| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.04 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A correlação genética entre características pode surgir essencialmente porque um ou vários genes podem influenciar simultaneamente várias características. No contexto do melhoramento de plantas, a sua determinação é útil para avaliar se, ao selecionar uma característica com interesse económico, não se está a prejudicar outra de igual importância, bem como para realizar seleção indireta de uma característica difícil de avaliar.
Neste trabalho faz-se a avaliação de correlações genéticas entre características de elevada importância económica (rendimento, peso do bago, volume do mosto do bago e características de qualidade do mosto), em variedades antigas de videira, para uma melhor objetivação da seleção policlonal e clonal. Os dados foram obtidos em populações experimentais de clones de 12 castas, onde são avaliadas amostras representativas da variabilidade genética intravarietal. Para a estimação das correlações genéticas entre características foram ajustados modelos lineares mistos multivariados.
As correlações genéticas revelaram-se mais fortes que as correlações fenotípicas, o que reforça a necessidade da implementação da metodologia seguida neste trabalho, isto é, o ajustamento de modelos lineares mistos multivariados que permitem estimar as correlações de origem genética e não genética.
Em geral, observaram-se correlações genéticas fracas entre rendimento e características de qualidade do mosto. Para a grande maioria das castas estudadas será possível fazer seleção policlonal e clonal satisfazendo vários critérios em simultâneo.
Genetic correlation between traits may arise essentially because one or several genes may simultaneously influence several traits. In plant breeding, its determination is useful to assess if, by selecting a trait with economic interest, another one equally important is not harmed, as well as to do indirect selection of a trait that is difficult to evaluate. This work evaluates the genetic correlations between traits of high economic importance (yield, berry weight, volume of the must of the berry and quality traits of the must) in ancient grapevine varieties, to better target polyclonal and clonal selection. All the data were obtained from experimental populations of clones of 12 varieties, where representative samples of intra-varietal genetic variability are evaluated. For the estimation of genetic correlations between traits multivariate mixed linear models were fitted. The genetic correlations proved to be stronger than the phenotypic correlations, which reinforces the need for the implementation of the methodology followed in this work, that is, the fitting of multivariate linear mixed models to estimate genetic and non-genetic correlations. In general, weak genetic correlations were observed between yield and must quality traits. For the majority of the varieties studied it will be possible to perform polyclonal and clonal selection satisfying several criteria simultaneously.
Genetic correlation between traits may arise essentially because one or several genes may simultaneously influence several traits. In plant breeding, its determination is useful to assess if, by selecting a trait with economic interest, another one equally important is not harmed, as well as to do indirect selection of a trait that is difficult to evaluate. This work evaluates the genetic correlations between traits of high economic importance (yield, berry weight, volume of the must of the berry and quality traits of the must) in ancient grapevine varieties, to better target polyclonal and clonal selection. All the data were obtained from experimental populations of clones of 12 varieties, where representative samples of intra-varietal genetic variability are evaluated. For the estimation of genetic correlations between traits multivariate mixed linear models were fitted. The genetic correlations proved to be stronger than the phenotypic correlations, which reinforces the need for the implementation of the methodology followed in this work, that is, the fitting of multivariate linear mixed models to estimate genetic and non-genetic correlations. In general, weak genetic correlations were observed between yield and must quality traits. For the majority of the varieties studied it will be possible to perform polyclonal and clonal selection satisfying several criteria simultaneously.
Descrição
Mestrado em Engenharia Agronómica. Universidade de Lisboa, Instituto Superior de Agronomia
Palavras-chave
Vitis vinifera L. variabilidade intravarietal modelos lineares mistos multivariados seleção policlonal intra-varietal variability multivariate mixed linear models polyclonal selection
Contexto Educativo
Citação
Ferreira, S.M.P. Correlações genéticas entre características culturais e enológicas em variedades antigas de videira. Lisboa: ISA-Universidade de Lisboa, 2022. Dissertação de Mestrado
Editora
Instituto Superior de Agronomia. Universidade de Lisboa
