2.1 Isolation and characterization of the pathogen

The diseased plants were collected from an open-field crop in Nanning, Guangxi Province, China (longitude E 108.33, latitude N 22.84). The infected tissue samples were rinsed thoroughly with sterile distilled water, cut with a sterile scalpel into approximately 0.5-cm pieces, surface disinfected with 75% ethanol for one minute, and then rinsed 3–4 times with sterile distilled water. Finally, all sterilized samples were transferred to potato dextrose agar (PDA) amended with 100 mg/L streptomycin and incubated at 28 °C in the dark for 3 days. Emerging colonies were transferred to new PDA two to three times until single colonies were obtained. Cultures on PDA were used for the observation of colony characteristics. The shape, length and width of 50 conidia of three representative isolates were observed by a Nikon Ni-U. The shape and external morphology of the isolated strain were also confirmed by scanning electron microscopy (SEM) following the techniques of Liao et al. (2020). The colony was suspended in 90% ethanol and then placed on a stub. The specimens were sputter-coated with gold and then scanned and photographed using SEM (Hitachi SU8100, Tokyo, Japan).

2.2 Pathogenicity test of the isolated strain on roselle seedlings

A representative isolate (GXRST29) was selected for expansion culture for the pathogenicity test. Spores were collected from 7-day-old PDA cultures maintained at 28 °C and suspended in sterilized distilled water at 10⁶-10⁷ conidia/mL, as determined using a hemocytometer. The resulting suspension was filtered through four layers of sterile cheese cloth to remove mycelial fragments. Five roselle plants at the 5-leaf stage were artificially inoculated by root dipping into the conidial suspension. The experiment was conducted three times, and negative controls were treated with sterile water. All treated plants were placed in an incubator at 28 °C with a 16 h light period for 25 days. Plants were observed every two days after inoculation for symptom appearance. Isolations were made from any resulting lesions on the inoculated plants, and the morphological and cultural characteristics were compared with those of the original fungal isolates used as inoculum. The reisolated fungi were further identified using molecular methods.

2.3 Molecular identification of the pathogenic bacteria

Total DNA was extracted from fresh aerial mycelia grown on PDA plates using a Fungal Genomic DNA Kit (Coolaber, Beijing, China) following the manufacturer’s instructions. The internal transcribed spacer (ITS) rRNA region was amplified using the primer pairs ITS1/ITS4 (ITS1: 5′-TCCGTAGGTGAACCTGCGG-3′, ITS4: 5′-TCCTCCGCTTATTGATATGC-3′) (Chehri, 2014). DNA amplification was performed in a final volume of 20 μL of reaction mixture containing 10 μL of 2 × Taq Master Mix (Vazyme, Nanjing, China), 0.3 μM of each forward and reverse primer and 50 ng of DNA template. The thermal cycling program for PCR amplification was as follows: predenaturation at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 40 s, annealing at 58 °C for 1 min, and extension at 72 °C for 2 min, with a final step of 72 °C for 5 min. Amplified products were visualized by gel electrophoresis in a 1% (w/v) agarose gel. PCR products were purified with a Gel DNA Extraction Kit (Vazyme, Nanjing, China) and cloned into a linearized pEASY-T1 cloning vector (TransGene Biotech, Beijing) according to standard procedures, followed by sequencing by a commercial sequencing service provider (Genecreate, Wuhan, China). Sequences were compared with those in the NCBI database (https://www.blast.ncbi.nlm.nih.gov). Multiple sequence alignment was performed using DNAMAN software. The maximum likelihood (ML) method was used to perform phylogenetic analysis using MEGA X (Kumar et al., 2018), with 1000 bootstrap replications according to Sharma and Kumar (2021), and evolutionary distances were computed using the Kimura 2-parameter method (Kimura, 1980).

2.4 Laboratory determination of pathogenicity

The growth rate method was used to measure the effect of eight fungicides on the roselle root rot pathogen. Media containing a series of gradient concentrations of agents were prepared with PDA, and sterile water was used as a negative control. A mycelial plug of the pathogen was collected by a sterilized punch with a diameter of 0.7 cm and placed in the center of each reagent-amended medium. All medium-containing plates were incubated at 28 °C in the dark for six days. The cross method was applied to measure the diameter of the colony, and the growth inhibition rate was calculated by the following formula: rate of growth inhibition (%) = (colony diameter of control - colony diameter of treatment)/colony diameter of control) × 100% (Kong et al., 2018). Each treatment was repeated at least three times. The biopesticides used in this study were as follows: Paenibacillus polymyxa, 200 million live bacteria per gram (Henan Golden Key Agricultural Science and Technology Co. LTD, Henan, China); Bacillus subtilis, one billion living spores per gram (Baoding Kelufeng Biochemical Technology Co., LTD, Hebei, China); Bacillus amyloliquefaciens, one billion living spores per gram (Huaiyang Shuoda Biological Technology Co., LTD, Jiangsu, China); and 80% ethylicin emulsion (Zhejiang Pinghu Agricultural Pharmaceutical Factory, Zhejiang, China). The chemical pesticides used in this study were as follows: 50% carbendazim wettable powder (Shanghai Yuelian Chemical Co., LTD, China); 70% methyl thiobacillam wettable powder (Yunfa Chemical (Shanghai) Co., LTD, China); 80% mancozeb wettable powder (Dow Yinong Agricultural Technology (China) Co., LTD); and 98% oxazoline soluble powder (Tianjin Lvheng Chemical Co., LTD, China).

2.5 Statistical analysis

All data were statistically analyzed according to the appropriate procedure for one-way analysis of variance using SPSS 17.0 software. Each measure was expressed as the mean ± standard deviation (mean ± SD). Student-Newman-Keuls (S-N-K) post hoc test was performed to control for significant differences between groups, and significance was considered for P < 0.05. The IC₅₀ was used to calculate the median inhibitory concentration and was obtained from the linear regression equation.

3.1 Characteristics of roselle root rot disease

Field experiments revealed that the incidence of roselle root rot was higher in wet, acidic, hardened continuous cropping land. Diseased plants were occasionally found at the seedling stage. However, during the flowering and fruiting stages, the disease severity increased with a high temperature and humidity environment. In the early stage of the disease, plant leaves wilted, similar to symptoms observed under drought conditions, but could recover in the morning and evening. In the late stage, plant wilting was aggravated, and leaves began to turn yellow as they wilted. The primary and lateral roots of the infected plants were discolored and rotted, and the xylem became black, with necrosis extending along the main root to the junction of the rhizomes, in addition to vascular tissue browning, wilting and death (Fig. 1a, 1b).

Close

Fig. 1 Characteristics of roselle root rot. a, Characteristics of diseased roots of adult plants. b, Characteristics of diseased adult plants. c, Control plants. d, Diseased seedlings infected by F. solani.

Export to PPT

3.2 Isolation and morphological identification of the pathogen

The diseased samples were collected from three different diseased plots for pathogen isolation and identification. A total of 18 fungal cultures were consistently isolated from the basal stems and roots of diseased roselle plants, and 29 pure strains derived from single spores were obtained; these were named and numbered as shown in Table 1. Colonies of the fungal isolates grown on PDA at 28 °C showed dry surfaces, compact texture, cottony or fluffy mycelia, and raised surfaces with wavy or curled edges (Fig. 2a). The aerial mycelium expanded rapidly and covered the medium surface, initially being white, then turning pale yellow after six days of growth. Microscopic observations revealed that microconidia were generally single-celled, hyaline and ovoid with sizes of 5.13 to 15.12 × 2.50 to 4.20 μm. Macroconidia were falciform with 3 to 4 septae and sizes of 19.08 to 24.35 × 4.5 to 8.00 μm (Fig. 2b-d). Based on the morphological and culture characteristics, the fungus was identified as Fusarium solani (Leslie and Summerell, 2006).

Close

Fig. 2 Morphological characteristics of F. solani. a, Colony characteristics of F. solani. b, Characteristics of conidia and hyphae of F. solani under a light microscope. c, Morphological characteristics of conidia of F. solani under a light microscope. d, The morphological features of F. solani characterized by SEM. Scale, b-c, 50 μm; d, 30 μm. H, hypha; C, conidium.

Export to PPT

3.3 Molecular identification of the pathogen

The mycelia of the isolated pure strains were collected for DNA extraction for further characterization, and their internal transcribed spacer rRNA regions (ITSs) were amplified using the primer pair ITS1/ITS4. Agarose gel electrophoresis showed that the amplified products were approximately 500 bp (Fig. 3). The amplicons were recovered from the gel, cloned and sequenced. The sequencing data were aligned with the NCBI nucleotide database (https://blast.ncbi.nlm.nih.gov/Blast.cgi) by using BLAST. BLAST alignment showed that 29 of the isolates were homologous to F. solani (Table 1), and the sequences included a partial ITS1 sequence, 5.8S rRNA gene, complete ITS2 sequence and partial large subunit ribosomal RNA gene sequence. The multiple alignment showed that the rRNA sequences of F. solani were highly conserved among different host plants (Fig. 4). The 29 purified strains described above could be divided into four categories according to the differences in the sequences of the conserved regions of F. solani and were deposited in GenBank under accession numbers OL314654, OL314655, OL314656 and OL314657. A BLAST search showed that these sequences were 99.62%, 99.81%, 86.14% and 99.81% homologous to the published ITS sequences of F. solani identified in GenBank (MN653249, MN165653, MT937227 and MN960010, respectively). A representative isolate (GXRST29) was selected to construct a phylogenetic tree to further reveal interspecies relationships. Based on molecular data, the fungus was confirmed to be F. solani and had the closest phylogenetic relationship to F. solani of olive (KU528858.1) (Fig. 4).

Close

Fig. 3 Alignment of ITS sequences from F. solani.

Export to PPT

Close

Fig. 4 Phylogenetic analysis of F. solani based on ITS sequences.

Export to PPT

Five roselle plants at the 5-leaf stage were artificially inoculated by root dipping into a 10⁶-10⁷-mL⁻¹ spore suspension of the isolated strain GXRST29 for pathogenicity testing. The experiment was conducted three times, and the negative controls were inoculated with sterile water. Compared to the control, the symptoms began with the inhibition of growth, leaf yellowing, plant dwarfing and root rotting by three days after inoculation (Fig. 1c, 1d). All plants exhibited symptoms similar to those on diseased plants observed in the field seven days after inoculation. Fungi that were steadily reisolated from the root and stem tissues of inoculated symptomatic plants were confirmed as F. solani by morphological and molecular identification. In contrast, the noninoculated seedlings had no root rot and produced an abundance of new roots. No fungi were isolated from them, and the same results were obtained in all three repetitions of the pathogenicity tests. Thus, these results satisfied Koch’s postulates, and F. solani was the main cause of roselle root rot.

3.4 The effect of fungicides at different concentrations on mycelial growth of Fusarium solani

In the present study, eight different fungicides, including four biological fungicides (Fig. 5) and four chemical fungicides (Fig. 6), were used to investigate their inhibitory effects against F. solani on roselle. The results revealed that the four biofungicides showed different degrees of inhibition of F. solani, and the inhibition rate increased with increasing mass concentrations. There were significant differences in the colony diameter, which was significantly decreased compared with that of the CK (P < 0.05) and differed between treatments with different mass concentrations. The different biological fungicides exhibited different inhibitory effects on F. solani (Table 2). A comparison of the inhibition rates of the four biological fungicides with different mass concentrations on F. solani revealed that Bacillus amyloliquefaciens showed the best fungistatic effect. When the concentration of the agent was 1 mg/mL, the inhibition rate was 50.09%, and when the concentration of the agent was 2.5 mg/mL, the inhibition rate was 100%; the fungistatic activity was significantly different among different concentrations (P < 0.01). Following ethylicin treatment, when the concentration was 16 mg/mL, the fungistatic rate was 55.39%, and when the concentration was 80 mg/mL, the fungistatic rate reached 93.19%. When the concentration of Bacillus polymyxa was 60 mg/mL, its inhibition rate was 100% (Table 2). This result indicated that the diameter of the colonies varied significantly among different concentrations of acetoillin (P < 0.01); however, the fungistatic rate was lower than that of Paenibacillus polymyxa after treatment with a high-quality concentration of ethylicin. In this study, the inhibitory effect of Bacillus subtilis was the worst; when its mass concentration was 80 mg/mL, the inhibition rate was only 74.67% (Table 2). Combined with the virulence regression equation and the IC₅₀ value, the effects of the four biofungicides against F. solani were comprehensively evaluated as follows: Bacillus amyloliquefaciens > 80% ethylicin > Paenibacillus polymyxa > Bacillus subtilis (Table 2).

Close

Fig. 5 Analysis of biocides against the mycelial growth of F. solani. A, Bacillus amyloliquefaciens. The concentrations of the agent from a to f were CK (sterile water), 0.1 mg/mL, 0.5 mg/mL, 1 mg/mL, 2 mg/mL and 2.50 mg/mL. B and C show Paenibacillus polymyxa and Bacillus subtilis, respectively, and the concentrations of the agent from a to f were CK (sterile water), 10 mg/mL, 20 mg/mL, 40 mg/mL, 60 mg/mL and 80 mg/mL. D, 80% ethylicin. The concentrations of the agent from a to f were CK (sterile water), 4 mg/mL, 8 mg/mL, 16 mg/mL, 32 mg/mL and 80 mg/mL.

Export to PPT

Close

Fig. 6 Effects of chemical fungicides on the mycelial growth of F. solani isolated from roselle. A, carbendazim. The concentrations of the agent from a to f were CK (sterile water), 0.0025 mg/mL, 0.005 mg/mL, 0.01 mg/mL, 0.02 mg/mL and 0.03 mg/mL. B, thiophanate-methyl. The concentrations of the agent from a to f were CK (sterile water), 0.005 mg/mL, 0.01 mg/mL, 0.025 mg/mL, 0.05 mg/mL and 0.075 mg/mL. C, carmazine. The concentrations of the agent from a to f were CK (sterile water), 0.01 mg/mL, 0.1 mg/mL, 0.5 mg/mL, 1 mg/mL and 1.5 mg/mL. D, hymexazol. The concentrations of the agent from a to f were CK (sterile water), 0.05 mg/mL, 0.1 mg/mL, 0.2 mg/mL, 0.4 mg/mL and 0.5 mg/mL.

Export to PPT

As shown in Fig. 6, the four chemical fungicides exhibited different degrees of inhibition of F. solani of roselle, and the inhibition rate increased with increasing mass concentration. There were significant differences in colony diameter, which was significantly lower than that of the CK (P < 0.01) and differed between treatments with different mass concentrations. However, the antibacterial effect of the four chemical fungicides was stronger than that of the biological fungicides, and a relatively significant fungistatic effect could be achieved under treatment with low concentrations of chemicals. For example, when the mass concentration of carbendazim was only 0.005 mg/mL, its antibacterial effect reached 42.53% (Table 3). The virulence regression equation revealed that the IC₅₀ of hymexazol (0.0050 mg/mL) was lower than that of carbendazim (0.0082 mg/mL). However, when the carbendazim concentration was 0.03 mg/mL, the fungistatic rate reached 100%, while when the hymexazol concentration was 0.5 mg/mL, the fungistatic rate only reached 94.68% (Table 3). In addition, although the IC₅₀ of thiophanate-methyl (0.0243 mg/mL) was higher than that of oxazim (0.0050 mg/mL), when its mass concentration was 0.05 mg/mL, the fungistatic rate was 77.32%, and when its mass concentration was 0.075 mg/mL, the fungistatic rate reached 100% (Table 3). This indicated that thiophanate-methyl presented a better inhibitory effect on F. solani than hymexazol at low concentrations. Thus, based on the comprehensive evaluation by using the inhibition effect, toxicity regression equation and IC₅₀ of the four chemical fungicides against F. solani, the effects of the four chemical agents against F. solani were as follows: carbendazim > thiophanate-methyl > hymexazol > carmazine.

4.1 F. Solani was the main pathogen that caused root rot on roselle

Root rot is an important disease affecting the growth and development of roselle, which could lead to a decline in the yield and quality of roselle calyces and serious economic losses. The disease has become an important obstacle in roselle planting (Nalim et al., 2011). However, there have been few reports on the characteristics of the pathogen causing roselle root rot in Nanning, Guangxi. In the present study, the pathogen of roselle root rot was isolated, and its morphological characteristics, pathogenicity and laboratory toxicity were determined. This was the first time that F. solani was confirmed to be the main pathogen causing roselle root rot in Nanning, Guangxi. The traditional classification of fungi is mainly based on their morphological, physiological and biochemical characteristics (Leslie and Summerell, 2006). However, due to the large diversity of fungi, traditional taxonomic methods often lead to false positive or false negative results, which creates certain difficulties for the classification and identification of fungi. The application of molecular biology technology based on nucleic acid sequencing provides a simpler and more convenient, sensitive and specific method for the classification of fungi (Chehri, 2014). In this study, the pathogen that causes roselle root rot was identified on the basis of morphology combined with an analysis of the specific ITS sequence of its rRNA. This method has been widely used in the isolation and identification of the pathogens causing dragon fruit rot (O Donnell, 2000), strawberry root rot (Mariscal et al., 2017), wolfberry root rot (Uwaremwe et al., 2020) and bitter gourd Fusarium wilt (Liu et al., 2017).

4.2 Biofungicide use is one of the main methods for Fusarium control

Fusarium is a worldwide-distributed soil-borne fungal disease that causes fusarium wilt; this disease can lead to vascular lesions and plant death and seriously reduce crop yield and quality. At present, dozens of crops, such as cotton (Zhu et al., 2020), banana (Kristle et al., 2019), tomato (Ye et al., 2020), and potato (Hussein et al., 2020), have been harmed by Fusarium, leading to serious effects on the development of the fruit and vegetable industry in China. In addition to breeding disease-resistant varieties, biological and chemical control are the main methods to inhibit pathogenic Fusarium. Bacillus biofungicides present good effects in the biocontrol of soil-borne and aboveground diseases. For example, Bacillus subtilis is an effective biocontrol agent against root rot pathogens of eggplant and tomato (Abdel et al., 2016; Madhi and Jumaah, 2020). Paenibacillus polymyxa showed a good control effect on cucumber fusarium wilt and fusarium head blight (Imen et al., 2016; Li and Chen, 2019).

In this study, four biological fungicides, Bacillus subtilis, Bacillus amyloliquefaciens, Paenibacillus polymyxa and acetoallicin, were analyzed for their inhibitory effects on F. solani isolated from roselle. The results showed that the four fungicides exhibited a certain inhibitory effect on F. solani, and the inhibitory effect increased gradually with increasing concentrations. Bacillus amyloliquefaciens showed a significant advantage over the other three biofungicides in its inhibitory effects against F. solani, with an IC₅₀ of 1.10 mg/mL. This was consistent with the control effect of Bacillus amyloliquefaciens on watermelon Fusarium wilt (Zhao et al., 2018). Bacillus subtilis exhibited the worst inhibitory effect on F. solani, with an IC₅₀ of 46.78 mg/mL. However, the control effect of Bacillus subtilis on strawberry root rot caused by F. oxysporum was better than that of other biological fungicides (Zhang et al., 2012). This indicates that different biological fungicides have different control effects on different crops, and this discrepancy might result from the differences in the mechanism of action of the different biofungicides.

Previous studies have suggested that Bacillus polymyxa can secrete enzymes that dissolve the cell wall components of pathogenic fungi (such as chitinase, protease, pectin, and cellulase) to cause mycelial lysis, mycelial cell malformation, spore wall fragmentation, and protoplasm overflow, rendering pathogens unable to infect host plants (Dahaieh et al., 2018). Bacillus amyloliquefaciens could control Fusarium wilt by inducing antioxidative enzymatic activities in maize seedlings (Paola et al., 2011). The biocontrol bacteria of sorghum wilt not only increased the activities of the related antioxidant enzymes, such as superoxide dismutase (SOD), peroxidase (POD), and ascorbic acid (APX), but also enhanced the activity of the defense enzymes phenylalanine ammonia lyase (PAL) and polyphenol oxidase (PPO) by inducing H₂O₂ reactive oxygen signaling. Moreover, the enzyme activities of urease, phosphatase, polyphenol oxidase and cellulase in soil were significantly improved (Dukare and Paul, 2021). This further indicates that different biological fungicides have different control mechanisms against F. oxysporum.

4.3 Chemical fungicides are also an effective method to control Fusarium

In addition, chemical fungicides are often used for the control of pathogenic Fusarium. Mancozeb showed a better control effect on ginger root rot (Veerubommu et al., 2013), and the antimicrobial effect of thiophanate-methyl on dry bean root rot was stronger than that of other chemical fungicides (Abd-El-Khair et al., 2019). In this study, the activities of carbendazim, thiophanate-methyl, hymexazol and mancozeb against F. solani of roselle were studied. The results showed that carbendazim had the highest activity against F. solani, followed by thiophanate-methyl, while mancozeb had the weakest effect. Previously, the inhibitory effect of carbendazim on F. solani was shown to be better than that of mancozeb (Rather et al., 2020), which was consistent with the results of our study. Zhang et al. (2021) reported that the control efficiency of thiophanate-methyl on saffron bulb rot was significantly outperformed by hymexazol, which was consistent with the results of our study. Guan et al. (2021) found that hymexazol exhibited a better inhibitory effect than carbendazim on F. oxysporum of bitter gourd. In the present study, although the IC₅₀ of hymexazol (0.0050 mg/mL) was lower than that of carbendazim (0.0082 mg/mL), the inhibition rate of carbendazim reached 82.61% at a lower concentration of 0.02 mg/mL, and when the concentration was 0.03 mg/mL, the rate reached 100%. The inhibition rate of hymexazol was only 94.68% when its concentration was 0.5 mg/mL. This indicated that the sensitivity of roselle F. solani to carbendazim was higher than that to hymexazol at low mass concentrations.