Uso de índices de reflectância foliar no monitoramento do patossistema Microcyclus ulei x seringueira
Carregando...
Data
2017-12-15
Autores
Bevenuto, João Alberto Zago [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O maior problema fitossanitário na heveicultura brasileira é a doença conhecida por mal das folhas, causada pelo fungo Microcyclus ulei, cujos sintomas ocorrem nas folhas jovens, causando intensas desfolhas, diminuindo a produção de látex e até mesmo a morte em plantas muito suscetíveis. A espécie Hevea brasiliensis possui um hábito deciduifólio completo que é um caráter importante, ou seja, o desfolhamento uniforme proporciona a recuperação da copa com folhas sadias e a redução do inóculo de patógenos que são descartados com as folhas velhas, onde as estruturas reprodutivas estão localizadas. Os objetivos deste trabalho foram: utilizar ferramentas de sensoriamento remoto para confrontar o efeito das variáveis climáticas em diferentes períodos do ano sobre interferência do mal das folhas em seis anos comparando o avanço da doença com os índices de vegetação, Índice de Vegetação com Diferença Normalizada (NDVI), Índice de Vegetação Ajustado para os Efeitos do Solo (SAVI) e Índice de Área Foliar (IAF), calculados para o período de cada ano estudado, analisar o efeito deciduifólio natural ou ocorrência do mal das folhas; e verificar a troca de folhas dos clones de seringueira em estudo comparando com a fenologia. A área utilizada para o estudo localiza-se no município de Registro, Estado de São Paulo, Brasil, na Fazenda Umuarama com plantio monoclonal de seringueira, foram avaliados seis clones diferentes: IAN 873, IAN 717; RRIM 600, Fx 3864, Fx 2261 e Fx 3844, por meio das imagens do satélite Landsat 5 utilizado os índices de vegetação NDVI, SAVI e IAF. Foram utilizados para análise estatística modelos lineares generalizados com a distribuição gama e função de ligação logarítmica tendo como fatores clones e índices de reflectância. Os modelos foram ajustados incluindo-se as covariáveis continuas mensurada no campo: folhas caídas totais e folhas caídas doentes Esses modelos foram comparados através do critério de informação de Akaike corrigido – AICC, para identificar o melhor modelo, verificada mediante desvios por graus de liberdade (scale deviance). Para comparações entre fatores foi utilizado foi o teste de Tukey–Kramer. Realizou a correlação de Spearman entre os índices de vegetação: NDVI, SAVI e IAF. Concluiu-se que as ferramentas de sensoriamento remoto são aplicáveis nos efeitos que as variações climáticas sobre a influência da doença mal das folhas. Os índices NDVI, SAVI e IAF foram significativos para a queda foliar da seringueira. Verificou também através dos índices a fenologia da seringueira nos períodos de troca de folhas: desfolha, reenfolha e densidade de copa. Ocorreu interação estatística significativa sobre o efeito dos clones, sua fenologia e índices de vegetação nos pixels da imagem. Obteve-se bom ajuste dos modelos dos índices com as imagens de satélites e as covariáveis dos dados de campo. A correlação de Spearman mostrou-se significativa entre os índices por valores dos pixels. Os índices são ferramentas de grande valia para estudos e análises sobre plantios florestais.
The biggest phytosanitary problem in Brazilian heveculture is the disease known as leaf blight caused by the fungus Microcyclus ulei. Whose symptoms occur in young leaves, causing severe defoliation, reducing the production of latex and even the death in susceptible plants. Rubber tree has a deciduous habit which is an important character. Uniform defoliation of clones provides a reduction in the interior of pathogens, whose locations to reproduction occur in the old leaves. The objectives of this work were: Using remote sensing tools to compare the effect of climatic variables in different periods of the year on leaf mischief interference in six years comparing the disease progression with Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI) and Leaf Area Index (LAI) vegetation indices, calculated for the period of each year studied, verifying the natural deciduous effect or occurrence of leaf blight; Estimate an exchange of leaves of the rubber tree clones under study, comparing them with phenology. The area used for the study is located in the municipality of Registro, São Paulo State, Brazil, at Umuarama Farm with monoclonal rubber plantation, IAM 873, IAN 717; RRIM 600, Fx 3864, Fx 2261 and Fx 3844, using Landsat 5 satellite images used in NDVI, SAVI and IAF vegetation indices. They were used for statistical analysis generalized linear model with logarithmic range distribution and binding function having as clones reflectance factors and indices. Set the same models including whether the continuous covariates measured in the field: total fallen leaves sick and fallen leaves. Compared these models through the Akaike information criterion corrected - AICC to identify the best model, verified through the deviations by degrees of freedom (scale deviance). Was compared between factors was used was the Tukey-Kramer test. Spearman correlation was performed between the vegetation indexes: NDVI, SAVI and LAI. It was concluded that remote sensing tools are applicable in the effects that climatic variations on the influence of evil leaf disease. The NDVI, SAVI and IAF indexes showed the leaf fall effect of the rubber tree. Also verified through the indices the phenology of the rubber tree in the periods of leaf change: defoliation, re-foliage and cup density. There was a statistically significant interaction on the effect of clones’ phenology and vegetation index in the pixels of the image. It obtained good adjustment of the models of the contents with the satellite images and field data covariates. The Spearman correlation coefficient was significant between the indexes through the values of the pixels. The indices are valuable tools for studies and analyzes on forest plantations.
The biggest phytosanitary problem in Brazilian heveculture is the disease known as leaf blight caused by the fungus Microcyclus ulei. Whose symptoms occur in young leaves, causing severe defoliation, reducing the production of latex and even the death in susceptible plants. Rubber tree has a deciduous habit which is an important character. Uniform defoliation of clones provides a reduction in the interior of pathogens, whose locations to reproduction occur in the old leaves. The objectives of this work were: Using remote sensing tools to compare the effect of climatic variables in different periods of the year on leaf mischief interference in six years comparing the disease progression with Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI) and Leaf Area Index (LAI) vegetation indices, calculated for the period of each year studied, verifying the natural deciduous effect or occurrence of leaf blight; Estimate an exchange of leaves of the rubber tree clones under study, comparing them with phenology. The area used for the study is located in the municipality of Registro, São Paulo State, Brazil, at Umuarama Farm with monoclonal rubber plantation, IAM 873, IAN 717; RRIM 600, Fx 3864, Fx 2261 and Fx 3844, using Landsat 5 satellite images used in NDVI, SAVI and IAF vegetation indices. They were used for statistical analysis generalized linear model with logarithmic range distribution and binding function having as clones reflectance factors and indices. Set the same models including whether the continuous covariates measured in the field: total fallen leaves sick and fallen leaves. Compared these models through the Akaike information criterion corrected - AICC to identify the best model, verified through the deviations by degrees of freedom (scale deviance). Was compared between factors was used was the Tukey-Kramer test. Spearman correlation was performed between the vegetation indexes: NDVI, SAVI and LAI. It was concluded that remote sensing tools are applicable in the effects that climatic variations on the influence of evil leaf disease. The NDVI, SAVI and IAF indexes showed the leaf fall effect of the rubber tree. Also verified through the indices the phenology of the rubber tree in the periods of leaf change: defoliation, re-foliage and cup density. There was a statistically significant interaction on the effect of clones’ phenology and vegetation index in the pixels of the image. It obtained good adjustment of the models of the contents with the satellite images and field data covariates. The Spearman correlation coefficient was significant between the indexes through the values of the pixels. The indices are valuable tools for studies and analyzes on forest plantations.
Descrição
Palavras-chave
NDVI, SAVI, IAF, índice de vegetação, Hevea, mal das folhas