Floristic and structural variation of weeds in eucalyptus plantations as influenced by relief and time of year

The objective of this work was to analyze the floristic variation and phytosociological structure of weeds as influenced by relief and time of year in eucalyptus plantations in Santana do Paraíso and Guanhães - MG. The total area sampled for each locality was approximately 10 ± 3 hectares, comprising three types of relief: lowland, slope, and upper area. In each type of relief, 10 plots of 1 m² were sampled, corresponding to 30 plots per locality, where they were randomly allocated in a zigzag. The taxonomic identification was performed in four assessments, corresponding to the months of November and March, comprising two ratings each season, always at the same points, and geo-referenced using the Global Positioning System (GPS). A total of 3,893 individuals, 18 families and 61 species, were identified in Santana do Paraiso and a total of 1,166 individuals, 13 families and 58 species, in Guanhães. In both localities, the most representative families in terms of wealth were: Poaceae, Asteraceae, and Fabaceae. Galinsoga parviflora was the most abundant species. The Vernonia polyantes was identified only in the lowlands, while Arrabida florida was identified in the slope and upper area. On the other hand, Emilia coccinea, Sida rhombifolia, S. paniculatum and Spermacoce latifolia were common to all three environments. Commelina benghalensis was present only in the month of March, while G. parviflora was present only in the month of November. It was concluded that the floristic and phytosociological variation of weeds in eucalyptus plantations is influenced by the type of relief and time of year, which should guide the management practices used in the culture. Objetivou-se analisar a variação florística e a estrutura fitossociológica de plantas daninhas em função do relevo e da época do ano em cultivos de eucalipto nos municípios de Santana do Paraíso e Guanhães-MG. A área total amostrada para cada localidade foi de aproximadamente 10 ± 3 hectares, compreendidos em três tipos de relevo: baixada, encosta e parte alta. Em cada tipo de relevo, foram amostradas 10 parcelas de 1 m², correspondendo a 30 parcelas por município, onde foram distribuídas aleatoriamente em zigue-zague. A identificação taxonômica foi efetuada em quatro avaliações, correspondentes aos meses de novembro dos anos de 2009 e 2010 e março de 2010 e 2011, sempre nos mesmos pontos, os quais foram georreferenciados por meio do Sistema de Posicionamento Global (GPS). Foi identificado, no município d