ABSTRACT. The continuous gains in cycles of recurrent selection have raised disagreement among researchers. Two hundred and forty-two full-sib families were obtained from CIMMYT and Piranão populations and evaluated in a simple lattice design in two environments to estimate the response to selection in the 11 th cycle of the UENF reciprocal recurrent selection program. No genotype x environment interaction was observed for most of the traits evaluated, indicating that a single recurrent selection program can be conducted for both sites. The population studied exhibited wide genotypic variability and heritability estimates ranged from 33 to 73%, indicating prospects of selection gain for the following cycles. The predicted genetic gain in yield was 12.90%. It is possible to conclude that the populations studied may be promising for the achievement of new selection cycles, which provides a continuous concentration of favorable alleles and the production of hybrids for the North and Northwest regions of the state of Rio de Janeiro.Keywords: genetic gains, selection index, interpopulation hybrids, longevity of recurrent selection.Resposta a seleção no 11º ciclo de seleção recorrente recíproca entre famílias de irmãos-completos de milho RESUMO. A contínua obtenção de ganhos em ciclos de seleção recorrente tem sido motivo de controvérsia entre pesquisadores. 242 famílias de irmãos-completos foram obtidas entre as populações CIMMYT e Piranão, e avaliadas em delineamento látice simples, em dois ambientes para investigar as estimativas da resposta à seleção no 11º ciclo do programa de seleção recorrente recíproca da UENF. Não houve interação genótipos por ambiente para a maioria das características avaliadas, indicando que um único programa de seleção recorrente pode ser conduzido para ambos os locais. A população avaliada revelou ampla variabilidade genotípica, com estimativas de herdabilidade variando de 33 a 73%, indicando perspectivas de ganho de seleção para os próximos ciclos. O ganho genético predito para produtividade foi de 12,90%. Conclui-se que as populações avaliadas são promissoras para a consecução de novos ciclos de seleção, proporcionando concentração contínua de alelos favoráveis e a produção de híbridos para as regiões Norte e Noroeste Fluminense.Palavras-chave: ganhos genéticos, índice de seleção, híbridos interpopulacionais, longevidade da seleção recorrente.

ABSTRACT.We examined the effect of incorporation of molecular markers on variability between and within populations in order to maximize heterotic effects and longevity of a maize reciprocal recurrent selection program. Molecular variability was quantified by inter-simple sequence repeat (ISSR) markers between and within the maize populations Cimmyt and Piranão in the 10th cycle of a reciprocal recurrent selection program. Forty-two S 1 progenies of each population were analyzed, these being families of full-sibs selected according to their agronomic traits. Thirteen primers were selected, which produced 140 bands; 114 of them were polymorphic and 26 monomorphic. Based on UPGMA grouping analysis and by genetic distances, it was possible to identify "contaminant" progenies. These progenies belong to the Piranão or Cimmyt groups, but cluster in the opposite heterotic group. Identification of "contaminant" progenies is relevant for selection, because, besides identifying genotypes that should be eliminated at the recombination stage, it allows increased heterosis expression in crosses between more genetically distinct individuals. After the elimination of the "contaminant" progenies and those that were allocated between the heterotic groups, a new statistical analysis was carried out, which demonstrated increased genetic distances

-

ABSTRACT. We estimated the genetic gains of the 12th cycle of reciprocal recurrent selection for maize traits of agronomic interest. We used 23 ISSR molecular markers in an attempt to maximize genetic variability among and within populations based on selection of S 1 progenies. To this end, 138 full-sib families were evaluated in a randomized block design in two environments (the municipalities of Campos dos Goytacazes and Itaocara, in the State of Rio de Janeiro, Brazil), with replications within sets. Direct selection for grain yield was used for the selection of the families. To assess genetic diversity among and within populations, we examined plants produced from part of the S 1s seeds from the parents that originated the 42 full-sib families that were selected from the agronomic traits. Direct selection for grain yield provided good gains for the traits evaluated, with estimated improvement of -0.87 days for days to flowering, 0.35 plants, 1.79 ears per plot, 0.58 g per 100-grain weight, 308.21 g ear weight per plot, and 261.83 kg/ha grain yield. Application of molecular markers at the stage of superior progeny selection led to increased genetic distance among populations, which is a very important factor for maximizing the ©FUNPEC-RP www.funpecrp.com.br Genetics and Molecular Research 11 (3): 3398-3408 (2012) Maize reciprocal recurrent selection 3399 utilization of heterosis and providing greater longevity to the reciprocal recurrent selection program.

Correlation and path analysis for yield components of supersweet cornIn breeding programs, information on the correlation between characters is essential to improve the simultaneous selection of characters. However, the measurement and interpretation of the magnitude of a correlation can lead to mistakes in the selection strategy. The objective of this study was to evaluate, using path analysis, the direct and indirect components between primary production and yield of supersweet corn ear and identify the characteristics that most contribute to ear yield (basic variable). The variables used explained 94.77% of the variance in ear weight (R 2 ). The path analysis showed that only two characters, grain volume (0.2637) and ear volume (0.2536), had a direct effect on production. It appears, therefore, that although the majority of the characters present high correlation estimates, these Em programas de melhoramento genético, informações sobre a correlação entre caracteres são de grande importân-cia para se aperfeiçoar a seleção simultânea de caracteres. Contudo, a quantificação e a interpretação da magnitude de uma correlação podem resultar em equívocos na estratégia de seleção. O objetivo deste trabalho foi avaliar, por meio da análise de trilha, as relações, direta e indireta, entre os componentes primários de produção e a produtividade de espiga de milho superdoce, e identificar os caracteres que mais contribuem para a produtividade de espiga (variável básica). Verifica-se que as variáveis utilizadas explicaram 94,77% da variação do peso de espiga (R 2 ). Pela análise de trilha, apenas dois caracteres, o volume do grão (0, 2637) e o volume de espiga (0, 2536), apresentaram efeito direto na produção. Constata-se, portanto, que, apesar de a maioria dos caracteres apresentarem altas estimativas de correlação, essas ocorreram por efeitos indiretos de outros caracteres. Assim, tanto para seleção direta, quanto para seleção indireta, essas características são eficientes no aumento do peso médio da espiga de milho superdoce. Neste caso, a melhor estratégia seria a seleção simultânea de caracteres, enfatizando-se as características cujos efeitos indiretos são maiores. É oportuno salientar que o volume de grão e volume de espiga apresentaram maiores herdabilidades, quando comparados com peso de espigas, ou seja, 91,92, 88,6 e 80,52%, respectivamente. A alta herdabilidade estimada (91,92 e 88,6%) é um indicativo de possibilidades de elevados ganhos genéticos na seleção.Palavras-chave: coeficiente de trilha, efeitos diretos e indiretos, resposta correlacionada.