MODELAGEM DA PRODUTIVIDADE DE SOJA EM FUNÇÃO DE VARIÁVEIS METEOROLÓGICAS NO CENTRO-SUL DO PARANÁ

• Ano de defesa: 2018
• Autor(a) principal: Ferreira, Rodrigo
• Orientador(a): Kawakami, Jackson
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Centro-Oeste
• Programa de Pós-Graduação: Programa de Pós-Graduação em Agronomia (Mestrado)
• Departamento: Unicentro::Departamento de Agronomia
• País: Brasil
• Palavras-chave em Português: CSM-CROPGRO-Soybean; regressão múltipla; simulação; modelo estatístico
• Palavras-chave em Inglês: CSM-CROPGRO-Soybean; multiple regression; simulation; statistical model
• Área do conhecimento CNPq: CIENCIAS AGRARIAS::AGRONOMIA
• Link de acesso: http://tede.unicentro.br:8080/jspui/handle/jspui/1189

The present study brought two distinct approaches for predicting soybean yield in the Central Region of Parana State through the mathematical crop modeling. It was used information from the soybean trials conducted by the Agraria Foundation for Agricultural Research (FAPA) in Entre Rios District-Guarapuava-PR, Candoi-PR and Pinhao-PR, for the soybean cultivars BMX APOLO RR, PRODUZA IPRO and TMG7262 RR, sowed in three seasons during five years (2012/13 to 2016/17). The first approach deals with the development of a statistical model for soybean yield as a function of weather variables, evaluating its capacity to estimate yields at different sowing times. The second approach brings the CSM-CROPGRO-Soybean mechanistic model and the objective was to calibrate and evaluate the capacity of this model to simulate the development and yield of three soybean cultivars in the environmental conditions of the Central Region of Parana State. The statistical model was able to estimate soybean yield at different sowing times throughout the study years. Due to the techniques used in its development, its estimates are more restricted to environments similar to those of model development. About CSM-CROPGRO-Soybean, the results of this model show that the genetic coefficients of the three soybean cultivars were well calibrated, resulting in simulated developmental characteristics in agreement with the observed values. The model evaluation also showed agreement between simulated and observed data for soybean yield. Both models presented adequate capacity to simulate the development and soybeans yields for the Central Region of Parana State and can be used as tools to help understanding the effects of the environment on soybean yield.