Simulação da produtividade de mandioca para o estado do Rio Grande do Sul

• Ano de defesa: 2013
• Autor(a) principal: Gabriel, Luana Fernandes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Agronomia; UFSM; Programa de Pós-Graduação em Agronomia
• Palavras-chave em Português: GUMCAS; Modelagem; Segurança alimentar; Cultura energética; Modeling; Food security; Energy crops; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA
• Link de acesso: http://repositorio.ufsm.br/handle/1/5076

Cassava plays an important role for food security in the tropics, where it is the third main food source. In Rio Grande do Sul State, Brazil, cassava is a subsistence crop for small farms, who use cassava for human and livestock food. The objective of this dissertation was to adapt and test a process-based model for the cassava crop in subtropical environment of the State of Rio Grande do Sul. We made three modifications in the model compared to the original version in Matthews and Hunt (1994): we considered a third independent clock in the cassava development for the onset of starch accumulation, we replaced the rate of leaf appearance sub-model, by the Wang e Engel model modified for cassava, and we modified the leaf senescence sub-model. Calibration of the model was performed by estimating parameters using the trial and error approach (Matthews and Hunt, 1994) that minimized the root mean square error between observed and estimated values, with total of 16 parameters were calibrated. A total of 25 independent datasets from experiments conducted in four sites of Rio Grande do Sul State (Santa Maria, Glorinha, Vera Cruz and Rio Pardo) were used for testing the model performance and a sensitivity analysis was performed running the GUMCAS model for 17 locations throughout the State of Rio Grande do Sul. In general, the calibration and modifications introduced in the GUMCAS model resulted in good simulations of some key ecophysiological processes such as leaf development and growth as well as of storage roots yield for a cassava genotype adapted to the subtropics. The model was able to capture different environmental conditions accross the Rio Grande do Sul State and with some adaptations for inputting data was able to simulate different management practices such as planting date, plant spacing, plant density, partial above-ground pruning during the growing cycle and two growing cycles as well as the effect of extreme weather events such as hail.