The 17 strains for this study were isolated from healthy leaves of variety of Nacional type cacao with more than 50 old years. Two localities from Guayas province: Balao (2°30'29.5''S, 79°46'34.8''W) and Naranjal (2°40'35.2''S, 79°38'21.2''W) and one from Azuay province: Molleturo (2°30'49.2''S, 79°26'11.2''W) were sampled. The endophyte strains showed higher percentage of growth inhibition (PGI) against M. roreri and M. perniciosa and it was also found that they were not pathogenic of fruits and leaves of T. cacao (Villavicencio-Vásquez, 2018). Strains of M. perniciosa (CIBE-MP22) were isolated from Vinces and M. roreri (CIBE- A12.1) from the Amazon region (Table 1).

Dual cultures of endophytes-pathogens in plates containing potato dextrose agar (PDA) (4 g potato, 20 g dextrose and 15 g agar) medium, at pH 6.2 incubated at 26° C in darkness were performed (Condori et al., 2016). Two replicate plates were prepared for each strain of endophyte fungi. After 13 days of incubation, ten discs (5 mm) from the interaction zone of each two reply were extracted and plated in another PDA-plate, incubated as described before for seven days, and observed for the growth of the endophyte or the pathogen (Fig. 1). Mycoparasitism of each strain was determined based on the percentage of success in the survival (growth) of the pathogen with respect to the total of discs of the two replicates (% mycoparasitism = N° discs with pathogen growth /20 discs x 100).

The fungal metabolites were obtained in liquid Czapeck medium, containing 2 g NaNO₃, 5 g KH₂PO₄, 0.5 g MgSO₄, 10 mg FeSO₄, 3 mg ZnSO₄, 1g Yeast Extract and 60 g Glucose per liter at pH 7. Three flask containing 150 ml of medium were each inoculated with four mycelial discs of the endophyte of a 7-day-old colony in PDA. Three replicates were used for each treatment (strain). The cultures were incubated at laboratory room temperature (« 25 ± 2 °C) in an orbital shaker (110 rpm/21 d). After 21 days growth, mycelia were collected and the liquid sterilized in a vacuum system through a 0.22 jm filter (Millipore). The test was carried out with two concentrations (50 % and 75 %) of crude extract fungus diluted with PDA medium and poured into Petri dishes and plates containing media mixed with sterile distilled water were included as control. Once the filtrate solidified, agar discs (5 mm) of the pathogens (7-day-old) were placed at the center of the Petri dish. Three replicate plates of each concentration for each endophyte metabolite were incubated at 26 ° C in the darkness. The percentage of inhibition of mycelial growth of the pathogens were recorded as the difference between the average radial growth in the presence and absence of fungal filtration using the formula proposed by (Ezziyyani et al., 2004).

Table 1

Figure 1

Where: R1 is the radius of control pathogen and R2 is the pathogen with diluted crude extracts.

The ability of each strain to colonize healthy cacao tissue was studied. Seven days old cultures in Petri dishes of each of the 17 strains were photographed. Next, each plate was scraped and 5 mL of 0.05 % Tween 20 was added to aid mycelium transference. The mycelium was macerated, and the concentration adjusted to 1 0⁶-10⁷ CFU ml^-1 with sterile distilled water. For this test, detached leaves from seedlings of varieties of Nacional type cacao were obtained from the Instituto Nacional de Investigaciones Agropecuarias, INIAP (2°15'15''S, 79°49'W) in Guayas, Ecuador. Detached leaves were chosen to be as similar as possible in terms of size and age (two months old). Then were disinfected with 2 % sodium hypochlorite for two minutes and washed with sterile distilled water (Bañuelos-Balandrán and Mayek-Pérez, 2008). The leaves were uniformly sprayed with a volume of 1.5 mL of the inoculum solution. The leaves sprayed with sterile distilled water served as control. Each leaf was incubated in a plastic bag containing wet paper towels to maintain 100 % relative humidity (Tahi et al., 2007). The leaves were incubated for seven days in dark at 26 °C at an incubation chamber. After incubation, the inoculated leaves were washed with tap water, then three discs of 5 mm from each side of the central vein were extracted and disinfected as described above. Each leaf disc was cultured in Petri dishes containing 2 % malt extract agar (12.75 g maltose, 2.75 g dextrin, 2.35 g glycerol, 0.78 g peptone and 15 g agar) according to the methodology described previously (Arnold and Herre, 2003). After the third day of incubation, all mycelium that emerged from each leaf disc were transferred to a new PDA-plate. Three leaves and 18 leaf-discs for each treatment were used.

The identification of the colonies was based on three macroscopic characteristic: color, shape and texture of the colonies, before inoculating and after the posterior reisolations. Digital photographs were taken at each stage. The percentage of colonization of each endophyte was determined (% =N° discs with the re-isolated-endophyte-strain/18 total discs seeded * 100)

The mycoparasitism, PGIs obtained in the crude extracts and the re-isolation of the endophyte in the detached leaves, were compared using a non-parametric analysis of variance (Kruskal-Wallis) with a level of significance of 0.01, since after several transformations they did not present normal distribution using Shapiro-Wilks test. All analyses were carried out using the statistical package InfoStat (Di Rienzo et al., 2008).

The figure 2 shows the percentage of survival (growth) of M. roreri and M. perniciosa after the confrontation with the endophytes. Ten endophytes strains prevented the growth of both pathogens (Ec027, Ec035, Ec067, Ec084, Ec098, Ec107, Ec113, Ec120, Ec151 and Ec157). The other seven endophytes showed variable response to the pathogens (Fig. 2).

Figure 2

The crude metabolites of 11 and seven out of the 17 strains were able to inhibit the growth of M. roreri and M. perniciosa at both concentration, respectively (Fig. 4). The strain Ec107 at the 75 % (v/v) of crude extracts showed the greatest inhibition (60 %) against M. roreri (Figs. 3A and 4A), whereas the same extracts had the opposite effect on M. perniciosa, that is, it stimulates the growth of the pathogen, showing the negative values (Fig. 4B). The second more inhibitory strain (Ec035) against M. roreri at the same concentration, showed the highest inhibition against M. perniciosa (47 %) (Fig. 3C and Figs 4A-4B). In the figures 3B and 3D, the controls are shown. On the other hand, crude metabolites extracts of six strains always stimulate the growth of the pathogens at both concentrations, Ec084 (Daldinia eschscholtsii) against M. roreri, and five other including Ec084 against M. perniciosa (Fig. 4A-4B). Seven out of ten strains with the highest mycoparasitic capacity of the mycelium against both pathogens, also inhibited the growth of both pathogens through their crude metabolites (Ec027, Ec035, Ec067, Ec098, Ec151 and Ec157).

Figure 3

Figure 4

Nine out of the 17 endophytes were not recovered from the detached leaves (Fig. 5). Three strains achieve the infections of the leaves and were recovered from 100 % of healthy-leaf-discs: Ec059 (X. feejeensis), Ec098 (L. theobromae) and Ec151 (L. theobromae), showing the highest host colonizing capacity. Six out of seven strains whose metabolites inhibited the pathogens at both concentrations, and showed grater mycoparasitism by confrontation of the colonies, were recovered from the leaf discs of the host (Fig. 5). These strains were Ec098, Ec151, Ec157, Ec120, Ec067 and Ec027, however, as best strains should be considered the first three with re-isolation percentages between 80 and 100.

Figure 5

In the present study, the strains with the best simultaneous results in the three variables analyzed were Ec098, Ec151 and Ec157, all belonging to L. theobromae. These strains reinfected the host tissue between 80 %-100 % of the times, and maintained inhibitory effects both at the level of interaction of the colonies (mycoparasitism) and by their metabolites against M. roreri and M. perniciosa. On the other hand, the strain Ec035 (L. theobromae) stood out for its inhibitory effects in the trials of mycoparasitism and metabolites against both pathogens; and Ec107 (C. gloeosporioides) for inhibiting in 60 % the growth of M. roreri with 75 % (v/v) metabolites. However, at the same time, Ec107 metabolites stimulate the growth of M. perniciosa (Fig. 4). Neither of the two last strains could be recovered from the host.