Eficiência do modelo CERES-Maize na simulação do desempenho de híbridos de milho

• Ano de defesa: 2008
• Autor(a) principal: Pereira, Antônio Marcos de Andrade Rezende
• Orientador(a): Von Pinho, Renzo Garcia
• Banca de defesa: Souza, João Cândido de, Soares, Antônio Alves
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: UNIVERSIDADE FEDERAL DE LAVRAS
• Programa de Pós-Graduação: DAG - Programa de Pós-graduação
• País: BRASIL
• Palavras-chave em Português: CERES-Maize; Simulação; Milho; Desempenho; Simulation
• Área do conhecimento CNPq: CNPQ_NÃO_INFORMADO
• Link de acesso: https://repositorio.ufla.br/handle/1/4011

The aim of this work was to verify the efficiency of the CERES-Maize model to simulate some agronomic characteristics in different corn hybrids. One experiment was installed at Federal University Lavras, Lavras, MG; and consisted of a complete randomized block design with three replications, five hybrids (AG 7000, AG 8060, DKB 199, GNZ 2004 and P30F90) and three sowing dates (11/24/2006, 12/19/2006 and 01/13/2007). The data collected were: flowering and physiologic maturity date, number of grains per square meter, grain mass, dry matter and productivity. These data were compared to the simulated ones by using the root mean square error (RSME), deviation percentage (PD) and the agreement index (D). The results showed that the corn sowed in January, independent of the hybrid; presented smaller values of number of grains per square meter, grain mass, dry matter and productivity than those hybrids sowed in November and December. Based on the comparison tools used in this experiment, CERES-Maize model was very efficient to simulate flowering and physiologic maturity date. The values of RSME obtained were less than 10% and D value higher than 0,80. For booth characteristics, the highest PD value was -11%. The simulation of number of grains for square meter, grain mass and productivity was considered good with obtained values of RSME smaller than 20%. For these simulated characteristics, higher PD values were observed in the last sowing time. It was evident; in this case, that environmental conditions, not favorable for the development of the hybrids, did affected the efficiency of the simulation. Ceres-Maize model showed to be an excellent tool for simulation of corn hybrids agronomic characteristics.