Original paper 66 ~ M 11 Open Access Maydica electronic publication - 2021 Maydica Introduction Northern leaf blight of maize (NLB), caused by the fun- gus Exserohilum turcicum (Pass) K.J. Leonard & Suggs, is one of the most prevalent disease of maize. In its most common manifestation, this leaf disease leads to elliptical and elongated leaf lesions, grayish green in color, with lengths ranging from 2.5 to 15 centimeters (Kimatiet al.2005). It has high genetic variability (Bor- chardt et al. 1998; Ferguson and Carson, 2004, 2007), although little variation in pathogenicity ( De Rossi 2015). The infectious phase starts in the oldest leaves after silking. The infections prior to this are linked to genot- ypes that are more sensitive and subject to yield loss (Pataky1992). In the United States and Canada, the estimated yield Selection of sweet corn genotypes for resistance to Northern Leaf Blight using longitudinal analysis Gabriel Moreno Bernardo Gonçalves*¹, José Crossa², Juan Burgueño², Jocarla Ambrosim Crevelari¹, Flavia Nicácio Viana¹, Vivane Mirian Lanhellas Gonçalves¹, José Arantes Ferreira Júnior¹, Messias Gonzaga Pereira¹ 1 Universidade Estadual do Norte Fluminense Darcy Ribeiro – Laboratório de Melhoramento Genético vegetal, Campos dos Goytacazes, RJ, Brazil 2 Biometrics and Statistics Unit, International Maize and Wheat Improvement Center (CIMMYT), Apdo. Postal 6-641, 06600, México DF, México Corresponding author: E-mail: gabriel.agrobio@gmail.com Keywords: repeated measure, line x tester, shrunken2, corn disease, Exserohilum turcicum Abstract Northern Leaf Blight (NLB) of maize (Zea mays) is one of the most common maize diseases. The economic dama- ge it causes is frequently linked to inadequate crop management and susceptible cultivars. To select sweet corn genotypes with a higher level of resistance to NLB, 18 single-cross hybrids, six populations, four interpopulation hybrids, and two controls were evaluated in two field trials. Genotypes were evaluated three times, and results were analyzed according to linear mixed models and repeated measures over time, using days to silk as a cova- riate. The entry effect was sliced into single-crosses, populations, controls and interclasses. Single-crosses weresli- ced in line, tester, and line by tester.Results indicate higher ear yield (EY) and greater severity in early genotypes, thereby validating the use of covariates. In all analyses, the selection accuracy was high, above 0.86. For NLB, population and simple hybrid means varied from one environment to another, while simple hybrids were always higher for yield.In the joint analysis of NLB, there was a significant difference between Entry and Entry x Location, Line x Location, Control x Location, and Interclass x Location interactions. For yield, only Entry, Single-Cross, Population, Interclass, Entry x Location, Single-Cross x Location, and Interclass x Locationeffects were significant. The P8HS population, and the interpopulation and simple hybrids from this population, showed less severity in the joint analysis. As a result, the breeding program should focus on lines derived from the P8HS population to select genotypes more resistant to NLB and with high EY. Abbreviations BLUE – Best Linear Unbiased Estimation BLUP – Best Linear Unbiased Prediction C8HS – SH2-8HS-CIMMYT sweet corn population CGZ – Campos dos Goytacazes-RJ/Brazil CSH – SH2-CIMMYT8 sweet corn population CV – coefficient of variation DTS – days to silk EY – ear yield GxE – genotype x environment ITA – Itaocara-RJ/Brazil, municipalitie MSe – mean square error MSg – mean square genotypes NLB – northern leaf blight P8HS – SH2-8HS-Piranão sweet corn population PSH – SH2-Piranão sweet corn population RRS – reciprocal recurrent selection SC – single-crosses SE – standard error sh2 - shrunken-2 gene