Classificação da qualidade dos experimentos com cana de açúcar
Arquivos
Data
2010-02-04
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Viçosa
Resumo
O coeficiente de variação como medida de precisão experimental tem sido utilizado para diversas culturas. O presente trabalho teve como objetivo geral avaliar o coeficiente de variação e acurácia seletiva como estatísticas candidatas a precisão experimental em ensaios com cana de açúcar. Objetivos específicos: (i) Propor faixas de classificação do coeficiente de variação em experimentos com cana de açúcar para diversas variáveis utilizadas por pesquisadores da cultura; (ii) estabelecer limites adequados para o coeficiente de variação, com aplicabilidade em experimentos com cana de açúcar, considerando as variáveis mais utilizadas nos ensaios, através do critério de classificação proposto por Costa (2002), levando em consideração as diferenças de produtividade ao longo dos cortes; (iii) Avaliar o coeficiente de variação e a acurácia seletiva como medidas de precisão experimental em experimentos com cana de açúcar. Para esse estudo foram consideradas algumas das variáveis mais utilizadas por pesquisadores desta cultura. Os dados foram obtidos de extensa revisão bibliográfica em revistas científicas, banco de dados do programa de melhoramento de cana de açúcar da Universidade Federal de Viçosa, dissertações de mestrado, teses de doutorado e relatórios técnicos. Foram propostas faixas de classificação segundo o método de Costa (2002), o qual se baseia no uso da mediana e do pseudo-sigma. As variáveis altura e brix apresentaram as menores faixas de classificação de CV, seguidas pelo número de colmos, Pol % da cana e tonelada de brix por hectare. As variáveis tonelada de colmo por hectare, tonelada de pol por hectare e peso da parcela apresentaram as maiores faixas de classificação do coeficiente de variação. Todas as variáveis apresentaram faixas de classificação específica, evidenciando a necessidade de se considerar, na classificação dos CV, a natureza da variável estudada. Para os limites de classificação por cortes os dados utilizados foram obtidos do banco de dados do programa de melhoramento de cana de açúcar da Universidade Federal de Viçosa separados por cortes 1º, 2º e 3º. Foram propostos faixas de classificação dentro de cada corte. Para isso utilizou-se o método proposto por Costa et al. (2002). Houve discordância quanto às faixas de classificação por corte para cada variável, acarretando em faixas de classificação especifica. Este resultado evidencia a necessidade de se considerar, na classificação dos CV, a natureza da variável estudada e os cortes. A variável TPH e TCH apresentaram maiores limites nas faixas de classificação por cortes. Com o objetivo de propor uma avaliação da acurácia seletiva como candidata a precisão experimental em ensaios do programa de melhoramento genético da cana de açúcar da Universidade Federal de Viçosa. Foram realizadas análises de variância para os dados de produtividade de cana de açúcar de 111 ensaios, e coletados os valores das estatísticas: quadrado médio do genótipo, quadrado médio do erro, média geral do ensaio, valor do teste F para genótipo e coeficiente de variação experimental. A seguir, foram estimadas as estatísticas, coeficiente de variação genotípica (CVg), coeficiente de variação experimental (CVe), herdabilidade (h2), acurácia seletiva (AS). Valores de F para as cultivares de cana de açúcar devem ser iguais ou superiores a 5,05 pra se obter uma acurácia tida como ideal (≥ 0,90). Experimentos com valores do F abaixo de 1,17 devem ser descartados, pois não atendem a uma acurácia mínima. Concluiu-se que o coeficiente de variação é uma estatística recomendada para avaliar a precisão experimental em experimentos com cana de açúcar com médias semelhantes ou de mesmo corte, pois para cada corte existe limites de classificação diferenciado. A acurácia seletiva apresentou-se como uma candidata viável para classificar experimentos advindos do programas de melhoramento genético da cana de açúcar por levar em consideração os atributos número de repetições e as variações ambiental e genética como ficou demonstrado neste trabalho.
The coefficient of variation as an experimental precision measure has been used for several crops. The present work had as general objective to evaluate the coefficient of variation and the selective accuracy as candidate statistics of experimental precision in essays with sugar cane. Specific objectives: (i) to propose classification ranges of coefficient of variation (CV) in experiments with sugar cane for several variables commonly used by this crop researchers; (ii) to establish adequate limits for the coefficient of variation, with applicability in experiments with sugar cane, considering the must used variables in the essays, through the classification criterion proposed by Costa (2002), taking into consideration the productivity differences along the cuts; (iii) to evaluate the coefficient of variation and the selective accuracy as experimental precision measures in experiments with sugar cane. The data were gotten from exhaustive bibliographic revision in scientific journals, from the sugar cane improvement program database of the Universidade Federal de Viçosa, master degree dissertations, doctor degree thesis and technical reports. Classification ranges were proposed according to the method of Costa (2002), which is based in the use of the median and the pseudo-sigma. The variables height and brix presented the smallest classification ranges of CV, followed by the number of stalks, Pol % of cane and ton of brix per hectare. The variables ton of cane per hectare (TCH), ton of pol per hectare (TPH) and weigh of the plot presented the larges.t classification ranges of the CV All variables presented specific classification ranges, making evident the necessity of considering, in the CV classification, the nature of the studied variable. For the limits of classification by cuts the used data were gotten from the sugar cane improvement program database of the Universidade Federal de Viçosa separated by cuts 1st, 2nd and 3rd. Classification ranges were proposed within each cut. The method proposed by Costa et al (2002) was used, which is based in the use of median and pseudo-sigma. There was disagreement to the specific classification ranges. This result makes evident the necessity of considering, in the classification of CV, the nature of the studied variable and the cuts. The variable TPH and TCH presented wider limits in the classification ranges by cuts. As the objective to propose a selective accuracy evaluation as candidate to the experimental precision classification in essays of the sugar cane genetic improvement program of the Universidade Federal de Viçosa. Analysis of variance were made for the sugar cane productivity data of 111 essays, and collected the values of the statistics: genotype mean square, error mean square, overall mean, value of the F statistic for genotype, and the experimental coefficient of variation. The following statistics were also estimated: genotypic coefficient of variation (CVg), experimental coefficient of variantion (CVe), coefficient of genetic determination between cultivars in the level of plot (h2) and selective accuracy (SA). For a safe inference, the F values for sugar cane cultivations must be equal or superior to 5.05, experiments with F values under 1.17 must be discharged, therefore they don t attend to the minimum accuracy. The SA presents itself as a possible candidate to classify experiments coming from the sugar cane improvement genetics program for taking into consideration the attributes, repetitions number, environmental variation and genetics, as became demonstrated in this work.
The coefficient of variation as an experimental precision measure has been used for several crops. The present work had as general objective to evaluate the coefficient of variation and the selective accuracy as candidate statistics of experimental precision in essays with sugar cane. Specific objectives: (i) to propose classification ranges of coefficient of variation (CV) in experiments with sugar cane for several variables commonly used by this crop researchers; (ii) to establish adequate limits for the coefficient of variation, with applicability in experiments with sugar cane, considering the must used variables in the essays, through the classification criterion proposed by Costa (2002), taking into consideration the productivity differences along the cuts; (iii) to evaluate the coefficient of variation and the selective accuracy as experimental precision measures in experiments with sugar cane. The data were gotten from exhaustive bibliographic revision in scientific journals, from the sugar cane improvement program database of the Universidade Federal de Viçosa, master degree dissertations, doctor degree thesis and technical reports. Classification ranges were proposed according to the method of Costa (2002), which is based in the use of the median and the pseudo-sigma. The variables height and brix presented the smallest classification ranges of CV, followed by the number of stalks, Pol % of cane and ton of brix per hectare. The variables ton of cane per hectare (TCH), ton of pol per hectare (TPH) and weigh of the plot presented the larges.t classification ranges of the CV All variables presented specific classification ranges, making evident the necessity of considering, in the CV classification, the nature of the studied variable. For the limits of classification by cuts the used data were gotten from the sugar cane improvement program database of the Universidade Federal de Viçosa separated by cuts 1st, 2nd and 3rd. Classification ranges were proposed within each cut. The method proposed by Costa et al (2002) was used, which is based in the use of median and pseudo-sigma. There was disagreement to the specific classification ranges. This result makes evident the necessity of considering, in the classification of CV, the nature of the studied variable and the cuts. The variable TPH and TCH presented wider limits in the classification ranges by cuts. As the objective to propose a selective accuracy evaluation as candidate to the experimental precision classification in essays of the sugar cane genetic improvement program of the Universidade Federal de Viçosa. Analysis of variance were made for the sugar cane productivity data of 111 essays, and collected the values of the statistics: genotype mean square, error mean square, overall mean, value of the F statistic for genotype, and the experimental coefficient of variation. The following statistics were also estimated: genotypic coefficient of variation (CVg), experimental coefficient of variantion (CVe), coefficient of genetic determination between cultivars in the level of plot (h2) and selective accuracy (SA). For a safe inference, the F values for sugar cane cultivations must be equal or superior to 5.05, experiments with F values under 1.17 must be discharged, therefore they don t attend to the minimum accuracy. The SA presents itself as a possible candidate to classify experiments coming from the sugar cane improvement genetics program for taking into consideration the attributes, repetitions number, environmental variation and genetics, as became demonstrated in this work.
Descrição
Palavras-chave
Precisão experimental, Classes de coeficientes de variação, Seleção, Experimental precision, Classes of coefficients of variation, Selection
Citação
COUTO, Maurício Farias. Rating the quality of experiments with sugarcane. 2010. 63 f. Dissertação (Mestrado em Estatística Aplicada e Biometria) - Universidade Federal de Viçosa, Viçosa, 2010.