Dazomet application suppressed watermelon wilt by the altered soil microbial community

Fusarium wilt disease causes severe decline of watermelon yield and quality. Researches have been reported that soil fumigation with dazomet can help control crop disease. Firstly, we discovered that the dazomet application suppressed watermelon wilt in field experiment compared to the control group. While the importance of microbial community in regulating plant health has been rising up, we therefore focused on examining the soil microbial diversity at six different sampling times after dazomet application by using Illumina MiSeq platform. Remarkably, our research results showed that some beneficial microbial genera have been altered, and these beneficial microbial genera have dominated the entire community, such as Nitrolancea, Pseudomonas and Penicillium after dazomet application. Instead, the relative abundance of Fusarium genus and the pathogen FON (Fusarium oxysporum f. sp. niveum, FON) had the decreased. As there was a significant accumulation of AP (available soil phosphorus) after dazomet application, we noticed that the beneficial microbes as Bacillus, Nitrolancea, Paenibacillus and Penicillium have significant positive correlation with AP but negatively related to morbidity. Together, these results demonstrate that the altered soil microbial community structure by dazomet application is critical to suppress watermelon Fusarium wilt. Thus, our results will drive investigations aimed to deploy interaction of microbiota contribute and plant immunity.


Materials and methods
Experimental design. This study was conducted at Gaoqiao Scientific Research Base of the Hunan Academy of Agricultural Sciences in the city of Changsha (112°58′42ʺ E, 28°11′49ʺ N), Hunan Province in China in 2018 and 2019. The soil was sandy loam. The trial crop was watermelon cultivars zaojia 8424, which was provided by Xinjiang Farmer Seed Technology Co., Ltd. China. The dazomet was provided by Beijing Sino Green Agri-Biotech Co., Ltd. Six greenhouses (30 m × 6 m) with the same background, which were cultivated watermelon under monocropping system for five years, were selected. Three of them were treated with dazomet as three replicates, others were as control group. The routine cultivation managements in all the greenhouses were the same. Every March before transplanting the watermelon seedlings, 6 kg (98% C 5 H 18 N 2 S 2 ) of dazomet were applied to one greenhouse, which was then tilled the soil by a rotary immediately after spraying. Controlling the depth of tillage soil 0-20 cm to ensure that dazomet was evenly mixed into the tillage layer. As soon as the soil temperature is above 8 °C, film mulching was set up to maintain the fumes of dazomet into the soil to kill most of the soil organisms, as well as to maintain the soil moisture content at approximately 40% for the germination and growth of weeds and pathogens. After 20 days, the film was uncovered and the greenhouse was kept ventilated. Then 15 days later, the watermelon seedlings nutrition bowl was cultivated and transplanted into the greenhouse. We planted the watermelon in the greenhouse with 50-60 cm plant spacing to enable pruning the climbing vines. We designed six different sampling times as following: 1 (March 6th, 2018, before dazomet treatment), 2 (April 24th, 2018, watermelon seedling stage), 3 (May 3rd, 2018, Fusarium wilt symptom appearance), 4 (March 6th, 2019, before dazomet treatment), 5 (April 22th, 2019, watermelon seedling stage), 6 (April 29th, 2019, Fusarium wilt symptom appearance). For each replicate, nine independent soil samples within depth of 0-20 cm in the shape of "S" from each greenhouse were pooled. Three greenhouses within same treatment regarded as three independent replicates. DAZ represents dazomet treatment group and CK represents the control group without dazomet application but using same conventional planting system. All the soil samples from greenhouses were packed into sealed sterile bags separately and brought back to the laboratory. After removing the plant roots and stones from the samples, we sieved them with a 20-μm mesh, and then divided each sample into three parts. Two of them were placed in sterile centrifuge tubes, stored at − 80 °C for sequencing analysis and Q-PCR test. While the other was used for measuring the soil properties, stored at room temperature. We have collected total of 36 samples in six different sampling times.
Field disease investigation. The incidence of Fusarium wilt was calculated during the whole watermelon onset period (Started from plants with rotted, discolored root and the vascular bundle became brown until the whole plant died). The disease incidence (%) = (number of infected plants/total number of surveys) × 100%.
Determination of soil physical and chemical properties. The soil characteristics are listed in Supplementary Table S1. Soil pH was determined in a soil: water ratio of 1:2.5 (wt./vol) using a pH meter (BPH-220, Bell Instrument Equipment Co. Ltd., Dalian, LN, China). To extract the water-soluble salts from the soil, samples of 1 mm sieved and air-dried soil weighing 20.00 g were placed in a 250 ml Erlenmeyer flask, 100 ml of distilled water was added (water: soil ratio of 5:1). Then put it into a dry triangular bottle after shaking for 5 min which was used for the determination of salt. A total of 30 ml of the soil leachate was placed in 50 ml of burnout solution. The solution temperature was measured, and then the conductivity of the solution was determined using a conductometer. The soil organic matter (SOM) was determined by oxidation with potassium dichromate by DF-101S heat collecting constant temperature magnetic stirrer (Gongyi yuhua instrument Co., Ltd, Gongyi, HN, China). Total P and K and available P and K concentrations in the soil were determined by ICP-AES (PerkinElmer 2100DV, PerkinElmer, Waltham, MA, USA) after the soils were digested using concentrated HNO 3 -HF-HClO 4 . Total nitrogen (N) and available nitrogen (AN) in the soil were determined by the Kjeldahl method and the alkali diffusion method, respectively (China Agricultural Technology Extension Service Center, 2014).
Soil microbial diversity analysis. Total genomic DNA was extracted from the soil samples using the E.Z.N.A Soil DNA kit (Omega Bio-tech, Norcross, GA, USA) according to manufacturer's protocols. The final DNA concentration and purity were determined using a Nanodrop 2000 UV-Vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA), and the DNA quality was checked by 1% agarose gel electrophoresis. Distinct regions of the 16S rRNA gene (V3-V4) and ITS1 were amplified by PCR (ABI Geneamp 9700, Applied Biosystems, Inc., Carlsbad, CA, USA) using specific primers (16S: 338F (5′-ACT CCT ACG GGA GGC AGC AG-3′), 806R (5′-GGA CTA CHVGGG TWT CTAAT-3′); ITS1F (5′-CTT GGT CAT TTA GAG GAA GTAA-3′), ITS2R (5′-GCT GCG TTC TTC ATC GAT GC-3′)), separately. The PCRs were conducted using the following programme:  Data analysis. Raw FASTQ files were demultiplexed, quality-filtered by Trimmomatic and merged by FLASH with the following criteria: (i) The reads were truncated at any site receiving an average quality score < 20 over a 50 bp sliding window, and the truncated reads shorter than 50 bp were discarded; (ii) exact barcode matching, primers were exactly matched, and reads containing ambiguous bases were removed; (iii) sequences with over 10 bp of overlap were merged according to their overlap sequence. The singletons were removed for further analyses. Operational taxonomic units (OTUs) were clustered with a 97% similarity cut-off using UPARSE Version 7.1 (http://drive 5.com/upars e/) and chimeric sequences were identified and removed using UCHIME 14 . The taxonomy of each 16S rRNA gene sequence was analyzed by the RDP Classifier algorithm against the Silva (SSU123) 16S rRNA database using a confidence threshold of 70%. The taxonomy of each ITS region sequence was analyzed by the RDP Classifier algorithm against the UNITE (8.0) ITS rRNA database using a confidence threshold of 70%. The diversity analysis for the sequencing data were performed on the free online platform of Majorbio Cloud Platform (www.major bio.com) based using the Qiime2 software 15 (https ://qiime 2.org/). For example, the sobs index of rarefaction curve, alpha diversity (student's t-test shannon index) and beta diversity of (NMDS, nonmetric multidimensional scaling analysis) with ANOSIM statistical analysis to compare the composition between treatments on OTU level. Moreover, the significant difference of microbial community species in different sampling groups was tested by Kruskal-Wallis H test on OTU, phyla and genera level respectively. The FDR (Falsely Discovery Rate) and Tukey's method 16 were used to analyze the multiple test correction for p-value > 0.95. Oneway ANOVA test was used to analyze significant differences of two groups. Differences between two groups were analyzed by student's t test. Correlation heatmap analysis of the correlation coefficient between environmental factors and selected species was determined by MeV (Multi Experiment Viewer) software (http://mev.tm4.org).
Other statistical analysis was performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). The figures of the microbial diversity indices and relative abundance of functional profiles were prepared using Microsoft Office 2010 (Microsoft Corporation, Redmond, WA, USA)and Adobe Illustrator CS5 (Adobe Systems Incorporated, San Jose, CA, USA) (https ://www.adobe .com/cn/produ cts/illus trato r.html).

Results
Effects of dazomet application on watermelon Fusarium wilt. The disease incidence was significantly reduced after dazomet application compared to the control group in our filed experiment, indicating that the continuous application of dazomet could effectively suppress the occurrence of watermelon Fusarium wilt. As well as the plant morphological comparison between the diseased plants and healthy plants showed in Fig. 1A. Furthermore, the statistical analysis results showed that the morbidity of watermelon Fusarium wilt in the control group were above 98% in year 2018 and year 2019, but the disease incidences were significantly decreased after treatment with dazomet compared with control in both two years (Fig. 1B). Even more, the disease incidence in 2019 (34.17%) was much lower than that in 2018 (79.82%) after treatment with dazomet showed in Fig. 1C.
Effects of dazomet application on soil properties of watermelon field. Results from the soil physicochemical analyses showed that there were no significant differences in the parameters, expect for EC (soil electrical conductivity) and AP (available soil phosphorous) (Supplementary Table S1). We observed that the electrical conductivity of dazomet treated soil was lower in 2019 than that of the previous year. Furthermore, we noticed that the available phosphorus content has significantly increased in the soil after dazomet application.
Comparative analysis of soil bacterial community structure in different sampling times. Bacterial sequence data evaluation and OTU analysis. The depth analysis of the 36 soil samples sequencing data indicated that the raw sample sequence was composed of 2,062,735 reads, clean sample sequnce with 1,491,552 reads, the average reads per sample was 41,432, and the total number of OTUs detected was 8599. The rarefaction curves of sobs index reflects the evenness of the community in the samples tends to be consistant ( Fig. 2A). The student's t-test shannon index showed that the alpha diversities of bacterial have significant differences between CK and dazomet treatment (Fig. 2B). And we found that the structures of bacterial communities have dynamics through all sampling times based on beta diversity analysis performed by NMDS (Fig. 2C). The venn diagram showed that there were 850 overlaped bacterial OTUs in all groups of samples (Fig. 2D). In addition, we found eight core OTUs had significantly dynamics based on Kruskal-Wallis H test of all sampling times. Therefore, we used the student's t-test of shannon index to detect these eight significant differences OTUs between treatments for every sampling time (Fig. 2E) (Fig. 3B). Further, we used student's t-test to detect their significant differences between treatments for every sampling time respectively showed in Fig. 3C. Our results indicated that the Nitrospirae had decreased significantly in DAZ group in year 2019, and all these three phyla have significantly decreased in DAZ in the 5 th sampling time compared with CK group. Moreover, we selected 8 distributions of genera as biomarkers with significant differences based on Kruskal-Wallis H test of multiple comparisons in all sample groups to detect their dynamics. For instance, these were Acidipila, Aquicella, Bacillus, Nitrolancea, Nitrospira, Paenibacillus, Pseudomonas and Streptomyces (Fig. 4B). Notably, we identified that the community abundance of Aquicella and Nitrospira have significantly decreased after dazomet application compared to the control from year of 2019. And we spotted that the percent of Pseudomonas and Nitrolancea have significantly accumulated after dazomet treatment (Fig. 4C). www.nature.com/scientificreports/ Comparative analysis of soil fungal community structure in different sampling times. Fungal sequence data evaluation and OTU analysis. The depth assessment analysis of the 36 soil sample sequencing data indicated that the raw sample sequence was composed of 1,318,001 reads, clean sample sequnce with 939,816 reads, the average reads of samples was 26,106 and the total number of OTUs detected was 1832. The rarefaction curves of sobs index reflects the evenness of the community in the samples tends to be consistant (Fig. 5A). The student's t-test of shannon index showed that the alpha diversities of fungal community have significant differences between CK and DAZ treatment (Fig. 5B). And we found that the structures of fungal communities have dynamics performed by NMDS on the relative abundances of OTUs during all sampling times (Fig. 5C). The venn diagram showed that there were 50 overlaped fungal OTUs in all groups of samples (Fig. 5D). At the same time, we found eight core funagl OTUs had significantly dynamics based on Kruskal-Wallis H test of all sampling times. Then, we used the student's t-test of shannon index to detect these seven significant differences OTUs between treatments for every sampling time. Notably, the significance test of core fungal OTUs indicated that the proportion of OTU2707(p__Ascomycota;c__Eurotiomycetes;o__Eurotiales;f__ Dynamic changes of dominant fungal community at different sampling times. In order to investigate the effect of dazomet treatment, we monitored the changes in the soil fungal communities over two years by community bar plot analysis both on phylum and genus levels respectively (Figs. 6 and 7). Based on statistical analysis showed in Fig. 6B,C, we found that the dominant phyla of Ascomycota and Mortierellomycota have significant changes after dazomet application, but their changes were on the contrary way. And at the genus level, we noticed that the percent of community abundance of Chaetomium, Fusarium, Mortierella and Penicillium have altered significantly (Fig. 7B). Notably, Penicillium has rapidly increased significantly after the dazomet application compared to the control group. However, the Mortierella and Fusarium had the decreased tendency after dazomet treatment compared to control (Fig. 7C).

Quantitative detection FON dynamics at different sampling times by real-time PCR. For fur-
ther verification the FON content in the soil at different sampling times, Q-PCR results showed that detection of pathogen FON have constantly decreased after dazomet treatment. In year of 2019, the content of FON had significantly declined in the DAZ group compared to CK (Fig. 7D). www.nature.com/scientificreports/

Discussion
In this experiment, our results showed that the disease incidence was significantly reduced after dazomet application compared to the control group. Moreover, the morbidity of watermelon Fusarium wilt was significantly decreased in year of 2019 than that in 2018 after dazomet application. Thus, this confirmed that dazomet application can not only effectively suppress the occurrence of Fusarium wilt in apple orchards 7 , cucumber 17 , ginger 6 , gooseberry 5 but also in watermelon wilt.
Since several researches have emphasized the role of soil microbial communities in plant health and growth 9 . For example, Xin's study used Arabidopsis mutants showed that the plant immune system was needed to maintain the normal growth of commensal bacteria 10 . Chen's research demonstrated one of Pseudomonas piscium directly interfered with the molecular machinery of Fusarium graminearum in wheat resulting in reduced virulence and fungal growth 18 . Due to the numerous commonalities in mechanisms and principles of both types of plant-microbe interactions, scientists from both fields have started to join forces to foster an integrated view of the molecules. A nice example of this was recently demonstrated the importance of microbe-microbe interactions associated in Arabidopsis used a number of computational and microbial techniques to characterize this soil Kruskal-Wallis H test was used to test the significance of the dynamic changes of bacterial community relative abundance at different sampling times on genus level. The X-axis represents the species, the Y-axis represents the average relative abundance of enriched communities. (C) Student's t-test bar plot showed significant differences of the main bacterial genera between treatment and CK. DAZ dazomet treatment, CK control. Numbers after letters indicate different sampling times. 1 (March 6th, 2018, before dazomet treatment), 2 (April 24th, 2018, watermelon seedling stage), 3 (May 3rd, 2018, Fusarium wilt symptom appearance), 4 (March 6th, 2019, before dazomet treatment), 5 (April 22th, 2019, watermelon seedling stage), 6 (April 29th, 2019, Fusarium wilt symptom appearance). Data are expressed as the means ± SD (n = 3). * 0.01 < P ≤ 0.05, ** 0.001 < P ≤ 0.01.

Scientific Reports
| (2020) 10:21668 | https://doi.org/10.1038/s41598-020-78839-5 www.nature.com/scientificreports/ microbes and plant roots interkingdom 19 . In agree with the theory, our significance analysis of Shannon index and beta diversity analysis results indicate that the structure of not only the bacterial but also the fungal communities have altered significantly after the application of dazomet. Therefore, we can infer that the application of dazomet may active plant defense activity by shifting the soil microbial community. Surprisingly, our comparative analysis of soil microbial community structure in different sampling times results demonstrated that the relative abundance of some dominant beneficial microbial have increased as well. For example, the bacterial of Bacillus, Pseudomonas, Nitrolancea, Streptomyces and the fungi genus Penicillium. On the contrary, the relative abundance of Fusarium genus and the pathogen FON had the decreased tendency after dazomet treatment compared to control. More specifically, numerous studies have indicated that most species of Bacillus promote crop yields and significantly affect the root inoculation and impact of plant growth promoting rhizobacteria on nutrient element contents 20 . Many of the polyketides and lipopeptides produced by Bacillus and Paenibacillus species have been described as antimicrobial agents that can be potentially applied www.nature.com/scientificreports/ as sustainable bio-organic products in medicine against human pathogens and in agriculture for controlling plant pathogens [20][21][22] . Some Nitrospira species have been reported to be involved in nitrification and nitrogen metabolism. Nitrolancea is a nitrite oxidizing bacteria that converts nitrite to nitrate 23 . Pseudomonas was recognized as one of an important beneficial microorganism to plant immunity [24][25][26] . For instance, Chen 18 and colleagues demonstrated that the Pseudomonas piscium reduced virulence and the plant pathogenic fungus Fusarium graminearum growth by directly interferes. De Cal 2 reported that Penicillium oxalicum effectively control of F. oxysporum f. sp. melonis on melon and watermelon, respectively long time ago. Taken together, the enrichment of those beneficial microbes and declined pathogen in the soil is likely the reason why the watermelon Fusarium wilt disease was suppressed after dazomet application.
In addition, we noticed that the available phosphorus content has significantly increased in the soil after dazomet application. Therefore, we hypothesized that either the soil microbial community or the plant immune system might cause the AP increase. While growing evidence supports the idea that the modulation of plant immunity can reshape soil microbiota under different environmental conditions 27,28 and plant-associated microbial communities promote plant nutrient uptake as well 25,29 . We made a correlation analysis between AP, disease incidence and soil microbial community as well. Indeed, we found that the beneficial microbes as Paenibacillus, Bacillus, Nitrolancea and Penicillium have significant positive correlation with AP but negatively related test was used to test the significance of the dynamic changes of fungal community relative abundance at different sampling times on phylum level. (C) Student's t-test bar plot showed significant differences of the main fungal phyla between treatment and CK. The X-axis represents the bacterial phylum, the Y-axis represents its average relative abundance. DAZ dazomet treatment, CK control. Numbers after letters indicate different sampling times. 1 (March 6th, 2018, before dazomet treatment), 2 (April 24th, 2018, watermelon seedling stage), 3 (May 3rd, 2018, Fusarium wilt symptom appearance), 4 (March 6th, 2019, before dazomet treatment), 5 (April 22th, 2019, watermelon seedling stage), 6 (April 29th, 2019, Fusarium wilt symptom appearance). Data are expressed as the means ± SD (n = 3). * 0.01 < P ≤ 0.05, ** 0.001 < P ≤ 0.01.

Scientific Reports
| (2020) 10:21668 | https://doi.org/10.1038/s41598-020-78839-5 www.nature.com/scientificreports/ to disease incidence (Fig. 8). Moreover, the recent findings identified the root microbiota drive integration of phosphate stress and immunity 30 . They represented an important advance in showing that plant responses to phosphate deficiency are directly linked to immunity and depend on an intact root microbial. Similarity, our results enhanced the thought that plant immune system may coordinate recognizing microbial with soil phosphate during microbiome assembly. Besides, Kei Hiruma 31 demonstrated an excellent mechanism of how the fungus Colletotrichum tofieldiae transfers the macronutrient phosphorus to Arabidopsis shoots. Therefore, these results may explain why dazomet management could inhibit watermelon Fusarium wilt disease in our study.

Conclusion
In conclusion, this experiment demonstrates that the altered soil microbial community structure by dazomet application is critical to suppress watermelon Fusarium wilt. Our data also suggest two possibilities for the dynamic changing of soil microbial community structure. One reason may the enriched AP lead to stimulate plant immune system to avoid disease by assembling beneficial microorganisms therefore the new community structure has been established. Other is that dazomet application first caused the changes of AP, likely beneficial the surviving of some specific microbes that positively correlated with it which served as a layer protecting plant from pathogen attack at the same time. In addition, our results will drive investigations aimed to deploy interaction of microbiota contribute and plant immunity.