Identificação de cultivares de café por meio da classificação por tipo e de técnicas de análise multivariada
Arquivos
Data
2007
Título da Revista
ISSN da Revista
Título de Volume
Editor
Resumo
Este trabalho teve como objetivo identificar e separar 16 cultivares de cultivares de Coffea arabica L, por meio de técnicas de análise multivariada, usando a avaliação dos defeitos do grão de café. Os frutos avaliados eram provenientes da Fazenda Experimental de Varginha, região Sul de Minas Gerais, colhidos na safra 2002/2003. As amostras foram preparadas no Pólo de Tecnologia em Pós-Colheita do Café da UFLA, sendo lavados, descascados e secados em terreiro de concreto. Após o beneficiamento, os grãos foram acondicionados em latas de alumínio e armazenados à 15 oC. Para a classificação por tipo, utilizou-se 300g de amostra de café beneficiado, na qual determinaram-se a presença, a quantidade e o somatório da equivalência dos defeitos intrísecos. Os dados foram submetidos ao software SAS para a determinação das variáveis canônicas, e da medida da similaridade por meio da Distância de Mahalanobis. Os resultados permitiram observar diferenças entre as cultivares, sendo possível separá-las em função dos defeitos analisados. Os defeitos observados que mais contribuíram para a separação das cultivares foram o chocho, o concha, o brocado e o ardido.
This work aimed to identify and to separate 16 cultivars the Coffea arabica L, through techniques of analysis multivariada, using the evaluation of the defects of the coffee bean. The appraised fruits were coming of Varginha Experimental Farm, in South Minas Gerais State - Brazil, were harvested in 2002/2003 crop. The samples were prepared in the Coffee Post Harvest Technology Center in UFLA, where were washed, peeled and dried in a concret yard. After the benefaction, the grains were packaged in aluminum cans and stored at 15 oC. For the classification for type, it was used 300g of sample of benefited coffee, they were determined the presence, the amount and the sum of the equivalence of the defects intricacies. The data were submitted to the SAS software for the determination of the canonical variables, and of the measure of the similarity through the Distance of Mahalanobis. The results allowed to observe differences among them cultivate, being possible to separate them in function of the analyzed defects. The defects observed that more contributed to the separation of the cultivars were the doddering, the shell, the brocade and the burning.
This work aimed to identify and to separate 16 cultivars the Coffea arabica L, through techniques of analysis multivariada, using the evaluation of the defects of the coffee bean. The appraised fruits were coming of Varginha Experimental Farm, in South Minas Gerais State - Brazil, were harvested in 2002/2003 crop. The samples were prepared in the Coffee Post Harvest Technology Center in UFLA, where were washed, peeled and dried in a concret yard. After the benefaction, the grains were packaged in aluminum cans and stored at 15 oC. For the classification for type, it was used 300g of sample of benefited coffee, they were determined the presence, the amount and the sum of the equivalence of the defects intricacies. The data were submitted to the SAS software for the determination of the canonical variables, and of the measure of the similarity through the Distance of Mahalanobis. The results allowed to observe differences among them cultivate, being possible to separate them in function of the analyzed defects. The defects observed that more contributed to the separation of the cultivars were the doddering, the shell, the brocade and the burning.
Descrição
Trabalho apresentado no Simpósio de Pesquisa dos Cafés do Brasil (5. : 2007 : Águas de Lindóia, SP). Anais. Brasília, D.F. : Embrapa Café, 2007.
Palavras-chave
Café Defeitos Cultivares Qualidade Análise multivariada, Coffee Defects Cultivars Quality Multivariate analysis
Citação
Mendonça, Luciana Maria Vieira Lopes; Pereira, Rosemary Gualberto Fonseca Alvarenga; Ferreira, Daniel Furtado; Borém, Flávio Meira; Mendonça, José Marcos Angélico de. Identificação de cultivares de café por meio da classificação por tipo e de técnicas de análise multivariada. In: Simpósio de Pesquisa dos Cafés do Brasil (5. : Águas de Lindóia, SP : 2007). Anais. Brasília, D.F. : Embrapa - Café, 2007. (1 CD-ROM), 5p.