Antagonistic potential of certain soilborne fungal bioagents against Monosporascus root rot and vine decline of watermelon and promotion of its growth

Monosporascus cannonballus responsible for cucurbits Monosporascus root rot and vine decline, is worldwide spread notably in Tunisia. The most appropriate strategies to suppress disease development are those able to reduce the ascospores population using eco-friendly approach treatments. Seven soilborne fungal isolates were tested in vitro (by dual confrontation technique) and in vivo in the greenhouse as potential bioagents against three virulent M. cannonballus isolates. In vivo experiments were divided into two assays, preventive and curative treatments. Trichoderma viride and T. harzianum exhibited high inhibitory activities against M. cannonballus mycelial growth with values more than 90%, followed by Aspergillus niger (87.89%) and Paecilomyces victoriae (80.44%). Furthermore, these two Trichoderma spp. when applied preventively and curatively in in vivo trials, reduced significantly disease incidence (8.33% and 16.67-20.83%), root disease index (0.79-0.8 and 1.25-1.17), and reduced also ascospores index (1.5-1.54 asc/g of peat) and (2.54-2.42 asc/g of peat), respectively, in comparison with control treatments. Moreover, T. viride and T. harzianum enhanced the growth development of watermelon plants treated preventively and curatively in the greenhouse. They significantly improved different horticultural measurements with mean values of plant height (76.75-79.83 cm, and 81.83-80.92 cm), root volume (2.39-2.22 cm3, and 1.84-1.88 cm3), above grounds fresh weight (16.07-16.57 g, and 12.84-14.93 g) and dry wt. (2.49-2.6 g, and 2.66-2.70 g), underground fresh wt. (0.725-0.654 g, and 0.717-0.690 g) and dry wt. (0.147-0.214 g, and 0.156-0.152 g). Based on current results, it appears that Trichoderma spp. could be employed in soil treatments to promote watermelon plant growth and development.


Introduction
Several fungal species cause worldwide plant lesions, rots, loss of secondary; tertiary and feeder roots. These are associated with sudden and uniform collapse of entire fields 1-2 weeks prior to harvest, resulting in total crop loss. The main species associated with these syndromes is M. cannonballus (Cohen et al., 2000). Indeed, the onset of root infection occurs during early stages of growth, followed by wilting and death of plants later in the season (Cohen et al., 2012). Monosporascus root rot and vine decline (MRRVD) is particularly severe in arid and semi-arid worldwide cucurbits production. According to Boughalleb et al., (2010); Ben Salem et al., (2013); Rhouma et al., (2018), this disease is prominent in several melon and watermelonproducing areas in Tunisia, and can infect and produce perithecia in different cucurbit roots (Mertely et al., 1993).
Investigations on the biology of M. cannonballus demonstrated that its ascospores function as the only known survival structures in soil (Stanghellini et al., 2000). Furthermore, Waugh et al., (2003) pointed that one melon plant infected by M. cannonballus could support the production of approximately 400.000 ascospores. Theses authors added that fields are considered problematic when the soil is infected with two ascospores/ g of soil which could be associated with significant crop losses, and concluded that M. cannonballus is a monocyclic pathogen. Consequently, this pathogen has a great potential to maintain and/or increase its inoculum build up in the cucurbits rhizosphere (Cohen et al., 2012). Waugh et al., (2003) reported that management of M. cannonballus can be accomplished in case of early detection and quantification of its primary inoculum. In several studies, ascospores were extracted from soils through a physical method based on a sucrose centrifugation technique (Stanghellini and Rasmussen, 1992;Boughalleb et al., 2010).
The most appropriate strategies used to suppress plant disease development were those able to reduce the size of pathogen population (Fry, 1982). Control of MRRVD is currently based on integrating different approaches (Cohen et al., 2012;Ben Salem et al., 2015a). Farmers used to apply fungicides treatments (Cohen et al., 2007;Ben Salem et al., 2015b); however, these chemical methods cause hazards to human health and increase environmental pollution. Therefore, alternatives strategies are required for plant diseases control. Biological control is the best alternative and eco-friendly approach for such treatments, defined as total or partial destruction of pathogen populations by other organisms, which occur routinely in nature (Rojo et al., 2007). For example, the use of Trichoderma spp. (Mennatoullah et al., 2010;, and Chaetomium spp. (Sales et al., 2007) against MRRVD presented high efficacy when tested under in vitro and in vivo conditions. Reda et al., (2008) revealed that beneficial bacteria are also able to inhibit M. cannonballus growth and induce resistance in melon. The objective of the present investigation was to screen certain soilborne fungal antagonist's for abilities to reduce M. cannonballus growth under in vitro and in vivo conditions.

In vitro antifungal potential of fungal bioagents against pathogenic M. cannonballus
Antifungal activities of the seven fungal antagonists on radial mycelial growth of the three pathogenic M. cannonballus isolates was determined by dual confrontation technique on Potato dextrose agar (PDA) according to Boughalleb-M'Hamdi et al., (2017).
Two discs plugs (0.5 cm diameter) of each pathogen and antagonist (4 days-old culture) were transferred separately to a single PDA plate (9 cm diameter). The antagonist plug was placed on one side of the plate (about 2 cm from the edge of the plate towards the center), while the pathogen plug was placed at the other side of the plate opposite to the antagonist plug, leaving a distance of 5 cm between the two plugs. A plug of PDA medium was used as control treatment, while the pathogen plug was placed at the other side. Three replicates (two plates / replicate) for each individual treatment were conducted and the plates were incubated at 28 ± 2°C for five days. The percent of inhibition of pathogen radial mycelial growth was evaluated according to the formula of Hmouni et al., (1996): Where: Cn is the diameter of radial growth of the pathogen in the presence of the antagonist, whereas, C0 is the diameter of growth of the pathogen in the control treatment.

In vivo antifungal potential of the fungal bioagents
In vivo experiments were divided into two assays of preventive and curative treatments at March, 2015 in the greenhouse. The first preventive assay was carried out by dipping roots of watermelon seedlings (cv. Crimson sweet) 15 days old, into a flask containing a conidial suspension of the different antagonists (3×10 8 cfu\ml each) for 30 min. 24 h before adding 50 ml (9×10 6 cfu\ ml) of each pathogenic isolate, separately. For curative treatments, watermelon seedlings were treated with each antagonist separately 7 days after inoculation of each pathogen, by adding 10 ml of fungal antagonist's suspension to each pot (3×10 8 cfu\ ml). Watermelon seeds were sown in nursery seed trays, with 18 plants per each treatment having 3 replicates. The soil substrate used in this in vivo experiment consisted of a mixture of peat and vermiculite (1:1), which was autoclaved twice at 120°C. The pots were then placed in a greenhouse for 60 days. Two controls were performed; one by inoculating the plants with the pathogen only (positive control), while the other with dist. water (negative control). The experimental design was a randomized complete block design (RCBD), and the entire experiment was repeated twice.
Inoculation of the soil substrate with M. cannonballus ascospores only was performed as reported by Stanghellini et al., (2000); Aleandri et al., (2017), with some modifications. M. cannonballus isolates were obtained from twomonth-old PDA agar cultures, perithecia were washed and then ascospores were sieved (32 µm). Ascospores concentration was adjusted with dist. water (5 ascospores/g of peat). All growth parameters were measured 2 months after inoculation.
The number of symptomatic plants and the total number of plants evaluated in each treatment were used to estimate the disease incidence (DI) of MRRVD, by using the following formula: DI (%) = (Total no. of symptomatic plants/ Total no. of plants) x 100 in reference to Ben Salem et al., (2015a).
Watermelon plants were carefully removed after 2 months, the root system was then gently washed in tap water. Roots were inspected visually for evidence of root necrosis, and for observing roots bearing perithecia of M. cannonballus containing single spored asci. Each root system was rated for the severity of M. cannonballus lesions using a root disease index (RDI) which is an adapted scale from Novel Research in Microbiology Journal, 2018 Aegerter et al., (2000), where 0 = no symptoms; 1 = few lesions (covering <10% of root) and secondary root rot is slight; 2 = rot of secondary roots or lesions covering approximately 25% of the root; 3 = lesions covering at least 50% of the root and dead secondary roots; and 4 = general root rot where most of the root is affected.
Soil samples treated with seven fungal antagonists and inoculated with three M. cannonballus isolates separately, were air-dried at room temperature and sieved through a 2-mm mesh before their ascospores quantification was accomplished. M. cannonballus ascospores were extracted by a method adopted from Boughalleb et al., (2010). Initially, sub-samples were sieved through a 250 µm sieve. A 20-g subsample was placed in 200 ml of water, agitated for 5 min. and then passed through two superposed sieves (75 and 30 µm). The collected material was washed and centrifuged at 2000 g for 4 min. The supernatant was discarded and then 30-40 ml of 50% sucrose solution was added to the pellet and then centrifuged again for 2 min. at 2000 g. After centrifugation, the supernatant was passed through a mesh of 30 µm. The materials retained were distributed in Petri dishes. This suspension was stored at 4ºC until being analyzed. The ascospores characteristics and count were done under a stereomicroscope (Nikon SMZ 1000) at a magnification of ×60. After the initial (Pi = 5 asc/g peat) and final (Pf) ascospores count, the following formula was applied to determine the percentage of the ascospores index (AI) according to Ferreira, (2011): After determination of the fresh wt. of above ground (stem + leaves) and underground (root) portions, plant samples were placed in an oven at 60°C for 48 h to determine the dry wt. (Heitholt, 1989). The height of the plant was measured (cm) using a flat ruler. Root volume (cm3) was determined by the immersion method as described by Musick et al., (1965), through comparing the levels of water before and after immersing the whole root in a known volume of this water.

Statistical analysis
Data were analyzed by ANOVA using SPSS version 20.0 statistical software (SPSS, SAS Institute, USA). Differences between treatments were determined by Duncan multiple range test at 5% of significance level.

In vitro antifungal efficacy of bioagents against M. cannonballus isolates on PDA
The seven antagonistic fungal isolates exerted high significant reduction (<0.01) on radial mycelial growth of M. cannonballus isolates after five days of incubation. The linear decrease of growth of all the pathogenic isolates ranged from 95.16% (MT41/T. harzianum) to 47.25% (MT3/ P. purpurascens) ( Table 1). Statistical analysis revealed high significant interactions between M. cannonballus isolates and the antagonists (<0.01).
The two Trichoderma spp. showed a good ability to limit the mycelial growth of all M. cannonballus isolates in vitro. In fact, the mycelial growth of the three M. cannonballus isolates decreased in presence of T. viride and T. harzianum with values ranging between 91.93 and 92.11%, respectively (Table 1, Fig. 1). Moreover, in vitro assay revealed that A. niger possessed a good antifungal potency with mycelial inhibition rate between 88.7% (MT3) and 87.89% (MT4), followed by Paecilomyces victoriae (80.44%) and P. purpurascens (48.15%) ( Table 1). This antagonistic potency was not only on the mycelial growth reduction, but also on the microscopic hyphal aspect. Compared to controls, M. cannonballus isolates treated with Trichoderma spp. and A. niger exhibited a mycelium with strong lyses, induction of mycelial cords via anastomosis between hyphal filaments and mycelium winding (Fig. 1).
Novel Research in Microbiology Journal, 2018

In vivo antifungal potential of the fungal bioagents on watermelon plants infested with M. cannonballus in the greenhouse
Statistical analysis indicated that plants inoculated with M. cannonballus isolates and treated preventively and curatively by the seven antagonistic fungi was highly significant (<0.01). However, no difference was detected between M. cannonballus isolates (≥0.05). Watermelon plants seemed healthy with no symptoms of M. cannonballus infection (disease incidence = 0%), when treated with T. viride, T. harzianum, C. globosum and A. glaucus for MT3 isolate; T. viride, T. harzianum, P. purpurascens and A. niger for MT4, and P. purpurascens, C. globosum and A. niger for MT41 isolate, when used as preventive treatment (positive control = 100%; negative control = 0%).
However, when plants were treated curatively, the antagonists showed varied antifungal activity with mean value of disease incidence 34.72% (ranging between 0 (MT4/ Paecilomyces victoriae) and 100% (MT4/ A. glaucus), and 40.28% (ranging between 0 (MT3/ T. viride) and 75% (MT3/ Paecilomyces victoriae; MT3/ P. purpurascens)). These obtained results indicated that T. viride and T. harzianum applied curatively reduced significantly disease incidence, recording the lowest value of 25% compared with positive control (100%) and negative control (0%) ( Table 2). These findings were confirmed after 2 months by above ground symptoms on watermelon plants infested with M. cannonballus isolates (Fig. 2).  Observing the infested roots treated curatively, T. viride and T. harzianum showed the most significant reduction of disease severity index with values ranged between 1.25 (0.63 (MT4) -2 (MT41)), and 1.17 (0.88 (MT3) -1.5 (MT41)), respectively, whereas, positive control value = 3.71 (Table 3). In addition, perithecia of this pathogen were not observed on roots treated curatively by these two Trichoderma spp. However, M. cannonballus isolates were pathogenic on watermelon plants in the presence of some antagonists such as; C. globosum, A. niger, A. glaucus and P. victoriae with root severity mean values of 2.33, 2.54, 2.17 and 2.42, respectively. These infested watermelon plants showed roots with typical symptoms of MRRVD including; lesions, rots, loss of secondary, tertiary and feeder roots, in Novel Research in Microbiology Journal, 2018 addition to production of perithecia (Fig. 3).  Root disease index scale of 1-4; where 0 = no symptoms; 1 = few lesions (covering <10% of root), secondary root rot slight; 2 = rot of secondary roots or lesions covering approximately 25% of the root; 3 = lesions covering at least 50% of the root and dead secondary roots; and 4 = general root rot, most of the root affected.

In vivo potency of fungal biocontrol agents on ascopsores populations of M. cannonballus
Application of antagonists preventively and curatively reduced the ascospores population levels (<0.01) (Tables 4 and 5). Results revealed and confirmed the efficiency of both T. harzianum and T. viride isolates by decreasing significantly the ascospores densities which varied from 1.42 (MT41) -1.69 (MT3) asc/g of peat, and between 1.29 (MT41) -1.73 (MT3) asc/g peat, respectively (positive control = 5.5 asc/g of peat). The lowest reduction of ascospores number was registered on plants treated by T. harzianum and T. viride with means of -260.68, and -243.47%, respectively. The effect of the other antagonists varied between 3.28 (Paecilomyces victoriae) -4.23 (P. purpurascens) asc/g peat. The decrease of ascospores index ranged from -52.54 to -18.17%, respectively (positive control = 8.5%).

In vivo potency of fungal biocontrol agents in promoting growth parameters of watermelon plants infested with M. cannonballus
The interaction between M. cannonballus and the seven antagonists was significant (p<0.05). However, there were no significant differences between M. cannonballus isolates. Growth promotion results for watermelon plants treated preventively are presented in (Tables 4 and 5). All treatments differed significantly from positive control. The best treatment was the combination of T. harzianum and T. viride with M. cannonballus isolate (MT4), which showed significant increase in growth parameters of aboveground parts, compared with the other treatments. The values of the aboveground fresh wt. were about 18.138 and 16.5 g, plant height were of 79 and 83 cm, however, aboveground dry wt. with MT3 were (2.65 and 6.52 g). After 9 weeks in the greenhouse, the root system was collected from all treatments and checked. Both MT3 and MT4 isolates produced fewer roots of infested plants. The underground fresh, dry wt. and root volume values for plants treated with T. viride ranged between 0.688 g (MT41) and 0.758 g (MT3), from 0.138 g (MT3) to 0.125 g (MT4 and MT41), and from 2.025 cm3 (MT3) to 2.825 cm3 (MT41), for the three growth parameters, respectively. For T. harzianum, results revealed low difference compared with the previous values. The improvement rates of the three growth parameters compared with the negative and positive controls were 98% -260% for underground fresh wt., 87.5 % -400% for underground dry wt., and 102% -400 % for root volume, respectively. Watermelon plants treated curatively showed an increase of the above and underground growth parameters compared with the positive control, however, there were a slight difference for plants treated preventively. Indeed, T. harzianum presented a good improvement of the above (14.93 g) and underground (0.717 g) fresh wt., and above and underground dry wt. (2.7 and 0.156 g, respectively). Trichoderma treatments were very effective against M. cannonballus infested watermelon plants; the severity of infection was reduced and the growth parameters of these plants improved as well.

Discussion
The control of soilborne pathogens was difficult as they produce viable structures such as ascospores which were resistant to adverse environmental conditions (Cohen et al., 2000). However, Rojo et al., (2007) pointed that the misuse of fungicides to manage these pathogens caused enormous problems to ecosystem and human's health.
Novel Research in Microbiology Journal, 2018 cannonballus isolates treated with three different antagonistic fungal spp.; small letters are for means comparison of the different antagonist's in the same column. b Duncan's Multiple Range Test is for percentage of M. cannonballus ascospores index mean in comparison among the seven antagonistic spp. for the different M. cannonballus isolates; capital letters are for comparison of means in the same row (Means of 6 plants per each replicates of three). c Probabilities associated with individual F tests. nd: not determined. After the initial (Pi = 5asc/g peat) and final (Pf) ascospores count, the following formula was applied to determine the percentage of the ascospore index (AI): AI (%) = (1-Pi/Pf) x100 Biological control involves the use of one living organism to control another, and this management technology has received much attention in recent times. The number of biocontrol agents (BCAs) registered for use is relatively low, although their application was successful and proved to cause enhancement in crop growth (Ben Salem et al., 2016).
In the current study, in dual culture in vitro assays using several fungal genera such as; al., (2006). Zhang et al., (1999) reported that T. virens exhibited in vitro antifungal activity by inhibiting mycelial growth of M. cannonballus and other soilborne pathogens such as Didymella bryoniae, Macrophomina phaseolina and Phomopsis cucurbitae. T. album isolates significantly suppressed the growth of M. cannonballus and it subsequently overgrew the pathogen (Zhang et al., 1999), while, Bacillus megaterium was less inhibitive (Mennatoullah et al., 2010).
Current in vivo assay results were in accordance with other studies such as Sanz et al., (1998), who reported that Trichoderma spp. exhibited high antifungal activity against Monosporascus sp. and Acremonium cucurbitacearum. According to Zhang et al., (1999), T. virens colonized the root systems of muskmelon plants, significantly reduced M. cannonballus colonization of roots, and suppressed disease severity of seedlings by seed treatments. In another study of El-Kolaly and Abdel-Sattar, (2013), treatments with T. harzianum and T. ressei have not only reduced the incidence of MRRVD, but also reduced the M. cannonballus root invasion, suggesting that BCAs were limiting the pathogen infection. Ben Salem et al., (2016) pointed that among six antagonists evaluated for biocontrol potential against M. cannonballus, only T. viride and T. harzianum significantly reduced disease incidence and severity index after a preventive treatment through soil drenching. Indeed, success of the preventive applications could be attributed to the hyperparasitism of the BCAs in the plant rhizosphere, which inhibited the root infection by soilborne pathogens and reduced their inoculums build up (Rini and Sulochana, 2007). Trichoderma spp. was demonstrated to have potential in M. cannonballus disease management for in vitro and in vivo assays (Pastrana et al., 2016;Boughalleb-M'Hamdi et al., 2018).
The increasing numbers of studies have contributed to unveiling the molecular basis of the plant-Trichoderma interaction, and the beneficial effects of Trichoderma spp. to plants. Some selected Trichoderma strains were shown to have direct positive effects on plants such as; increasing their growth potential and nutrient uptake, fertilizer use efficiency, percentage and rate of seed germination, in addition to stimulation of plant defences against biotic and abiotic stresses. It has been reported by Segarra et al., (2009)  activate induced systemic resistance (ISR) in plants, a mechanism triggered after root colonization by nonpathogenic rhizobacteria or fungi and is regulated by a specific signal transduction cascade.
Trichoderma spp. are also known to produce a large number of antibiotics including; trichodermin, trichodermol, polyketides, peptaiboils, sesquiterpenes, and steroids, all these active compounds are known to promote plant growth besides having biocontrol potential (Harman et al., 2004). Later, Müller et al., (2013) added that these fungi are prolific producers of a number of secondary metabolites with pharmaceutical and biotechnological significance that involve; nonribosomal peptides, peptaibols, poliketides, pyrones, siderophores, beside volatile and non-volatile terpenes. For these reasons, they are major sources of many biofungicides and biofertilizers (Kaewchai et al., 2009). The germination of M. cannonballus ascospores and subsequent attachment to roots, occurs exclusively only in the cucurbits rhizosphere. According to Stanghellini et al., (2010), the interaction of M. cannonballus with susceptible cucurbits roots appears to be strongly related to the microbial composition in the rhizosphere.

Conclusion
Trichoderma spp. applied preventively and curatively showed significant effect on watermelon plants infested with M. cannonballus, and could be recommended for biocontrol use. T. viride and T. harzianum allowed not only the protection of plants, but also the improvement of the agronomic parameters including better axial growth and greater root biomass. Based on the current results, it is deduced that tested T. viride and T. harzianum could be employed in soil treatments as BCA's to induce cucurbits systemic resistance, through a specific signal transduction cascade. The systemic resistance induction of cucurbits by Trichoderma spp. against M. cannonballus is a subject of future research.