Alleviating Continuous Monocropping Obstacle in Melon: Biological Elimination of Phenolic Acid


 Background: Melon (Cucumis melo L.) is one of the most important fruit crops grown in China. However, the yield and quality of melon have significantly declined under continuous cropping. Phenolic acids are believed to be associated with the continuous monocropping obstacle (CMO) and can influence plant microbe interactions. Coumaric acid (CA) is one of the major phenolic acids found in melon root exudates. The objectives of this study were to estimate the elimination of CA by the soil bacterium K3 as well as its effects on mitigating melon CMO. CA degradation was investigated by monitoring the CA retained in the growth medium using high performance liquid chromatography (HPLC). The effects of CA and K3 on rhizosphere soil microbial communities were investigated by the spread plate method and Illumina MiSeq sequencing. Furthermore, the effects of CA and K3 on melon seedling growth were measured under potted conditions. The changes in soil enzymes and fruit quality under K3 amendment were examined in a greenhouse experiment. Result:The results suggest that the addition of CA had the same result as the CMO, such as deterioration of the microbial community and slower growth of melon plants. HPLC and microbial analysis showed that K3 had a pronounced ability to decompose CA and could improve the soil microbial community environment. Soil inoculation with K3 agent could significantly improve the fruit quality of melon.Conclusion: Our results show that the effects of K3 in the soil are reflected by changes in populations and diversity of soil microbes and suggest that deterioration of microbial communities in soil might be associated with the growth constraint of melon in continuous monoculture systems.


Background
Intensive farming practice exacerbates the continuous monocropping obstacle (CMO) of crops, and successive cropping of conspeci cs renders soil conditions less suitable for their growth [1]. Potent allelochemicals, phenolic acids are believed to be associated with CMO and have a negative in uence on microbial communities and soil functions [2]. Phenolic acids from root exudates have been identi ed in many plants, such as tobacco [3], peanuts [4], wheat, and watermelon [5,6] .
Minor changes in the composition and quantity of root exudates (such as phenolic acids) can cause signi cant changes in microbial populations in the rhizosphere.. Phenolic acids have an important effect on the growth of plant pathogens, which has been con rmed by many studies. Benzoic acid (0.1 mmol·L − 1 ) could promote the mycelial growth and sporulation of the peanut root rot pathogen Fusarium sp. in vitro [7]. The secretion of p-hydroxybenzoic acid, ferulic acid, and cinnamic acid from the roots of watermelon stimulated Fusarium wilt pathogen Fusarium oxysporum F. sp. niveum spore germination and growth [5]. In addition, root exudates from many crops have been shown to be potentially autotoxic.
High levels of phenolic acids (150 µg·g − 1 and 75 µg·g − 1 ) signi cantly inhibited the growth of peanut plants (such as plant height, root length, and fresh weight) [4]. The addition of phenolic compounds to soil signi cantly reduced the leaf area, plant height and dry weight of cucumber. A similar phenomenon has also appeared in melons. With the increase of the continuous cropping period of melons, the phenolic acid content of the rhizosphere soil and the incidence of fusarium wilt increased signi cantly, and the growth rate of melons decreased signi cantly [8].
Phenolic acid secreted by roots could not only impact the microbial biodiversity and abundance of rhizosphere soil but also manipulate biological and physical interactions between plants and soil microorganisms [9]. The soil microbial community is thought to be responsible for biological processes needed to maintain a healthy soil and suppress plant diseases [10], such as mineral nutrition cycling, organic matter turnover, soil structure formation, and toxin removal [11]. High microbial diversity and appropriate composition play a pivotal role in maintaining soil health and promoting plant growth by preventing pathogen invasion and establishment [12], antagonizing pathogen growth, competing with pathogens for nutrients [13], and modulating the host immunity [14]. In addition, a higher proportion of soil bacteria indicates better soil quality and higher soil nutrient content.
It has been illustrated that biocontrol bacteria could adjust the ratio between soil bacteria and fungi [15]. The rhizosphere microbial community structure that was destroyed by continuous inoculation was able to be repaired with biocontrol agents, as shown by the number of bacteria in rhizosphere soil, which increased markedly under biocontrol applications compared with a control (CK), but the density of pathogen and fungal increased dramatically in the rhizosphere of CK plants [16]. It has been found that microbial community structure was ameliorated by bio-organic fertilizer, which was fermented from Paenibacillus polymyxa, Bacillus subtilis, Penicillium sp., and Aspergillus sp. [8]. Furthermore, consistent overall effects were observed with the biocontrol amendments Rhizoctonia solani and Trichoderma virens on continuous cropping soil of potato, including increased bacterial population [17].
Many studies have shown that allelochemicals and autotoxins are easily degraded by microbes [18]. Thus, the effects of a vast diversity of microorganisms on the fate of plant phenolic compounds (and other potential allelochemicals) found within the soil should not be underestimated. However, under continuous cropping, relatively little is known about the detailed effects of phenolic acids on inhibiting the growth of soil bacteria. There is still no experimental evidence for the ability of biocontrol bacteria to decompose phenolic acids.
Consecutive monoculture of melon (Cucumis melo L.) is widespread in China and responsible for a continuous decline in fruit yield and quality and increased crop susceptibility to diseases [8]. In this study, we assessed the amelioration of K3 on continuous cropping soil of melon and melon plants grown in both pots and in eld conditions. Previous studies on melon root exudates have found that coumaric acid (CA) is the main phenolic acid in melon root exudates [19]. We determined the changes of soil microbes after the amendment of K3 and CA. The elimination of CA by K3 was also measured. The overall goal of this study was to understand the mechanism behind the ability of K3 to overcome the obstacles of the continuous cropping of melon.

Results
Effects of CA on soil microorganisms and melon growth The addition of CA solutions had strong effects on these soil microorganisms (Table 1). In the MB medium, fungal populations were signi cantly higher than those in the CK, whereas the bacterial counts were not signi cantly different compared with CK. The bacteria-to-fungi ratio in CAG treatment was 57.80% lower than that in CK group. In the soil matrix, the addition of CA leads to an increase in the number of bacteria in the soil, and a decrease in the number of fungi. There is also a signi cant decrease in the ratio of bacteria to fungi due to the addition of CA in the soil, but the amplitude is lower than in the MB medium. In soil containing exogenous CA, slow growth of melon plants was observed compared with that of CK1 (Fig. S1, Table 2). The plant height, shoot fresh weight and root fresh weight in the CAG1 treatment were reduced by 31.84%, 33.97%, and 40.91%, respectively, compared with CK1. A lower blade number and root length were also observed in the CAG1 treatment, equivalent to a decrease of 8.77% and 19.13%, compared with CK1 respectively. The addition of CA resulted in signi cantly lower root vigor of muskmelon seedlings than the control group CK1 (Table 2).  (Table S1). At the phylum level, the fungal community was dominated by members of Ascomycota and Zygomycota (Fig. 2b). The most noticeable difference between continuous cropping soil (CK1, T2) and non-continuous soil (CK0) was that the relative abundances of Glomeromycota were signi cantly (P < 0.05) reduced by 99.88% and 99.92%, respectively, while Zygomycota abundances were signi cantly (P < 0.05) increased by 2.89 and 2.74 times, respectively. The abundances of Ascomycota and Basidiomycota in the T2 treatment differed signi cantly (P < 0.05) from that of CK1, equivalent to a 1.52-fold increase and a 92.53% decrease, respectively. Thus, the Ascomycota community was markedly promoted, whereas the Basidiomycota community was inhibited in the presence of K3 (Fig. 2b). At the genus level, the dominant fungi Mortierella, Conocybe, Chaetomium, and Fusarium showed a signi cant change are listed in Table  S2. Conocybe were not found in non-continuous soil (CK0), but were detected in the CK1 and T2 treatment, whereas the relative abundance of Mortierella increased signi cantly (P < 0.05) in CK1 and T2.

The Effect Of K3 On The Soil Microbial Diversity
In general, the OTUs of bacteria and fungi in continuous cropping soil (CK and T2) were lower than those of healthy soil (CK0) (Fig. 3). After the T treatment, the bacterial OTUs of continuous soil increased, and the OTUs a liated to fungi decreased. The unique bacterial OTUs treated by T2 were higher than those of CK1 by 90.00%. Moreover, higher Chao1 and ACE indices of the T2 treatment were shown than those under CK1 ( indicating that the fungi changed from peak abundance state (CK1) to low abundance state (T2) ( Table 3).
From the perspective of fungal community diversity, the Simpson index of the T2 treatment was signi cantly (P < 0.05) lower than that of CK1, demonstrating that the dominant species in CK1 played key roles in the community resulting in a lower community diversity. The prevalence of the dominant species in the fungal community was decreased in the T2 treatment increasing the diversity of fungal communities. The actual effect of K3 on continuous cropping of melon Amelioration on CMO of melon seedlings by K3 addition in pots All of the biomass and agronomic traits of seedlings in CK1 showed signi cantly lower values than in other treatment groups (Fig. S2, Table 4). The plant height, blade number, root length, and fresh weight of shoots and roots in the CK1 treatment groups differed signi cantly (P < 0.05) from those of CK0 resulting in decreases of 51.54%, 18.18%, 85.51%, 60.14%, and 50.00%, respectively.  (Table 4). A signi cant (P < 0.05) increase of the plant weight and blade number was observed in the T2 treatment (Fig. S2), reaching to 1.62-fold and 1.55-fold respectively. The root length in the T2 treatment did not differ signi cantly (P > 0.05) from the CK treatment, resulting in a 1.58-fold increase. The fresh weight of shoots and roots in the T2 treatment signi cantly (P < 0.05) improved by 1.41-fold and 1.50-fold, respectively, relative to the CK1 treatment group ( Table 4).
Effect of K3 on rhizosphere microorganism quantity of melon in the greenhouse Microbial densities, including bacteria and fungi, are shown in Fig. 4. The number of soil bacteria peaked in April and subsequently exhibited a downward trend as the melon plants grew. The densities of soil fungi gradually increased from February to April, and after a drop to the lowest values in June, they rebounded in July. The levels of soil bacteria signi cantly (P < 0.05) increased during the whole growth period of melon after amendment with K3 agent, except in May. The densities of soil fungi in the K3 treatment were lower than those of CK, although they did not signi cantly decrease except for in February. In the root zone soil of melon, the bacteria-to-fungi ratio was higher for the K3 treatment than the CK treatment.

Improvement in soil enzyme activities by K3 under continuous cropping conditions
The soil enzyme activity in the K3T treatment increased substantially compared with CK in the whole melon growing period (Fig. 4). The sucrose activity (Fig. 4a) of the K3-inoculated soil was signi cantly (P < 0.05) higher in February, April, and May, equivalent to increases of 96.11%, 234.15%, and 54.16% compared with that of the CK, respectively. However, no signi cant difference was observed in other months. Higher urease activity was observed in the K3T treatment group. The urease activities (Fig. 4b) in the K3T treatment were signi cantly (P < 0.05) increased by 20.06%, 8.58%, and 18.62% in April, June, and July, respectively, compared to that of the CK. Neutral phosphatase activity (Fig. 4c) showed increases with melon growth, with the lowest and highest values appearing in February and June, respectively.
Signi cant (P < 0.05) increases in neutral phosphatase activity of 9.84%, 10.31%, and 14.12% were found in April, May, and June, respectively, after inoculation with K3, compared with CK. Increases in PPO activity ( Fig. 4d) were parabolic and reached a maximum in April, and the PPO activity in the K3T treatment was signi cantly (P < 0.05) higher than that of CK in April, resulting in a 43.77% increase.

Improvement effect of K3 on melon fruit quality in greenhouse
Soil inoculation with K3 agent could signi cantly improve the fruit quality of melon (Table 5). Single fruit weight markedly increased by 10.19%, although this increase was not signi cant (P > 0.05) compared with CK. The amount of soluble sugar, soluble solids, and vitamin C were signi cantly (P < 0.05) increased by 20.26%, 21.21%, and 10.14%, respectively, compared with those of the CK. The NO 3 -N in fruits was signi cantly (P < 0.05) reduced by 50.86% under the K3T treatment compared with that of the CK.

Discussion
Due to the high complexity of rhizosphere, the catabolism of root exudates and its stimulating effect on microbial activity were studied by evaluating the role of a single low molecular weight organic substance in a simple system [20,21]. The results showed that phenolic compounds in root exudates were effective regulators for microbial assembly in rhizosphere [22]. In this study, the allelopathic role of CA on soil microorganisms and the elimination of CA by K3 were explored, which will help to elucidate the mechanisms underlying the problem of natural regeneration failure in melon elds and could guide us toward more optimal approaches of soil disease control during continuous monoculture.
Soil microorganisms are ubiquitous, play important functions related to soil fertility, play a key role in plant growth and health, and provide the rst line of defense against root infections by soil pathogens [23,24]. The change of soil microbial community structure may lead to the change of microbial function performed by the community, which will affect the growth of plants [25,26]. Therefore, the deterioration of soil bacterial communities causes poor plant performance in continuous planting systems.
Based on cultivation-dependent methods, previous researches have proved that phenolic acid can change the number of culturable soil microorganisms [27]. In addition, compared with other types of compounds, root exudates rich in phenolic compounds have a more signi cant impact on the soil microbial community [28,22]. According to reports, in some cases, phenolic acid from root exudates can inhibit or promote the biochemical and physiological processes of microorganisms. Microbial metabolism is an important determinant of the degree and duration of phenolic acids in soil [18].
Different concentrations of phenolic acids have different effects on microbial growth [29].According to reports, in some cases, phenolic acid from root exudates can inhibit or promote the biochemical and physiological processes of microorganisms [30]. Benzoic acid (BA) amendment of the soil was found to signi cantly reduce the ratio of bacteria to fungi [7]. Continuous application of p-coumaric acid to the rhizosphere of cucumber has the same result [31]. Qu's research has shown that bacterial diversities decreased with the increasing concentration of BA and 3-phenylpropanoic acid [32]. The addition of 4 µg·kg − 1 BA or 8 µg·kg − 1 3-phenylpropanoic acid phenols signi cantly (P < 0.05) reduced the soil microorganism density and soil microbial diversity, and extrapolated that BA and 3-phenylpropanoic acid destroyed the balance between the bacteria and fungi, and led to the accumulation of pathogenic microbes and reduction of rhizosphere growth promoting bacteria [3]. In this study, adding CA to the liquid medium under co-cultivation conditions signi cantly increased the number of soil fungi, reduced the number of soil bacteria, and reduced the ratio of bacteria to fungi, which is consistent with previous research results.
Under eld conditions, phenolic acid can be continuously produced and secreted into the surrounding soil by roots. This could reinforce the effects of the microbial utilization of phenolic acid in the rhizosphere and lead to a shift in soil microbial communities [7]. Since the elds are continuously monocultured, plant root exudates and the identical regimes of the elds could lead to disruption of soil microbial community composition and structure [33]. The application of microbial agents has been found to be an e cient method to improve the deteriorated soil microbial environment caused by continuous cropping. Previous research found that after the application of effective microbial agents pathogens decreased while the antagonists increased. Moreover, the abundance of bacteria related to elemental cycling and plant growth promotion increased substantially [34].
Plant health in natural environments depends on interactions with complex and dynamic communities comprising macroorganisms and microorganisms [35,36]. In the pot experiment of the current study, which explored the improved effect of different K3 application dosages on continuous soil, the results obtained by plate count showed that the densities of soil bacteria markedly increased after amendment with K3 agent; as a result, the bacteria-to-fungi ratio increased. The same trend was observed in the experiments of greenhouse and liquid medium co-culture. To gain further insight into the effect of treatment on root microbial diversity, we performed β-and α-diversity analyses. The OTU distribution (βdiversity) analysis showed that the bacterial OTUs increased and the fungal OTUs decreased in the continuous monocropping soil treated by K3 (T2 treatment). In the T2 treatment, the diversity of bacteria increased, accompanied by a decrease in the fungal community compared with CK1. To pinpoint differences within the rhizosphere microbiota, the microbial diversity within each sample (α-diversity) was calculated. This calculation showed that bacterial communities in the T2 treatment were signi cantly more diverse and complex than those in CK1 and that the Simpson diversity for the T2 treatment was signi cantly lower than that of CK1. One interpretation of our results is that the addition of K3 caused some bacteria with ecological advantages to proliferate and thus the original dominant fungus lost its advantage. From the perspective of community structure, the relative abundance of the dominant composition of bacteria and fungi changed substantially at the phylum level; thus, K3 amendment of soil strongly stimulated or inhibited speci c bacterial and fungal phyla.
In regard to bacterial phyla, most studies have shown that the relative abundance of Acidobacteria is markedly negatively correlated with soil pH [37]. The results of this study showed that the addition of K3 reduced the relative abundance of Acidobacteria, which suggests that the K3 decomposed the phenolic acid in the continuous monocropping soil to increase the soil pH, thus reducing the abundance of Acidobacteria, which are adapted to the acidic environment. Proteobacteria and Bacteroidetes have been reported to be more abundant in the rhizosphere of the microbial remediation soil than diseased soil [38], whereas, Proteobacteria can suppress pathogens [15]. This indicates that the relative abundance of Proteobacteria is positively correlated with soil health.
For the fungal community, further analysis at the genus level revealed that Mortierella was highly enriched in the CK1 and T2 treatment, indicating that they grow better in the continuous cropping melon soil. Interestingly, the relative abundance of Fusarium belonging to the Ascomycota signi cantly decreased by 84.61% under the T2 treatment, whereas Chaetomium signi cantly increased by about 6.49fold. Chaetomium spp. is widely distributed in natural soil and has potential biocontrol effects on many plant pathogens and nematodes [39]. In contrast, Fusarium, in particular, harbors notorious pathogenic plant fungi with a wide variety of hosts and infection strategies [40]. The present results indicate that the fungal environment for plant growth was optimized after K3 treatment. From this perspective, we can speculate that K3 as a biocontrol agent and acts as a soil guard to maintain a healthy soil microbial structure by promoting the growth of bene cial bacteria and inhibiting the propagation of pathogenic fungi.
Biodegradation as a means of biological control usually depends upon the oxidative activities of microorganisms and has played an important role in eliminating environmental compounds [41]. Inoculation of bene cial microorganisms could degrade phenolic allelochemicals as an energy source, and detoxify these substances [42]. It therefore represents a promising way to reduce the accumulation of phenolic acid and alleviate their stimulation for soil pathogenic fungus growth. Nevertheless, biodegradation e ciency, persistence, and competitiveness of the microbial agents in soil restrict their applications. Therefore, screening appropriate microorganisms for soil phenolic acid degradation may be necessary for a successful control of soil-borne pathogens in continuously cropped soils. Some Grampositive species, such as Bacillus pumilus [43], B. subtilis [44], and Lactobacillus plantarum [45] have the genes encoding phenolic acid decarboxylase (PAD). In B. subtilis, several phenolic acids speci cally induce expression of PAD, which converts these antimicrobial compounds into vinyl derivatives [46]. Soil is a complex and dynamic system that supports plant growth. Soil microorganisms are fundamental for soil health and provide ecosystem services that are essential for plant production [47]. In the continuous cropping soil environment, plant growth and development are restricted by microbial deterioration and represent a major constraint to sustainable agricultural production [48]. Bene cial microbes can directly inhibit pathogens by producing antimicrobial compounds. However, bene cial microbes can also inhibit pathogens indirectly by stimulating the plant's immune system, a phenomenon called induced systemic resistance [49]. Several researchers have reported the effects of bene cial bacteria for improving plant growth under a normal as well as a stressful environment [50]. Bene cial microbes optimized the rhizobacteria of peanut, which mitigated the CMO [34]. Similarly, bio-organic fertilizers supported by P. polymyx and B. subtilis effectively suppressed Fusarium wilt disease in melon caused by continuous cropping [8]. Biocontrol bacteria could alleviate crop diseases effectively and ensure the normal growth of cucumber by inoculating Bacillus spp., Bacillus amyloliquefaciens, and other bacteria for disease regulation [51]. In line with previous research, the pot experiment in the current study was conducted with continuous monocropping soil treated by K3, in which horticultural characteristics and biomass of melon were signi cantly improved.
The soil enzymes, sucrose, urease, and neutral phosphatase play essential roles in the cycling of elements. PPO has frequently been reported to participate in plant defense against pests and pathogens [52]. It is generally believed that continuous cropping causes a decrease in soil enzyme activity. The results of the greenhouse experiment in the present study showed that the K3 agent signi cantly increased the activity of urease, neutral phosphatase, and PPO, especially increasing the sucrose activity.
At the same time, the quality of melon fruit was greatly improved; the amount of soluble sugar, soluble solids, and vitamin C signi cantly increased, accompanied by signi cant reduction in NO 3 -N. These results show that the inoculation of K3 can alleviate CMO effectively and improve melon growth.
The present study demonstrated that the K3 agent has a signi cant effect of alleviating melon CMO by decomposing phenolic acids and improving rhizosphere microbial communities. This was re ected by the increased biomass of seeding plants, the changed horticultural characteristics of melon, the increased soil enzyme activity, and the improved quality of fruit after inoculation with K3. The current study provides a theoretical and practical foundation for the use of K3 agent in other crops.

Material And Methods
Effects of CA on soil microorganisms and melon growth

Effects of CA on soil microorganisms
To determine the effect of CA on microbial (bacteria and fungi) abundance of the healthy soil (previous crops were wheat), 4 g of the healthy soil (fresh weight equivalent) was added to 36 mL of sterilized water and shaken on a rotary shaker at 28°C and 180 rpm for 30 min to create a soil suspension. MB (mineral basal medium consisting of 1.0 g NH 4 NO 3 , 0.5 g MgSO·7H 2 O, 0.5 g (NH 4 ) 2 SO 4 , 0.5g NaCl, 1.5 g K 2 HPO 4 , 0.5 g KH 2 PO 4 , and 5 g glucose) liquid medium amended with 100 μg·mL -1 CA (CAT), and the MB liquid medium was taken as the CK. Two mL of the above prepared soil suspension was added to the above two different MB liquid mediums and co-cultured in a shaker at 28°C and 180 rpm for 18 h. Serial dilutions of fermentation liquor were made, and 0.1 mL aliquots were spread on the surface of the plates.
Luria broth (LB) agar and Rose Bengal agar medium [53] were used to evaluate the population of bacteria and fungi, respectively. Plates were incubated at 25-30°C for 24-72 h. After culturing, colonies were counted and expressed as number per milliliter of medium.
Inhibitory effects of exogenous CA on the growth of melon seedlings Melon seeds (C. melo L.) cultivar "Tianbao" were purchased from an agricultural market in Yangling, Shaanxi Province, China.All seeds were surface sterilized with 0.7% sodium hypochlorite solution for 5 min and rinsed three times in sterile distilled water. After the seeds were kept at 55°C for approximately 30 min, the they were germinated in the incubator at 37°C. When the sprouts emerged, the seeds were sown in the matrix soil. At the one-and-a-half-leaves stage, seedlings that showed consistent growth were selected to be transplanted to the pots, and one seeding was planted in each pot.
Pot experiments were carried out in a tissue culture room on the North Campus of Northwest A&F University from 8 June 2018 to 2 July 2018.To assess the induced toxicity of exogenous CA to melon seedlings, the experimental soil was collected from a eld that had never been planted with melon (previous crops were wheat). The experiment was arranged in a completely randomized design with six replicates. A 20-mL aqueous solution of 10 μg·mL -1 CA was added to 500 g of soil (dry weight equivalent) and mixed to obtain the CAG1 treatment (400 μg·kg -1 soil CA). Soil with an equal volume of distilled water was used as the CK1. Soil moisture was maintained at 20% (absolute water content of soil) of the water holding capacity.
The number of soil bacteria and fungi was determined as above. At the seedling stage, an appropriate amount of water was added to the pots to loosen the soil. Six plants were uprooted, and the soil adhering to the melon seedling roots was shaken off slightly. The roots were rinsed carefully until there was no soil left, and the root surface was then dried with absorbent paper. This was followed by measurement of the fresh weight of shoots and roots. The plants were placed for drying in an oven for 15 min at 105°C and then placed at 70°C for 3 days. The dry weight of each plant, the length and width of the rst true leaf, plant height, and root length were measured directly. Phenylalanine ammonia lyase (PAL) activity, peroxidase activity (POD), polyphenol oxidase (PPO) activity, root activity (TTC method) of melon seedling leaves were determined with reference to Cao Cuiling's method [54].
Decompositionof CA byK3 The Bacillus K3 used in this study was previously isolated and identi ed by our laboratory, Northwest A&F University (Yangling, China). The K3 agent was prepared by fermentation of K3; the number of effective bacteria was higher than 3.01 × 10 10 cfu·g -1 . The bacterial and fungal libraries for MiSeq sequencing were constructed following previously described protocols. The gene-speci c primers 515F (5′-GTG CCA GCM GCC GCG GTA A-3′) and 806R (5′-GGA CTA CHV GGG TWT CTA AT-3′) were used to amplify the V4 region of bacterial 16S rRNA genes while primers ITS1F (5′-CTT GGT CAT TTA GAG GAA GTA A-3′) and ITS2 (5′-GCT GCG TTC ATC GAT GC-3′) targeted the ITS1 region of the fungal internal transcribed spacer (ITS) [55]. Bacterial 16S rRNA and fungal ITS sequences were ampli ed using the ABI 2720 Thermal Cycler (Thermo Fisher Scienti c). All ampli cations were conducted in a 25 μL mixture including 12.5 μL of 2×KAPA HiFi HotStart ReadyMix, the forward and reverse primers at 0.2 μM nal concentration, 2.5 μL of template DNA (5 ng·μL -1 ), and nuclease-free water up to 25 μL. The PCR conditions were 94°C for 2 min, followed by 25 cycles of 30 s at 94°C, 30 s at 55°C, and 1 min at 72°C, with a nal extension of 10 min at 72°C. The obtained PCR products were puri ed using Agencourt AMPure XPPCR Puri cation Beads (Beckman Coulter, Brea, CA, USA) and quanti ed by an Invitrogen Qubit3.0 spectrophotometer (Thermo Fisher Scienti c). The puri ed amplicons were pooled in equimolar concentrations and employed for library construction. To detect the microbial (bacteria and fungi) abundance in the rhizosphere soil, 2 g of the rhizosphere soil from the pot experiment was taken from each treatment. The rhizosphere soil of melon was obtained by collecting soil from the root area (10-20 cm from the taproot) of the greenhouse plant using a shovel. Soil samples were mixed thoroughly and transported in plastic bags to the laboratory. The method for determining the number of bacteria and fungi was the same as 2.1.1 After culturing, colonies were counted and expressed as number per gram (fresh weight) of soil.
Soil samples were collected from the top 20 cm of the soil surface in the greenhouse trial, air-dried, homogenized, and sieved to 100 mesh before analysis. The activity of sucrose, urease, neutral phosphatase, and polyphenol oxidase (PPO) of soil were measured following the method used by Gao [56].Single fruit quality was measured directly. The amount of soluble sugar was determined using the anthrone colorimetric method, soluble protein was evaluated using the Coomassie blue staining method, vitamin C was determined using the molybdenum blue colorimetric method, and nitrate nitrogen was measured by nitration of salicylic acid colorimetry. Soluble solids were assessed by a hand-held 2WAJ Abbe refractometer (Shanghai Optical Instrument Factory, Shanghai, China).

Statistical analysis
Microsoft Excel 2007 was employed to process the data. Statistical analysis was performed using SPSS 23.0 (SPSS, Inc., Chicago, IL, US). Comparisons among treatments were performed using one-way analysis of variance (ANOVA) or t tests in IBM, with signi cance at P < 0.05. MiSeq sequencing data analyses were conducted with packages in R. To estimate the taxonomic (operational taxonomic units (OTUs) and metagenomic species level and community diversity, the diversity index was calculated by "diversity" in vegan R package. All of the values are expressed as mean ± standard error.