Vegetal material and essential oil extraction

We used lime peel (Citrus limon), epazote above-ground plant parts (Dysphania ambrosioides) and cinnamon bark (Cinnamomum zeylanicum). Limes were collected in orchards in Yautepec Morelos, Mexico (18° 53’ 09’’ north and 99° 03’ 38’’ west). Epazote above-ground parts and cinnamon barks were purchased at the Central de Abasto (wholesale market) in Cuautla, Morelos, Mexico (18° 48’ 44’’north and 98° 57’ 21’’ wet). The vegetal species were identified at the Universidad Autónoma del Estado de Morelos Herbarium (number of specimens 34057-lime and 34058-epazote).

Vegetal material was used to obtain EOs by hydrodistillation (^{Díaz-Cedillo et al., 2013}). One kilogram of vegetal material was placed in an Italian-type distiller and kept on boiling (97 °C) for 2 h. The EOs obtained were stored in amber jars at 4 °C and then their yield was gravimetrically determined. Furthermore, lime EOs were purchased at the company Hersol (State of Mexico, Mexico), epazote AE at the company Aceites y Esencias (State of Mexico, Mexico) and cinnamon bark AE at the company dōTERRA (Mexico City, Mexico). According to the manufacturer’s specifications, the EOs were 99 % pure industrial-grade.

Chemical characterization of essential oils using CG-EM

The commercial EOs and the ones obtained by hydrodistillation were diluted in chloroform in a 1:10 v/v. Then, 1 µL of the dissolution was injected in a gas chromatograph (GC) SCION 456-GC (Bruker Daltonics, Billerica, USA) adapted to an EVOQ mass triple quadrupole detector and a PAL-COMBI xt automatic sample injector (Bruker Daltonics, Inc.). The mass spectrometer (EM) was set to a 50 to 500 mass-charge^-1 range using the MS Workstation version 8.2 software (Bruker Daltonics, Inc.).

The GC-EM was equipped with a BR-1ms capillary column 30 m long, 0.25 mm internal diameter and 0.25 µm thick (Bruker Daltonics, Inc.). The injector and detector temperature was 220 °C and 280 °C, respectively. Helium (He) was used as carrier gas at a 1 mL min^-1 flow rate. The initial oven temperature was 55 °C for 1 min and then increased to 155 °C at heat speed rate of 20 °C min^-1; this rate was maintained for 2 min and then increased to 255 °C at heat speed rate of 10 °C min^-1 and a total analysis period of 20.14 min per sample.

The detected compounds were identified by comparing their retention time and mass spectrum with data from standards included in NIST (National Institute of Standards and Technology, MD) team’s library.

Fungus isolate and crop conditions

The A. alternata isolate was obtained from the fungi collection of the Laboratory of Tecnología Postcosecha de Productos Agrícolas (Ceprobi, Morelos, Mexico). The species that had already been morphologically and molecularly identified was grown in potato-dextrose-agar culture (PDA, Bioxon, Mexico) for 14 days at 28 °C.

Inhibition test of mycelial growth

The EOs inhibitory effect was determined using the agar-dilution method reported by ^{Chen et al. (2014)} which involves mixing essential oil concentrations (0.25, 0.5, 1.0 µL mL^-1) with 0.1 % (v/v) Tween 20 (Hycel, Mexico City, Mexico) in 150 mL of PDA medium. Then, Petri dishes 90 mm in diameter were filled with 25 ml of the mixture (6 replications per treatment) and left to cool at room temperature (25±2 °C). PDA and PDA-Tween were used as controls.

A PDA disk 100 mm in diameter containing A. alternata was grown in the center of a Petri dish and incubated at 28±2 °C. Mycelial growth was measured daily using a vernier caliper in order to evaluate the diameter reached by the mycelium over time. The test ended when the mycelium completely covered the Petri dish containing the control treatment. The results were reported as mycelial growth inhibition index (IM), according to the following equation (1), where C_C denotes the control’s growth, and C_T , the growth in the treatment.

(1)

Treatments reaching 100 % inhibition were regrown in PDA culture but without adding essential oil to evaluate the fungistatic or fungicide effect of the EOs used.

Inhibition test for conidial germination

The effect of EOs on conidial germination was evaluated using the technique described by ^{Bautista-Baños et al. (2008)}. A conidial solution (10⁵ conidia mL^-1) of a 14-day old monosporic culture of A. alternata was prepared and added to six disks containing PDA (25 µL disk^-1) which then were placed on glass slides. Later, 10 µL of EOs were added to each disk (0.25, 0.5, 1.0 µL mL^-1) incubated at 28±2 °C within sealed Petri dishes. Germination was carried out after 2, 4, 6 and 8 h. A conidium was considered germinated when a germ tube was observed regardless of its length (^{Costa et al., 2015}).

After the incubation period, a drop of lactophenol blue solution was added to each disk and 100 conidia per disk were observed under an optical microscope 40x objective (Nikon alphaphot-2 YS2-H, Japan). The percentage germination was calculated (G_(%)) using the following equation (2), where E_G denotes the number of germinated conidia, and T_E the total conidia.

(2)

Adjusting experimental data to mathematical models

Experimental data of A. alternata mycelial growth in PDA culture in terms of time, with and without EOs, were adjusted to the model of ^{Baranyi and Roberts (1994)} to estimate the maximum growth rate and apparent lag phase time using the following equation (3). Where D_(t) denotes the diameter of the mycelium (cm) at any time t (días); D_o , the diameter at t=0, µ_max (cm day^-1), the maximum growth rate; and λ (days), the apparent lag phase time.

(3)

The model was developed using DMFit version 3.5 (Microsoft Excel add-in) (Institute of Food Research, Norwich, UK). To evaluate the model, it was estimated the Mean Square Root Error (MSRE) and the determination coefficient (R²). Models with MSRE values near to cero and R² near to one were considered to be the best (^{Tremarin et al., 2015}). Growth data (experimental and predicted by the model) were used to plot surface graphs using SigmaPlot 12.0 (Systat Sofware Inc., CA, USA).

The conidial percentage germination was adjusted to the logistic model (Equation 4) to estimate the parameters, the increase in the percentage germination and the time required to reach 50 % germination. Experimental units containing EOs that not allowed fungal germination were omitted in the analysis. For this purpose, we used the statistical software InfoStat version 2016 (Universidad Nacional de Córdoba, Argentina). Where P denotes the germination percentage; P_max , the maximum percentage germination (100 %); k(h^-1), the increase in the rate of percentage germination; r (h), the time required to reach 50 % germination; and t (h), the time (^{Dantigny et al., 2007}).

(4)

Statistical analysis

Data of the volatile compounds of the six different EOs identified through GC-EM were analyzed using principal component analysis (PCA) to obtain the variability of the studied systems with the software package InfoStat version 2016; mycelial growth and conidial germination at the end of the growth period were analyzed using a completely randomized design in a factorial arrangement with 6 replications. The factors used were 1) type of essential oil (lime, epazote and cinnamon); 2) source (hydrodistilled or commercial); and 3) concentration (0.25, 0.5 and 1.0 µL mL^-1). Tukey’s test (p≤0.05) to compare the medians, and the InfoStat version 2016 software for data analysis was used. The experiment was performed in duplicate.

Chemical composition of essential oils

Results show differences in the composition of the EOs evaluated regarding the vegetal material and the method used to extract them (by hydrodistillation or commercial). Cinnamon, lime and epazote EOs obtained by hydrodistillation yielded 0.49, 0.33 and 0.015 %, respectively.

Considering the type of vegetal material from which the EOs were extracted, it was observed that lime and epazote oils contained terpenes mainly (D-Limonene, ρ-Cymene and Ascaridole), while the predominant compounds in cinnamon oil were aldehydes (cis-Cinnamaldehyde) (Table 1). However, considering the method used, differences in the content of the detected compounds were noted.

Table 1. Summary of the main volatile organic compounds of three essential oils detected by CG-EM. 

AE-essential oil; AR-relative area; TR-retention time; GF-functional group; A-aldehyde, E-ester, T-terpene

As for D-Limonene, it was observed 7.26 % more content in the lime AE obtained by hydrodistillation compared with the commercial AE. In epazote AE, two main components were identified: ρ-Cymene in non-commercial oil (22.40 %) and Ascaridole in commercial oil (27.30 %). Finally, the content of cis- Cinnamaldehyde in commercial cinnamon oil was 4.20 % higher than that obtained by hydrodistillation; this compound was the highest in both EOs.

Principal component analysis (PCA)

For PCA, compounds representing an area greater than 1 % were used in the chromatographic analysis. According to ^{Terrádez (2002)} there is no denifite rule regarding the number of principal components (PC) to be used, but it is important to keep in mind that one of the objectives of the analysis was to reduce the number of variables. In this regard, we selected the PCs that contributed at least 5 % of the total variance and whose proportion of accumulated variance explained at least 70 % of the same (^{Mora-Aguilera and Campbell, 1997}). The results were grouped in five PCs that explain 100 % of the variability in the chemical composition of the EOs used (Table 2). However, given that PC1 and PC2 explained 74.0 % of the variability, the results were described in PC1 and PC2 terms.

Table 2. Principal Component Analysis (eigenvalues) among the compounds of three different essential oils. 

Figure 1A, shows the three groups formed considering the type of vegetal material. The first group corresponds to lime oil obtained by hydrodistillation; the second group, to cinnamon oils (commercial and non-commercial); and the third group, to lime (commercial) and epazote (commercial and non-comercial) oils. In PC1 we observed that essential cinnamon oil was different from epazote and lime oils. In PC2, differences were found among epazote (comercial and non-comercial) and lime (commercial) oils, and cinnamon (commercial and non-commercial) lime oils obtained by hydrodistillation.

On the other hand, significant values were observed in PC1 due to the presence of monoterpenic compounds, while PC2 was positively correlated with aldehydes and esters (data not shown). This is clearly appreciated in Figure 1B with three groups formed: a) the first group included D-Limonene and L-β- Pinene, which correspond to lime oil composition; b) in the second group we observed aldehydes, esters and non-cyclic monoterpenes such as cis-Cinnamaldehyde, compounds that are present in cinnamon oil; and c) the third group, showed oxigenated monoterpenes as Ascaridole, compounds that are present in epazote oil.

Figure 1. Distribution of essential oils (A) and compounds identified (B) according to eigenvectors from principal component analysis calculated for PC1 and PC2 (74 % variance). 

Mycelial growth inhibition

Statistically significant differences were observed (p<0.0001) in A. alternata in vitro mycelial growth (at day 9-time when the control treatment reached its maximum growth) by the effect of the essential oil type, source and concentration (Table 3). Results showed that A. alternata did not grow on the PDA medium to which cinnamon essential oil (commercial and non-commercial) was added at concentrations of 0.5 and 1.0 µL mL-1, and which had a fungicidal effect on the fungus. In contrast, commercial lime essential oil at 0.25 µL mL^-1 concentration did not inhibit A. alternata growth, while, at the same concentration, comercial lime and epazote oils showed limited inhibition (4.06 and 3.42 %, respectively). A similar effect was observed with commercial lime oil at a concentration of 0.5 µL mL^-1 (3.18 % inhibition).

Table 3. Mycelial growth and percentage inhibition of A. alternata incubated for 9 days in PDA medium to which AEs were added. 

CM-Mycelial growth; DE-standard deviation; medians with the same letter are significantly different (p > 0.05) by Tukey’s test.

Inhibition of conidial germination

The percentage germination of A. alternata conidia to which EOs were directly applied showed statistically significant differences (p<0.0001), as a result of the treatments, when they were evaluated after 8 h (data not shown). No germination was observed in treatments containing cinnamon essential oil (0.25, 0.5 and 1.0 ml^-1), non-commercial lime and epazote oils (0.5 and 1.0 µL mL^-1) and commercial lime and epazote oils (1.0 µL mL^-1).

Treatments showing the highest germination inhibition percentage were non-commercial lime (92.50 %) and epazote oils (93.33 %) at a concentration of 0.25 µL mL^-1, while at a concentration of 0.5 µL mL^-1 it was the commercial lime and epazote oils (91.0 % and 94.17 %, respectively) that showed the highest percentage germination inhibition; this suggests that the results depend on the oil concentration used.

Mycelial growth model

Experimental data of A. alternata mycelial growth on PDA at different concentrations of EOs, in terms of time were adjusted to the model of Baranyi and Roberts. Mycelial growth tended to decrease as the concentration of EOs in the culture medium increased (Figure 2).

Figure 2. Surface graph of the effect of different concentrations of essential oils on A. alternata mycelial growth. Essential oils obtained by hydrodistillation. a. lime. b. epazote. c. cinnamon. Commercial essential oils. d. lime. e. epazote. f. cinnamon. Experimental data (), data predicted by the model (). 

Based on data of the evaluated treatments, we developed 14 models (Table 4). The average determination coefficient (R²) was 0.97, while the root-mean-square error (RMSE) was 0.15. In the models developed, we observed that as the concentration of essential oil increased, the maximum growth rate decreased (µ_max) and the time of the apparent lag phase increased (λ).

Table 4. Parameters obtained by adjusting experimental data of mycelial growth of A. alternata incubated in PDA and AEs to the model of Baranyi and Roberts. 

O-AE obtained;C-AE commercial;µ_max -maximum growth rate; λ- apparent lag phase time; SE-standard error; R²-determination coefficient; RMSE- square root mean error

Conidial germination model

Experiment data of conidial germination were adjusted to the Logistic model. Using the evaluated treatments 5 models were developed (Table 5). The average RMSE was 1.98. The increase in EOs concentration affected germination, and as the concentration of essential oil increased, there was a lower rate of germination increase k (h^-1) and more time was required to reach 50 % germination r(h).

Table 5. Parameters obtained by adjusting experimental data of conidial germination of A. alternata incubated in direct contact with AEs to the logistic model. 

O-AE obtained;C-AE commercial; k-rate of increase in percentage germination; r-time required to reach 50 % germination; SE-standard error; RMSE-square root mean error