Aloe vera is a crop of great economic importance and relevance to agriculture and industries around the world, and Mexico stands out as the main producing and exporting country of aloe raw material (^{Álvarez et al., 2012}). Tamaulipas is considered the state with the highest production rate in the country and a production surface area on the rise, although the presence of phytopathogenic microorganisms has been found (^{Rubio et al., 2020}). One of the limitations that increase losses and affect yields is the attack of Fusarium spp., Penicillium spp. and Pectobacterium spp. (^{Jiménez, 2015}), one of which, F. oxysporum, is considered as one of the phytopathogenic fungi in aloe that causes rotting of the stem base, yellowing, wilting of the leaves and therefore, the death of the plant. On the other hand, the genus Erwinia is reported in several areas of the world as the causal agent of bacterial rot in roots, stem and leaves of Aloe spp. (^{Rubio et al., 2020}).

Diseases caused by phytopathogens are controlled using chemical products and the most common active ingredients against Rhizoctonia solani, Fusarium solani and F. oxysporum are Carbendazim, Mancozeb, Fosetyl-Al, Hymexazol, Chinosol and Metalaxyl-M + Mancozeb (^{Mannai et al., 2018}). However, the use of agrochemicals has caused several controversies due to their high costs, toxicity levels and environmental damages (^{Nicolopoulou-Stamati et al., 2016}). Due to this, searches for biological control alternatives have been carried out, since they produce no residual effects or contamination (^{Pappas et al., 2020}). There is an important group of microorganisms that display antagonistic effects on other phytopathogenic microorganisms (^{Sánchez-León and Bustos, 2020}), including Bacillus subtilis and Trichoderma spp., which characteristically display antagonistic effects (^{Fiorentino et al., 2016}). For example, Trichoderma harzianum and T. koningiopsis have been used for their action mechanisms against Fusarium solani (^{Miguel-Ferrer et al., 2021}). It also has an antagonistic effect against pectolytic phytopathogenic bacteria, since it is able to inhibit them by 32.4% (^{Astorga-Quirós et al., 2014}). This may be due to its enzymatic properties, since it secretes proteases, chitinases and glucanase, which damage the cell wall of the phytopathogenic fungus, causing a lysis that facilitates mycoparasitism (^{Ribeiro et al., 2019}). On the other hand, Bacillus spp. can induce growth in plants, even when their root systems have been affected by Fusarium spp. (^{Castro et al., 2019}). Antagonism has also been reported in the confrontation of B. subtilis with Fusarium spp. (^{Solano-Báez et al., 2021}), since it produces antibiotics and enzymes and also helps in the solubilization of phosphates and the biological fixation of nitrogen (^{Miljaković et al., 2020}). Due to this, the aim of this investigation was to determine the antagonistic effect of native strains of Bacillus spp., T. asperellum and T. harzianum against F. oxysporum and Dickeya chrysanthemi under in vitro conditions.

The tests were carried out in the microbiology lab of the Central Integral de Laboratorios (CILO), in the Faculty of Engineering and Sciences of the Autonomous University of Tamaulipas in August, 2019. Native, previously identified microorganisms were used, which were a part of the ceparium of the microbiology laboratory: T. asperellum, T. harzianum, F. oxysporum and D. chrysanthemi, which were isolated from plots with Aloe barbadensis plants.

In order to obtain Bacillus isolations, samples were collected from the soil using the five of gold method, and five samples were taken from 1 kg of soil from two Aloe vera plantations in Padilla, Tamaulipas (23° 59’ 26.1” N, 98° 56’ 39.0” W). From each sample, 10 g of soil were taken and resuspended in 90 mL of sterile distilled water; serial dilutions were performed until 1×10^-3 and it underwent a thermal treatment at 85 °C for 15 min. Next, 100 µL were extracted and inoculated in a potato-dextrose-agar (PDA) medium, incubated at 27± 1 °C for 24 to 48 h, after which bacterial cultures of varied morphologies were obtained, which were chosen for their inhibition characteristics.

Each species was described macroscopically according the morphology of the cultures: shape, edges, elevation, surface, size, color, consistency, and a photograph was also taken; in addition, its cell morphology was determined by Gram’s staining, with the type of cell walls, after 24 and 48 h of incubation, in order to observe the microscopic morphology at 100X (^{Calvo and Zúñiga, 2010}). Once the presence of bacterial endospores was verified, a sample was taken and placed in nutritious agar (AN) for 24 h to 27± 1 °C. Tests were carried out for maltose, xylose, mannitol, citrate use, starch hydrolysis de starch, gelatin hydrolysis, growth in NaCl at 7% and growth in anaerobic agar in order to identify B. subtilis cultures following the methodology proposed by ^{Reinoso et al. (2006}). To determine the bacterial species following the methodology by ^{Schaad et al. (2001}), the bacteria were tested to see if they were able to hydrolyze gelatin, starch, as well as using sugars, citrate and alcohols.

The antagonistic ability of Bacillus spp. (B4 and B5) against F. oxysporum was evaluated with two treatments (B4 and B5) with five repetitions each at two incubation temperatures. In order to do this, the methodology by ^{Corrales et al. (2011}) was used, with modifications, which consisted in planting F. oxysporum in the center of the Petri dishes with PDA medium and AN. For the case of Bacillus spp., it was places in the four cardinal points. In the case of the control, only F. oxysporum was inoculated. Afterwards, they were incubated at 25± 1 °C and 27± 1 °C. Next, the mycelial growth of F. oxysporum was measured every 24 h until it filled the Petri dish and using the data obtained, the percentage of root growth inhibition (PICR) was calculated, and the formula used by ^{Ávila et al. (2020}) was used. Based on the results, the scale proposed by ^{Corrales et al. (2011)} was established. In addition, the number of spores was determined by taking a piece of PDA medium with F. oxysporum, which was then placed in a test tube with 1 mL of sterile water. It was shaken for 3 min, then he spores were counted in a Neubauer chamber; the results underwent a comparison of means.

For T. asperellum and T. harzianum, 5 mm of active mycelium was placed 4 cm away from F. oxysporum. For the controls, F. oxysporum was planted in the center of the Petri dishes, incubated at 25± 1 °C for 10 days and measurements were taken of the radial growth of the fungi every 24 h. The antagonism of T. asperellum and T. harzianum was evaluated with the percentage of inhibition, with the formula used by ^{Ávila et al. (2020}). In addition, the level of antagonism was determined according to the scale used by ^{Bell et al. (1982}).

To evaluate the antagonistic ability of Bacillus spp. against D. chrysanthemi, two treatments, B4 and B5, were established. The methodology consisted in using a 5 mm hole puncher to punch holes into the Petri dishes with AN medium. On the side, sterile AN was mixed with bacteria from the treatments (B4 and B5) at a concentration of 1×10⁶ UFC and the holes in the four cardinal points were filled. Next, the dishes were incubated at 25± 1 °C for 24 h, a suspension of D. chrysanthemi was prepared at a concentration of 1×10⁶ UFC in a spray bottle and sprayed onto each treatment. The variable to evaluate was the inhibition halo and growth was measured every 24 h for one week. On the other hand, to measure the antagonistic ability of T. harzianum and T. asperellum (TJ and TP) against D. chrysanthemi, the methodology by ^{Corrales et al. (2011}) was used, with modifications, which consisted in placing the D. chrysanthemi isolations in the center of the Petri dishes and the Trichoderma spp. isolations in the four cardinal points. For the control, D. chrysanthemi was only inoculated in the center of the Petri dish and they were all incubated at 25± 1 °C. Measurements of the radial growth of the D. chrysanthemi cultures were taken every 24 h until the Petri dish was full. Later, the formula used by ^{Ávila et al. (2020}) was used to calculate the percentage of inhibition and it was located on the scale by ^{Bell et al. (1982}).

The statistical analysis of the antagonistic confrontations between Bacillus spp. and F. oxysporum; T. asperellum and between T. harzianum and D. chrysanthemi and F. oxysporum were submitted to a covariance analysis test (ANCOVA). In the case of Bacillus spp. against D. chrysanthemi, an analysis of variance was used (ANOVA), as was the case for counting spores from the confrontations between Bacillus spp. and F. oxysporum with a temperature of 27± 1 °C. In order to find differences between treatments, Tukey’s comparison of means (P<0.05) was applied using the SAS statistical package SAS (version 9.0).

Twelve bacterial cultures with a variety of morphologies were isolated and two bacteria were selected (B4 and B5) for their inhibition characteristics, which displayed macroscopic characteristics of the Bacillus genus, producing whitish, large, extended and irregular cultures. They were Gram positive, with a bacillar shape, with motility and the position of the spore is central. The temperature at which they developed was 45 °C with an adequate growth, yet no growth was observed at 65 °C; they displayed an adequate growth rate at a pH of 5.7 and they developed in 7% NaCl. In the oxidative and/or fermentative glucose metabolism test, they displayed no aerobic growth, yet they did present glucose degradation. The tests for maltose, mannitol, arabinose, xylose, the use of citrate, starch hydrolysis, gelatin hydrolysis and oxidase were positive, indicating that the bacterium Bacillus spp. belongs to the species subtilis.

The results of both B. subtilis strains (B4 and B5) against F. oxysporum indicated no significant differences between isolations and both reduced the growth radii (RCA) of F. oxysporum. Nevertheless, in the incubation temperature of 25± 1 °C, B4 presented 49.6% inhibition and it extended around the fungus, displaying competition for space and nutrients. Regarding the scale used, it presented itself in the middle classification. On the other hand, B5 obtained a percentage of antagonism of 39.7% with a low classification (Table 1) and its growth was slow. However, F. oxysporum displayed a less populated mycelium and a less intense color than the control.

In the incubation at 27± 1 °C (Table 1), B5 proved to have an increase in the degree of inhibition with 43.0% and a middle classification, according to the scale proposed by ^{Corrales et al. (2011}) because the temperature had an influence on the production of metabolites, since a murky liquid was found, and with a bad odor. On the other hand, B4 reduced its degree of inhibition to 38.7% with a low classification and it produces metabolites that helped reduce the growth of the mycelia. In the latter experiment, B4 and B5 presented a reduction in the production of cottonlike mycelia in the top frontal part (Figure 1). These results were lower than those obtained by ^{Corrales et al. (2011)}, who reported percentages of 40 to 70%. This variability can be explained by each strain having a different adaptability, and perhaps the synthesis of secondary metabolites depends on the incubation temperature so that, depending on the temperature, some metabolites may be produced with a greater or lower effect on that pathogen. In such results, Bacillus spp. behaves differently, which may be due to what was described by ^{Sánchez et al. (2015}), who mentioned that there is a wide phenotypic, biochemical, serological and genetic diversity, even amongst the members of the same species.

According to the conidial concertation results, a significant difference was observed between the treatments and the control, since the control displayed an average of 12.6 spores per dilution, unlike the B5 treatments which obtained 1.3 spores, and B4, 2.6 spores, which proves that the treatments caused a reduction in the concentration of spores and the color of F. oxysporum was reduced and paler. The effect of the antagonism of B. subtilis could be due to its large production of bioactive compounds, among which cyclic lipopeptides (CLP) such as surfactin, iturin and fengycin stand out (^{Cochrane and Vederas, 2016}). Additionally, Bacillus spp. produces antibiotics and enzymes (^{Miljaković et al., 2020}).

Figure 1 Antagonistic effect of B. subtilis strains (B4 and B5) against F. oxysporum (FLL); A) Isolation B4 at a temperature of 25± 1 °C, it surrounds F. oxysporum displaying competition for space and nutrients; B) Control of F. oxysporum; C) Isolation B5 at a temperature of 27± 1 °C surrounding a F. oxysporum, displaying a production of metabolites. 

In the analysis of covariance of de T. asperellum (TJ) and T. harzianum (TP) against F. oxysporum, there were no significant differences. The confrontation with TJ displayed a growth of 6.9 cm and an inhibition of 70.5% (Table 2), although TP displayed a growth of 6.6 cm and an inhibition of 69.2% and both treatments displayed competition for space and nutrients. Furthermore, the growth radii of F. oxysporum decreased as the growth radii for T. asperellum and T. harzianum grew, which suggests an antagonism. Additionally, both isolations covered half of the surface of the Petri dish on day three of the evaluation, due to its rapid growth, competition for space and nutrients. Likewise, they are classified as class two where Trichodermaspp. grows two thirds of the surface of the medium or, in other words, this fungus covers 65% of the Petri dish (Figure 2). These results were better than those reported by ^{Rodríguez-García and Wang-Wong (2020}), who confronted native T. asperellum strains with F. oxysporum and reported a level of inhibition of up to 67%, which was their best treatment, since native strains proved to be the most aggressive against the phytopathogen. The results may be due to what was reported by ^{Andrade-Hoyos et al. (2019}), who mentioned that Trichoderma sp. has the ability of performing mycoparasitism, which, in turn, may be due to the production of lytic enzymes, proteases and the ability to produce lysis in the hyphae of the phytopathogenic fungus, penetrating it and feeding off it, since the lethicins it produces degrade its cell wall. It also produces chitinase and the -1, 3-glucanase activities (^{Li et al., 2018}), degrading the fungal cell wall, which limits the growth of the pathogen (^{Coban, 2020}). The previously mentioned biocontrol mechanisms helped the antagonist make better use of the nutrients in the medium and deprive the pathogen from using these resources (^{García-Espejo et al., 2016}). Trichoderma sp. can control a wide range of phytopathogenic fungi such as R. solani, F. solani, Sclerotinia sclerotiorum, Fusarium (^{Martínez-Martínez et al., 2020}).

Figure 2 A) Trichoderma harzianum (1) surrounding F. oxysporum (2) expressing competition for space and nutrients in a PDA medium; (B) T. asperellum (1) surrounding F. oxysporum (2) expressing competition for space. 

In the analysis of variance of the evaluation of B. subtilis against D. chrysanthemi no significant differences were found. B4 was found to have developed a halo around it with a width of 4.4 mm in width, and B5, a 4.3 mm wide halo, both caused by the metabolites produced by the bacteria to inhibit D. chrysanthemi, displaying antagonism. In addition, B. subtilis began colonizing and expanding, in a display of competition for space and nutrients, overgrowing D. chrysanthemi (Figure 3). These results are opposite to those by ^{Sneha and Anuradha (2017}), who evaluated B. subtilis against Pectobacterium carotovorum, and found no inhibition halos. However, the results obtained in this experiment were quite lower than those reported by ^{Gerayeli et al. (2018}). In their evaluation of Bacillus spp. against P. carotovorum, they reported halos of 10 mm in width. In those results, Bacillus spp. behaves differently, probably due to a report by ^{Sánchez et al. (2015}), who mentioned that in the members of a same species there may be a wide phenotypic, biochemical, serological and genetic diversity. On the other hand, ^{Sneha and Anuradha (2017)} proved that Bacillus spp. has an antagonistic effect on pectolytic bacteria such as P. carotovorum.

Figure 3 A) Halo of inhibition from the growth of D. chrysanthemi as an effect of isolation B4; B) Isolation B4 overgrows bacteria D. chrysanthemi, expressing competition for space and nutrients. 

Regarding the analysis of covariance performed on Trichoderma spp. against D. chrysanthemi, no significant differences were observed between treatments. TJ displayed a percentage of 41.9 and TP, 40.2% and both surrounded D. chrysanthemi, which showed signs of competition for space and nutrients. Initially, the bacterium has a creamy appearance, yet with time its appearance became dry, its growth stopped and there was even a slight growth of mycelia until the day of the evaluation. According to the scale by ^{Corrales et al. (2011}), both treatments have a middle antagonism classification. These results were higher than those reported by ^{Astorga-Quirós et al. (2014}), who obtained a 32.3% antagonism when evaluating Trichoderma spp. against Pseudomonas marginalis, a pectolytic bacterium that causes soft rot.

The native T. asperellum and T. harzianum strains resulted antagonistic, with inhibitions of 39.7 to 49.6% against F. oxysporum, and in the confrontation with D. chrysanthemi, inhibition rates of 40.2 to 41.9% were found. B. subtilis reached a middle level of antagonism against F. oxysporum, despite being bacteria confronted against a fungus, and it produced an inhibition halo of 4.4 mm against D. chrysanthemi.