Isolates’ origin

Isolate 303CS of F. oxysporum f. sp. ciceris, race 5 -GenBank access number KJ000584- used on this study was provided by the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pesqueras. As biological control agents, the following strains -taken from the Centro de Ciencias de Sinaloa (CCS) collection- were used: bacteria of the genus Bacillus sp., T442, T3141 strains; bacteria of the genus Pseudomonas sp., T162, 7A1, and 751 strains; fungi of the genus Trichoderma sp., HRG-050, HRG-060, and HRG-083 strains. These strains were previously isolated in chickpea crops, infected with the disease (Table 1).

Table 1 Geographic location of the places of origin of biological control microorganisms, using in vitro and greenhouse tests. 

Isolate activation

Bacteria were cultured using an inoculated loop sampling method in a Petri dish. Pseudomonas strains were cultured in a medium based on Agar FLO (20 g peptones mix, 14 g agar-agar, 15 g K₂HPO₄, 15 g MgSO₄, and 1.0 L distilled H₂O); Bacillus was cultured in a nourishing agar medium (5 g peptone gelatine, 3 g meat extract, 15 g agar-agar, and 1.0 L distilled H₂O); and, fungal strains and Foc race 5 were grown using a portion inclusion method (^{Leslie and Summerell, 2006}). To carry out the experiments, 1 cm diameter dishes of each strain were placed on the center of the Petri plates, with a PDA culture (4.0 g potato infusion, 20 g dextrose, 15 g agar, and 5.6±0.2 final pH). The plates were incubated at 28 °C, for 24 h and 5 d, for bacteria and fungi, respectively.

In vitro antagonism tests against Foc, race 5

Strain antagonistic activity was studied using the dual culture technique (^{Ezziyyani et al., 2004}), in Petri plates with a PDA culture medium. Confrontations between bacteria strains versus Foc race 5 were arranged as follows: first, a 1 cm wide phytopathogen disk grown in PDA was placed on one side of the plate; 3 d later, a bacterial strain sample was put in the center of the plate, using an inoculating loop. The portion method was used to culture antagonistic fungal strains (^{Howell, 2003}): a 1 cm wide disk containing Foc race 5 was placed on one side, and, 5 d later, an antagonist disk of the same size was placed at the other end of the plate. The inoculated plates were incubated 7 d at 28 °C. The strains’ antagonistic capacity was evaluated using the percentage inhibition of radial growth (PICR), using the following formula:

PICR=[(R1-R2)/100] (Suárez et al., 2008)

where

PICR is the inhibition percentage of phytopathogen mycelium growth.

R1 is the highest radial growth (pathogen-control mycelium radius).

R2 is the lowest ratio (pathogen mycelium radius on dual culture).

The mycoparasitism degree exercised by Trichoderma strains was determined based on the surface of the Petri dish that it colonized when the phytopathogen was present, according to the scale of ^{Bell et al. (1982)} (Table 2).

Table 2 Scale to determine the antagonism level proposed by ^{Bell et al. (1982)}. 

The confrontations were observed using an optical microscope, in order to determine the antagonistic mechanism used by Trichoderma on the pathogen fungi. Additionally, the total amount of haustoria generated by Trichoderma strains during their confrontations with Foc were quantified 5 d after they were cultured. For this purpose, five 1-cm² samples were randomly taken from the Petri dish where the confrontation between the antagonist and Foc had taken place, and they were observed under the microscope.

The experimental design was completely random, with four repetitions per treatment (Table 3); each repetition contained a confrontation (experimental unit). The data were subject to an ANOVA, and to a mean comparison Tukey test (p≤0.05), using SAS 9.0 (^{SAS Institute, 2002}).

Table 3 Treatments for the evaluation of biological control of the wilt in greenhouse chickpea plants. 

Inoculum production

Microorganism inoculum production was carried out using 500 mL flasks, using a inoculating loop to sample the biological material, under aseptic conditions, inside a laminar flow bell, and using as culture medium for Pseudonomas: 300 mL of tryptic soy broth (17 g casein peptone, 3 g soy peptone, 5 g NaCl, 2.5 g K₂HPO₄, 2.5 g dextrose, and final pH of 7.3±0.2); for Bacillus: 300mL of nurturing broth was use (5 g peptone, 3 g meat extract, and final pH 6.9±0.2); 300 mL of dextrose potato agar (PDA) were used to produce Trichoderma. Once the flasks were inoculated with their respective strains, they were incubated during 5 d at 28 °C, stirring them at 150 rpm. After this, their content was put on 50 mL test tubes (Falcon) and they were centrifuged at 3800 rpm for 20 minutes, at room temperature; the biomass was later harvested. Then, Sinaloa 92 white chickpea seeds were treated with a 1×10⁸ ufc mL^-1 concentration with the bacterial strains, using the Mc Farland 139 method (^{Ortigoza and Ruiloba, 1998}). The turbidity method was used to standardize bacterial suspensions. The seeds with the fungal strains were treated with a concentration of 1×10⁸ conidium mL^-1, using the Neubauer chamber method (^{Ferron, 1981}).

Greenhouse experiments

Inoculated seeds were sown on pots containing 12 L of substrate (Peat Moss). Before the sowing, the substrate (sterilized in autoclave) was inoculated with F. oxysporum f. sp. ciceris (Foc) race 5, using a 1×10⁸ conidium mL^-1 concentration. The experimental design was completely random, with five repetitions per treatment (Table 3); each repetition had 10 pots, and each pot had two plants (20 plants per repetition). The seeds were sown on December 2013; they were watered 15 d after the sowing, and ^{Murashige-Skoog (1962)} nourishing solution was used.

The following response variables were evaluated: plant vigor (development and sagging percentage of the plant); wilt (chlorophyll percentage); and root rot (necrosis percentages with subjective visual assessment scales) (Table 4). In addition, the following elements were evaluated: greenness (chlorophyll content on the leaves), using a SPAD 502 (Spectrum Technologies, Inc.); plant height (cm), from the stem base to the top of the plant; wet biomass (g); and dry biomass weight (g) of plants dehydrated in an oven at 80 °C, for 24 hours. Additionally, the stem diameter (mm) was evaluated using a digital Vernier caliper, while yield (g plant^-1) was calculated using a digital scale.

Table 4 Subjective scale to calculate the vigor, foliage wilt, and root rot, in chickpea plants grown in greenhouses. 

Data analysis

A normality and homogeneity of variance test was carried out using the inhibition percentage data; the means were compared with the Tukey test (p≤0.05), using SAS 9.0 (^{SAS Institute, 2002}). The data obtained from the greenhouse treatments of the non-parametric variables (plant vigor, foliage wilt, and root rot), along with their replication per treatment, were analyzed using the Kruskal-Wallis test. Parametric variable data (height, stem diameter, chlorophyll content, fresh weight, dry weight, and yield) were subject to an ANOVA test and a Tukey (p≤0.05) mean difference test, using SAS 9.0 (^{SAS Institute, 2002}).

Antagonistic activities of bacterial and fungal strains

Statistical analysis showed that in vitro radial growth of F. oxyspourm f. sp. ciceris (Foc) race 5 diminished 80.07 % to 38.89 % (p≤0.05) when it was evaluated in relation to biological control microorganisms (Table 5). The highest Foc PICR (80.07 %) was observed in its confrontation with the HRG-060 strain and it was different (p≤0.05) to the PICR of other antagonist strains. On that matter, ^{Michel-Aceves et al. (2005)} report 16.40 % to 77.80 % inhibitions, when in vitro isolated native Trichoderma are evaluated against F. oxysporum f. sp. lycopersici and 13.10 % and 94.40 % inhibitions when are evaluated against S. rolfsii. With regard to bacterial strains, the highest PICR was observed in T442 strain, which had a 59.26 % inhibition of Foc.

Table 5 In vitro mycelial growth inhibition of Foc race 5 against native microorganisms. 

Different letters indicate statistically significant differences (Tukey; p≤0.05).

Three days after inoculation, the HRG-060 Trichoderma strain had colonized 100 % of the surface of the medium where Foc had grown; HRG-050 and HRG-083 colonized the surface on the fourth day, as a result of which they held the 1^st degree of antagonism in the scale of ^{Bell et al. (1982)}. According to ^{Guédez et al. (2012)}. This indicates that these strains are highly aggressive, even more aggressive than Trichoderma spp. strains (^{Sánchez García et al., 2017}), which colonized 100 % of the medium where the phytopathogen of F. solani, F. oxysporum, and F. verticillioides grew, 10 d after inoculation.

During its confrontation with Foc, the HRG-060 strain formed 123±15 haustoria cm^-2 per confrontation, while the HRG-050 and HRG-083 strains formed 49±12 and 58±9 haustoria per confrontation, respectively. The highest amount of haustoria formed by the Trichoderma HRG-060 strain highlighted its higher efficiency against F. oxysporum f. sp. ciceris, since it facilitated the entrance of hyphae and the mycoparasitism of Foc. This proves that Trichoderma strains have a high mycelial growth inhibition capacity, particularly strain HRG-060 which formed more haustoria than the other fungal strains.

The control mechanism exercised by Trichoderma strains over Foc was the result of mycoparasitism: Trichoderma’s hyphae stuck to Foc’s hyphae and phialides, formed haustoria, and penetrated the hypha. These results are similar to those observed by ^{Kexiang et al. (2002)} who reported that mycoparasitism eliminated the pathogen’s mycelial growth.

There are different types of hypha interaction for Trichoderma (such as parasitism), and they are considered to be potential bioregulators of soil fungi (^{Infante, 2009}): Trichoderma hyphae coil up and penetrate F. oxysporum f. sp. cubense and penetrate Phythium sp. and R. solani hyphae. ^{Rivero (2008)} evaluated four Trichoderma isolates against Alternaria padwickii, Bipolaris oryzae, Curvularia lunata, and Phoma sp., and found out that it had a high competitive capacity, with two or more kinds of hypha interaction, except in the case of Phoma sp.

Protective effect of biological control microorganisms

The F. oxysporum f. sp. ciceris isolate used in our study is highly aggressive: it causes a 93.27 % rot in the positive control plant roots (Table 6).

The vigor of T7 (Trichoderma sp. HRG-060) treatment plants was greater than with other treatments (Table 6); meanwhile T3 (Pseudomonas sp. T162) plants showed the lowest vigor (p≤0.05). In the case of other treatments, the plants showed higher values than control and negative control. Treatment plants whose seeds were inoculated with Trichoderma (T6, T7, and T8) showed lower wilt, but they were less wilted (p≤0.05) that T5 (positive control) and T11 (Benomyl) plants. The lower root rot percentage occurred in T7 (Trichoderma sp. HRG-060) plants, and it was 41 % and 19 % lower than the symptomatology showed by T5 (positive control) and T11 (Benomyl) plants, respectively (Table 6). Additionally, the presence of the disease in T2, T6, T8, T9, and T10 plants was similar to its presence in T11 plants, and lower than in T5 plants (Table 6).

Table 6 Protective effect on the fusarium wilt of chickpeas, under greenhouse conditions. 

Means with different letters in each variable indicate statistically significant differences (Tukey; p≤0.05).

The greatest protective effect against fusarium wilt in the greenhouse bioassay was provided by the Trichoderma HRG-060 strain treatment (Table 5), which also had the best PICR (80.07 %). The PICRs of the HRG-050 and HRG-083 strains were 71.29 and 67.13 %, respectively, but there was no beneficial protective effect in chickpea plants inoculated with Foc. This information does not match the recommendations made by ^{Sánchez-García et al. (2017)} who mention that strains with good in vitro control also have that effect in vivo in inoculated plants, and recommend in vivo tests as an adequate tool to estimate biological control in plants.

The good results obtained from PICR means, from Trichoderma native strains, can be the result of the antagonists’ coadaptation with pathogen fungi, because they originated in similar climates and habitats (^{Andrews, 1992}). Therefore, there might not have been enough recognition between Foc and Trichoderma HRG-050 and HRG-083 strains. This recognition would have occurred through the lectin-carbohydrate interaction which agglutinates cells that interact with pathogen’s cell surface and created a favorable atmosphere for mycoparasitism (^{Infante et al., 2009}); therefore, the level of mycoparasitism was not the same as in the in vitro tests. This information confirms that coadaptation between the antagonist fungus and the pathogen fungus is required for control (^{Baker and Cook, 1974}). Therefore, carrying out plant tests is necessary to prove the true potential of the biological control agents. In all treatments, the pathogen was once more isolated from the necrotic tissue of all inoculated plants and it was morphologically confirmed to be the same as the inoculated strain. No symptoms were observed in the non-inoculated controls.

Plant growth promotion by bacterial and fungal antagonists

Plants in T7 (Trichoderma sp. HRG-060) were the highest: 26 and 57 % higher than control plants (T1) and positive control (T5), respectively (Table 7); the greatest stem diameter was shown by T4, T7, and T10. The chlorophyll content in T9 (Bacillus sp. T442) and T4 (Pseudomonas sp. 751) plants was greater than in other treatments (p≤0.05). The wet biomass weight was higher (p≤0.05) in T9 plants (Bacillus sp. T442) that in T1 (control), T5 (positive control), and T11 (Benomyl). In other treatments -except T2 (Pseudmonas sp. 7AI) and T3 (Pseudomonas sp. 162)-, the wet biomass weight was greater, in relation to T1 (control) and T11 (Benomyl) plants, when T9 (Bacillus sp. T442) plants had a greater wet biomass weight. Likewise, dry biomass weight showed a significant increase in T9 (Bacillus sp. T442) plants, in relation to plants in the remaining treatments.

Plants in T9 (Bacillus sp. T442) had the greatest chlorophyll content, wet biomass weight, and dry biomass weight values, but T7 (Trichoderma sp. HRG-060) plants showed a significant increase (p≤0.05) in grain yield, in relation to other treatments (Table 7); there was a significant relation (R²=0.91) between disease control in the plant and grain production yield. In our study, a greater (wet and dry) biomass production and a greater quantity of chlorophyll does not entail a greater grain yield. On that matter, ^{Howell (2003)} and ^{Mohiddin et al. (2010)} report that Trichoderma species promote an increase in plant yield and contribute to disease resistance or tolerance, when they interact with plants. ^{Mohiddin et al. (2010)} point out that Trichoderma is located inside and outside the rhizosphere, where it can colonize and protect the roots. Trichoderma does not only protect the plant, but it also helps its development; according to ^{Stefanova (2007)}, treating tobacco seeds with Trichoderma reduces external pollutants (such as Rhizopus stolonifer).

Table 7 Effect of inoculating seeds with bacterial and fungal microorganisms with the chickpea plant growth parameters. 

Means with different letters in each variable are statistically significant (Tukey; p≤0.05).

Various substances that promote plant growth are produced by certain rhizosphere microorganisms -including bacteria from the Bacillus genus, and fungi from the Trichoderma genus- and they can have a direct or indirect influence on the metabolism and physiology of the plant (^{Bhattacharyya and Jha, 2012}), because they synthesize and excrete phytostimulant substances, such as plant growth regulators and volatile organic compounds that reinforce plant immunity (^{Ahmad et al., 2008}; ^{Ambreen and Shahida, 2010}). Therefore, studying the interaction between the Trichoderma HRG-060 strain and the chickpea plant and subsequently influencing greater grain yield are essential.