Materials and methods

Trials were carried out in the Hernando Bertoni Research Center, Paraguayan Institute of Agricultural Technology, (CIHB-IPTA), Caacupé, Paraguay.

The isolates used, from the wheat spikes, were morphologically identified following ^{Klaubauf et al., (2014}) and confirmed with the specific molecular marker MoT3 (^{Pieck et al., 2016}). Previous evaluations in materials of known reactions of susceptibility and resistance (Caninde 11 and Milan) confirm their capability for infection.

Inoculum were obtained following the method suggested by ^{Marangoni et al. (2013}). The pieces of filter paper, 0,5 cm in diameter with preserved isolates, were planted in Petri dishes with Oat meal Agar (OA), and incubated for 12 days at 25 °C with light periods of 12 hours. Later, in order to force sporulation, the mycelia were crushed on the dish and incubated under constant light for 72 hours.

Forced inoculations were carried out on a minimum of 6 and a maximum of 10 spikes per genotype in the phenological stage of anthesis, using a 1/5 HP 58 PSI Oilless Airbrush Compressor Kit, at a concentration of 5.10⁴conidia.mL^-1. Approximately 0,3 mL of inoculum was used for each spike. The inoculated plants were kept under controlled temperature (28±2 °C) and humidity conditions (85±5 %) for a period of 72 hours for both trials.

SHWs are materials with a broad variability in terms of their reactions to different diseases. Considering differential reactions in P. oryzae, even between old and new isolates reported by ^{Cruz et al., (2016}), the maximum value of reaction (MVR) was used to classify the SHWs and guarantee the reproducibility of the test. The option of pondering the material according to the MVR over the total of spikes evaluated helps discard biases due to fortuitous escapes to the infection (that is, susceptible plants with no reaction). This classification criterion is robust for the selection of materials according to their response to the infection, since only the genotypes with a low MVR display resistance.

Two independent evaluations were carried out on synthetic wheats. In a first evaluation, 64 genotypes of SHWs (Table 1), provided by the International Maize and Wheat Improvement Center (CIMMYT), were inoculated with a monosporic isolate (P14ATae039) of the fungus P. Oryzae.

In a second evaluation, based on the maximum values of reaction (MVR) for wheat blast on the spike, 18 materials were selected for a second evaluation. The chosen genotypes displayed MVR between resistant and moderately resistant, and as a form of susceptibility control, other susceptible reaction materials were included. In order to cover a broader range of pathogenic susceptibility, two virulent isolates, P14ATae039 and P14YTae031, were used in the second inoculation.

The reaction to the disease was evaluated 15 days after inoculation (DAI), using the scale proposed by ^{Chavez et al. (2017}), modified for the MVR as follows: 0 = No infection, 1 = Up to 10% of the spike necrotized (resistant), 2 = Up to 40% of the spike necrotized (moderately resistant), 3 = Up to 60% of the spike necrotized (moderately susceptible), and 4 = 100% of the spike necrotized (susceptible). After this period, the spikes of seven genotypes that displayed resistance or moderate resistance in the first reading, along with a susceptible control, were kept in the conditions described above, to observe the development of the spike infection for both isolates for a 30-day period. For the progress of the disease, we used a totally randomized design; each spike represents an experimental unit in which the reaction was observed at 8, 15, 22 and 30 DAI. The calculation of the value of the area under the disease progress curve was carried out using the function AUDPC, implemented in the package agricolae for R version 3.4.2. Analyses of the AUDPC underwent an ANOVA and the comparisons between averages were performed by Test: LSD Fisher Alfa=0,05.

Results and discussion

The data of the first infection and classification of the 64 SHWs based on their maximum values of reaction to an isolate (P14ATae039) are shown in Table 2.

Based on the first evaluation of synthetic wheats, over 10% of the SHWs were classified as Resistant (MVR 0-1), 34% were classified as moderately resistant or moderately susceptible (MVR between 2 and 3), and over 55% as susceptible (MVR 4).

Like in ^{Gul Kazi et al. (2012}) and ^{Das et al. (2015}), the reactions to fungal diseases using characteristics have been differentiated in this study. The correct phenotyping of these materials is crucial for the selection of the SHWs that can be used as parents in the breeding programs and/or in different studies of molecular characterization and genic expression, differential to the interaction of wheat and P. oryzae.

The MVR of the preselected 18 SHWs, with two isolates (P14ATae039 and P14YTae031), is graphed in Figure 1.

The data shown in Figure 1 indicates that only one of the genotypes (TS29) was resistant to both isolates; seven were classified as intermediates (TS77, TS53, TS101, TS78, TS32, TS22 and TS63), since they presented the resistant to moderately resistant reaction to one of the isolates, but moderately resistant to moderately susceptible for the other. This interaction trait was reported by ^{Chávez and Kohli (2018}), where moderately resistant varieties for one isolates showed a moderately susceptible for the other. This type of reaction specificity may probably be explained by the genetic flow between isolates from different hosts, which gives the isolates a virulence variation (^{Gladieux et al., 2018}).

The data displayed in Figure 2 shows the differences in the average progress of the disease amongst the genotypes identified as resistant or moderately resistant. Only three of the genotypes, TS49, TS53 and TS77, kept their resistance or moderate resistance for the 30-day period. The materials TS29, TS73 and TS101 changed their reaction from moderate resistance to moderate susceptibility, while the synthetic TS78 became susceptible at the end of the 30 days. The SHW TS105, used as a control, was susceptible throughout its development. Considering the differential reactions presented by the materials to the different isolates, the progress of the disease was analyzed in a second instance and is shown in Figures 3a and 3b. The results show a significant interaction between the synthetic wheats evaluated and the two isolates of P. oryzae, which confirms the results obtained by ^{Cruz et al., (2016}), in which the Overland and RonL cultivars were resistant to an isolate, but susceptible to the other.

Figure 1 Maximum values of reaction of 18 SHWs to the forced infections of two isolates of P. oryzae. Reactions to the isolate P14ATae039 (P039) and reactions to the isolate P14YTae031 (P031). Caacupé, 2018. 

Wheats viz. TS29 and TS49 were relatively stable in maintaining their resistance or moderate resistance, and other viz. TS73, TS78 and TS105 were moderately susceptible or susceptible to both isolates. However, the remaining SHWs viz. TSHs TS53, TS77 and TS101, have a significantly different reaction between the two isolates (Figure 1).

Synthetic wheat TS101 was resistant to isolate P031, yet susceptible to P039; wheats TS53 and TS77 were between moderately susceptible and susceptible to P031, yet resistant to P039 (Figure 3).

This result is consistent with those obtained by ^{Chávez et al. (2018}), who observed an interaction of the varieties with different isolates of P. oryzae. This interaction of genotypes with different isolates extends to a greater number of isolates under study (data not reported here).

Figure 2 The average disease progress at 8, 15, 22 and 30 DAI in eight selected synthetic hexaploid wheat genotypes. Caacupé, 2018. 

Figure 3 Evolution of the reaction to the infection and its development against the isolate P14YTae031 (P031) (a) and P14ATae039 (P039) (b) in eight synthetic hexaploid wheats at 8, 15, 22 and 30 DAI. Caacupé, 2018. 

In order to study the progress of the disease in the preselected wheats, the area under the disease progress curve was calculated for both isolates, Table 3.

If we analyze the data presented in Table 3, on average, and considering the wheat by isolate interaction, the progress of the disease for both isolates is significantly different. In a similar study on four isolates of P. oryzae,^{Lule et al. (2014}) did not find significant differences in the AUDPC in rice when interacting with its host, but they did when comparing isolates.

The data of the area under the disease curve confirm the classification of the materials based on the maximum value of infection, without considering the materials with differential reactions. Only synthetic wheats TS29 and TS49 changed their classification based on the maximum value of infection and area under the disease curve.

The TS29, classified as moderately resistant due to its maximum value of infection, was considered moderately susceptible, since it presented a greater progress of the disease after the evaluation of 15 DAI.

The situation with the TS49 was the complete opposite of the reaction of the TS29, since it had a much slower development to reach a lower AUDPC. This selection strategy using AUDPC for this pathogen was also used by ^{Prabhu et al. (2003}) to classify the reactions of six elite lines of rice in a breeding program. ^{Jeger and Viljanen-Rollinson (2001}) agree that the progress of the disease is a more accurate and precise estimate for the selection of materials for a breeding program or genetic studies, helping identify the most stable materials in time.

Considering the genetic variability within the different collections of Ae tauschii crossed to obtain the SHWs, this type of reactions is to be expected. Similar results were reported by Guz Kazi et al., 2012 and ^{Masood et al. (2016}) in their studies on the reaction to diseases as well as in agronomical traits and in quality in general.

In a study related to the complete wheat genome, ^{Jighly et al., (2016}) mapped regions known for the resistance to multiple fungal pathogens in a collection of SHWs. Aside from identifying the presence of combinations previously reported in bread wheats, they were able to observe interactions for various diseases individually. Their assertion that the use of markers for the introduction of multiple resistances to diseases from SHWs can be an important source for the creation of elite varieties is very valid for the case in question, where sources of resistance are very limited.

It is worth mentioning that no molecular markers have yet been identified for the resistance to P. oryzae that validate their presence in the SHWs. Although TS29 and TS49 can be considered parents for the breeding programs and can be combined with each other, genotypes such as TS49 are the most valuable to provide a higher level of resistance and low progress of the disease until the end of the infection.

Conclusion

Taking advantage of the wide variability that exists in the hexaploid synthetic wheats, new sources of resistance to wheat blast, from different genetic origins were identified. These genetic resources represent a move forward in the search for more variability for resistance to P. oryzae in wheat, which is currently very narrow and may serve as a source to transfer its resistance to future varieties worldwide.