**2.3 Genotypic characterization**

Every isolate was streak-plated over PDA and incubated for 4 days under the conditions described above. Following incubation, 5 mL sterile water was added and the surface of the colony was rubbed with a wire loop in order to free the fungal elements. The resulting suspension was added to 100 mL of culture medium for *Colletotrichum* (Martinez Culebras et al., 2000), and incubated 48 h in the dark at 28-30 ºC in orbital motion (Orbital Shaker, Forma Scientific, Inc.) at 180 rpm. The mycelium was placed on a nylon filter (200 µm pore diameter) and dried with absorbent paper until all moisture was removed. Once dry, it was extended forming a layer as thin as possible and was dried in stove at 37 + 0.2 ºC until constant weight. Then it was ground to thin powder in a mortar.

The same procedure was applied to two strains belonging to the NITE Biological Resource Center (Japan) collection: *C. kikuchii* NBRC 6711 and *C. sojina* NBRC 6715.

In order to extract total DNA, Di Conza et al. (2007) protocol was followed. DNA was quantified through absorbance reading in a spectrophotometer, and its quality determined by means of electrophoresis in 0.8% w/v agarose gel (Sambrook et al., 1989). RAPDs were carried out following Williams et al. (1990), using 20 oligonucleotides (FAGOS/Ruralex, Argentina) (Table 3).

The reaction mixture was prepared for a 50 L total volume with these components: 2.5 mM magnesium chloride, 125 M of each dNTPs (INBIOHIGHWAY), 1 M oligonucleotide, 5 U *taq* DNA polymerase (INBIOHIGHWAY) and 20 ng DNA per reaction. Amplification was carried out using a MJ Research Thermal Cycler under the following conditions: 1 cycle of 5 min at 95ºC, 40 cycles comprising 75 s at 94ºC, 90 s at 36ºC and 150 s at 72ºC, and 1 final cycle of 10 min at 72ºC. Amplification products were separated by electrophoresis in 1.5% w/v agarose gel with 0.5X TBE 0.089M Tris-borate, 0.002 M EDTA. The run time was 180 min, with a 100V constant voltage. The gel was stained with ethidium bromide, and the molecular weight marker was 100-bp DNA Ladder (Promega). Band profiles obtained were photographed and analyzed with Gel Doc XR System (BIORAD-Life Science Cat. # 170- 8170) using the Quantity One Software.

All RAPD reactions were done in duplicate.

A matrix with 0 and 1 corresponding to absence and presence of band, respectively, for each one of the images resultant from each oligonucleotide used, was constructed, giving a total of 18 binary matrices with order mx21, where 18 represents the amount of nucleotides considered, m the bands obtained and 21 the total fungi analyzed. From these matrices, distance matrices (21x21) were obtained with the Jaccard coefficient, designed for asymmetric binary variables (Everitt & Hothorn, 2009; Johnson & Wichern, 1998) (Table 4).

Then, an average distance matrix (21x21), essential to apply the cluster technique, was calculated with the 18 distance matrices (Abonyi & Feil, 2007; Bolshakova & Azuaje, 2003; Peña, 2002).

The choice of clustering algorithm depended on the type of data and the purpose of using the technique. It was considered convenient to apply more than one clustering algorithm for a database and then compare the agreement between the results.

Therefore, a partition (FANNY) and a hierarchical (AGNES) algorithm were selected. FANNY computes "diffuse" clusters, giving each item a degree of belonging to the cluster. The algorithm combines the optimum choice of the amount of clusters (k) and the parameter r, linked to the degree of diffusivity or *membership exponent*.

Every isolate was streak-plated over PDA and incubated for 4 days under the conditions described above. Following incubation, 5 mL sterile water was added and the surface of the colony was rubbed with a wire loop in order to free the fungal elements. The resulting suspension was added to 100 mL of culture medium for *Colletotrichum* (Martinez Culebras et al., 2000), and incubated 48 h in the dark at 28-30 ºC in orbital motion (Orbital Shaker, Forma Scientific, Inc.) at 180 rpm. The mycelium was placed on a nylon filter (200 µm pore diameter) and dried with absorbent paper until all moisture was removed. Once dry, it was extended forming a layer as thin as possible and was dried in stove at 37 + 0.2 ºC until

The same procedure was applied to two strains belonging to the NITE Biological Resource

In order to extract total DNA, Di Conza et al. (2007) protocol was followed. DNA was quantified through absorbance reading in a spectrophotometer, and its quality determined by means of electrophoresis in 0.8% w/v agarose gel (Sambrook et al., 1989). RAPDs were carried out following Williams et al. (1990), using 20 oligonucleotides (FAGOS/Ruralex,

The reaction mixture was prepared for a 50 L total volume with these components: 2.5 mM magnesium chloride, 125 M of each dNTPs (INBIOHIGHWAY), 1 M oligonucleotide, 5 U *taq* DNA polymerase (INBIOHIGHWAY) and 20 ng DNA per reaction. Amplification was carried out using a MJ Research Thermal Cycler under the following conditions: 1 cycle of 5 min at 95ºC, 40 cycles comprising 75 s at 94ºC, 90 s at 36ºC and 150 s at 72ºC, and 1 final cycle of 10 min at 72ºC. Amplification products were separated by electrophoresis in 1.5% w/v agarose gel with 0.5X TBE 0.089M Tris-borate, 0.002 M EDTA. The run time was 180 min, with a 100V constant voltage. The gel was stained with ethidium bromide, and the molecular weight marker was 100-bp DNA Ladder (Promega). Band profiles obtained were photographed and analyzed with Gel Doc XR System (BIORAD-Life Science Cat. # 170-

A matrix with 0 and 1 corresponding to absence and presence of band, respectively, for each one of the images resultant from each oligonucleotide used, was constructed, giving a total of 18 binary matrices with order mx21, where 18 represents the amount of nucleotides considered, m the bands obtained and 21 the total fungi analyzed. From these matrices, distance matrices (21x21) were obtained with the Jaccard coefficient, designed for asymmetric binary variables (Everitt & Hothorn, 2009; Johnson & Wichern, 1998) (Table 4). Then, an average distance matrix (21x21), essential to apply the cluster technique, was calculated with the 18 distance matrices (Abonyi & Feil, 2007; Bolshakova & Azuaje, 2003;

The choice of clustering algorithm depended on the type of data and the purpose of using the technique. It was considered convenient to apply more than one clustering algorithm for

Therefore, a partition (FANNY) and a hierarchical (AGNES) algorithm were selected. FANNY computes "diffuse" clusters, giving each item a degree of belonging to the cluster. The algorithm combines the optimum choice of the amount of clusters (k) and the parameter

a database and then compare the agreement between the results.

r, linked to the degree of diffusivity or *membership exponent*.

constant weight. Then it was ground to thin powder in a mortar.

Center (Japan) collection: *C. kikuchii* NBRC 6711 and *C. sojina* NBRC 6715.

**2.3 Genotypic characterization** 

Argentina) (Table 3).

Peña, 2002).

8170) using the Quantity One Software. All RAPD reactions were done in duplicate. AGNES makes agglomerative hierarchical clusters, the distance between clusters (k) being analyzed by Ward method.

Silhouette Coefficient (SC) and Agglomeration Coefficient (AC) were used for FANNY and AGNES, respectively, to evaluate cluster quality (Kaufman & Rousseeuw, 1990). Data were processed with R version 2.10.1 Software (Torgo, 2003).


Table 3. Oligonucleotides used in RAPD reaction


Table 4. Contingency table summarizing the coincidences in the presence/absence of bands between pairs of fungi. Letters i and j: correspond to the same fungus or to different fungi; 1: presence of bands; 0: absence of bands; a, b, c and d are the frequencies1.

*b c <sup>J</sup> abc* . No 0-0 matches in numerator or denominator.

 1 coefficient of distance is given by:

Phenotypic and Genotypic Variability in

Fernández et al., 1991).

**Oligonucleotide** 

selected for RAPD

**3.2 Genotypic characterization** 

and OPA-06 were not satisfactory.

by Ellis (1971), Crous & Braun (2003) and Solheim (1929).

Figure 5 shows band profiles obtained with oligonucleotide OPA-14.

**Maximum number of bands** 

*Cercospora kikuchii* Isolates from Santa Fe Province, Argentina 103

Similar morphologic structures were observed in the 19 isolates. The direct examination of the injury showed brown pigmented, fasciculated and septate conidiophores (200-300 µm long per 4-5 µm wide), with simpodial growth and conidiogenous cells integrated, terminal or intercalary, with thickened and darkened conidiogenous loci (Figure 3). Single, long, acicular, hyaline, pluriseptate (12 to 26 septa) conidia with truncate bases and subacute apices (170 to 190 µm long per 3 µm wide at the base) were observed (Figure 4) as described

Conidia formation process involves internal and external walls of the conidiogenous cells, so when the conidio arises, a scar appears in its origin, on the conidiogenous cell (Cai, 2004;

A total of 90 bands were obtained, 100% of them were polymorphic. Amplification size ranged between 107 bp and 2750 bp, an average of 51 amplified fragments/oligonucleotide being produced. OPA-01 was the oligonucleotide which produced the least amount of bands (1) and OPA 12 the most (23) amplified fragments (Table 5). Amplifications with OPA-02

> **Minimum number of bands**

Total 926 926

Table 5. Total number of bands and polymorphic bands obtained with the oligonucleotides

A FANNY cluster analysis with k= 6 and r =1.2 (optimum combination) was carried out. Few distances near 0.2 were observed (Fig 6). C21 and C22 (isolated from Margarita and

OPA-01 19 1 49 49 1846-168 OPA-03 17 9 49 49 1437-202 OPA-04 16 6 46 46 1451-156 OPA-05 19 7 48 48 1485-107 OPA-07 14 5 49 49 1599-197 OPA-08 16 3 58 58 2009-253 OPA-09 14 7 53 53 1561-134 OPA-10 18 4 56 56 1797-114 OPA-11 13 5 45 45 1422-212 OPA-12 23 7 55 55 1883-200 OPA-13 17 8 51 51 1495-156 OPA-14 15 6 44 44 1999-309 OPA-15 16 7 53 53 2730-241 OPA-16 18 8 50 50 2252-215 OPA-17 16 4 60 60 2673-213 OPA-18 18 9 51 51 1655-229 OPA-19 18 9 58 58 2750-208 OPA-20 16 7 51 51 1572-242

**Total bands** 

**Number of polymorphic bands** 

**Pair of bases (bp)** 
