RESISTANCE OF CUCURBITA SPP. TO FUSARIUM SOLANI FOR USE AS ROOTSTOCK

Fusarium solani f. sp. cucurbitae (Fsc) is a soil pathogen and the adoption of resistant rootstocks is an effective method of control. Hybrids of Cucurbita spp. are the main rootstock used for watermelon. This study aimed to evaluate the virulence of Fsc race 1 (Fsc1) in three cucurbits and check the reaction of Cucurbita spp. genotypes to the fungus for its use as rootstock. Four experiments were performed. The virulence of Fsc1 in melon, watermelon and Cucurbita spp. was evaluated in experiment I. In experiment two and three the severity of the pathogen in 19 cucurbits was analyzed. And experiment four evaluated the compatibility of these genotypes as a rootstock for watermelon. Cucurbita spp. proved to be more resistant to Fusarium solani f. sp. cucurbitae race 1 than the melon. The high frequency of plants resistant to Fsc1 was found in the genotypes BGC622, BGC620, BGC567, BGC530, BGC186, BGC381, BGC692, BGC082, ES0061 and ES0062. In addition, the evaluated genotypes may be used as watermelon rootstocks, except for the ES530 strain, which was incompatible with the cultivar BRS Opara, but could be used in pumpkin breeding programs, as well as other resistant genotypes.


INTRODUCTION
Cucurbits stand out in importance from the nutritional, cultural and economic point of view, being grown mainly for human and animal food, with a significant participation in food production. In 2017, the world production of fruits of this family totaled more than 262 million tons, distributed as follows: watermelon (45.1%), cucumber (31.9%), melon (12.5%) and 10.5% constituted for pumpkins and squashes (FAO, 2017).
In Brazil, the production of melons and watermelons in 2018 totaled more than 2.8 million tons (IBGE, 2019a). There is no information on the national production of pumpkins and other cucurbits, but in the country the agricultural establishments in the survey of 2017 recorded production exceeding 417 thousand tons of pumpkin, 'jerimuns' and squash, and more than 637 thousand tons of the other cucurbits, for example, zucchini, loofa sponge, chayote, gherkin and cucumber (IBGE, 2019b).
The continuous and intensive cultivation of these species in the same place can cause an increase in the density of soilborne pathogens, affecting the quality and productivity of these crops. Among these pathogens, Fusarium solani f. sp. cucurbitae (Fsc), present in practically all areas of cultivation of cucurbits in Brazil, regardless of being irrigated or rainfed, as it is easily disseminated by contaminated seeds, tools and implements, as long as environmental conditions favor the occurrence of the pathogen. Under conditions of high humidity, for example in rainfed cultivation after the rainy season, associated with temperatures between 20 ºC and 28 ºC, that favor the occurrence of the pathogen (REGO; CARRIJO, 2000).
According to Martyn et al. (1996), most cucurbits are susceptible in the seedling phase to Fsc, which produces resistance structures called chlamydospores, surviving in crop residues, soil and infected seeds, even under adverse environmental conditions for long periods. Two races of this pathogen have been identified based on the specificity of the tissue in which they act. Race 1 infects hypocotyl and branches, causing cortical rot, as well as in ripe fruits, causing dry rot, whereas, race 2 affects only the fruits (TOUSSON; SNYDER, 1961).
Chemical control of Fusarium solani is difficult, although there are products with efficient action, which have restrictions on use, as there is no record recommending its use in cucurbit crops at the Ministry of Agriculture, Livestock and Food Supply. However, other management measures are being studied, such as: crop rotation, solarization. (PÉREZ-HERNÁNDEZ; PORCEL-RODRÍGUEZ; GÓMEZ-VÁZQUEZ, 2017), antifungal essential oil (CRUZ et al., 2015;NASCIMENTO;VIEIRA;KRONKA, 2016) and grafting.
Some cucurbits are suitable for use as rootstocks for watermelon, cucumber or melon. Among them, the following can be cited: Cucurbita spp., Lagenaria siceraria, Citrullus spp., Benincasa spp. and Luffa spp. However, the most popular ones are the rootstocks of Cucurbita spp., especially hybrids of C. maxima x C. moschata, which have been widely used in watermelon, as they provide nonspecific but efficient protection to a wide variety of pathogens, which colonize the root system, and tolerance to abiotic stresses (PICÓ et al., 2017). In addition, some species of the genus Cucurbita spp. and C. lanatus var. citroides exhibit greater root development and rusticity, favoring better use of water and nutrients, and increased productivity, without affecting the quality of the grafted cultivar's fruits (SANTOS et al., 2016). The use of genetic resistance for soilborne pathogens is one of the most important technological advances in agriculture. In this context, the adoption of grafting can be a short-term and viable strategy to solve problems related to soil pathogens. Based on genotypes from the Active Cucurbit Germplasm Bank (BGC) and from Embrapa Semiárido breeding programs, Cucurbita spp. and C. lanatus var. citroides were selected for resistance to soil pathogens and with potential for watermelon rootstock, considering increases in fruit productivity and quality (GAMA et al., 2013;SANTOS et al., 2014;SANTOS et al., 2016).
In view of the grafting potential in commercial watermelon crops in Brazil, due to problems related to soil-dwelling pathogens, it is necessary to identify sources of resistance to such pathogens in the germplasm of Cucurbita spp. In this context, the objective of this work was to evaluate the virulence of Fsc race 1, in three cucurbits, and the reaction of Cucurbita spp. to Fsc1 aiming at its use as rootstock.

MATERIALS AND METHODS
The four experiments were conducted in a greenhouse at Embrapa Semi-arid, in the municipality of Petrolina-PE, Brazil, from March 2017 to June 2018. The genotypes came from the Active Cucurbit Germplasm Bank for Northeast Brazil and belonged to the breeding collection of Cucurbita spp. (BGC) of Embrapa Semi-Arid. An isolate of Fsc1, from the Phytopathology Laboratory of Fungi Collection of Embrapa Semi-Arid, obtained from pumpkin plant with symptoms of stem rot, was used. The genotypes were planted in pots containing 500 mL of a commercial substrate for vegetables, in randomized blocks with 20 replicates (Experiment 1) and 10 replicates (Experiments 2, 3 and 4), the experimental unit consists of one plant.
In experiments 1 and 3, inoculation occurred at 10 days after planting, when the plants had one true leaf, while in experiment 2, inoculation occurred at 45 days after sowing, when the plants had three or four true leaves. In the inoculations, there was a difference of age of the plants in the experiments 2 and 3, to verify if the reaction of the plants to Fsc1 was influenced by their stage of development. a) Experiment 1 To assess the virulence of Fsc1, three species of cucurbits, previously established as susceptible, were inoculated with the fungus: two genotypes of Cucurbita spp. (BGC010 and BGC814), two of melon (cultivar ES0442 and strain ES0371.04) and two of watermelon (ES31654/2 and ES31662/2). For inoculation, the Fsc1 isolate was grown in Potato Dextrose Agar (PDA) medium for 15 days, at 25 °C and 12 h photoperiod. Before inoculation, the collar of each seedling was injured by a set of three disinfected 3-mm pins. Then, a disk of culture medium 5 mm wide, containing structures of the pathogen, was deposited under a cotton ball moistened with sterile water (to promote a humid chamber at the inoculation site), fixing them with aluminum tape around the collar of the plant. The inoculum remained in contact with the injured tissue for 72 h. After inoculation, the plants were kept in a greenhouse at a temperature of 29.9 ± 3 °C and 58 ± 2% relative humidity (experiments 1 and 3) and at a temperature of 27.4 ± 4 °C and 74 ± 1% relative humidity (experiment 2). The inoculation method and procedures adopted were similar for all inoculations.
At eight and twenty two days after inoculation, in experiments 2 and 3, respectively, the following parameters were measured: the length and width of the lesions, using a millimeter ruler; the depth of the lesion, assessed by a scale of scores from 1 to 3 (score 1 = superficial lesion; 2 = intermediate lesion and 3 = deep lesion, reaching the conducting vessels) and the severity of the disease was also assessed, adopting a scale of scores from 1 to 5, adapted from Preisigke (2014): 1 = plant without symptoms -highly resistant (HR); 2 = lesion with less than 50% of the circumference of the hypocotyl, without of depression, without soaking or moisture within the lesion, as well as without strangulation of the hypocotyl -resistant (R); 3 = lesion with more than 50% of the circumference of the hypocotyl, but without strangulation, depressed or not, with soaking or moisture -moderately resistant (MR); 4 = destruction of the cortex and strangulation of the hypocotyl -susceptible (S); and 5 = dead plant -highly susceptible (HS). The assessments of severity in experiments 2 and 3 were carried out when more than 50% of BGC010 plants (susceptibility previously known) have reached the score 4 (susceptible). c) Experiment 4 To evaluate the compatibility of Cucurbita spp. resistant to Fsc1 as a rootstock for watermelon, seven genotypes were evaluated (ES567, ES620, ES530, ES0061, ES0062, ES0007 and ES0012) as a rootstock for the BRS Opara watermelon cultivar. Additionally, this cultivar was evaluated without grafting and under autografting, totaling nine treatments, with 10 repetitions and the experimental unit consisting of one plant. It is of fundamental importance in the selection of the rootstock that it is compatible with the cultivar to be grafted.
The planting of the genotypes used as rootstock was carried out four days after the planting of the scion cultivar, due to the vigor of the pumpkins. At nine days after sowing, when the plants had the first true leaf, grafting was carried out using the approach method. This method consisted of making a diagonal cut from top to bottom in the rootstock hypocotyl and from bottom to top in the hypocotyl of the cultivar to be grafted. Then, the two seedlings were joined and fixed with devices made with aluminum foil (replacing the grafting clips). After this procedure, the grafted plants were transferred to plastic cups of 150 mL with small holes at the bottom, containing the commercial substrate for vegetables Tropstrato ®.
The seedlings were kept in a greenhouse, at 25 ºC ± 1.0 and 65% RH ± 1.0, for seven days, when "weaning" was performed (cutting the root system of the grafted cultivar and eliminating the aerial part of the rootstock). As a preventive phytosanitary control, 0.25 mL.L -1 of Difenoconazole was applied. The following day after weaning, the surviving plants were placed in a pot with a capacity of 10 liters of soil, containing soil + cattle manure, in a 3:1 ratio.
The plants were evaluated for: seedling survival index after weaning (SIW), by counting the seedlings that survived 24 hours after the elimination of the root system of the scion and the aerial part of the rootstock, and estimating the percentage in relation to the total grafted; seedling survival index after transplantation (SWT), by counting the plants that remained alive up to 30 days after transplanting, and estimating the percentage in relation to the total of plants that were submitted to transplantation; and total survival index, by the sum of SIW and SWT.
The development of plants was also evaluated through the variables: length of the main branch, with the use of measuring tape; precocity, considering the period from sowing to the anthesis of the first male and female flower; and dry mass of plants, where the plants (root + branch + leaves + flowers) were placed in paper bags and dried in an oven, at 63 ºC, for 120 h. The dry mass was determined on an electronic scale with an accuracy of 0.001 g.
The data obtained in the evaluations were subjected to analysis of variance and the means were compared by the Tukey test, at 5% significance level. In experiments 2 and 3 Pearson's correlation coefficients were estimated between the variables: length, width and depth of the lesion and disease severity scale. Descriptive statistics was applied to group the genotypes in resistance classes and to analyze the frequency of plants within each score. For the analysis of variance, Sisvar ® software version 5.6 (FERREIRA, 2015) was used and Pearson's correlation coefficient was estimated by SPSS ® (IBM SPSS Statistics for Windows, version 20.0, 2011, USA).

RESULTS AND DISCUSSION
When inoculating different species of cucurbits with the fungus, it was observed that the melon was more sensitive to the inoculum of Fsc1 than the genotypes of Cucurbita spp. but the severity of the watermelon inoculum was similar to that of melon and Cucurbita spp. (Figure 1), demonstrating the virulence of the pathogen and the wide range of hosts within the cucurbits.
Corroborating with the present study, Nagao, Sato and Ogiwara (1994), when submitting genotypes of Cucurbita spp. and Cucumis sativus to different isolates, observed that cucumber was less susceptible to Fsc1 than the species of the genus Cucurbita spp. These results suggest the existence of different reactions of cucurbit species to Fsc1. Thus, genotypes can respond differently to different isolates, depending on the variability in the degree of virulence of Fsc1 populations, the environment (temperature, abiotic stresses, among others) and the age of the plant. to Fusarium solani f. sp. cucurbitae race 1. Bars with the same letters do not differ according to Tukey's test at the 5% probability level. Score scale from 1 to 5 (where: 1 = highly resistant plants; 2 = resistant; 3 = moderately resistant; 4 = susceptible and 5 = highly susceptible plant).
In 26 isolates of Fsc1 from watermelon, melon and cucumber at different growth stages, grown in different locations, great genetic diversity was observed among the isolates, with the formation of 10 similarity groups, with no correlation with geographic origin or species (ALYMANESH et al., 2009). In addition, it must be considered that there may be genetic variability, for resistance or tolerance to Fusarium solani, between and within species, as found by Anjos et al. (2018) in genotypes of Capsicum spp.
( Table 1) shows the results of the severity of the Fsc1 lesion in the genotypes of Cucurbita spp. in experiments 2 and 3. For the analyzed variables, different responses were found between the genotypes as a function of the experiment. In general, the lesion length was higher in the plants evaluated in experiment 2, than in those inoculated in experiment 3. Shorter length of the lesion on the plant collar was observed, in experiment 2, in BGC381, BGC249, BGC082, BGC567, BGC530, BGC830, BGC814 and BGC622, and the last-mentioned stood out with 0.84 cm, as well as BGC962 in experiment 3 (0.16 cm). On the other hand, greater lesion length was observed in the genotypes BGC685 (experiment 2), and BGC249, BGC381, BGC830 and ES0030 (experiment 3). Approximately 82% and 77% of the genotypes had lesion length with intermediate size when inoculated with Fsc1 in experiments 2 and 3, respectively.
The length of the lesion is an important parameter in the evaluation of Fsc1, as it is one of the characteristic symptoms of the disease and after artificial inoculation there is the appearance of lesions in the plant tissue with longitudinal cracks, especially in the collar. Although the inoculation in experiment 3 occurred in plants with younger age, in general, they showed shorter lesion length. Different isolates can produce different amounts of toxic metabolites, being expressed, indirectly, by the intensity of the induced symptoms. This does not apply to the present study, since the same Fsc1 isolate was used. Antonio et al. (2019) when evaluating different genotypes of melon inoculated with Fusarium solani found lesions ranging from 1.09 to 4.10 cm, being classified as moderately resistant and highly susceptible, respectively.
It was found that, in experiment 2, the accessions BGC530, BGC830, BGC814, BGC622, BGC082, BGC010, BGC249 and BGC567 stood out with a smaller lesion diameter, varying from 0.34 to 0.44 cm (Table 1), whereas BGC685 had larger diameter (0.86 cm). In experiment 3, a smaller diameter of the lesion was observed in 70.6% of the genotypes, ranging from 0.26 cm to 0.35 cm. In this trial, BGC010, BGC495 and BGC 814 exhibited the largest diameter of the lesion, ranging from 0.44 cm to 0.50 cm. In kiwi fruits, lesions with a diameter of 20 mm were observed nine days after inoculation with Fusarium solani, which later covered the whole fruit (YANG et al., 2018).
The diameter of the lesion caused by Fsc1 in plant tissues is a characteristic that can define the severity of the disease, because the greater its width, the greater the possibility of collar constriction, which may result in plant collapse. However, in Cucurbita spp. there are no reports available that associate the width of the lesion with the severity of the disease in seedlings inoculated with the studied fungus, as both plants classified by scale of grades as resistant and those classified as susceptible exhibited similar lesion widths.
The deeper the lesion on the plant stem caused by the inoculation of Fsc1, the greater the possibility of it reaching the vascular system, which, when damaged or broken, compromises the flow of water and nutrients, resulting in wilting, yellowing and death of the plant. In the recent literature, there are no studies evaluating the depth of the lesion in the collar of the plant caused by Fsc1 in Cucurbita spp. However, Wang et al. (2014) observed that, during storage, sweet potato roots with Fusarium solani developed slightly concave lesions on the surface, but when cutting these roots, they found that the lesions could extend to the center of the organ.
In ripe melon fruits inoculated with four Fsc1 isolates from different locations, it was observed that the maximum depth of the lesion varied from 3.0 cm to 3.4 cm (CHAMPACO; MILLER, 1993). Also, according to these authors, an Fsc1 isolate can be weakly pathogenic in seedlings and highly aggressive to fruits. However, most cultivated cucurbits are sensitive to the fungus in the seedling phase (ABDUL -HASAN; HUSSEIN, 2016). In addition, melon appears to be more sensitive to Fsc1 than Cucurbita spp. Considering that the tissues found in the stem of pumpkin seedlings are more lignified than those of most commercial melon seedlings, probably those that have greater difficulty for penetration than melon seedlings, which are more easily degraded and more vulnerable to the action of the fungus.
The results of Fsc1 severity and plant frequency regarding resistance in the genotypes of Cucurbita spp. in two experiments are found in (Table 2). For the plants inoculated in experiment 2, it was found that the accessions BGC530, BGC082, BGC249, BGC622, BGC830 and BGC010 showed the lowest severity of the disease (average score from 2.0 to 2.4) ( Table 2). The BGC685 access showed the highest average score (5.0) for disease severity, with 100% of plants showing a high susceptibility reaction to Fsc1. In 14 cultivars of zucchini, grown in different cycles, for 20 months, in bags with pearlite infested with Fsc1, it was found that, although the severity of the disease was significantly reduced after 14 months, all genotypes were highly susceptible to the disease (PÉREZ-HERNÁNDEZ; PORCEL-RODRÍGUEZ; GÓMEZ-VÁZQUEZ, 2017). Table 2. Severity (S) and frequency of reaction of Cucurbita spp. to Fusarium solani f. sp. cucurbitae race 1 in two inoculation periods: experiment 2 and experiment 3. Also in experiment 2, a group of medium severity (scores 2.4 to 3.8) was formed by 54.5% of the genotypes. A frequency of plants highly resistant to Fsc1 was found in BGC530 (3.7%), BGC814 (4.2%) and BGC830 (15.8%) ( Table 2).
In experiment 3, 73.7% of the genotypes were classified as resistant to Fsc1, with a score of 1.9 to 2.4 ( Table 2). In that same experiment, it was possible to observe that BGC620 and ES0062 stood out for showing 100% of resistant plants. Low frequencies of highly resistant plants were observed in ES0061 (7.1%), BGC010 (10%), ES0030 (10%) and BGC249 (28.6%). The promising genotypes with high frequency of resistant plants should be submitted to a selection program under inoculation, followed by self-fertilization to obtain homozygosity of this character. Thus, they can be used as rootstocks for watermelon.
The cultivation of watermelon with the use of resistant rootstocks not only promotes the control of soil pathogens, but can also increase fruit productivity and quality. This increase in watermelon yield can be attributed both to the control of the disease and to the improvement in plant growth, due to the characteristics of vigor and rusticity of the root system, allowing the absorption of a greater amount of nutrients (SANTOS et al., 2016). Other genotypes resistant to Fsc have already been identified in different cucurbits suitable for watermelon rootstock in Citrullus colocynthis and hybrids of C. colocynthis x C. lanatus var. citroides (BORGI et al., 2009). C. lanatus var. citroides (BOUGHALLEB et al., 2016) and Cucurbita spp. (BOUGHALLEB et al., 2008).
It was observed that when inoculating the plants in experiment 3, six of the genotypes showed the same behavior observed when they were inoculated in experiment 2, remaining resistant (BGC249, BGC622 and BGC530) or susceptible (BGC814, BGC830 and BGC010). On the other hand, the genotypes BGC381 and BGC685, which showed a susceptibility reaction in experiment 2, showed a resistance reaction in experiment 3 ( Table  2). The greater or lesser susceptibility of a genotype to Fsc1 can vary according to the host-pathogenenvironment interaction. And as already noted by Boughalleb et al. (2005), genotypes may respond differently to different isolates, due to the variability in the degree of virulence of FSC1 populations. Moreover, the environment (temperature, stress etc.) and age of the plant can also interfere in the plants' reaction to the presence of the pathogen.
Thus, in the present work, although the plants of Cucurbita spp. were at different ages at the time of inoculation, four true leaves in experiment 2 and one in experiment 3, probably, this was not the factor that most contributed to the lower severity of the disease in experiment 3, since the tender and succulent tissues, exhibited by the seedlings in this experiment, would be more sensitive to the action of the fungus, contrary to what was observed in the results presented. In addition, the same isolate was used in both experiments. Thus, the divergence regarding the resistance or susceptibility reactions observed for the same genotype in the two inoculations can be attributed to environmental conditions, where the temperature of 27.4 ± 4 °C and 74 ± 1% RH (experiment 2) favored the action of the fungus during the inoculation period, while at a temperature of 29.9 ± 3 °C and 58 ± 2% RH it promoted a reduction in the severity of Fsc1 (experiment 3). According to Rego and Carrijo (2000), temperatures between 20 ºC and 28 ºC associated with high humidity are conditions that favor the occurrence of the pathogen. Similar results to that observed in the present study were found in soybean, in Paraná, in two inoculation periods: July / 2002 and September / 2002, where there was a difference in the intensity of the symptoms of F. solani f. sp. glycine between the two periods of experimentation, and the severity of the disease in the second one was lower than that of the first experiment (KLINGELFUSS; YORINORI; DESTRO, 2007), demonstrating the effect of environmental factors on this characteristic.
For Cucurbita pepo grown in greenhouses, 97.2% of the plants exhibited symptoms of SCF and, later, 70.8% of these died at an average temperature of 29 ºC (PÉREZ-HERNÁNDEZ; PORCEL-RODRÍGUEZ; GÓMEZ-VÁZQUEZ, 2017). Thus, occurrence of high temperatures, humidity and mild night temperatures, as well as plants stressed by weather conditions, nutritional problems and / or attack of other pathogens, associated with damaged tissue, are factors contributing to the incidence of Fsc.
Pearson's correlations between the variables showed different behavior depending on the experiment. In experiment 2, even with the low magnitude in the correlation coefficient, the scale of scores used to assess the severity of the disease was significant and positive only with the depth of the lesion. However, the length showed a positive and significant correlation with the width and depth of the lesion (Table 3). Table 3. Correlation between the length, diameter and depth of the lesion, and the score scale used for the evaluation of Fsc1severity in Cucurbita spp. genotypes submitted to two inoculation periods: experiment 2 (E2) and experiment 3 (E3).

Experiment
Diameter ( In experiment 3, the score scale adopted for disease severity exhibited significant and positive correlations with the length (high correlation), the width and depth of the lesion. Nagao, Sato and Ogiwara (1994) evaluated the severity of Fsc1 by a scale of scores in genotypes of Cucurbita spp. and Cucumis sativus, associating susceptibility to the disease with severe infection of the hypocotyl, root discoloration and plant death. In melon germplasm, the reaction of susceptibility to Fsc1 was also assessed by a scale of scores, associated with the visual aspect of the lesion size and stem constriction (GUIMARÃES, 2016). Galon (2013), when evaluating 33 genotypes of cucurbits submitted to two distinct inoculants of Fsc1, considered as susceptible plants those that exhibited severe wilting, chlorosis and leaf necrosis and death.
Thus, it can be inferred that the score scale used in the present work proved to be effective in distinguishing susceptible and resistant plants, resulting in the optimization of the process of selection of resistant plants by trained people, especially when the goal is to evaluate a large number of plants. However, measuring the width and length of the lesion per plant, despite taking time, making the selection process slow and costly, is a more objective assessment.
According to Goto; Santos and Cañizares (2003), in the seedling phase, the low survival rate of the grafted plants, the excessive growth and the rupture of the grafting point indicate incompatibility of the rootstock with the scion. In the post-transplanting phase, the lack of compatibility is expressed with lack of growth and defoliation, premature death of the plant and poor callus formation, characterized by excessive development below or above the grafting region.