Collecting and obtaining isolates

In 2009-2011, leaves, stolons and strawberries showing typical anthracnose symptoms were collected in commercial plots of Festival, Albión and Camino Real strawberry varieties in the Maravatío and Zamora valleys. To capture greater variability in the samples, they were taken from plots separated by a distance of 1 to 3 km, approximately.

The sensitivity test was conducted at the Phytopathology Laboratory of the Centro Regional Universitario Centro-Occidente (CRUCO), Universidad Autónoma de Chapingo, in Morelia, Michoacán, and the pathogen’s molecular identification was performed at the Laboratory of Biotechnology and Seed Pathology for Postgraduate Students in Genetic Resources and Productivity (PREGEP), Colegio de Postgraduados, in Montecillo, State of Mexico.

Obtaining monoconidial isolates

From strawberries, leaves and stolons showing symptoms, portions of tissue approximately 2 cm were selected from the area showing symptoms, washed with water, disinfected with a 1 % sodium hypochlorite solution for 2-3 minutes, and rinsed with sterile distilled water. They were then dried in a laminar flow cabinet and grown on potato-dextrose-agar (PDA) (Difco, Becton, Dickinson and Company, Sparks, MD 21152, USA). From the stock cultures, monoconidial cultures were prepared using serial dilutions; the conidia were placed on 2 % water agar (WA) and incubated for 10-18 hours at room temperature. Later, using a dissection needle, the germinated conidia were extracted one by one and placed in Petri dishes containing PDA. The monosporic isolates (Table 1) were stored at 4o C and later used in fungicide sensitivity tests.

Table 1. Colletotrichum acutatum isolates according to cultivar, structure and location where strawberries were collected in 2009-2010. 

Molecular identification

Monoconidial isolates from the Maravatío (8 isolates) and Zamora (11 isolates) valleys were selected. DNA extraction was carried out following the protocol of ^{Doyle and Doyle (1990)} with minor modifications. Later, through PCR, the ITS1, 5.8S and ITS2 regions of the ribosomal DNA were amplified using ITS5/ITS4 primers (^{White et al., 1990}) in a C1000 TouchTM Thermal Cycler (Bio-Rad, USA) as follows: initial denaturation at 95 °C for 4 min, followed by 35 cycles at 95° for 2 min, 55 °C for 1 min, 72 °C for 2 min, and a final extension at 72 °C for 2 min. Later, the PCR products were cleaned using a QIAquick purification kit (Qiagen Inc., Chatsworth, CA, USA), according to the manufacturer’s instructions. Sequencing was performed in Macrogen, Maryland, USA; then the sequences were edited in Bioedit Sequence Alignment Editor v.7.0.5 (^{Hall, 1999}), and consensus sequences were generated. The sequences were then compared with the sequences stored in the Genbank database of the National Center for Biotechnology Information (NCBI) using the BLASTNucleotide (^{Zhang et al., 2000}) program; we took into consideration the highest value resulting from the comparisons.

Fungicide application

Commercial formulations of thiophanate methyl from the benzimidazole group (Cercobin M^®, BASF Mexicana, Zapopan, Jalisco) at 70 % concentration were used in the tests, as well as azoxystrobin, which belongs to the group of quinone outside inhibitors (QoI) (Amistar^®), at 50 % concentration of the active ingredient (Syngenta Agro S.A de C.V. México, D.F.).

Stock solutions and amended culture medium preparation

Stock solutions of each fungicide were prepared by mixing the active ingredient with sterile distilled water. Using the concentrated solution, 1:10 serial dilutions were prepared to logarithmically obtain the desired concentrations. The solution containing the corresponding fungicide concentration was added to the medium when it cooled to 50-60 °C, constantly stirring to ensure uniform homogenization, and immediately placed in 15 mL Petri dishes.

Mycelial growth test

Six doses of each fungicide were evaluated: 100, 10, 1, 0.1, 0.01 and 0.001 µg mL^-1 of azoxystrobin, and 2000, 1000, 100, 10, 1 and 0.1 µg mL^-1 of methyl thiophanate, plus an absolute tester to which no fungicide was added. From the margins of 6-day old active colonies of each of the monoconidial isolates, we obtained disks 0.5 cm in diameter, which were placed in the middle of 90 x 15 mm Petri dishes containing culture medium to which the fungicide had been added. Three replications (Petri dishes) of each dose were prepared. The dishes were incubated in darkness at room temperature (20±4 °C) for five days. After this period, the diameters of the colonies were measured perpendicularly in two directions to obtain the average.

Conidia germination test

The same six doses of each fungicide were evaluated, plus a tester to which no fungicide was added. Three replications of each dose (50x15 mm Petri dishes) were used. Monoconidial suspensions were prepared at concentrations of 10³ to 10⁴ conidia mL^-1. The suspensions were grown in 2 % WA culture medium to which fungicide had been added, only as a germination surface, prepared in the same way as in the mycelial growth inhibition stage. Six hours after being sown, the number of germinated and non-germinated conidia in each replication of a 100-conidia random sample were counted. A conidium was considered germinated when the length of its germ tube was 2 times greater than the diameter of the conidium.

Estimation of effective doses 50 (ED₅₀) and data analysis

Based on the two measurements of the diameter of the colony during the mycelial growth tests, we obtained the average, and after the three replications (Petri dishes) —which were used to calculate growth or germination compared to the control per each fungicide dose and isolate, we adjusted the model and estimated the ED₅₀ using the log- Logistic model with upper asymptote using the NLMIXED procedure of SAS (Statistical Analysis System version 9.3). The model is expressed as follows:

Where y denotes relative mycelial growth or relative conidial germination, α is the upper asymptote (maximum y value possible), β the slope at the inflection point, and ED₅₀ the dose at which growth/germination is reduced by 50% (Schabenberguer and Pierce, 2001; ^{Rebollar-Alviter et al., 2007}). Given that not all the initial values resulted in convergence for all the isolates, different sets of values were explored at the beginning of the iterations in order to achieve convergence. The adjustment to the model was confirmed considering the residuals pattern for each isolate (^{Rebollar-Alviter et al., 2007}). Once ED₅₀ was estimated, an exploratory data analysis was conducted, and the sensitivity distributions were illustrated with frequency histograms.

Before estimating the statistical moments, normality tests were conducted on the original data, and the ED₅₀ values were converted to 10-base logarithms. The estimated moment, the dispersion of basic statistics, the bias or misalignment parameters and the normality tests were conducted using the Proc univariate included in SAS software (version 9.3). Once the general sensitivity distribution of the two strawberry producing areas was explored, the data from each area were individually analyzed to determine significant differences between the distributions in each strawberry producing valley in terms of the sensitivity of C. acutatum isolates. This was done because the two production areas may exert different levels of selection pressure due to their contrasting technology levels. The comparison was made using the Kolmogorov-Smirnov test and the NPAR1WAY (SAS version 9.3) procedure to determine whether the distributions came from a single population or different populations.

Morphological and molecular identification of isolates

All the pathogen isolates associated with strawberries showing anthracnose symptoms and damage collected in the two strawberry producing areas of Michoacán proved to be Colletotrichum acutatum. The isolates were characterized by forming white colonies with dark gray centers and concentric orange circles both at the center and at the edges. The conidia were smooth, unicellular, aseptate, spindle-shaped and measured 10-14 x 4-6µm, and similar to those reported by ^{Damm et al. (2012)}. The sequences obtained by PCR amplification of the region included in the Internal Transcribed Spacer (ITS) of the rDNA of the 20 isolates were 99 % identical to the AM991131 and AY266405 accessions placed in the NCBI’s GenBank.

ED₅₀ estimation

The log-Logistic model properly described the data of the dose-response to the fungicides used in this experiment for each of the isolates used. An example of ED₅₀ estimation in the Ms12 isolate in response to mycelial growth using azoxystrobin is shown in Figure 1. This model has proved to be very useful for estimating ED₅₀ when studying fungicide resistance and for detecting the hormesis phenomenon in these compounds (^{Schabenberger and Pierce, 2001}; ^{Rebollar-Alviter et al., 2007}; ^{Flores and Garzon, 2013}).

Figure 1. Example of the Log-Logistic model adjustment with an asymptote greater than the relative mycelial growth using azoxystrobin in a Colletotrichum acutatum isolate representative from the strawberry producing area in Maravatío, Michoacán. 

Distribution of the sensitivity of C. acutatum isolates to thiophanate methyl

The sensitivity of C. acutatum isolates to thiophanate methyl at the mycelial stage ranged from one 0.28 to 9.72 µg mL^-1 dose of ED₅₀ with an average dose of 2.39 µg mL^-1 for the Maravatío valley, but it fit a normal distribution when the data were converted to the base-10 logarithm. The distribution of sensitivity in Zamora had a normal distribution and thus no data transformation of any type was required (Figure 2); its values ranged from 1.39 to 2.99 µg mL^-1, and were similar to the ones obtained for Maravatío. In the case of conidia, their ED₅₀ doses showed a wide distribution ranging from 0.9 to 396.4 µg mL^-1, with a mean of 20.9 µg mL^-1 for the Maravatío valley; the values for the Zamora valley were lower and ranged from 0.43 to 63.32 µg mL^-1, with an average of 10.9 and a median of 5.0 µg mL^-1 (Table 2). In this case, there were no significant differences between the sensitivity distribution in both valleys for mycelium (Pr>Ksa=0.13), but there were significant differences for conidia (Pr>Ksa=0.0004). However, considering the mode of action of thiophanate methyl, mycelial growth is the process that most precisely reflects the isolates’ sensitivity, since benzimidazoles act by joining the subunits of the β-tubulin (TUB2) protein, which affects the microtubule assembly, and therefore interferes with processes such as mitosis and cytoskeletal formation. This is directly manifested in cell division (^{Ishii, 2012}); point mutations at this site result in a lack of effectiveness in the field (^{Ma and Michailidis, 2005}).

Figure 2. Sensitivity distribution of Colletotrichum acutatum to azoxystrobin fungicide for conidia germination test. A= Maravatío valley; B= Zamora valley. Sensitivity distribution to thiophanate methyl (mycelium), C= Maravatío valley; D= Zamora valley. 

Table 2. Basic statistics and moments of ED₅₀ (µg mL^-1) values for thiophanate methyl in Colletotrichum acutatum mycelium and conidia obtained from strawberries collected in Maravatío and Zamora, Michoacán. 

^xStandard deviation.

In studies conducted by ^{Peres et al. (2004)} using C. gloesporioides and C. acutatum obtained from citrus, the authors found that in the case of benomyl (benzimidazole), C. acutamum isolates were moderately resistant, since their growth was inhibited by approximately 55 % at DE₅₀= 0.1 µg mL^-1 and 80 % at DE₅₀=1 µg mL^-1. The latter isolates showed point mutations in the TUB2 gene. ^{Chung et al. (2006)} studied different Colletotrichum species and reported that in C. acutatum isolates they estimated ED₅₀ between 1 and 10 µg mL^-1 of thiophanate methyl, indicating that they were more resistant than C. gloeosporioides. ^{Kim et al. (2007)} also reported using ED₅₀ doses between 3.8 and 8.8 µg mL^-1 in C. acutatum isolates obtained from chili in concentrations higher than 100 µg mL^-1, but found no point mutations in the TUB2 gene. The results of our study show that, based on mycelial growth, the isolates from both strawberry producing areas showed ED₅₀ no higher than 3 µg mL^-1. For this reason, it may be considered that C. acutatum populations have moderate resistance to benzimidazole fungicides such as thiophanate methyl, and more likely to benomyl, because of they share the same mode action, as previously shown (^{Peres et al., 2004}; ^{Valero et al., 2010}). These results suggest the need to implement benzimidazole resistance management strategies in the study areas given the stability found in some fungi such as Botrytis cinerea (^{Hollomon, 2015}); among these strategies are the use of fungicide mixtures or alternating fungicides of different mode of action and not using them for eradication purposes (after symptom development) (^{Brent and Derek, 2007}).

Distribution of the sensitivity of C. acutatum isolates to azoxystrobin

The distribution of the sensitivity to azoxystrobin of C. acutatum isolates in mycelium ranged between 0.04 to 0.36 µg mL^-1 ED₅₀, with a median of 0.12 µg mL^-1 in the Maravatío valley, while in the Zamora valley, the ED₅₀ values ranged from 0.07 to 0.99 µg mL^-1, with the same median value. For conidia, the estimated ED₅₀ ranged from 0.01 to 0.56 µg mL^-1 and 0.006 a 0.15, with a median of 0.05 and 0.04 µg mL^-1, for Maravatío and Zamora, respectively (Table 3; Figure 2). Converting the ED₅₀ values to base-10 logarithms satisfactorily normalized the distribution curve, except for the isolates from Zamora in the mycelium sensitivity test. The Kolmogorov-Smirnov test used to compare the two sensitivity distributions (for conidia and mycelium) did not show significant differences between the two study areas (Pr>Ks=0.58 for conidia; Pr>Ks=0.79 for mycelium).

Table 3. Basic statistics and moments of ED₅₀ (µg mL^-1) values for azoxystrobin in Colletotrichum acutatum mycelium and conidia obtained from strawberries collected in Maravatío and Zamora valleys. 

^xStandard deviation.

^{Mondal et al. (2005)} reported ED₅₀ base values of 0.09 to 0.99 with a median of 0.4 µg mL^-1 of azoxistrobin for C. acutatum mycelia obtained from citrus in Florida. Based on these results, they selected discriminatory doses of 1 ug mL^-1. Later, ^{Forcelini et al. (2016)} determined that azoxystrobin sensitivity in C. acutatum isolates obtained from strawberry in Florida was 0.30 µg mL^-1. According to the authors, isolates showing inhibition up to ED₅₀=3 µg mL^-1 were considered sensitive; those with up to 100 µg mL^-1 were considered moderately resistant, and those with sensitivity higher than 100 µg mL^-1 were considered completely resistant. Moderately resistant isolates were associated with point mutations at codon 129 of the cytochrome bc1 gene, where the aminoacid phenylalanine changed to leucine (F129L), while in the highly resistant isolates, the mutation was associated with the change from the aminoacid glycine to alanine (G143A). Based on these studies, we may conclude that all the isolates tested in this study are sensitive to azoxystrobin, since the data show no evidence of resistance in the two strawberry producing areas studied based on ED₅₀ distribution, although this active ingredient and other molecules of the same group have been in the Mexican market for over 10 years. However, given that this group is considered to be high risk because it induces resistance in fungi, it is necessary to implement resistance management strategies (^{Hollomon, 2015}) in order to avoid changes in sensitivity distribution in the short and long term. The implementation of resistance management strategies must take into account general principles that determine the development of resistant isolates (^{van den Bosch et al., 2014}).