Obtaining the Forl strain

The Forl strain was previously isolated from the rhizosphere of tomato plants grown in a commercial greenhouse in the Yaqui Valley, Sonora. The fungus was distinguished by comparing its colonial and morphological characteristics using specialized literature (^{Leslie and Summerell, 2016}; ^{Nelson et al., 1983}), which was identified molecularly with specific primers according to ^{Hirano and Arie (2006)} (data not published). A copy of the isolation is kept as Forlmx-19 in the microorganism bank of the Ecological Chemistry Lab in the microorganism bank of the Ecological Chemistry Lab in the Research Center for Food and Development, Civil Association (CIAD A.C.), Hermosillo, Sonora.

Obtaining L. tridentata extracts and content of nordihydroguaiaretic acid (NDGA)

Preparation of the extracts. Extracts were produced using four solvents: dichloromethane (DCM), methanol (MeOH), ethanol (EtOH), and water (H₂O) using young L. tridentata leaf and stem samples from the Sonora desert (^{Peñuelas et al., 2015}). For the DCM and MeOH extracts, 90 g of dry and crushed L. tridentata leaves were used to for separate extractions, with each solvent in a Soxhlet system at 39 and 64.7 °C, respectively, for 15 h of reflux, followed by rotatory evaporation. For the EtOH and aqueous extracts, 1 kg of fresh material was softened in 10 L of an ethanol:water solution 7:3 (v/v) for 8 days at room temperature (25 °C) and in the dark. After time it was filtered using gauze and then paper (Whatman #5). The filtrate underwent rotatory evaporation; the solid phase is considered as EtOH extract, and the supernatant was frozen at -45 °C and freeze-dried to be considered as the aqueous extract. The NDGA content was quantified in the four extracts obtained using reverse phase high performance liquid chromatography (RP-HPLC), in a HPLC system (Hewlett Packard series 1100, Germany) connected to a diode arrangement detector (DAD) in the Ecological Chemistry Laboratory in CIAD A.C. Hermosillo, Sonora. We used the methodology proposed by Hyder et al. (2002) with an ODS C-18 hypersil column and as a mobile phase: A: H₃PO₄ 1 % and B: 85 % acetonitrile-14 % H₂O-1 % H₃PO₄ in a 60/40 proportion, with Isocratic elusion, and a flow of 1 mL min^-1. The detection was carried out at a wavelength of 283 nm. For the quantification, a calibration curve was created with the NDGA standard (purity≥90 %, Sigma Aldrich, United States). The volume of injection of the NDGA standard and the samples was 25 μL.; the data were analyzed using the software Chemstation (Hewlett Packard, Germany).

Determination of the percentage of inhibition of Forl by L. tridentata extracts

We used the poisoned medium method (^{Guerrero-Rodríguez et al., 2007}), which consists in pouring the extracts in the agar dextrose potato (ADP) culture medium in sterile, 5 cm petri dishes; after the medium has solidified, a sample of the vigorous growth margin of the 12-day old strain was planted by poking the center. The period of incubation lasted 7 days at 28 °C.

Variables evaluated

The mycelial growth was recorded every day starting on day 3 of incubation, measuring in mm the two crossed diameters; monitoring ceased when the control reached 100% of the Petri dish. The inhibition was determined using the formula of the percentage of inhibition = (dc-dt/dt)x100, where dc is the average diameter of the mycelial growth of the control, and dt is the diameter of the mycelial growth with the different extracts (^{Lira-Saldivar et al. 2003}).

Experimental design

A 4x5 factorial analysis was used, where A were the extracts, and B, the concentrations (Table 1). The experimental design was completely random, and an ANOVA was used, along with a Tukey average comparison test (p<0.05) in the software Statgraphics 5.1.

Table 1 Experimental design of the in vitro iotests of the L. tridentata extracts 

v dichloromethane w methanol, x ethanol, y water, negative control: ADP mixed with acetone:water

Biotest in vivo

This was carried out under greenhouse conditions in the Experimental Transfer and Technology Center 910 of the Sonora Institute of Technology (ITSON) in the Yaqui Valley, Sonora. One hundred commercial cocktail tomato seedlings var. F1 Prolyco2 with 35 days of growth were used. The agronomic management of the crop (fertilization, irrigation, and cultural handling) was carried out according to the Sonora Technical Agricultural Agenda (^{SAGARPA, 2015}).

Inoculation of Forl. The streak plate method was previously used to culture the Forl strain in Roux bottles with 250 ml of ADP with 3.5 ml of tartaric acid at 10 %, previously sterilized and gellified according to ^{NOM-111-SSA1-1994} and ^{Camacho et al. (2009)}. They were incubated at 28 °C for 7 days until the mycelium developed. Glass and 20 al 0.02 % tween beads were placed on the agar; using soft movements, the conidia were moved to a sterilized beaker. They we re counted in a Neubauer chamber, and adjusted to a concentration of 10⁷ conidia mL^-1 with distilled water. The roots of the plantlets were submerged for 30 min in this inoculant solution after having cut an incision into their secondary roots with a scalpel. They were then transplanted into 9 L pots containing clay loam soil from the Yaqui Valley, Sonora.

Application of treatments. To achieve the adhesion of the different concentrations of the extracts evaluated, the agricultural adherent® (Técnica Mineral S.A. de C.V) was used, at 1 L ha^-1. The extracts were reactivated in acetone at 100 ml ml^-1 (extract DCM and EtOH) and acetone:water (extract MeOH and H₂O), the agricultural adherent was added and mixed until a suspension was obtained. Every 7 days after transplanting, uncovering the top section of the root, the pertinent extract was applied to it.

Variables evaluated

The following measurements were taken, 49 days after inoculation (dai): chlorophyll index with Minolta SPAD 502 (between 11:00 and 14:00 h) in chlorophyll units (CU), plant height with measuring tape (from the base of the stem to the apex) and number of leaves. After the emergence of the 5th true leaf, each plant was evaluated on a scale of 1 to 5 (Table 2), depending on the intensity of the disease (^{Marlatt et al. 1996}). Once classified, the severity index was calculated using the formula by ^{Towsend and Heuberguer (1943)}: Severity= (∑nv/ 5N) x 100, where n= plants per each category; v= value of each category; N= plants in experimental units. The data were normalized with a natural logarithm before. The incidence was calculated by the percentage of plants that showed some symptom of Forl infection. Plants were considered survivors if they remained alive after the completion of the experiment.

Table 2 Scale to evaluate the severity of Folr in the tomato crop 

^{Marlatt et al. (1996)}

Experimental design

A totally random design was applied with 10 treatments and 10 repetitions (Table 3). The concentrations of the treatments were defined according to the results of the CMI performed previously. An ANOVA was carried out, as well as a multiple range test using the software Statgraphics plus version 5.1.

Table 3 Treatments applied in the in vivo biotest 

u commercial fungicide, v dichloromethane, w ethanol, x methanol, y water.

Determination of the percentage of Forl inhibition by the effect of applying L. tridentata extracts

Table 4 shows the activity of the extracts on the mycelial growth of Forl. The L. tridentata extract with DCM displayed an inhibition of 98 % at 750 ppm; 94 % for MeOH at 2000 ppm , and 97 % for EtOH at 500 ppm, which indicates that the compounds extracted with EtOH had a greater fungicidal power over Forl in vitro. These results were used as a basis to determine the doses for the treatments in vivo, considering the concentration of each extract (p≤0.05) the completely inhibited the mycelial growth of the fungus. The concentration was increased to 3000 y 6000ppm of the extract H₂O, since it presented little effect on the growth inhibition of Forl.

The highest inhibition of the hydroalcoholic extract of L. tridentata has been reported at 7000 ppm for Aspergillus flavus and 5000 ppm for Penicillium sp. (^{Moreno-Limón et al., 2011}). The antifungal effect of L. tridentata is attributed to phenolic compounds, because polyphenolic extracts inhibited to the maximum with 0.7 ppm in Pythium sp., Colletotrichum truncatum, C. coccodes, Alternaria alternata, F. verticillioides, F. solani, F. Sambucinum and Rhizoctonia solani (^{Osorio et al., 2010}).

Among the phenolic compounds of L. tridentata, NDGA stands out since it is extracted from the resin in leaves and stems (^{Lira et al., 2003}). The amount of this compound extracted is directly related to the degree of polarity of the solvent (^{Moreno-Limón et al., 2011}). In this way, this study quantified 68, 64, 6.07, and 2.8 mg g^-1 of NDGA in the extracts of DCM, EtOH, MeOH, and H₂O, respectively, correlating the concentrations quantified in the extracts with the effect on the mycelial growth shown in Table 4.

Table 4 Effect of the L. tridentata extracts on the mycelial growth of Forl 

v dichloromethane, w methanol, x ethanol, y water.

Different letters represent significant differences between treatments (Tukey p≤0.05).

Giving a standard recommendation for the use of extracts in the control of phytopathogenic microorganisms, since there is not only the influence of the basic triangle host-pathogen-weather, but also of factors such as solvents and conditions of extraction, latitude and altitude of the area of sampling of L. tridentata (^{Lira-Saldivar et al., 2003}), variability in breeds and pathogenicity among the special forms of F. oxysporum Fol and Forl; furthermore, in vivo one must consider the effect of the biodiversity of organisms in the soil and the surroundings.

Biotest in vivo

Chorophyll index. The content of chlorophyll in the leaves is shown in Table 5. All the extracts displayed higher levels (p<0.05) in comparison with the absolute control and the commercial fungicide. The treatments worth pointing out are those with H₂O 6000, DCM 3000, and MeOH 4000 ppm with 49, 48, and 45 units of chlorophyll, respectively, which are values higher than those reported for healthy tomato plants (35 and 37 UC) (^{Mercado-Luna et al., 2010}). The remaining treatments presented values within this range, except for the control and the positive control, which showed the damage caused by Forl.

The UCs provide an indirect and quantitative measure of the amount of N in the plant, as well as the effect of weather variables (^{Rincón-Castillo and Ligarreto, 2010}; ^{Vázquez et al., 2012}). The plants treated with L. tridentata extracts maintained an overall state of well-being, even after having been infected with Forl.

Plant height. Table 5 shows the values for plant height in each treatment; all the extracts were statistically higher (p≤0.05) to the absolute control and commercial fungicide, from the first week to the end of the experiment. Treatments H₂O 4000, MeOH 4000, and DCM 3000 ppm, presented plants 49, 48, 42 % taller than plants that received the commercial fungicide. Increases in plant height were reported for tomato plants (144 and 148 %) by ^{Díaz-Díaz et al. (2013)} with L. tridentata extracts in concentrations of 3000 and 2000 ppm with the use of MeOH and acetone as solvents. The absolute control (0 ppm) showed general growth delay caused by the typical symptomatology of Forl: invasion of the root, premature loss of lower leaves and rot of the radicle. The Neem (Azardiachta indica) and L. tridentata extracts favor height (up to 35 % in comparison to the control) in tomato plants when they have been infected with F. oxysporum (^{Hadian et al., 2011}).

Leaf number. All the treatments presented a higher number of leaves (p≤0.05) than the absolute control, particularly the treatments MeOH 4000, DCM 3000, H₂O 4000 ppm, with 52, 50, and 45 % more leaves than the commercial fungicide, respectively (Table 5). ^{Díaz-Díaz et al. (2013)} recorded a 25 % increase in dry weight when applying extracts of L. tridentata by 3000 and 2000 ppm in tomato plants inoculated with P. capsici, which accounts for a greater vegetative growth, as in the results presented in this investigation; the use of L. tridentata as an infusion is also reported as a stimulant tomato plant growth in combination with lanoline and cocoa butter, since it increases dry (77 %) and fresh weights (100 %) (^{Recinos, 2010}). No signs of toxicity were observed in the doses used, as was the case in the aqueous extracts of L. tridentata at 30 % and after 5 days of application (^{Rojob, 2008}). The infestation by Forl implies the wilting of the lower leaves towards the top (^{Fernández-Herrera et al., 2013}), and therefore the lower levels of infestation helped the plants generate and preserve a higher number of leaves than the absolute control and the commercial fungicide.

Table 5 Vegetative development of tomato plants inoculated with Forl and treated with L. tridentata extracts at 49 dai 

u commercial fungicide, v dichloromethane, w ethanol, x methanol, y water. Different letters represent significant differences between treatments (p≤0.05).

Severity. Significant differences (p≤0.05) were found between the treatments with MeOH at 4000 and DCM at 3000 ppm, which displayed severity indices of 55 and 57 % (respectively) in comparison with the 94% of the absolute control and H₂O 6000 ppm. ^{Díaz-Díaz et al. (2013)} reported reductions of up to 83 % in severity when evaluating tomato plants infected with P. capsici and treated with L. tridentata extracts and acetone at concentrations of 2000 ppm. The results shown (Figure 1-A) indicate a reduction of 45% in the severity in MeOH 4000ppm, and a relation is established between the variables of vegetative growth and. There is a similar behavior between treatments DCM 3000, MeOH 4000, and H₂O 4000 ppm, which gave lower severity indices and higher levels of chlorophyll, height, and number of leaves than the commercial fungicide, taking into account the first report on the severity of Forl on tomato plants that have been treated with L. tridentata extracts.

The extract DCM 3000 ppm presented the lowest severity index and a higher vegetative development than the commercial fungicide, as well as an inhibition of 100% at 1000 ppm in the Forl in vitro biotest and a higher NDGA content. In regard to this, ^{Martins et al. (2012)} stated that the antifungal activity of L. tridentata can be attributed to three compounds: NDGA, kaempferol, and quercetin, which are highly compatible with the solvent dichloromethante, which is more available due to the polarity of the three compounds. ^{Hanaa et al. (2011)} and ^{Villa-Martínez (2015)} used aqueous extracts of neem and willow (Salix babylonica) in tomato plantlets, which reduced the incidence of the disease of wilting by F. oxysporum in tomato plantlets by up to 30%, by the increase of the defensive antioxidant enzymes secreted by the tomato plants, such as peroxidase, catalase and the dismutase superoxidase.

Incidence. Statistically significant differences (p≤0.05) were obtained; 100 % of the plants treated with the control and H₂O 6000 ppm treatments registered symptoms of Forl infection (Figure 1-B); however, when applying the extracts DCM 3000 and MeOH 4000 ppm, the incidence reached only 40 and 50 %, respectively. No reports were found on in vivo tests for the control of Forl with L. Tridentata extracts; however, in carrot plants, there were reports of a 5 % incidence of blight caused by A. dauci when applying a hydrosoluble extract of L. tridentata mixed with chitosan (^{Hernández-Castillo et al. 2006}).

Figure 1 Percentage of severity (A), incidence (B) and survival (C) of plants inoculated with Forl. 

Survival. At 49 dai, one out of every 10 plants treated with the control and H₂O at 6000 ppm survived, whereas with extracts DCM 3000 and MeOH 4000 ppm favored the survival of 60 % of the plants. With treatments EtOH 2000 and MeOH 6000 ppm, 50 % and the commercial fungicide, 40 % (Figura 1-C). These reports show a reduction in the aggressiveness of Forl due to the effect of the L. tridentata plant extracts, which may be due to the phenolic substances present in the extracts react chemically with the sensitive systems of the enzymes, making them catalytically inactive with this pathogen, therefore favoring the survival of the plants (^{Tappel and Marr, 1954}).