Wheat blast or brusone caused by Magnaporthe oryzae pathotype Triticum (MoT) is one of the most serious problems for wheat production in the tropical/subtropical region of South America. The disease appeared for the first time in northern Paraná, Brazil in 1985 (^{Igarashi et al., 1986}) and was later detected in Paraguay in 1989 (^{Cunfer et al., 1993}). However, the first blast epidemic was reported in 2002 and caused production losses greater than 70% in early-sown fields (^{Viedma and Morel, 2002}). The environmental conditions that favor blast epidemics are temperatures between 18 and 25°C and high relative humidity during heading and flowering, usually during the El Niño phenomenon (^{Kohli et al., 2011}; ^{Cunha et al., 2017}).

Given that wheat blast is a disease that is difficult to control, its management must include the use of resistant wheat varieties along with chemical control practices to reduce production losses. However, it has been difficult to develop resistant varieties due to the low number of available resistance sources. Some varieties derived from CIMMYT’s Milan line have shown high levels of resistance to wheat blast (^{Kohli et al., 2011}). Recently, (^{Cruz et al., 2016}) determined that the 2NS translocation, derived from Triticum ventricosum, is responsible for conferring resistance to Milan and other derived varieties. For this reason, identifying different resistance sources and characterizing the reaction of new genotypes are considered of vital importance. This study was conducted to evaluate the reaction of Paraguayan wheat varieties to Magnaporthe oryzae infection, as well as that of several advanced lines developed in the country and the main foreign varieties currently sown in Paraguay.

The trial was conducted in the greenhouse of the Centro de Investigación Hernando Bertoni, Instituto Paraguayo de Tecnología Agraria, (IPTA) in Caacupé, Paraguay (south latitude 25°23’25.314”, west longitude 57°11’27.848”). The trial included 32 wheat varieties and four advanced lines provided by the IPTA’s Programa Nacional de Investigación de Trigo (Table 1), all of them currently sown in Paraguay. The varieties were sown on 12 April 2016 in 15x25 cm plastic pots containing a substratum made up of soil and leaf mulch at a 3:1 ratio. Two plants per pot were sown for each fungus isolate, and four pots were planted for each variety. Each pot was considered a replication. The greenhouse was kept at a temperature of 15 ± 2°C, and the agronomic management consisted of one 15-15-15 fertilizer application at a dose of 5 g per pot 15 days after emergence, plus one application of urea 45 days after emergence using the same amount. Irrigation was applied twice a week, and the plants were monitored on a weekly basis to detect pests. Magnaporthe oryzae pathotype Triticum (MoT) isolates used in the study were taken from infected wheat spikes collected in the field, as follows: P13-009 was collected in Capitán Miranda, Itapúa, in the 2013 cycle; P14-025, in Yhovy, Canindeyú, in the 2014 cycle; and P14-039, in La Flor farm, Alto Paraná, in 2014. These isolates were selected based on their virulence in previous trials and on their places of origin, which represent the main wheat production areas in Paraguay. The isolates were kept on filter paper at -18 °C as part of the isolates collection of the Pyricularia in Wheat Project. They were replicated in Petri dishes containing oatmeal flour agar culture medium and incubated for 10 days at 25°C and a 12-h photoperiod. The mycelium was later crushed using an L-shaped glass rod, and the dishes were exposed to continuous fluorescent light for three days to favor sporulation. Then, the spores were removed using a paint brush and sterile distilled water (^{Marangoni et al., 2013}). The conidia concentration was adjusted with a Neubauer hemacytometer at 5x10⁴ conidia ml^-1 by adding sterile distilled water to dilute it (^{Chávez et al., 2015}). Inoculation was performed when each variety reached stage 59 on Zadok’s scale, this is, when the spikes were completely out of the flag leaf sheath. Spikes were sprayed using a manual sprayer (three sprayings per spike). After the spikes were sprayed, the plants were kept in darkness for 24 h, at 80±5% moisture and 27±2°C temperature in a heated room in the greenhouse to favor infection. After that period, the plants were removed from the heated room and kept in the greenhouse, at the same temperature, 60 ± 5% moisture and 12-h photoperiod. Symptoms were observed and evaluated 15 days after inoculation. The scale proposed by ^{Tagle et al. (2014)} was adapted to evaluate infection severity on the spikes. This scale classifies symptoms as follows: 0 = no infection; 1 = small lesions, < 1.5 mm; 2 = intermediate lesions, < 3 mm; 3 = mixture of green and white glumes, without apparent necrosis, caused by a hypersensitive reaction; 4 = complete spike necrosis. Data were analyzed using Kruskal-Wallis’ test and the INFOSTAT (version 2016e) statistical program. Based on the infection means, on which the test is based, the genetic materials were classified as Resistant (0-1), Moderately resistant (1.1-2), Moderately susceptible (2.1-3) and Susceptible (3.1-4).

Significant statistical differences were found between the varieties and their interaction with the inoculated isolates (Table 2). For the P13-009 isolate, two varieties (Canindé 1, CD 116) were classified as resistant; three were moderately resistant (TBIO Toruk, TBIO Iguazu, TBIO Sintonía), and the rest were moderately susceptible and susceptible. For the P14-025 and P14-039 isolates only one variety was classified as resistant (Canindé 1), and two were moderately resistant (CD 116 and TBIO Sintonía). The rest were classified as moderately susceptible and susceptible. As for the Paraguayan varieties, Canindé 1 was the only one classified as resistant to the three inoculated isolates, while Itapúa 75 that was inoculated with isolate P14-025, and Canindé 3, with isolates P13-009 and P14-039, were moderately susceptible. The remaining varieties, as well as the four advanced lines evaluated, were susceptible. The interaction among varieties and the fungus isolates used is limited to intermediate cases, in which a variety classified as moderately resistant to an isolate shows a moderately susceptible reaction to another isolate, and vice versa (TBIO Iguazú, TBIO Toruk: MR P13-009; MS P14-025). This kind of interaction among wheat varieties and Magnaporthe isolates shows the specific reactions between the host and the pathogen that must be thoroughly studied in the future.

It should be noted that Canindé 1 was resistant to the three pathogen isolates, while CD 116, its sister line derived from the same genealogy, was resistant to one of the isolates (P13-009) and moderately resistant to the other two (P14-025 and P14-039). Considering that both varieties are derived from the Milan line, which contains the 2NS segment that confers resistance to the disease (^{Kohli et al., 2011}; ^{Cruz et al., 2016}), the difference in their reactions is due to their genetic background, which can vary during the process of selecting a progeny. It is also important to highlight the moderate resistance of the TBIO Sintonía variety to the three isolates used. The susceptible and/or moderately susceptible reaction of most wheat varieties sown in Paraguay could be a warning that strong epidemics could occur under favorable conditions. Evaluation under controlled conditions is valuable as a reference for genetic improvement activities aimed at selecting resistant materials to wheat blast mildew, but it is possible that the varieties may react differently under field conditions. The difference in the varieties’ performance under field conditions and when inoculated in the greenhouse has already been observed by ^{Igarashi (1990)} and ^{Urashima and Kato (1994)} and is attributed to the pathogenic variability of the fungus (^{Urashima et al., 2004}). In Paraguay, ^{Kohli et al. (2012)} reported the reaction of varieties Canindé 3, Canindé 11, Canindé 12, Canindé 13, Itapúa 40 and Itapúa 70 as moderately susceptible to susceptible in the field, while Canindé 1 and Itapúa 75 were moderately resistant. Except for Itapúa 75, a long-cycle variety, the reactions observed in the field coincided with those observed in this study under controlled greenhouse conditions. In Brazil, ^{Fronza et al. (2016)} observed that varieties QUARTZO, FUNDACEP RAICES, BRS 220 and IPR 85 are moderately resistant under field conditions. In the present study, these varieties were classified as moderately susceptible and susceptible to the three isolates. However, both studies agree on the resistance of variety CD 116. Considering that environmental conditions play an important role in the expression of the disease in the field, it is necessary to evaluate the resistance of genetic materials while controlling the spectrum of environmental conditions and widening the pathogen’s variability under controlled conditions.

This study shows the susceptibility of most of the wheat varieties sown in Paraguay, both local and foreign, which means that they would be vulnerable to severe wheat blast epidemics under favorable conditions. It also shows the first evidence of the interactions that may exist between wheat varieties and different isolates of the Magnaporthe oryzae pathotype Triticum fungus. For this reason, it is urgent to identify and introduce new resistance sources in the improvement program in order to increase the number of resistant varieties in the future.