Competição em testes de progênies de eucalipto e suas implicações na seleção e no melhoramento
Carregando...
Arquivos
Data
2009-07-03
Autores
Pavan, Bruno Ettore [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Objetivo deste trabalho foi a identificação das formas de competição em testes de progênies de eucalipto e a influencia desta nos parâmetros genéticos e na seleção. Foram usados dados de dois testes de progênies de polinização aberta de eucalipto, instalados no delineamento em blocos casualizados. O experimento 1 (EXP1) foi constituído por quatro testemunhas (clones) e 49 progênies e o experimento 2 (EXP2) por 44 progênies, ambos com seis repetições e parcelas lineares de 10 plantas. Em três idades (aos 2, 4 e 7 anos para o EXP1 e aos 3, 5 e 7 anos para o EXP2), avaliou-se o crescimento em altura (ALT), o diâmetro à altura do peito (DAP) e o volume comercial de madeira com casca (VOL), analisando-se os dados pela metodologia REML/BLUP. Para VOL aos sete anos de idade foi feita a análise de covariância para identificação da competição e suas formas através de sete covariáveis: índice de competição de Hegyi (IC), auto-competição (AT), alo-competição (AL), média da autocompetição (MAT), média da alocompetição (MAL) e media aritmética dos quatro (M4) e oito vizinhos mais próximos (M8). O efeito dessas covariáveis foi estudado individualmente e em todas as suas possíveis combinações, avaliando-se as alterações em todos os componentes de variância. Ainda aos sete anos foi efetuada a seleção para os três caracteres, com e sem o auxílio de covariáveis de competição e, para VOL, em todas as idades, foi simulada a seleção com e sem o auxilio de covariáveis de competição, comparando-se os parâmetros genéticos e a eficiência da seleção precoce em relação à idade adulta. A auto-competição parece causar menor variabilidade genética e erro entre parcelas, já a alo-competição interfere de forma oposta. A rotina de análise que apresentou melhores resultados foi a que incluiu as covariáveis MAT/IC. A competição intergenotípica causa...
The aim in this study was to identify the forms of competition in eucalyptus progeny tests and their influence on genetic parameters and selection. Data from two progeny tests of open pollinated families of eucalyptus were used. The plants were set up in a randomized plot design, in which experiment 1 (EXP1) comprised four control (clones) and 49 families and experiment 2 (EXP2) comprised 44 families. Both experiments consisted of six replications and linear plots of 10 plants. At three ages (2, 4 and 7 years old for EXP1 and 3, 5 and 7 years old for EXP2), growth was assessed in terms of height (H), diameter at breast height (DBH) and commercial volume of wood with bark (VOL). The data were analyzed using the REML/BLUP methodology. For VOL at seven years old, covariance analysis was performed to identify the competition and its forms, through seven covariables: Hegyi competition index (HCI), autocompetition (AUT), allocompetition (AL), mean autocompetition (MAT), mean allocompetition (MAL) and arithmetic mean of the closest four (M4) and eight (M8) neighbors. The effect from these covariables was studied singly and in all possible combinations, and the changes to all variance components were evaluated. Selection for three characteristics with or without the aid of competition covariables was also done after seven years old. For VOL, at all ages, selection with or without competition covariables was simulated by comparing the genetic parameters and early selection efficiency in relation to mature age. Autocompetition seemed to cause less genetic variability and error between plots, while allocompetition had the opposite effect. The analysis routine that presented the best results was the one that included the MAT and HCI covariables. Intergenotype competition caused selection bias among the eucalyptus open pollinated families and might have given rise to incorrect choice of genetic... (Complete abstract click electronic access below)
The aim in this study was to identify the forms of competition in eucalyptus progeny tests and their influence on genetic parameters and selection. Data from two progeny tests of open pollinated families of eucalyptus were used. The plants were set up in a randomized plot design, in which experiment 1 (EXP1) comprised four control (clones) and 49 families and experiment 2 (EXP2) comprised 44 families. Both experiments consisted of six replications and linear plots of 10 plants. At three ages (2, 4 and 7 years old for EXP1 and 3, 5 and 7 years old for EXP2), growth was assessed in terms of height (H), diameter at breast height (DBH) and commercial volume of wood with bark (VOL). The data were analyzed using the REML/BLUP methodology. For VOL at seven years old, covariance analysis was performed to identify the competition and its forms, through seven covariables: Hegyi competition index (HCI), autocompetition (AUT), allocompetition (AL), mean autocompetition (MAT), mean allocompetition (MAL) and arithmetic mean of the closest four (M4) and eight (M8) neighbors. The effect from these covariables was studied singly and in all possible combinations, and the changes to all variance components were evaluated. Selection for three characteristics with or without the aid of competition covariables was also done after seven years old. For VOL, at all ages, selection with or without competition covariables was simulated by comparing the genetic parameters and early selection efficiency in relation to mature age. Autocompetition seemed to cause less genetic variability and error between plots, while allocompetition had the opposite effect. The analysis routine that presented the best results was the one that included the MAT and HCI covariables. Intergenotype competition caused selection bias among the eucalyptus open pollinated families and might have given rise to incorrect choice of genetic... (Complete abstract click electronic access below)
Descrição
Palavras-chave
Forest improvement, Mixed models, Selection methods, Florestas, Melhoramento florestal, Modelo misto, Métodos de seleção
Como citar
PAVAN, Bruno Ettore. Competição em testes de progênies de eucalipto e suas implicações na seleção e no melhoramento. 2009. xi, 124 f. Tese (doutorado) - Universidade Estadual Paulista, Faculdade de Ciências Agrárias e Veterinárias, 2009.