Emprego do BLUP/REML na avaliação genética de linhagens elites de milho-pipoca

• Ano de defesa: 2010
• Autor(a) principal: Valente, Mágno Sávio Ferreira
• Orientador(a): Viana, José Marcelo Soriano
• Banca de defesa: Cruz, Cosme Damião , Sediyama, Carlos Sigueyuki , Furtado, Marcos Ribeiro
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Genética e Melhoramento
• Departamento: Genética animal; Genética molecular e de microrganismos; Genética quantitativa; Genética vegetal; Me
• País: BR
• Palavras-chave em Português: Análise dialélica; Modelos mistos; Acurácia
• Palavras-chave em Inglês: Diallelic analysis; Mixed models; Accuracy
• Área do conhecimento CNPq: CNPQ::CIENCIAS BIOLOGICAS::GENETICA::GENETICA QUANTITATIVA
• Link de acesso: http://locus.ufv.br/handle/123456789/4719

In breeding programs, it is important to use viable alternatives to reduce cost and time spent on population evaluations, combined with adequate statistical methodologies applying the maximum information available. Within this context, this work aimed to: i) study the genetic control of popping capacity and grain yield in popcorn; ii) predict the effects of general combining ability (GCA) and specific combining ability (SCA), applying mixed model analysis; iii) evaluate the performance per se of the lines used in hybrid crosses; and iv) evaluate the efficiency of Best Linear Unbiased-Prediction (BLUP) in predicting the performance of non crossed hybrids in diallelic analysis. Data from experiment evaluations of intra-population diallels of the Beija-Flor and Viçosa popcorn populations were used in the analyses. Eighteen S6 Beija-Flor elite lines were analyzed in ten environments and 29 S6 Viçosa elite lines in eight environments. Combined analyses were carried out, considering all the environments, with the additive model being adopted. The software ASREML was used in all analyses. GCA variability was verified both for grain yield and popping capacity. A significant SCA variability was verified in Beija-Flor for the two variables. In Viçosa, SCA was significant at 8% for grain yield while for popping capacity, the presence of additive effects was practically verified in its expression. Through the incorporation of sibling relationships in diallelic analysis, SCA values of combination of a line with itself (sjj) were predicted, without such cross being carried out. The use of BLUP/REML allowed the prediction of over 43 to 210% hybrid combinations, if compared only to the hybrid observed. Based on hybrid predictions, promising crosses were identified both for grain yield and popping capacity, being observed that the inclusion of a pedigree seems to make the analyses more accurate, as compared to the model not using this information. The performance per se and under S6 family crosses showed moderate correlations (r = 0.53) for popping capacity, showing that families with high popping capacity GCA can be selected, based on their performance per se. Aiming to obtain information on the efficiency in predicting S6 families based on performance per se and on S3 family crosses, a positive correlation (r = 0.76) was found for production between the GCA of top cross hybrids of the S3 families and the S6.family hybrids. When verifying BLUP efficiency in the prediction of non crossed hybrids, moderate to high Pearson and Spearman correlations were found among the predicted and observed values. Prediction of hybrids in the absence of dominance (popping capacity, r = 0.88) was more efficient than in the presence of dominance (grain yield, r = 0.53). The overall results showed that the use of mixed models is an efficient alternative in diallelic analysis, providing good precision in estimating genetic parameters, besides allowing the prediction of non crossed hybrids.