the composition of the soil microbial community, and diminished the relative abundance of fungi in the rhizosphere, particularly those linked to soil-borne illnesses, resulting in a transition of soil microorganisms from a “fungal type” to a “bacterial type”, which influenced the metabolic pathways of microorganisms, especially those connected to metabolic pathways of amino acids and cofactors, and fostering a more favorable soil environment for melon cultivation. The application of T. harzianum markedly improved the resistance of oriental melons to Fusarium oxysporum f. sp. melonis (FOM) and mitigated the challenges associated with successive cropping of oriental melons. A theoretical framework has been established for the dissemination and application of T. harzianum to attain environmentally sustainable prevention and control (Figure 7). In addition, the results of this study may be potentially affected by the confounding factor of the difference in soil initial microbial community composition; it is worth noting that in addition to direct antagonism, T. harzianum-induced plant systemic resistance may be another biocontrol pathway (alternative mechanism). This speculation needs to be further verified by transcriptomic analysis.