Onion (Allium cepa) is considered one of the basic ingredients in human diets and as a crop around the world (^{Joaheer et al., 2019}). In Mexico, it is the third most important vegetable, with a production of 85,104 t for the year 2018, planted in approximately 7,329 ha, leaving a revenue of de 26,029,376 dollars (^{FAO, 2020}). Baja California is the state that produces the most onion, with a surface of 3,443 ha, and the state of Puebla is fifth, with 2,480 ha and a yield of 21,371 t ha^-1 (^{SIAP, 2020}).

The most frequent diseases caused by fungi in this vegetable can affect foliage, roots and fruits during the stages of harvest and postharvest (^{Ji et al., 2018}). Some of the diseases with the highest incidence on the Allium species are caused by Rhizoctonia solani, Brotrytis allii, Aspergillus niger, Colletotrichum circinans and Fusarium spp. (^{Montes-Belmont et al., 2003}), the latter being the main devastator of this crop in recent years (^{Abdalla et al., 2019}). Basal rot caused by the genus Fusarium spp. is widely distributed around the globe and has become a limitation in onion and garlic producing areas (^{Kiehr y Delhey, 2015}). The main species within the genus Fusarium that harms the onion crop around the world are F. proliferatum, F. solani and F. oxysporum, reducing its yield by up to 50% (^{Haapalainen et al., 2016}), and it produces symptoms in the onion plant that include wilting, rotting of the roots and the basal lamina of the bulb (^{Sanogo and Zhang, 2015}). In Mexico, the genus Fusarium is responsible for diverse diseases in onion plantations (^{Montes-Belmont et al., 2003}). Hence its importance in this investigation.

Chemically synthesized products have been used for decades in the control of these diseases (Gan and ^{Wickings, 2017}). However, their use is related to the generation of resistance, environmental damage and the deterioration of human health (^{Andrade-Hoyos et al., 2019}). Therefore, biological control is considered an efficient and environmentally feasible practice for the development of sustainable agriculture (^{Pérez-Torres et al., 2018}).

The genus Trichoderma contains antagonistic species capable of controlling a wide number of fungi that affect plants of agricultural interest (^{Romero-Arenas et al., 2017}). Its success and use in agriculture are due to its action mechanisms such as the competition for space, mycoparasitism, antibiosis (^{Nawrocka et al., 2018}), and the production of volatile compounds (^{Hernández-Melchor et al., 2019}). Consequently, the aims of this investigation were to morphologically characterize fungal isolations of the genus Fusarium relates to onion bulbs in the rural town of La Soledad, Chietla, Puebla-Mexico and to determine the in vitro antagonistic ability of T. harzianum T-H4 on fungal isolations of the onion crop.

Isolation area. Plant tissues with rotting of the stem (basal lamina) or bulb and scarce radicle development were collected from a 3,144.3 m² “Crystal white” onion field with a background of high fungal incidence during the summer-autumn 2019 production season (18° 27’ 39.3258” N and -98° 37’ 11.2614” W). The field belongs to the town of La Soledad, in the municipal area of Chietla in the state of Puebla, with a warm desert weather (Bwh) and a mean rainfall of 700 mm (^{García, 2004}). Sampling was aimed at individuals with symptoms related to the genus Fusarium; all samples were kept in an ice box until they were transferred to the laboratory.

Isolation of Fusarium sp. from plant tissues. Thirty bulbs were selected to be disinfested with tap water and sodium hypochlorite at 10%. Pieces of bulb were placed in wet chambers, using polyethylene boxes and moist filter paper; the polyethylene boxes were incubated at 28 °C for 10 days. After the incubation time, the dead bulb tissue was cut into 0.5 cm² sections, disinfested with sodium hypochlorite at 1.5% and washed three times with distilled water. They were dried using sterilized paper and finally planted in Petri dishes with Potato Dextrose Agar (PDA) at a temperature of 28 °C with ambient light for three days. The cultures developed were isolated and purified using monospore cultures, following indications by ^{Morales et al. (2020)}.

Morphological characterization. The fungal cultures obtained were identified with an analysis of morphological characteristics related to the genus Fusarium with taxonomic identification keys by ^{Barnett and Hunter (2006}), in a microcultivation system under an optical microscope (Carl Zeiss, Jena, Germany) at a magnification of 1000x (^{Samson et al., 2014}). To evaluate the mycelial development, pieces of the promissory isolations were inoculated in Petri dishes with PDA, incubated in the dark at 28 °C for 10 days and the mycelial diameter was measured every 12 h to estimate the growth speed (cm), which was measured with the linear growth function y=mx + b (where ‘y’ is the distance, ‘x’ is time and ‘b’ is the constant factor) and expressed in centimeters per day (cm d^-1) (^{Zeravakis et al., 2001}). The diameter was measured using a digital caliper (CD-6 Mitutoyo), always in the same direction —in triplicate—, which was established at random for each repetition. Only the average was used to calculate the speed of mycelial growth per day.

Molecular identification. DNA was extracted from conidia, conidiophores and mycelia from the isolations obtained with morphological characteristics related to Fusarium, using the CTAB 2% method (^{Rivera-Jiménez et al., 2018}). The DNA was resuspended in 100 mL of sterilized HPLC water and quantified using spectrophotometry (Nanodrop 2 000 C, Thermo Scientific). Next, it was diluted to 20 ng mL^-1 and stored at -20 °C. For the amplification, primers ITS1 and ITS4 (^{White et al., 1990}) were used, along with 15 μL of the reaction mixture containing 0.18 μL of each primer, 0.18 μL of dNTP, 0.9 μL of GoTaqVR (Promega) polymerase DNA, 3 μL of DNA suspension and 10.56 μL of distilled water. Amplification was carried out in a Peltier (PTC-200, Bio-Rad) DNA thermocycler at 95 °C for 1 min; 35 cycles at 95 °C for 30 s; 55 °C for 1 min; an extension of 72 °C for 45 s; and finally, 72 °C for 5 min (^{Salazar-González et al., 2016}). The amplified PCR products were verified by electrophoresis in agarose gel at 1.5% (Seakem) and were purified with ExoSAP-IT (Affymetrix). Both chains were sequenced in the Colegio de Posgraduados with BigDye Terminator v3.1 (Applied Biosystems) in a 3 130 ×L genetic analyzer sequencer (Applied Biosystems, Foster City, CA) and the software Bioedit 7.0.4.1, following instructions by the manufacturer. The sequence obtained was deposited in the Gen Bank’s National Center for Biological Information (NCBI) database and compared with five accessions (Table 1) available in the NCBI database to confirm its identity (^{Nei and Kumar, 2000}).

In vitro pathogenicity tests. Koch’s postulates were implemented to confirm the pathogenicity of Fusarium sp. Certified onion seeds of the variety Crystal white with a percentage of germination of 90% were used after disinfecting. Six seeds were placed in each Petri dish and after germinating, discs, 5 mm in diameter with Fusarium sp. mycelia were placed in the center of the Petri dish, in triplicate with two repetitions; a disc, 5 mm in diameter with sterile agar-water was used as a control. The seeds were incubated inside a controlled-temperature (22 ± 2 ^oC) environmental chamber in complete darkness until 90% of the control group germinated. After this moment, a radicle length of over 5 mm was used as a germination criterion; the growth or elongation of the radicles were recorded in both groups (^{OECD, 1984}; ^{USEPA, 1996}). The data recorded underwent the Shapiro and Wilk test of normality (p<0.05) and were compared based on Student’s t test for paired samples (p<0.05) (^{Iturbide-Zuñiga et al., 2017}). Finally, small sections (approx. 1 × 1 cm) were taken from radicles with disease symptoms, sterilized on the surface with sodium hypochlorite at 1% for 10-30 s and rinsed three times with sterile distilled water for reisolation; the segments were then placed in a PDA medium and incubated in the dark at 28 °C for 10 days. Next, the isolation was observed under the optical microscope to confirm that it was strain CFbC using taxonomic identification keys (^{Barnett and Hunter, 2006}).

Evaluation of the antagonism of T. harzianum with Fusarium sp. in dual culture. The evaluation of the antagonism was carried out using strain T-H4 of T. harzianum, isolated from the root of Persea americana, the sequence of which was included in the data base of the National Center for Biological Information (NCBI) with accession number MK779064.1, which is in the laboratory of the Eco-Campus Valsequillo, of the Institute of Science, Benemérita Universidad Autónoma de Puebla (BUAP) (^{Andrade-Hoyos et al., 2020}). The dual culture technique was used, following ^{Andrade-Hoyos et al. (2019)} -in triplicate- to determine the percentage of radial growth inhibition with the formula PICR= [(R1-R2/R1) x 100] for each trial evaluated for a 10-day period. The data recorded underwent the Shapiro Wilk test (p<0.05) and compared based on Student’s t test for paired samples (p<0.05) in the IBM SPSS Statistics statistical package, version 25 (^{Iturbide-Zuñiga et al., 2017}). The experiment was performed twice for its validation. In order to complement the evidence of antagonism, each trial was compared and classified with the scale established by ^{Bell et al. (1982}): I) the growth of T. harzianum covered the entire surface of the medium and reduced the Fusarium sp. culture, II) the growth of T. harzianum covered at least two-thirds of the medium, III) T. harzianum and Fusarium sp. grew over 1/2 and 1/2 of the medium’s surface, without superposing each other, IV) Fusarium sp. grew over at least 2/3 of the medium and resisted the invasion of T. harzianum and V) the growth of Fusarium sp. covered the entire surface of the medium.

Identification. Using the proposed isolation strategy, only one culture related to the genus Fusarium was obtained. Strain CFbC displayed a radial growth with a speed of growth of 0.5294 ± 0.0833 cm d^-1, a range which agrees with the study by ^{Groenewald et al. (2006}). The culture of strain CFbC was abundant aerial mycelia, initially white, but after maturing they turned slightly cinnamon-brown, and on the reverse of the dish, a slightly orange color can be noticed (Figure 1a and b). Strain CFbC presented septated hyphae, hyalines, septated macroconidia with five septa, with a distinct curvature, a length of approximately 60 -120 µm (Figure 1c), with a thick wall, slightly arched in the ventral part, and abruptly arched in the dorsal part, with a foot-shaped basal cell and a filamentous apical section. The morphological characteristics mentioned above coincided with those described by ^{Barnett and Hunter (2006}), and ^{Leslie and Summerell (2006}) for Fusarium sp.

The amplification of the ribosomal ITS region gave a product of 533 pb of strain CFbC, which displayed an identity of 100% with the Fusarium incarnatum-equiseti species complex (Table 1) according to the Gen Bank data base. This sequence was included in the NCBI data base with accession number MN612793.1.

Pathogenicity tests. Eight days after the in vitro inoculation with strain CFbC, all plants began showing symptoms, with a percentage of germination of 95%. The pathogenicity test results are shown in Figure 2, where a radicle growth below 5 mm (Figure 2a) is presented, along with apical strangulation, necrosis in the radicle (Figure 2b), and colonization in seedlings at the seed germination level in dishes with agar-water (Figure 2c), results which are positive for the inoculation of Fusarium sp. In seedlings used as controls, no symptoms were observed (Figure 2d), which proves Koch’s postulates.

In the pathogenicity test, radicle growth decreased by 42.02%, displaying significant differences. Koch’s postulates were confirmed with the reisolation of Fusarium sp. from rhizomes of seedlings germinated 10 days after inoculation. This pathogen has been reported by ^{Dauda et al. (2018}), who first reported the regressive death of onion plants caused by Fusarium equiseti in Nigeria.

Figure 1 a) Culture of Fusarium sp. in a PDA medium, b) Reverse of culture in PDA, orange coloring, c) Short monophyllids and microconidia grouped into false heads, d) Septated macroconidia with five septa, e) Microconidia, f and g) Macroconidia with distinctive cuvatures, 40 X. 

Evaluation of the antagonism of T. harzianum on Fusarium sp., in a dual culture. Areas of interaction were observed between T. harzianum (T-H4) and Fusarium sp. (CFbC), in the form of competition for space and nutrients. The reduction in the growth rate in dual cultures is an indicator of the antagonistic capacity of Trichoderma (^{Guigón-López et al., 2010}). In this study, the growth of Fusarium sp. in confrontation of strain T-H4 was clearly inhibited as of five day (Figure 3a). When evaluating the PICR, significant differences (p≤0,0001) were found in Student’s t test, with a value of t=92.57. The PICR was 76.24% (Figure 4b), a value which falls into class II (Figure 4a) according to the scale established by ^{Bell et al. (1982}). ^{Michel-Aceves et al. (2005)} report an inhibition of 73% in radicle growth when evaluating the antagonistic effect of native isolations of Trichoderma spp. on the growth of F. subglutinans, similar to the results of this study. On the other hand, ^{Jagraj et al. (2018}) reported 75.9% for F. oxysporum against T. harzianum, higher than T. viride and T. koningii, with 67.7 and 55.6% respectively for strains isolated from tomato plants.

^{Reyes et al. (2008}) noted that one of the significant characteristics of Trichoderma is its high growth rate, and the secretion of secondary metabolites of a different nature, which slow down the growth rates of other competitors. In addition, mycoparasitism is another action mechanism of the genus Trichoderma. In this case, swellings were observed in macroconidia and hyphae, as well as the breaking of septum, the presence of granulations and the vacuolization of hyphae by T. harzianum (Figure 4b), reaffirming the high mycoparasitic capacity of strain T-H4. ^{Duarte-Leal et al. (2018}) observed mycoparasitism between the hyphae of T. asperellum (Ta.25), where it penetrated and caused the lysis of the hyphae of Fusarium oxysporum f. sp. ciceri (F-50), as reported in the present investigation. The results confirmed the notifications by ^{Gonzálezet al.(2012}) and ^{Hyderet al. (2017}) regarding the enzymatic degradation of the cell wall of the phytopathogenic fungi during the mycoparasitic action, which causes lysis and thus the disorganization of the cytoplasmic content which, according to these authors, is related to the action of the enzymes chitinase, β-1,3-glucanases, β-1,6-glucanases, α-1,3-glucanases and proteases excreted by Trichodermaspp. Additionally, ^{Leónet al.(2012}) observed strangulation and the lysis of the walls of the hyphae, followed by a disintegration and degradation of the cell walls of Fusariumsp. when confronting strains Tb111 andTc241 ofTrichodermaspp.

Figure 2 Germination of seeds inoculated with Fusarium sp. a) Minimal root growth and necrosis, b) Strangulation and necrosis, c) Colonization in the apex, d) Control without symptoms. 

The hyphal interactions that intervene in the parasitism are measured by the enzymatic activity of the antagonist, which allows the penetration, deformation, disintegration and death of the hyphae of the pathogen (^{Leónet al., 2012}), and therefore the strains with a high enzymatic potential tend to have a greater mycoparasitic effect (^{Gonzálezet al., 2012}); herein lies the importance of the selection of strains with various types of hyphal interactions.

Figure 3 Comparison of mycelial growth and PICR. a) Growth of Fusarium sp. in the presence and absence of T. harzianum b) Percentage of inhibition of radial growth (PICR) of Fusarium sp. 10 days after inoculation in the presence and absence of T. harzianum. *Means with different letters indicate significant differences with the Student’s t test (p≤0.05). 

Figure 4 a) Antagonism of strain T-H4 of T. harzianum against Fusarium sp., showing type II in the scale by^{Bell et al. (1982}), b) Mycoparasitism of T. harzianum on the hyphae and conidia of Fusarium sp., * swelling of the hypha, **curling of the hypha, *** granulations inside the hypha, **** swelling of the macroconidia. 

Antagonism tests reflect the capacity and genetic variability of the antagonist and the phytopathogen to resist antagonism and perform a preliminary selection of the most efficient antagonistic strains to be evaluated under field conditions, as well as to complement and determine its ability exercised for the biocontrol (^{Morales et al., 2020}).

Fusarium sp. was identified as being related to basal rot in bulbs from onions (Allium cepa) of the Crystal white variety, found in La Soledad, which belongs to the municipal area of Chietla, in the state of Puebla, Mexico. The T. harzianum strain T-H4 displayed an adequate antagonistic ability in vitro against Fusarium sp., with a PICR of 76.24%. It is necessary to verify the antagonistic capacity of T. harzianum T-H4 in the field against strain CFbC as well as its identification with the elongation factor of translation (EF1α) in order to recommend its use as an alternative for the onion farmers of the area.