Introduction

Fusarium head blight (FHB), or fusariosis, on small-grain cereals is a devastating disease throughout the world, decreasing the quality, weight, and carbohydrate and protein content of grain; thus compromising the yield and quality of grain production and the health of consumers through the production of associated mycotoxins (^{Bai et al, 2001}). The use of infected grains can furthermore modify bread production because the metabolites produced by the fungus inhibit the growth and enzymatic activity of yeast (^{Saric et al, 1997})· The main etiologie agents of FHB are Fusarium graminearum, F culmorum, F. avenaceum, F. poae, Microdochium nivale, and M. majus. The distribution and prevalence of these pathogens are largely determined by climatic conditions such as temperature and humidity In Argentina, the dominant fungal agent associated with FHB is F. graminearum (Schwabe), anamorph of Gibberella zeae (Schw) Petch (^{Galich, 1997}; ^{Gilbert and Tekauz, 2000}; ^{Lori et al, 2003}; ^{Kikot et al, 2011}). FHB epidemics in Argentina were reported in 1976, 1978, 1985, 1993, 2001, and 2012 (^{Moschini and Fortugno, 1996}; ^{Kikot et al, 2011}). Fusarium graminearum produces mycotoxins, such as the trichothecenes deoxynivalenol, its derivatives 3-acetyldeoxynivalenol and 15-acetyldeoxynivalenol, and nivalenol. The role of trichothecenes in plant disease is not clear, but ^{Proctor et al. (1995}), ^{Gang et al. (1998}), ^{Šíp et al, (2011)}, and ^{Malbran et al. (2012}) reported a positive correlation between aggressiveness and deoxynivalenol production by F culmorum and F graminearum.

Degrees of aggressiveness could result from qualitative and quantitative differences in the production of mycotoxins and the enzymes that degrade the plant cell wall (^{ten Have et al, 2002}; ^{Jenczmionka et al, 2003}; ^{Jenczmionka and Schafer, 2005})· But trichothecenes per se are related to the spread of the disease throughout the plant, rather than the establishment of FHB (^{Goswami and Kistler, 2005}; ^{Nicholson etai, 2007}).

The relevance of the selection of genotypes resistant to FHB in wheat is based on the rationale that host resistance is considered the most appropriate means of controlling the disease. Several types of genetic resistances to FHB in wheat are described; for example, type I is associated with resistance to initial infection and type II with resistance to the spread of infection within the plant; both were described by ^{Schroeder and Christensen in 1963} (^{Mesterházy et al., 1999}; ^{Nicholson et al, 2007}). To distinguish between these sources of resistance, researchers developed inoculation methods to test genotypes in breeding programs. Whereas spraying wheat heads with conidial suspensions allows the evaluation of type-I resistance, the inoculation of a single floret per head with conidia (i.e., a point inoculation) enables a monitoring of the spread of the disease among the spikes and, therefore, an assessment of type-II resistance (^{Bai and Shaner, 1994}; ^{Bai et al., 2001}).

The objective of this study was to identify new sources of resistance to FHB in Argentine wheat cultivars and lines for comparison with known resistant international cultivars. The hypothesis was that the characterization of resistance to FHB in local lines and cultivars and in international germplasm would provide relevant information regarding differences in genotypie performance against the disease.

Materials and Methods

Results and Discussion

The pathogenicity on the Argentine cultivars showed statistically significant differences in all the variables analyzed, except for incidence, whereas in the international cultivars significant differences were observed only for TGW (Table 1). ^{Cowger et al. (2009)} found significant differences for variables incidence, severity, percentage of damaged kernels, and deoxynivalenol content among winter-wheat cultivars.

Table 1 Statistical analysis of the pathogenicity tests on Argentine and international cultivars infected with Fusarium graminearum isolate. 

TGW: thousand-grain weight. *Controls

For the Argentine cultivars, the Pearson correlations were positive for the associations between incidence and severity (p≤0.01), incidence and the percentage of damaged kernels (p≤0.01), severity and the percentage of damaged kernels (p≤0.01) severity and TGW (p = 0.01), and the percentage of damaged kernels and TGW (p≤0.01). But the correlation was negative between percentage of damaged kernels and TGW, and between the ISK index and TGW. In a similar study, ^{Hernández (2010)} also observed a positive correlation between severity and the percentage of damaged kernels.

The symptoms showed a high association with the level of infection in the grains. Besides, the TGW was reduced due to an increment in incidence, severity, and the ISK index, in accordance with the increase in infected grains. In Table 2 it is shown that the grain weight was more affected in cultivars with higher susceptibility to the disease.

Table 2 Correlation among the variables analyzed in Argentine cultivars infected with Fusarium graminearum. 

TGW: thousand-grain weight. ISK index: incidence, severity, and damaged kernel.

The values above the diagonal of identity-redundant correlations, demarcated by dashes, express the Pearson correlation among variables, whereas the values below that diagonal indicate the significance (p≤0.05).

From the severity values obtained in the Argentine cultivars, a frequency histogram of FHB was constructed (Figure 1). Response to the disease of 21 cultivars (Cronox, LE 2357, Klein Proteo, Buck Huanchen, BioINTA 2005, Buck Puelche, Klein Castor, Buck AGP FAST, Arex, Klein Nutria, BioINTA 1005, Klein Carpincho, Klein Rayo, Buck Chacarero, Klein Zorro, Atlax, Buck 75 Aniversario, Buck Meteoro, BioINTA 1002, Klein Tigre, АСА 906, and Buck 55 CL) was not different (p > 0.05) to that of the resistant controls Sumai 3 and YMI 6-26B, and showed severity levels lower than 20 %. Baguette P 11 and Baguette 18, however, were susceptible since they showed a degree of severity above 50 %. ^{Mergoum et al. (2007}) characterized the behavior of different cultivars against FHB in North Dakota under conditions of natural and artificial infection throughout different years and observed significant differences among the cultivars.

Figure 1 Frequency histogram of the Argentine cultivars infected with Fusarium graminearum. The bars illustrate the degree of severity as a percent on the abscissa versus the relative frequency of the cultivars that belonged to the corresponding degree of severity plotted as a fraction on the ordinate. The respective severity-versus-frequency values of the control cultivars Sumai 3, YMI 6-26B, and ProINTA Granar, shown above the bars, is indicated by vertical lines demarcating their position within the scheme. 

Table 3 lists the values of incidence, severity and damaged kernels of the Argentine susceptible cultivars with severity levels lower than 15 %. The 11 cultivars included in this group and the Sumai 3 control had lower values for all the variables analyzed, as compared to the averages observed in either the susceptible or the international cultivars, the latter were previously assessed as susceptible. The Argentine susceptible cultivars are the following five that showed the highest disease severity: Baguette p 11 (61.9 %), Baguette 18 (53.6 %), Buck Mangrullo (48.5 %), Klein Guerrero (44.1 %), and BioINTA 3000 (42.5 %). The Argentine cultivars with greater resistance to disease (Table 3) had mean levels for the three variables analyzed that were lower than the averages for domestic and international cultivars.

Table 3 Argentine cultivars with higher resistance to FHB compared to the average values of susceptible Argentine and international cultivars. 

ARC: Argentine resistant cultivars; ASC: Argentine susceptible cultivars; 1С: international cultivars. ^†ProINTA Granar and Klein Don Enrique. ^⁋Baguette p 11, Baguette 18, Buck Mangrullo, Klein Guerrero, and BioINTA 3,000. ^§Nobeoka Bouzu Komugi, Catbird, and Frontana. ¹Average value group of five cultivars with high severities (42.4 % to 61.9 %). ²Average value group of five cultivars with high incidence values (71.3 % to 96.6 %). ³Average value group of five cultivars with high numbers of damaged kernels (45.8 % to 91.7 %).

The characterization and evaluation of the international wheat cultivars preselected for FHB resistance, was carried out in order to verify the response to the disease under Argentine climatic conditions and with a local strain of the pathogen. The evaluation of the specimens considered resistant to the disease, according to information provided from the place of shipment, was confirmed since no significant differences were observed among the cultivars.

A study of the advanced wheat lines having different growth cycles obtained from the INTA breeding program was also useful in the search for genotypes resistant to FHB. Disease progression was registered according to the Fusarium index integrating the values of incidence, severity, and the combination of the two, expressed as (incidence X severity)/100 (^{Hernandez, 2010}), and the lines with low values of the Fusarium index were considered resistant. The highest level of Fusarium index values were for the long-growth-cycle lines and the lowest level for the short-growth-cycle lines, whereas the values for the lines of intermediate growth fell somewhere in between. Despite certain degree of overlap between the adjacent groups, the average values of the three groups showed a clear-cut difference among the responses of those three categories of growth-cycle kinetics. All the lines showed infection levels (severity) below 20 %, except Т 00212, р 05375 (long growth cycle), Т 00219 (medium growth cycle) and Т 00224 (short growth cycle). The lines with higher resistance to FTB infection, and severity levels below 30% were characterized as moderately resistant (Table 4). According to ^{Jin et al. (2013}), for 363 lines from the USA, 43 % were susceptible, 32 % moderately susceptible, and 7-0 % resistant.

Table 4 Groups (according to growth cycle) of advanced lines from the INTA breeding program and characterized as resistant to Fusarium head blight. 

FI: Fusarium index.

This evaluation of the resistance to penetration by the pathogen into the spike (type I) in a high number of wheat genotypes, of three wheat groups with different origin, is relevant to the control of FHB and improvement in disease outcome, because the information allows characterizing wheat genotypes behavior against the pathogen under conditions similar to those in nature. Therefore, such a research enables the identification of new sources of resistance (^{Xu et al., 2001}; ^{Brennan et al., 2007}; ^{Vogelgsang et al., 2008}).

Our results suggest that the F. graminearum isolate used in assessing the response of different wheat genotypes to infection was highly infective and allowed evaluating the disease in a large number of lines and cultivars. The artificial inoculation used in the experiments insured contact between the pathogen and the host, which facilitated a successful infection and subsequent disease progression.

Conclusions

The Fusarium isolate used in this investigation was appropriate since the highly infective inoculum allowed us to analyze the progression of the disease in 210 national and international genotypes. We observed and recorded significant differences among the Argentine wheat cultivars in response to the disease, measured as severity, thousand-grain weight, and percent damaged kernels.

The group of international germplasm evaluated showed resistance against the disease since no significant differences were observed among the cultivars. With regard to the Argentine cultivars, new sources of resistance to FHB were discovered since these specimens were previously evaluated, but for other diseases of easier and faster detection than FHB.

The group of advanced wheat lines from INTA, Argentina, responded to FHB challenge with low levels of incidence and severity as expressed in the Fusarium index values. Among these lines, the best resistance to FHB occurred with the short-growth-cycle cultivars, followed by the intermediate- and finally the long-growth-cycle genotypes. This response gradient could indicate that the growth cycle per se affects the resistance to infection, probably as a result of a shorter time of exposure between anthesis and maturity, thus making those lines an especially preferable choice during years characterized by an increased threat of FHB infection.