Seleção genotípica e parâmetros genéticos da mamona influenciados por modelos fixos, aleatórios e mistos
Carregando...
Data
2019-07-22
Autores
Piedade, Gabriela Nunes da
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Os efeitos da interação genótipos x ambientes dificultam a recomendação de cultivares de mamona (Ricinus communis L.) com características agronômicas satisfatórias, devendo-se o melhorista atentar-se à correta avaliação dos dados. Objetivou-se neste estudo avaliar os diferentes modelos em experimentos de melhoramento de mamona FCA Porte Baixo e compará-los com a análise tradicional de experimentos agrícolas, no que se refere aos parâmetros genéticos obtidos e a ordenação dos genótipos avaliados. Para tanto, foram utilizados dados de experimentos desenvolvidos nos anos de 2014 e 2015 nos municípios de Botucatu, Ilha Solteira, Penápolis e São Manuel, na época de semeadura da safra, nos quais 20 linhagens S5 foram distribuídas em delineamento experimental de blocos casualizados com três repetições. Com os dados de produtividade média dos grãos, análises de variância conjunta (ANOVA), considerando os diferentes modelos: fixo, aleatório e misto (bloco aleatório e fixo, ambiente e linhagem fixo), foram realizadas, assim como foram obtidos os componentes de variância associados aos efeitos aleatórios e os componentes quadráticos associados aos efeitos fixos. Parâmetros utilizados em programas de melhoramento genético e o ordenamento dos genótipos, também foram calculados. Observou-se que os valores da estatística F, assim como sua significância, obtidos na ANOVA, apresentaram variações de acordo com o modelo utilizado. Houve diferença entre os componentes de variância e quadráticos, assim como entre os parâmetros genéticos estimados. As metodologias de análise tradicional, modelo aleatório e modelo misto, proporcionaram o mesmo ordenamento dos genótipos, exceto para o modelo onde bloco assume efeito fixo. Deve-se evitar o modelo em que todas as fontes de variação apresentam efeito aleatório, além do modelo em que apenas a fonte de variação bloco assume efeito fixo. Em contrapartida, quando ambiente assume efeito fixo, estimativas favoráveis são obtidas.
The effects of genotype x environment interaction make it difficult to recommend castor bean (Ricinus communis L.) cultivars with satisfactory agronomic characteristics, and the breeder should pay attention to the correct evaluation of the data. The aim of this study was to evaluate the different models in FCA Porte Low castor bean breeding experiments and compare them with the traditional analysis of agricultural experiments, regarding the genetic parameters obtained and the ordering of the evaluated genotypes. For this purpose, data from experiments developed in 2014 and 2015 in the municipalities of Botucatu, Ilha Solteira, Penápolis and São Manuel at the time of sowing were used, in which 20 S5 strains were distributed in a randomized complete block design with three repetitions. With the average grain yield data, joint variance analysis (ANOVA), considering the different models: fixed, random and mixed (random and fixed block, environment and fixed lineage), were performed, as well as the variance components were obtained, associated with random effects and the quadratic components associated with fixed effects. Parameters used in breeding programs and genotyping were also calculated. It was observed that the F statistic values, as well as their significance, obtained in ANOVA, presented variations according to the model used. There was a difference between the variance and quadratic components, as well as between the estimated genetic parameters. The methodologies of traditional analysis, random model and mixed model, provided the same ordering of genotypes, except for the model where the block assumes fixed effect. Avoid the model in which all sources of variation have a random effect, in addition to the model in which only the source of variation block assumes a fixed effect. In contrast, when environment assumes a fixed effect, favorable estimates are obtained.
The effects of genotype x environment interaction make it difficult to recommend castor bean (Ricinus communis L.) cultivars with satisfactory agronomic characteristics, and the breeder should pay attention to the correct evaluation of the data. The aim of this study was to evaluate the different models in FCA Porte Low castor bean breeding experiments and compare them with the traditional analysis of agricultural experiments, regarding the genetic parameters obtained and the ordering of the evaluated genotypes. For this purpose, data from experiments developed in 2014 and 2015 in the municipalities of Botucatu, Ilha Solteira, Penápolis and São Manuel at the time of sowing were used, in which 20 S5 strains were distributed in a randomized complete block design with three repetitions. With the average grain yield data, joint variance analysis (ANOVA), considering the different models: fixed, random and mixed (random and fixed block, environment and fixed lineage), were performed, as well as the variance components were obtained, associated with random effects and the quadratic components associated with fixed effects. Parameters used in breeding programs and genotyping were also calculated. It was observed that the F statistic values, as well as their significance, obtained in ANOVA, presented variations according to the model used. There was a difference between the variance and quadratic components, as well as between the estimated genetic parameters. The methodologies of traditional analysis, random model and mixed model, provided the same ordering of genotypes, except for the model where the block assumes fixed effect. Avoid the model in which all sources of variation have a random effect, in addition to the model in which only the source of variation block assumes a fixed effect. In contrast, when environment assumes a fixed effect, favorable estimates are obtained.
Descrição
Palavras-chave
Ricinus communis L., Análise, Efeito, Herdabilidade, Produtividade