Introduction

The fungus Ustilago maydis [(D.C.) Corda] is a basidiomycete fungus known in the world to cause the disease called common maize (Zea mays) coal, this infection leads to the formation of large tumors that when developed on the cob are known as ‘huitlacoche’ or ‘cuitlacoche’ and are used in the preparation of different dishes (^{Ruíz and Martínez, 1998}; 2001; ^{Pataky and Chandler, 2003}).

In Mexico, huitlacoche has been used since pre-Hispanic times as an exquisite food; however, the unattractive appearance of the disease in maize has attracted the attention of national and international diners, giving it added value (^{Bernal and Ramírez, 2011}). In addition to noting that this fungus predominates glucose and fructose, it also contains carbohydrates, proteins, fats, minerals and vitamins such as riboflavin, niacin, biotin, and folic acid that constitute its nutritional value (^{Kealey and Kosikowski, 1981}; ^{Lizárraga, 1995}).

Huitlacoche occurs in almost all maize-producing regions, although favorable climatic conditions for its development are very varied, some authors claim that the attack of this pathogen may be more severe in humid environments (relative humidity of 72-80%) temperate (temperature of 17-20 °C) (^{Villanueva et al., 1999}; ^{Martínez et al., 2005}), other authors indicate that the attack is more common in warm areas (temperature of 26-34 °C) and moderately dry (^{Agrios, 2005}).

The infection forms a dikaryon filament because of mating interactions. This process is controlled by locus a and b. The locus a regulates the formation of mating filaments or conjugation tubes, the transcription of proteins encoded by locus b comes into play after fusion to maintain the filamentous growth of infectious dikaryon and to promote the development of pathogens in the plant.

One aspect of mating in U. maydis is that there are multiple specificities in locus b with estimates ranging from 25 to 33 different types (^{Banuett, 1992}; ^{Banuett, 1995}; ^{Kronstad, 2003}). U. maydis presents a genetic variability in the virulence of the fungus, this has a lot of lines or biotypes with different physiological and pathogenic characteristics. The main characteristic of this fungus is to be heterothallic which causes new hybridization biotypes to emerge in each new sexual generation, so a single damage is possible to isolate a range of haploid lines that differ in their pathogenicity.

It has also been shown that the growth index, color, spore production of colonies and other characteristics in culture medium can be modified by environmental fluctuations such as temperature, nutrition, vitamins, growth substances and chemical agents (^{Martínez et al., 2000}; ^{Martínez et al., 2005}).

Current breeding programs are based on natural coal infection in field plots and removal of the most susceptible improvement lines in order to maintain blight resistance; however, numerous reports are available in the influence of the various inoculation techniques, the host development stage and environmental conditions for pathogenesis (^{Kealey and Kosikowski, 1981}; ^{Pataky, 1991}).

In the artificial production of huitlacoche there is the artificial inoculation technique on commercial hybrids of sweet maize and open pollination varieties (^{Martínez et al., 2000}) this technique has been developed since the XVIII century (^{Walter, 1935}). In different parts of Mexico, the pathogenicity of some strains of U. maydis and the susceptibility and resistance in maize varieties have been evaluated; however, for the state of Aguascalientes there is no information on the pathogenicity of wild strains, as well as susceptibility and resistance in maize varieties adapted in the estate.

Therefore, the objective of this research was to evaluate different strains of Ustilago maydis from Aguascalientes state to infect maize plants under controlled conditions with the artificial inoculation method.

Materials and methods

Results and discussion

In the severity index (IS), it can be observed that there are significant differences between treatments, where the treatments of the creole 1varieties (94.7 ±1.3), AS (90.7 ±9.3) and Asgrow (89.3 ±10.7) with cross 2 were those that had the highest values of severity index, while the treatments of the creole 2 variety with cross 2 (42.7 ±17.6) and with cross 1 (48 ±4.6) were the ones that presented the lowest values of IS. In Figure 1, huitlacoche galls formed in cob can be observed through artificial production of the different treatments evaluated.

Figure 1 Huitlacoche galls in cob through artificial production of the different treatments evaluated in the period 2014-2016. 

In the incidence percentage (IP), it is observed that there are significant differences between treatments, where the variety C3 with cross 2 and the variety C1 with the cross had a higher PI (100%), while the variety C2 with cross 2 obtained a lower IP (60 ±23.1) (Table 3).

In the results of the grams per infected cob (GMI) variables and in yield per hectare in tons (RHt) the behavior of the samples was the same, in both variables there are significant differences between treatments, where the treatment of the AW variety with cross 3 presented a higher yield per cob (230.9 ±39.4) and per hectare (13.2 ±3), while the treatment of the variety C1 with cross 1 presented the lowest yields per cob (32.9 ±11.6) and per hectare (1.6 ±0.8) (Table 3).

The yields obtained with the Asgro2 hybrid inoculated with cross 3 (13.2 t ha^-1) approximate the yields obtained by ^{Martínez et al. (2000)}, where they report that, in experimental and commercial open pit trials, they reached a maximum production of 14 t ha^-1. However, cross 1with INIFAP hybrid (3.7 t ha^-1) and variety creole 3 with cross 1 (5.1 t ha^-1) are close to those reported by ^{Madrigal et al. (2010)} where they obtained that the hybrid 30G40 presented the highest yield with 4 786.2 kg ha^-1.

Although, ^{Martínez et al. (2000)} obtained a better yield of the fungus per infected plant (279.7 g) compared to the results of this research (230.9 g), the difference between the two results can be influenced by the varieties used in both investigations. The incubation period of the fungus varies from 12 to 25 days, in greenhouse can present a faster incubation period (7 to 16 days) and slower in the field (17 to 19 days) (^{Thakur et al., 1989}; ^{Pope and McCarter, 1992}; ^{Valverde et al., 1993}; ^{Carroll, 1994}; ^{Pataky et al., 1995}).

This is consistent with what was obtained in the results for the GMI variable (Table 3) since the incubation and growth period of the fungus varied for all treatments, some had lower production and other varieties had optimal development of the fungus at 22 days of incubation.

The results obtained in creole seeds in the incidence percentage variable (PI) were 77.8-88.9%, while in the severity index variable in percentage (IS) it was between 56.9-81.8%, in turn the variable grams per infected cob (GMI) resulted in 59.2- 152.4 g and the yield per hectare in tons (RHt) was 3-8.1 t ha^-1 (Table 3), these results vary with respect to those reported by ^{Valdez et al. (2009)}.

Where they obtained in 15 genotypes of creole maize variations in the percentages of infection in cobs from 31 to 92% (average 72%), for the percentages of coverage of the fungus on the cob ranged from 46 to 97%, with an average value of 77% and huitlacoche weight per cob, it varied considerably on maize lines (80 to 450 g) and yields of approximately 15 t ha^-1.

^{Martínez et al. (2000)} evaluated 100 fungus isolations in an experimental hybrid, selecting 12 of them for their vigor, presenting on average 190 g of the fungus per infected plant, and 135.24 g per inoculated plant (8.11 t ha^-1), severity index of 36.82 and an incidence of 70.64%.

Of the 300 families of maternal half-siblings evaluated (1998) by inoculation with the mixture of the 12 selected isolations (1997), 16 susceptible families were chosen, which had on average 154.97 g of the fungus per infected plant and 112.88 g per inoculated plant (6.67 t ha^-1), 76.67% incidence and 34.82 severity index, as well as 14 families resistant (0% incidence) to the fungus attack. The results obtained in IP, IS, GMI and RHt in hybrid varieties inoculated with cross 2 exceed those obtained by these authors (Table 3).

Principal components analysis (PCA)

A principal components analysis (PCA) was performed, taking into account the results obtained from all measured variables: severity index (IS), incidence percentage (PI), yield per cob (GMI) and yield per hectare in tons (RHt), to compare the main characteristics of treatments of the varieties C1, C2, C3, AS, AW and IP with crosses 1, 2 and 3.

Where, it was observed that principal component one (CP1) and two (CP2) explained 91.1% and 8.3% of the total variance, respectively. The combination of states of two-components by a multivariate ordering Biplot can explain 99.4% of the accumulated variability of variables

(Figure 1) and this was considered to be high enough for reliable differentiation between treatment characteristics. El value of the cophenetic correlation (1 000) indicated that the variable reduction made was adequate.

In Figure 2, it can be seen that the treatments more related to each variable are located closer to the arrow that represents said variables, in this way it can be observed that there is a grouping of treatments, this grouping was between the uninoculated varieties (GSN: orange oval) and those that were inoculated (GI: red oval), in this way it can be observed that the uninoculated varieties had the lowest values in all variables analyzed: IS, PI, GMI and RHt.

Figure 2 Biplot of multivariate ordering of treatments for the production of huitlacoche in the period 2014-2016. Orange point= uninoculated treatments; red point= varieties inoculated with cross 1; blue point= varieties inoculated with cross 2; green point= varieties inoculated with cross 3; yellow point= determined variables. Orange oval= group formed by uninoculated treatments (GSN); red oval= group consisting of inoculated treatments (GI); yellow oval= group 1 (G1), blue oval= group two (G2); green oval= group 3 (G3). 

However, in the group consisting of varieties that if they were inoculated with different crosses (GI: red oval), it can also be observed that there is the formation of three groups, grouped mainly according to the cross used for inoculation.

Group one (G1: yellow oval) consists of treatments of the AW variety inoculated with cross 2 and C2, C3, AS, AW and IP varieties inoculated with cross 3, where, this group of treatments present a higher yield per cob (GMI) and per hectare (RHt), otherwise to group three (G3: green oval) consisting of the treatments of C1, C2 and IP varieties inoculated with cross 1 and C1 and C2 varieties inoculated with cross 2, which were the treatments that had the lowest values of GMI and RHt, as well as IS and PI.

However, group two (G2: blue oval) consisting of the treatments of C3, AS and AW varieties inoculated with cross 1, C3, AS and IP varieties inoculated with cross 2 and C1 variety inoculated with cross 3, was the group of treatments that presented intermediate values in all the variables analyzed, since they presented good GM1 and RH, but not larger than those of G1, while for IS and PI values, both G1 and G2 had similar values.

The crosses used are made up of only two strains for each cross, where G2 had the largest PI (93.3%) and G1 obtained the highest values for IS (84.4%), GMI (203.4 g) and RHt (11.4 t ha^-1). The largest infection percentage was obtained with the variety Bida-54 and the inoculum of ten strains (78.8%), followed by the inoculum with four strains in the same maize variety (52.8%). The Bengal variety, obtained an infection level of 27% and huitlacoche yield of 6.3 t ha^-1.

The significant differences between the results are consistent with the authors Christenses (1963); ^{Villanueva et al. (1999)}; ^{Pataky and Chandler (2003)}, where they mention that there is a pathogenic variation of the fungus, the infection is based on the ability of the pathogen, the susceptibility of the host and its interaction with the environment and that for each variety of maize there is a ratio of optimal temperatures and humidity for the development of U. maydis.

The strains evaluated from the state of Aguascalientes had high IS values compared to some authors mentioned, which agrees with Calderón (2010), where it reports that the IS value generated by each strain of huitlacoche is not entirely related to the variety or color of maize, indicating that even though the varieties of white maize are the most susceptible, this rather depends on the virulence of the strain.

Conclusions

The strains of the state of Aguascalientes (1B, 3B, 13B and 16B) had a severity index in high percentage, presenting a high pathogenicity and an option for the production of huitlacoche in the state of Aguascalientes. The hybrids had greater susceptibility with the three crosses evaluated compared to creole varieties. It is important to observe the behavior of each variety and with this establish a homogeneous incubation period in each variety to obtain a higher production of huitlacoche.