Tomato (Solanum lycopersicum) is one of the most produced vegetables in the world; Mexico is ranked ranks tenth in worldwide production. At the national level, the major producing states are Sinaloa, San Luis Potosí and Michoacán. In 2018, the area for cultivation to tomato was 49,415 ha, from which 7,737 ha were cultivated under the protected agriculture system, mainly under the shade mesh modality. The same year, in the state of Guerrero an area of 119.8 ha was also allocated to tomato under the same conditions (^{SIAP, 2018}). However, although tomato is grown under protected conditions, it is affected by diseases caused by Phytophthora infestans, Oidiopsis taurica, Rhizoctonia spp., Fusarium oxysporum, Oidium neolycopersici and Pseudomonas syringae pv. tomato, among others (^{Rodríguez-Alvarado et al., 2011}).

The Fusarium fungus is recognized for its phytopathogenic ability, which is usually associated with vascular wilting (^{Forero-Reyes et al., 2018}). The use of genotypes resistant or tolerant to the pathogen has been one of the tactics to minimize the fungus infection, since its presence, along with favorable environmental conditions, reduces tomato yields (^{Sánchez-Aspeytia et al., 2015}). For this reason, different genotypes are used in order to fulfill the producer needs, including production increases, tolerance to environmental conditions (high temperature), pests and diseases (fungi, viruses, nematodes) (^{Rodríguez et al., 2004}). Therefore, the objective of this study was to identify the microorganism associated with wilting, and asses its incidence in nine commercial tomato genotypes in the presence of the microorganism’s natural inoculum in four production systems under shade mesh conditions.

The study was conducted in Tepecoacuilco, Guerrero, Mexico, in the greenhouse “Valerio Trujano” of the Centro de Estudios Profesionales (CEP) of the Colegio Superior Agropecuario del Estado de Guerrero (CSAEGRO) (18° 17’ 54.5” N, 99° 27’ 38.6” W). Nine commercial saladette tomato genotypes were selected from the companies Harris Moran (eight genotypes) and Nunhems (one genotype) to be established under the climatic prevailing conditions in the zone (temperature: 20-30 ºC and warm-humid with summer rains): six genotypes with indeterminate growth (Cid, Sun 7705, Moctezuma, Cuauhtémoc, Ramsés and Aníbal) and three genotypes with determined growth (Palomo, Pony express and Toro), using fertigation, and under a shade house on an area of 1000 m². All the genotypes were established in four production systems. Bags were filled with red tezontle substrate that had been used for chili crops (serrano and jalapeño peppers), where Fusarium associated with wilting, had been detected. The pots were placed on black plastic to prevent direct contact with the soil and weed growing among the pots. In the case of the production system in soil, grey-black padded plastic was used. Before transplanting, quaternary ammonium at a dose of 5 mL L^-1 of water up to the field capacity was applied to disinfect the soil and the tezontle substrate.

The seeds of the genotypes were sown in substrate (jal soil: peatmoss: naturabono) that was previously sterilized. Nutrition in this stage was based on the application of Rootex (3 g L^-1 of water), Gro-Green (2 g L^-1) and Ultrasol inicial (15-30-15). The experiment started by transplanting seedlings 30 days after sowing. Chemical and biological products were applied during the vegetative and production stages of the crop for pest and disease management. As for crop nutrition, applications were made according to the phenological stages, and complemented with foliar applications of micronutrients (Bayfolan Forte, Aminofit, Poliquel multi, Gro-Green, Biozime, among others). Irrigation was automated and started by adding 0.5 L of water per plant (vegetative stage) up to 3.0 L of water a day (fructification).

In all the production systems, agronomic practices were carried out, including tutorate, weed control (manual), bud pruning and plant health. On the other hand, the lateral buds of the genotypes with indeterminate growth were removed to keep only the main stem; for the genotypes with determined growth this practice was omitted. For the experiment, an incomplete randomized blocks design with a divided plot array was used in the largest plot where the production systems were established: 1) padded soil; 2) 37x37 cm white-black polyethylene bags; 3) 40x40 cm bags; and 4) 40x45 cm bags. The small plot was sown with the nine genotypes and three replications per production system. Each replication of the genotype (32 plants) was randomly distributed in each production system (large plot). Each system consisted of 864 plants with 96 plants per genotype, giving a total of 3,456 plants in the four systems. Each experiment unit consisted of two double-row furrows, 4 m long separated by a distance of 1.2 m between furrows and 0.5 m among plants. To assess the yield, each experiment unit was harvested by taking 12 plants from 32 plants per replication. The tomato fruits were weighed to obtain the yield per genotype and per production system. Data of the incidence and yield were subjected to a statistical analysis using the Statistical Analysis System (SAS) software, where an analysis of variance and Tukey’s tests were done.

The wilt incidence was detected in all the plants of each production system. In addition, plants from each system showing wilt symptoms were selected to identify the associated microorganism. The plants were washed with running water, then cross-wise cuts of approximately 1 cm were made to stems and roots, from which the areas of the disease progression were selected, rinsed with sterile distilled water, submerged in 2% sodium hypochlorite for 60 s and rinsed again with sterile distilled water (triple wash). The cuts were sown in Petri dishes containing potato-dextrose-agar (PDA) culture medium. Humidity chambers were also carried out in Petri dishes and incubated at environmental conditions. When the presence of mycelium was detected, tips of hyphae were transferred for purification and morphological identification (^{Leslie and Summerell, 2006}).

The initial symptoms were foliage yellowing, loss of turgor at the apex and general wilting; subsequently, the plants showed yellowing and necrosis on the intermediate and basal leaves (Figure 1A and B). Finally, the plants became brownish due to the attached fruits. At the level of the stem neck, a dark brown lesion developed (Figure 1C), which, as the days went by, expanded toward the upper part of the stem and, when a cross-wise cut was made, it could be seen that the xylem had a light brown color (^{Fernández-Herrera et al., 2013}). It should be noted that the symptoms were observed in the production stage of the crop. The Fusarium genus that produced cottony reddish mycelium with formation of slightly curve macronidia, 2-3 septa and abundant microconidia, was consistently isolated from the samples sown in PDA medium and in humidity chambers (20 plants) (Figure 1D). Rhizoctonia solani was also isolated though sporadically (two samples) and mixed with Fusarium. There are reports of Fusarium found in tomato genotypes under greenhouse conditions in the state of Michoacán, a fact that suggests that the pathogen is present in the production systems (^{Rodríguez-Alvarado et al., 2011}).

The incidence evaluation was carried out 114 days after transplanting. The genotypes with indeterminate growth showing low incidence to wilt were Cuauhtémoc (11.7%) and Ramsés (12.7%) and the ones with high incidence were Cid, Moctezuma and Sun 7705 (20, 22.4 and 29.7%, respectively) in the four systems that were evaluated (Figure 2). The genotypes with determined growth (Palomo, Pony express and Toro) had low incidence of the 0.3 and 1.8%, respectively (Figure 2). The Toro genotype is resistant to F. oxysporum (^{Rodríguez-Alvarado et al., 2011}), and this information is in agreement with our results that showed that not only Toro showed lower incidence but also the other genotypes with determined growth habit. Although in this study tolerance was not identified at the species level, some level of the genotypes’ tolerance can be associated with the microorganism.

Figure 1.  Symptoms and microorganism associated with tomato plants wilt under shade mesh in Tepecoacuilco, Guerrero. A) General yellowish and wilt under a padded soil system; B) Black-white polyethylene bags with tezontle; C) Dark brown lesion at the stem base; and D) Fusarium sp. conidia. 

It is important to mention that, although the seed companies (Harris Moran and Nunhems) report that the seed is tolerant to the fungus, this does not mean it is not susceptible to the pathogen’s attack, given that infection by and susceptibility to F. oxysporum race 3 has been found in genotypes Cid, Anibal and Sun 7705 in San Luis Potosí (^{Hernández-Martínez et al., 2014}). It should also be highlighted that selecting genotypes tolerant to F. oxysporum is not the only parameter that should be considered when selecting the genotype to be sown, but also other parameters, including yield, tolerance to other diseases or pests and climatic conditions in the area where the crop is to be sown (^{Monge-Pérez, 2014}), as well as the genetic variability that the pathogen may have in the soil, since F. oxysporum f. sp. lycopersici variability has been reported in high tomato production areas (^{Ascencio-Álvarez et al., 2008}).

In the system with higher wilt incidence, 37x37 cm polyethylene bags (26%) with 224 withered plants were used. In this system, Anibal, Sun 7705, Cid and Moctezuma genotypes had the highest percentage of withered plants ranging from 39.5 to 49% of incidence (Table 1). An explanation for the high incidence is the substrate, which had been used for other crops of the same botanical family (serrano and jalapeño peppers), where the microorganism and wilt symptoms were observed. The wilt incidence in 40x45 cm bags (40 withered plants) was of 4.7% (Figure 3). These results may be explained by the greater root development, the use of higher volume containers, and therefore, a higher capacity of the plants to tolerate the pathogen’s attack compared to that of the plants in smaller bags (37x37 cm). In this regard, ^{Arizaleta and Pire (2008}) point out that the containers size has to do not only with the root development and growth, but also with the aerial development of the plants, in the case of coffee trees. On the other hand, ^{Oagile et al. (2016}) also conclude that the size of large containers in the development of tomato seedlings affects the plant’s height, number of leaves, and fresh and dry weight. These data support the results obtained in this study, since we observed that the genotypes sown in larger bags (40x45 cm) produced taller plants, had higher root fresh and dried weight (data not published), as well as lower wilt incidence. This system (40x45 cm) also produced the greatest number of withered plants of the Sun 7705 genotype.

While the highest level of incidence was observed in the production system where 37x37 bags were used, the estimated potential yield was of 74.3 t ha^-1, and in the production system where 40x45 cm bags were used (lower wilt incidence), the estimated potential yield was of 72.3 t ha^-1 (Table 1). These results show that regardless of the wilt incidence, the genotypes produced good yields under the conditions in the studied area. Also, wilt was observed in the phenological stage of the crop, so it did not significantly affect the yield.

Wilt symptoms in tomato crops were associated with the Fusarium genus. The genotypes with indeterminate growth Cuauhtémoc and Ramsés showed low wilt incidence caused by the microorganism in all the production systems, and in the case of the genotypes with determined growth, all of them had low incidence (0.2-1.8%). The Sun 7705 genotype had the highest wilt incidence and the lowest yield with a range of 59.7 - 70.6 t ha^-1.

Figure 2. Wilt incidence (Fusarium sp.) in a tomato crop with determined and undetermined growth established in four production systems under shade mesh in Tepecoacuilco, Guerrero. Cycle: Winter-spring 2017-2018. Equal letters indicate that there are no significant differences, according to Tukey’s test (P<0.5). 

Finally, in the production system evaluated, an effect on genotype performance was observed, as well as on the wilt incidence, since only a small number of withered plants were observed in the 40x45 cm containers. In this study, Fusarium natural infection was observed in tomato. However, it is important to conduct further studies in order to identify the microorganism using molecular techniques and evaluate the pathogenicity in all genotypes to determine their tolerance.

Figure 3. Percentage of wilt incidence (Fusarium spp.) in tomato plants per each production system under shade mesh in Tepecoacuilco, Guerrero. Cycle: Winter-spring 2017-2018. Equal letters indicate that there are no significant differences, according to Tukey’s test (P<0.5).