Popcorn germplasm resistance to fungal diseases caused by Exserohilum turcicum and Bipolaris maydis

Compared with the common corn, popcorn shows greater susceptibility to pests and diseases occurrence, being the use of resistant cultivars the most efficient strategy. In this regard, Germplasm Banks deserve special attention because they contain accessions that can be used as sources of resistance in breeding programs. The State University of Northern Rio de Janeiro (UENF) maintains a popcorn Germplasm Bank with accessions from tropical and temperate countries, including Diversity Centers for the species. In this study, we investigated the performance of 37 popcorn accessions landraces, hybrids, and inbreed lines ― from the Germplasm Collection of UENF’s Popcorn Breeding Program with respect to the fungal leaf diseases northern leaf blight (NLB) and southern leaf blight (SLB), PLANT BREEDING Article Popcorn germplasm resistance to fungal diseases caused by Exserohilum turcicum and Bipolaris maydis Railan do Nascimento Ferreira Kurosawa1, Marcelo Vivas1, Antonio Teixeira do Amaral Junior1, Rodrigo Moreira Ribeiro1*, Samila Barbosa Miranda1, Guilherme Ferreira Pena1, Jhean Torres Leite1, Freddy Mora2 1.Universidade Estadual do Norte Fluminense Darcy Ribeiro Centro de Ciências e Tecnologias Agropecuárias Laboratório de Melhoramento Genético Vegetal Campos dos Goytacazes (RJ), Brazil. 2.Universidade de Talca Instituto de Ciências Biológicas Talca Chile. *Corresponding author: rodrigo.moreira85@yahoo.com.br Received: Feb. 1, 2017 – Accepted: Apr. 10, 2017 caused by Exserohilum turcicum and Bipolaris maydis, aiming to select promising accessions for genetic resistance to be used in future crosses to generate new resistant cultivars for farmers. The dendrogram scattering by the UPGMA multivariate technique was efficient in discriminating resistant accessions. Satisfactory results were obtained for resistance to NLB in accessions L71, L75, L76, P7, and PARA-172, which can thus be considered remarkable sources of resistance. For resistance to SLB, the superior accessions that can be indicated as sources of resistance are ARZM-05083, ARZM-07049, and PARA-172.


INTRODUCTION
Popcorn (Zea mays L.) is a type of corn whose main characteristics are its hard and small kernels and the ability to expand due to the pressure formed inside the grains upon heating.When compared with cultivars of common corn, popcorn plants usually display greater susceptibility of pests and diseases occurrence (Arnhold 2008;Hallauer and Carena 2009;Ribeiro et al. 2016).Among the leaf diseases affecting popcorn, the common blotch (northern leaf blight, NLB), caused by Exserohilum turcicum (Pass.)Leonard and Suggs (syn.Helminthosporium turcicum Pass.), is characterized as one of the major (Harlapur et al. 2008;Hurni et al. 2015;Ishfaq et al. 2014;Sabato et al. 2013;Wang et al. 2014).Under favorable climatic settings and in susceptible hosts, the pathogen may cause loses in grain yield that range from 27 to 90% (Ferguson and Carson 2007;Muiru et al. 2010;Wang et al. 2010;2012).
Another disease, the southern leaf blight (SLB), caused by the fungus Bipolaris maydis (Nisik.)Shoemaker (synonym of Helminthosporium maydis Nisik.and Myiake), occurs in corn with different intensities, depending on the degree of resistance, the crop management system, and the climatic conditions during its growth (Ali et al. 2011;Altaf et al. 2016;Hussain et al. 2016).Southern leaf blight is one of the most important diseases affecting corn and represents a great threat to corn growing all around the globe (Altaf et al. 2016;Kump et al. 2011;Noor et al. 2015).Under certain conditions, this fungus dramatically reduces the stand and yield of crops, with losses that can be as high as 70% (Ali et al. 2011;Hussain et al. 2016).The factors determining the magnitude of losses are the growth stage of the crop, the susceptible varieties, and the planting time (Ali et al. 2011).
Several control measures are implemented to minimize the damage caused by spot blotches (NLB and SLB), e.g., the spraying of fungicides and the planting of resistant varieties.Furthermore, some practices can be adopted, such as crop rotation, the use of the adequate planting density and spacing, and the use of balanced fertilization and elimination of crop residues (Bergamin Filho and Amorim 2011).The use of cultivars with geneticresistance potential against the main popcorn diseases is the most viable and economic control measure (Ayiga-Aluba et al. 2015;Ferguson and Carson 2007;Ishfaq et al. 2014;Vieira et al. 2009b).Associated with this method, the knowledge of the type of control as well as of the heritability of the traits involved in the expression of resistance are highly important in the development of resistant cultivars.This is because these data provide guidance for more efficient selection in subsequent works of breeding programs (Vieira et al. 2009a).
In this scenario, germplasm banks take on an indispensable importance in regard to the congregation of genetic variability, which is necessary especially to give basis to breeding programs of plants to obtain superior genotypes (Newton et al. 2011;Quintal et al. 2012;Vivas et al. 2015).UENF has a popcorn Germplasm Bank with accessions from different eco-geographic regions of Brazil and abroad, obtained from grants and exchanges, which have the potential to be investigated for resistance to diseases.This investigation should start with the identification of sources of resistance to leaf diseases in popcorn, as a result of the high losses caused by leaf pathogens and of the economic importance of the crop.
That said, in this study we investigated the performance of 37 popcorn genotypes with respect to fungal leaf diseases caused by E. turcicum and B. maydis and to select promising genotypes with genetic resistance to be used in future popcorn breeding programs, aiming to generate new resistant cultivars for farmers.

Implementation of the experiment
Two experiments were implemented in the municipality of Campos dos Goytacazes, Rio de Janeiro, Brazil, during a 1st season (October 2013 to March 2014), with an average temperature of 26 °C, and a 2nd season (May to September 2014), with an average temperature of 22 °C.
The experiment was implanted as a randomized-block design with four replications, in both growing seasons.The experimental units consisted of simple 3.0 m row containing 16 plants each.The experimental field was prepared mechanically by harrowing followed by furrowing.Seeding was performed manually, with plants spaced 0.20 m apart and rows by 0.90 m.Cultivation practices included thinning the seedlings 15 days after emergence, weeding out, periodic irrigation maintaining the soil under field capacity, and topdressing applied 30 and 45 after emergence, using a 20-0-20 NPK formulation with 300 kg.ha -1 and 200 kg.ha - urea, respectively.The ears of each genotype were collected after visual observation of the moment when the bracts were dry, in each growing season, to prevent early or late harvests.The useful plot of the experiment consisted of six competitive plants, totaling 0.90 m 2 .

Evaluated traits
The performance of the genotypes concerning the leaf diseases was monitored by estimating the incidence and severity of the symptoms.To this end, two estimation methods were adopted: i. Measurement of incidence of the disease symptoms along the plant, expressed as a percentage; ii.Quantification of severity symptoms along the first leaf below the first ear, expressed as a percentage.Leaf diseases were manifested by a natural infection of the etiological agents on the field.Evaluations were performed by taking six competitive plants per plot, this is, only plants with correct spacing and in the center of the plot were evaluated.
The incidence of NLB and SLB symptoms based on the plant was estimated by using a diagrammatic scale adopted by Agroceres (1996).The scale presents an interval of 1 through 9, in which 1 = 0 % incidence; 2 = 0.5% incidence; 3 = 10% incidence; 4 = 30% incidence; 5 = 50% incidence; 6 = 70% incidence; 7 = 80% incidence; 8 = 90% incidence; and 9 = 100% incidence.To evaluate the severity of NLB in the leaf, six competitive plants were used per plot, and the evaluations occurred every seven days from flowering to the senescence stage.Considering that only three evaluations were conducted, and that our goal was simply to compare genetic materials, we believe that the analysis based on the average of the evaluations better reflects the potential of the evaluated genotypes.For this variable, the diagrammatic scale proposed by Lazaroto et al. ( 2012) was adopted, containing the severity intervals in percentage corresponding to 0.5%, 1.0%, 2.5%, 6.5%, 15.5%, 30.0%, and 54.0%, based on which the percentage of the leaf area compromised by the disease was estimated.The scale proposed by James (1971) was used to estimate the leaf severity of SLB, with severity intervals expressed by the following scale: 0%, 1%, 5%, 25%, and 50%.

Statistical analysis
The traits were statistically evaluated by analysis of variance, and when a significant difference was detected, Scott-Knott's mean clustering algorithm was employed at the 5% probability level.Phenotypic (rp), genotypic (rg) and environmental (re) correlation coefficients were estimated, and the significance level was evaluated by the t test.In the estimate of genetic diversity, Generalized Mahalanobis Distance was used as a dissimilarity measure, and subsequently the accessions were grouped based on the UPGMA method, using, for this, the traits incidence and severity of SLB and NLB.The adjustment between the matrix of distances and the dendrogram was estimated by the cophenetic correlation coefficient (CCC), developed by Sokal and Rohlf (1962).All analyses were performed using the computer resources of Genes software (Cruz 2013).

Resistance to Northern Leaf Blight (NLB)
There was a significant genotype effect for incidence of plants with symptoms in both growing seasons.Two groups were formed in the first, and four in the second growing season.Genotypes L71, L75, L76, P7, and PARA-172 were allocated in the group with highest resistance in both evaluated periods (Table 2).
For leaf severity, a significant effect of genotype was only observed in the second evaluation period.The means clustering tests also revealed the presence of more than one group in the second harvest (Table 2).Thirty-one genotypes were included in the group with the lowest means, with noteworthy values found for L71, L75, L76, P7, and PARA-172.In addition to these, L61, L66, P3, P8, ARZM-05083, BOYA-462, BOZM-260, URUG-298-Amarelo, URUG-298-Roxo, and IAC-125 were highlighted for showing values lower than 5% of their leaf area affected, considering the average between the first and second harvests (Table 2).P3, P7, and P8 are lines from the State University of Maringá (UEM); P3 was extracted from the CMS-42 compound of Embrapa Maize & Sorghum; P7, from the 'Zaeli' triple hybrid; and P8 from the IAC-112 hybrid (Silva et al. 2011).The resistance of line P3 is possibly related to the fact that it originates from the genetically bred population CMS-42 (Pacheco et al. 1998).The line P7 may have its expression of resistance because its genealogy -Zaeli -holds alleles favorable to NLB, as stated by Vieira et al. (2009b).The P8 line probably showed good performance because it originated from hybrid IAC-112, which has recognized resistance to NLB (Vieira et al. 2009b).Among genotypes developed by UENF that stood out -L61, L66, L71, L75, and L76 -, lines L61, L66, and L71 had good performance, likely because they originated from cultivar BRS Angela, developed by Embrapa Maize & Sorghum, after six cycles of intrapopulation recurrent selection (Pacheco et al. 2005).According to Bleicher and Balmer (1993), recurrent selection is efficient to increase the level of resistance to NLB in susceptible corn cultivars, confirming the assertion of Parteniani and Miranda Filho (1978).Based on this premise, Kasozi et al. (2015) observed that two S 1 recurrent selection cycles were effective in improving the "Longe5" population for yield and resistance to the damage caused by the maize weevil and to the main corn diseases, among them NLB.These authors achieved a 29% increase in grain yield and 15% reduction of the NLB disease from C 0 to C 2 , demonstrating the potential for improvement of these traits through recurrent selection.In an evaluation of the history of the NLB disease in seven cycles of recurrent selection, Ribeiro et al. (2016) concluded that selection was effective in maintaining the equilibrium of the disease in the improved population of popcorn UENF-14 and that the source of resistance was not lost with the advance of the selection cycles.These authors also stated that it is possible to combine genes for resistance to NLB with the other agronomic traits of interest.
The superiority of lines L75 and L76 can be attributed to the fact that they originated from cultivar Viçosa, which was evaluated under natural conditions of NLB infestation by Miranda et al. (2003) and was considered an expressive source of resistance.Nevertheless, Arnhold et al. (2008) evaluated S 1 progenies from Viçosa population for resistance to diseases and expression of agronomic features, and found superior resistance to NLB compared with the control IAC-112.
The breeds ARZM-05083, BOYA-462, BOZM-260, URUG-298-Amarelo, URUG-298-Roxo, and PARA-172 originate from CIMMYT, and genealogies or any expression of resistance to diseases for them are unknown.These accessions expressed similar resistance to that of commercial hybrid IAC-125, which is recognized for having moderate resistance to NLB (Cruz et al. 2014).Of these, PARA-172 stood four for showing an even superior resistance to that of hybrid IAC-125, which makes its use in breeding programs for introgression of resistance to NLB strategic.
Considering the performance of the genotypes in both growing seasons, lines L71, L75, L76, P7 and accession PARA-172 are considered superior.

Resistance to Southern Leaf Blight (SLB)
There was a significant effect of genotype in all evaluations performed, with three groups formed in the first harvest and four in the second, for incidence in the plant.Genotypes ARZM-05083, ARZM-07049 and PARA-172 had the lowest estimates for incidence of SLB in both evaluation periods (Table 3).For severity of SLB in the leaf, there was a significant genotype effect only in the second growing season, with three and four groups formed for the first and second periods, respectively (Table 3).The genotypes that expressed the highest levels of resistance to SLB, in both growing seasons, were L61, L66, L70, L76, P2, P3, P4, P7, ARZM-05083, ARZM-07049, BOZM-260, PARA-172, URUG-298-Amarelo, and IAC-125 (Table 3).
Lines L61, L66, and L70 are from recurrent selection (Pacheco et al. 2005), whereby it is concluded that the resistance shown was due to the combination of favorable alleles stemming from the implementation of recurrent selection.Of the accessions from UEM, the resistance of lines P2 and P3 is attributed to the fact that they are from the improved population originating from compound CMS-42 (Pacheco et al. 1998).The resistance observed in hybrid top cross IAC-125 was expected to occur, as described in the table provided by EMBRAPA about the behavior of corn cultivars available in the Brazilian market in the 2013/14 crop regarding the main diseases (Cruz et al. 2014).The breeds from CIMMYT -ARZM-05083, ARZM-07049, BOZM-260, PARA-172 and URUG-298-Amarelo -have shown to be materials of interest for breeding aiming at resistance to diseases.These cases are genotypes from temperate countries, like Argentina (ARZM-05083, ARZM-07049), Paraguay (PARA-172), and Uruguay (URU-298-Amarelo) and tropical country like Bolivia (BOZM-260).These findings indicate these genotypes can be recommended in crosses with plants that have desirable agronomic traits, given the expansion of the genetic basis of popcorn accessions and the development of new cultivars.In addition, the selection could be applied within some accession, in order to achieve improved lines for explore hybrids, for instance Silva et al. (2015) evaluated the genetic divergence among and within accessions of popcorn and found that accession BOZM-260 showed the highest proportion of heterozygous plants among the evaluated accessions and also exhibited allelic variation in all microsatellite loci analyzed.These authors also found that ARZM-05083 was genetically distant from the other genotypes assessed, which corroborates the strategy of use of popcorn germplasm from temperate climates to broaden the genetic base of breeding programs.
Considering the accessions performance in both growing seasons, for both incidence and severity, three accession can be indicated as the highest sources of resistance to SLB: breeds ARZM-05083, ARZM-07049, and PARA-172, all originating from CIMMYT.Special emphasis should be placed on genotype PARA-172, which expressed the highest degree of resistance for both NLB and SLB (Table 3), thus showing to be an excellent alternative in crosses aiming at resistance to popcorn leaf diseases or for the extraction of lines for the generation of superior hybrids.With this goal, UENF's Popcorn Breeding Program has acted intensely in the development of lines from superior germplasms; nine S 6 descendants of PARA-172 are already under growth in the field for the generation of individuals from the next inbred generation.

Cluster analysis
There was a good fit in the graphic representation of the clustering distances, since the cophenetic correlation coefficients (CCC) were 0.87 for the first and 0.78 for the second growing season.For the first harvest, the genotypes were gathered in three larger groups (Figure 1a),  (a) (b) one of which was formed by lines L51 and L52, that were the most susceptible to the two diseases evaluated here (Tables 2 and 3).The second group included 18 of the 35 remaining genotypes.A subgroup was formed in this group, consisting of L59, P8, L88, IAC-125, P10, P3, P7, P9, and PARA-172, which were the genotypes most resistant to both diseases (Tables 2 and 3).The third group contained 17 accessions, and except for P2 and L71, which showed resistance to NLB and SLB in the first harvest, the others were moderately resistant to the action of the fungus Bipolaris maydis.Four groups were formed for the second harvest (Figure 1b); the first included the genotypes of highest susceptibility -L53, L54, L80, L52, and CHZN-13134, whilst the second group allocated lines L55 and L65, which were susceptible to NLB in the second harvest and moderately resistant to NLB and SLB in the second harvest.It is thus unquestionable that these reactions facilitated the agglutination of these genotypes in the same group for both growing seasons.The third group contained 23 accessions, which were moderately resistant to NLB and SLB, except for the subgroup formed by lines P5, P1, L77, L51, and L59, which were all moderately resistant to NLB and susceptible to SLB.In the second group, a subgroup was formed containing accessions PARA-172, ARZM-05083, and ARZM-07049, which stood out for having consistently the highest levels of resistance to SLB in the second growing season, especially PARA-172, which revealed the highest estimates for severity to both diseases in the second growing season (Table 3).In this way, these are populations with high potential to be used as sources of resistance.Moreover, the consistent variability present between these populations -confirmed by the high genetic diversity estimates as estimated by ISSR markers (Amaral Júnior et al. 2011) -make them even more interesting for the production of superior segregants.
The efficiency of the UPGMA clustering method is thus proven in the discrimination of reaction to NLB and SLB.As such, it is an interesting multivariate statistical alternative in the indication of genotypes for breeding programs aimed at the introgression of genes providing resistance to NLB and SLB in segregating populations, or even to obtain superior hybrids.

Estimates of correlations
Elevated estimates for phenotypic and genotypes correlations were detected for incidence and severity of NLB in both growing seasons, similarly to what occurred for incidence and severity of SLB (Table 4).The magnitude of genetic correlations were superior to the phenotypic correlations, evidencing a strong correspondence of selection based on the phenotype with the genetic effects between incidence and severity.Thus, only one of the two traits would already be sufficient for the selection of resistant genotypes and, in this aspect, because of the greater practicality, it is recommended to use the incidence traits as the main variable in experiments conducted in the field, under natural inoculation conditions.Under controlled inoculation conditions, however, it is suggested to use severity as a criterion of selection.This is because in experiments undertaken in greenhouses, the distribution of diseases is dependent upon the inoculation being performed; on the other hand, for experiments in the field, the disease is typically distributed more evenly.Furthermore, incidence is a more practical variable to be evaluated.In turn, were observed genetic, phenotypic and environmental correlation estimates varying of low to high magnitude for incidence and severity of NLB and SLB, depending on the pair of traits assessed and the season in which this evaluation took place (Table 4).Therefore, there is no direct relationship between traits that provide resistance to these diseases.These results are probably because the mechanisms of resistance to these diseases have distinct origins.According to Chang and Peterson (1995), Bipolaris maydis is controlled by two recessive genes in corn, in which the dominant gene provides susceptibility to SLB.On the other hand, for E. turcicum, the genetic control of NLB can be achieved by quantitative resistance (by using several genes of small effects) or even by qualitative resistance (from few genes of high effect in reducing the intensity of the disease) (Vieira et al. 2009b).Ogliari et al. (2005) emphasize that these two types of mechanisms of resistance to E. turcicum may work separately or together, suggesting the existence of different mechanisms providing resistance to NLB and SLB, which will require greater accuracy in the development of breeding programs when aiming at the combination of these diseases in progenies under selection.

CONCLUSION
A good source of resistance was found to NLB through accessions L71, L75, L76, P7 and PARA-172, which can thus be considered remarkable sources of resistance.For resistance to SLB, the superior accessions that can be indicated as sources of resistance are ARZM-05083, ARZM-07049, and PARA-172.Thus, with respect to leaf fungal diseases caused by E. turcicum and B. maydis, promising genotypes with genetic resistance may be used in future crosses in popcorn breeding programs, in order to generate new resistant cultivars for farmers.

Figure 1 .
Figure 1.Dendrograms of 37 genotypes of popcorn by the UPGMA method using Generalized Mahalanobis Distance as a measure of genetic dissimilarity, for both growing seasons.(a) Growing season from October 2013 to March 2014 and (b) Growing season from May to September 2014.

Table 1 .
Description of popcorn genotypes from the Germplasm Bank of the State University of Northern Rio de Janeiro.

Table 2 .
Average of three assessments in percentage of incidence and severity of the fungal leaf diseases northern leaf blight (NLB), caused by E. turcicum, in the first and second popcorn growing seasons.
Means followed by the same letters represent a statistically homogenous group according to Scott-Knott's algorithm.Season 1 = growing season from October 2013 to March 2014 and Season 2 = growing season from May to September 2014.

Table 3 .
Average of three assessments in percentage of incidence and severity of the fungal leaf diseases southern leaf blight (SLB), caused by B. maydis, in the first and second popcorn growing seasons.
Means followed by the same letters represent a statistically homogenous group according to Scott-Knott's algorithm.Season 1 = growing season from October 2013 to March 2014 and Season 2 = growing season from May to September 2014.

Table 4 .
Estimates of correlations among incidence of NLB in the plant (ETP), severity of NLB in the leaf (ETF), incidence of SLB in the plant (BMP), and severity of SLB in the leaf (BMF) for two popcorn growing seasons.Significant at the 1 and 5% probability levels , respectively, according to the t test.Season 1 = growing season from October 2013 to March 2014 and Season 2 = growing season from May to September 2014.