In recent growing seasons, a severe foliar epiphytotic disease occurred in several maize hybrids in the northern coastal regions of Sinaloa during the months of November through January, including periods of relative humidity ≥ 90% for 13-17 h, and an average daily temperature between 18 and 26 °C. The symptoms of the disease developed initially in the lower leaves causing oval, long, elliptical, grayish-green or tan necrotic lesions of different sizes. Later, symptoms spread to the middle and upper leaves of the plants. Although yield losses are yet to be estimated, personal observations by the author indicate the disease has destroyed up to 65% of the foliage of some commercial maize hybrids grown in the northern coastal region of the state. Although maize leaf blight caused by Exserohilum turcicum (Et) is one of the most important foliar diseases in maize grown in Mexico, at the present time, there is not enough scientific evidence on the etiology of the disease, although maize growers and field advisors have shown interest in determining its etiology. Consequently, the main goal of the present study was to confirm the identity of the causal agent of maize leaf blight at the species level, based on morphometric and molecular data of the anamorph.

Fifteen symptomatic samples were collected from the same number of maize fields in the municipalities of Ahome and El Fuerte, Sinaloa, from December 16, 2013 through January 15, 2014. One sample from each field consisted in five symptomatic leaves collected from the corners and center of each field (n=25). Fragments of symptomatic leaves (5-6 cm long) were disinfested in a 0.5% sodium hypochlorite solution for 2 min, washed for 3 min in sterile distilled water and dried on Whatman No. 1 filter paper. Fragments were placed in a moist chamber, for 48 h at 25 °C. From conidia formed on the leaf tissue, monosporic cultures were obtained through direct transference of them with a dissecting needle to Petri plates with potato dextrose agar (PDA), under a dissecting microscope.

Mycelial growth of eight fungal monoconidial isolates were evaluated on the following culture media: potato dextrose agar (PDA), Czapek-Dox agar (CDA), glucose peptone agar (GPA) and Richard´s synthetic agar (RSA), and maize leaf extract agar (CLEA) (^{Dhingra and Sinclair, 1985}). After eight days of incubation, under 14-h light/10-h dark regime at 25 °C, the surface color of the fungal colonies in the different culture media were in the olivaceous and dark grayish-olive colors.

All isolates presented light olive pigmentation on the back of the colony in CDA and RSA, but no pigmentation was observed in other media. The radial mycelial growth rate of the isolates on CLEA ranged from 3.7 to 4.9 mm/day and was significantly (P≤0.05) greater than the growth on the other media. There was a significant interaction (P<0.0001) between the medium and the isolate on radial mycelial growth, indicating that growth rates of the isolates were medium-dependent. On PDA, isolates produced between 1-3 conidia at the tip of the conidiophore (Figure 1). Conidia were curved, spindle and elongated, with an average of five septa, and ranged 44 - 110 µm in length (85.3 µm average) and 10 - 21 µm in width (14.8 µm average). No differences in morphology were observed among the Et isolates obtained from symptomatic maize leaves, and the characteristics were consistent with previous reports (^{Tang et al., 2015}; ^{Shi et al., 2017}).

Figure 1 Phylogenetic analysis of isolates associated to maize leaf blight from Sinaloa. A) Maximum-likelihood tree (Log likelihood= -1276.64) based on the internal transcribed spacer (ITS) of four Exserohilum isolates (Setosphaeria refers to the teleomorph of Exserohilum). The tree was constructed with Mega 6.06 (bootstrap = 1000), using the Jukes-Cantor (JC) substitution model with invariant sites (+I). The isolates characterized in this study are in boldface. The corresponding sequence of Pleospora tomatonis was used as an out group. Et =E. turcicum. Database accession numbers of the sequences precedes the scientific names of organisms. Bootstrap values are shown as percentages. The scale bar indicates the expected number of amino acid substitutions. Morphology of E. turcicum from maize (isolate Et-2). B) Elliptical- shaped lesions of maize leaf blight on maize hybrid DK-300 artificially inoculated with E. turcicum in the greenhouse. C) Conidiophore single straight or bent bearing one to three conidia. D) Conidium club-shaped with basal cell swollen at the point of attachment. E) Scanning electron micrograph of a smooth conidium showing a protruding hilum in 8-day culture on PDA. 

For the observation of conidia under the electron microscope, fragments of 3.5 cm² of an 8-days-old PDA culture of a maize blight fungal isolate (Et3) were cut and immersed in a fixing solution (FAA: 10% formaldehyde, 5% acetic acid, 50% dilution of ethanol 96%, and 35% water). After 48 h, samples were washed with tap water for 10 min and immediately dehydrated by increasing concentrations of ethanol. Samples were dehydrated to a critical point with CO₂ in a desiccant glass BAL-TEC CPD030. Samples were mounted in aluminum sample holders over a conductive carbon strand, and covered with a fine gold layer in an ionizer Denton Vacuum Desk II (Hofstra Group, Ltd. Co; Boston MA, USA). Finally, electron-micrographs of conidia were taken under a scanning electron microscope JEOL JSM-53110 LV (^{Bozzola and Russell, 1999}). Following these procedures, images revealed smooth conidia with a protruding hilum (Figure 1).

For the pathogenicity tests, eight monoconidial fungal isolates were grown on PDA at 24 °C in darkness for 9 days. Three mL of sterile distilled water were added to the colonies and the mycelium with conidia were dislodged from the medium using an inoculation loop; 0.2 mL of the conidial suspensions were evenly distributed over a Petri dish 90 mm in diameter containing PDA medium, and incubated at the same temperature. The inoculated dishes were sealed with sealing film (Parafilm^®) and placed upside down at 24 °C in the dark for 9 days. Conidia were scraped from the surface of the medium using a sterile spatula and the conidial suspension was filtered through two layers of sterile cheese cloth to remove mycelium. Eight monoconidial fungal isolates associated with maize leaf blight were tested on the maize commercial hybrid DK-3000, which showed high incidence of the disease in the previous growing season. Each conidial suspension was adjusted to a concentration of 2 x 10⁴ conidia/mL with sterile distilled water and sprayed on four maize plants at the VS5-6 developmental stage. Control plants were sprayed with sterile distilled water. After inoculation, plants were placed in black plastic bags to ensure 100% relative humidity (RH) for 48 h and incubated at 12-15 °C. Inoculated and control plants were arranged in a completely randomized block design with four replications (4 pots 25 x 20 cm, each with one plant). Afterwards, plants were subjected to 100% RH for five consecutive days for 12 h daily. The pathogenicity of the isolates was determined 14 days after inoculation, considering the percentage of foliage area diseased (FAD) of plants in the various replications. The experiment was conducted twice in a greenhouse. Greenhouse temperature ranged 8 - 26 °C and 12 - 30 °C in the first and second experiment, respectively. In order to standardize data, percentages of FAD were arcsin transformed. The transformed data were analyzed using ANOVA and the mean separation was achieved following Tukey’s test (^{Little and Hills, 1973}). To fulfill Koch’s postulates, by the end of the experimental periods, the fungus was isolated from inoculated plants and its identity confirmed using morphological identification of twenty conidia per isolate. All eight isolates of Et tested were pathogenic on the maize hybrid DK-3000. Variation in virulence was detected among the isolates. Ten days after inoculation, maize leaves showed typical lesions and blight symptoms similar to those observed under field conditions (Figure 1). FAD in inoculated plants varied from 9.5 to 54.1% and from 9.0 to 26.9%, with significant differences (P=0.05) among isolates, in the first and second experiment, respectively. Control plants sprayed with sterile distilled water remained asymptomatic during the study.

For molecular characterization, the ITS rDNA region was amplified using the universal primers ITS1 (5’ TCCGTAGGTGAACCTGCGG 3’) and ITS4 (5’ TCCTCCGCTTATTGATATGC 3’). PCR was performed in a 25 μL volume containing 1 ng DNA template, 1.5 mM MgCl₂, 0.5 mM of each dNTP, 0.4 μM of forward and reverse primers, and 1 U of Taq DNA polymerase (Invitrogen, Brazil, Cat. No. 11615-050). The thermocycler was programmed for initial denaturation at 95 °C for 4 min, followed by 35 cycles of denaturation at 94 °C, annealing at 54 °C, extension at 72 °C, each for 1 min and a final extension at 72 °C for 5 min. PCR products were separated by agarose gel electrophoresis (1% w/v in 0.5 X TAE) and visualized by ethidium bromide staining. PCR products were purified using the QIAquick PCR purification Kit (Qiagen, Cat. No. 28106) and quantified using a Nanodrop 2000. PCR products were sequenced in both directions with an ABI 3730XL sequencer (Applied Biosystems, USA). Sequences were edited in CHROMAS Pro 1.6 (Technelysium Pty Ltd, South Brisbane, Queensland, Australia) and compared to sequences in the NCBI (National Center for Biotechnology Information) using the BLAST-N software and the Megablast algorithm. All sequences were deposited in GenBank under the following accession numbers KT253948-KT253952. MEGA 6.06 was used for alignment and phylogenetic analysis. Sequences were aligned together with reference sequences of Exherohilum/Setosphaeria species (^{Tang et al., 2015}), using the MUSCLE alignment program. Multiple alignments were subjected to a DNA substitution model analysis in MEGA, to select the model that best fits the data. A phylogenetic tree was constructed using the Jukes-Cantor (JC) model and the maximum likelihood (ML) method. The rate variation among sites was modeled by a proportion of invariant sites (I). Tree topology support was assessed by 1 000 bootstrap replicates.

Sequences from S. turcica (teleomorph of Et) and from different species of Setosphaeria/Exserohilum (E. oryzicola, S. monoceras, S. pedicellata) were used to conduct the phylogenetic analysis. S. turcica sequences included isolates obtained from maize (HF934950, KF278460; reference culture sequence, ATCC64835), some of these identified as formae specialis zeae (all KJ Genbank numbers except for KJ922752) and from sorghum identified as formae specialis sorghi (KJ922752) (^{Tang et al., 2015}). The maximum likelihood phylogenetic tree generated with the ITS rDNA sequences showed that the isolates from Sinaloa are closely related to S. turcica sequences obtained from maize and sorghum (^{Tang et al., 2015}). These sequences are clearly separated from other Exherohilum/Setosphaeria species (Figure 1). Isolates Et-6 and Et-7 were different from the rest of Et sequences included, but still grouped within S. turcica.

Cultural and morphometric characteristics of conidia, colony morphology, and sequencing of the ITS rDNA region of the anamorph indicated that E. turcicum [(Pass.) (Leonard and Suggs)] (syn. Helminthosporium turcicum Pass.) is the causal agent of the maize leaf blight in Sinaloa, as reported in other maize growing areas in the world (^{Shi et al., 2017}). The analysis performed does not allow to discriminate at the formae specialis level. Recent studies have demonstrated the utility of universally primed polymerase chain reaction (UP-PCR) technology to characterize the genetic diversity and phylogenetic relationships among the formae specialis of Et (^{Tang et al., 2015}). In future, this approach would be useful in the process of discriminating the existence of formae specialis in Sinaloa.

The severity of maize leaf blight was assessed in 115 commercial fields ranging from tasseling through kernel dough stage in the municipalities of Ahome and El Fuerte, Sinaloa in the 2013-2014 growing season. The fields inspected ranged from 10 to 50 ha. Each field was sampled in the 4 corners and the center for disease severity; ten plants were taken from each sampling point and one leaf was taken randomly from the plants totaling 150 leaves per field. The average percentage of foliar area diseased (FAD) was determined using a scale ranging 0-4, where 0=No detectable lesion, 1= a few lesions on the leaves (≤ 5%), 2= several small and large lesions on many leaves (5.1-10%), 3=lesions on many leaves (10.1-15%), 4=many enlarged lesions (15.1-20%) Ten symptomatic leaves from each field were placed in an ice chest (8-10 °C) and taken to the laboratory to determine whether E. turcicum was associated to the disease. The presence of the fungus on symptomatic tissue was confirmed by the observation of conidia under the stereo and compound microscopes.

The disease severity, reflected as percent FAD, varied according to hybrids and locations. Hybrid P3254W showed the highest disease severity, particularly in the sectors of Despensa-Bolsa de Tosalibampo No. 2, Olas Altas-Bachomobampo, San Isidro-Grullas Derecha, Ahome-Grullas Izquierda and Concheros-9 de Diciembre, with percentage FAD of 17.3, 13.3, 9.0, 7.6 and 3.5%, respectively. The same hybrid presented a reduced disease severity in sectors Aguila Azteca-El Guayabo, El Fuerte, 5 de Mayo-Sufragio, Santa Rosa-Los Tercos with AFA of 3.2, 0.8, 0.4, 0.0 %, respectively. Hybrids XR56 and Máximo exhibited AFA of 11.0% in sector Olas Altas-Bachomobampo. Where hybrid P3254W presented highest disease severity. The hybrids DK 2030, Sultán, Gorila, DK.2038, N1R01, P3254W, XR47, Garañón, MN1078, Caribú, DK3000, Máximo, DK 2036, Máximo and DS2301 exhibited an average percentage FAD ranging from 0.0 to 4.6% of AFA. The accumulated daily episodes, in hours, with relative humidity (RH) ≥90% in sector Despensa-Bolsa de Tosalibampo No2 from December 16, 2013, through January 01, 2014, where the hybrid P3254W showed the highest severity of foliar blight, reached 317 hr; , while in sector 5, de Mayo-Sufragio such time periods were of 184 hr and the same hybrid showed 0.4% FAD. In general, a gradual decrement of maize blight was observed, regardless of the hybrid, as distance from the shoreline and elevation increased s, from 4 to 100 km and from 0 to 80 masl, respectively.