Monitoramento da comunidade de plantas daninhas na cana-de-açúcar e da eficiência de controle químico por meio de técnicas de análise multivariada de dados
Carregando...
Data
2012-02-16
Autores
Squassoni, Vanessa Luzia [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Para minimizar os efeitos negativos da interferência das plantas daninhas, o controle químico é o método mais utilizado nos canaviais devido à praticidade e o menor custo. O processo de controle de plantas daninhas é governado por muitas variáveis, ou seja, é multivariado. Entre a escolha do tratamento químico e o resultado final de controle das plantas daninhas atuam vários fatores, o que explicam a grande variabilidade de resultados de controle que se obtém nos canaviais. Quanto maior for o conhecimento sobre a ação integrada desses fatores, melhor será a escolha do tratamento químico para cada unidade de área, daí a necessidade de se organizar os fatores envolvidos num banco de dados e analisá-los. Neste processo, dois pontos são fundamentais: o conhecimento da flora de plantas daninhas e da sua organização e o controle de resultados proporcionados pelos diferentes tratamentos químicos. Desta forma, utilizando técnicas de análise exploratória de dados através da análise de agrupamento e análise discriminante linear, objetivou-se com este estudo: organizar as comunidades de plantas daninhas em agrupamentos padrões; avaliar com que eficácia estão sendo tratados os principais tipos de infestação predominantes nesses canaviais; avaliar a eficiência dos principais tratamentos químicos aplicados, assim como a sua relação com o custo dos tratamentos empregados. Para isso, utilizou-se um banco de dados de até 250 registros de talhões de cana-de-açúcar, contendo informações de infestação e controle de plantas daninhas, e tratamentos químicos empregados, que foram fornecidos pela Usina Santa Cruz, localizada em Américo Brasiliense, SP. A coleta de dados foi realizada nos anos de 2009 e 2010, através de um sistema de monitoramento na cana soca e cana planta, e foram organizados em tabelas cadastrais e...
Chemical control is the most commonly used method on sugarcane fields to minimize the negative effects of weed interference, due to its convenience and low cost. The weed control process is ruled by many variables – i.e. a multivariate process. Several factors are present between choosing the chemical treatment and the final outcome of weed control – this may explain the great variability of control results that can be obtained from sugarcane fields. As knowledge of the integrated action of factors expands, the better are the choices for chemical treatment on each unit area – hence the need to organize the involved factors in a database and analyze them. Two are the key points during the process: knowledge of weed flora and their organization; and control of results provided by each chemical treatment. This way, by using exploratory data analysis techniques through cluster analysis and linear discriminant analysis, the objective of this study was to: organize weed communities into default clusters: evaluate how efficiently the main types of infestation that predominate in those sugarcane fields are being addressed; evaluate the efficiency of the main chemical treatments used, as well as the cost of the applied treatments. For this, we used a database of up to 250 registrations of sugarcane plots, with information about infestation and weed control and chemical treatment used – this information was provided by Santa Cruz Mill, which is located in the city of Américo Brasiliense-SP. Data collection was carried out during 2009 and 2010, through a monitoring system of plant cane and ratoon cane; data was organized into registration tables, to be later submitted to multivariate statistical analysis. Two large groups of plots were detected through cluster analysis: one with high infestation and other with... (Complete abstract click electronic access below)
Chemical control is the most commonly used method on sugarcane fields to minimize the negative effects of weed interference, due to its convenience and low cost. The weed control process is ruled by many variables – i.e. a multivariate process. Several factors are present between choosing the chemical treatment and the final outcome of weed control – this may explain the great variability of control results that can be obtained from sugarcane fields. As knowledge of the integrated action of factors expands, the better are the choices for chemical treatment on each unit area – hence the need to organize the involved factors in a database and analyze them. Two are the key points during the process: knowledge of weed flora and their organization; and control of results provided by each chemical treatment. This way, by using exploratory data analysis techniques through cluster analysis and linear discriminant analysis, the objective of this study was to: organize weed communities into default clusters: evaluate how efficiently the main types of infestation that predominate in those sugarcane fields are being addressed; evaluate the efficiency of the main chemical treatments used, as well as the cost of the applied treatments. For this, we used a database of up to 250 registrations of sugarcane plots, with information about infestation and weed control and chemical treatment used – this information was provided by Santa Cruz Mill, which is located in the city of Américo Brasiliense-SP. Data collection was carried out during 2009 and 2010, through a monitoring system of plant cane and ratoon cane; data was organized into registration tables, to be later submitted to multivariate statistical analysis. Two large groups of plots were detected through cluster analysis: one with high infestation and other with... (Complete abstract click electronic access below)
Descrição
Palavras-chave
Cana-de-açúcar - Cultivo, Erva daninha - Controle químico, Herbicidas, Sugar cane - Chemical control
Como citar
SQUASSONI, Vanessa Luzia. Monitoramento da comunidade de plantas daninhas na cana-de-açúcar e da eficiência de controle químico por meio de técnicas de análise multivariada de dados. 2012. xx, 88 f. Dissertação (mestrado) - Universidade Estadual Paulista, Faculdade de Ciências Agrárias e Veterinárias, 2012.