Validação de modelo fenológico de produtividade de cafeeiro no sul do estado de Minas Gerais

• Autor(a) principal: Miranda, Wezer Lismar
• Data de Publicação: 2015
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/10777
• topic: Fenologia do cafeeiro; Coffea arabica L.; Previsão de produtividade; Modelagem agrometeorológica; Coffee phenology; Productivity prediction; Agrometeorological modeling; Engenharia de Água e Solo

Coffee production is influenced by the variations of the climatic elements, by the occurrence of adverse conditions (frost, dry spells, etc.) and by plant physiology (bieniannial). The prediction for coffee productivity is very complex due to the intricate physiological mechanisms, cultivar diversity and management conditions. Considering the importance of predicting the productivity for market regulation, with the present study we aimed at validating a model for estimating coffee productivity, based on the use of phenological indexes, submitted to the edaphoclimatic conditions of the southern region of Minas Gerais, Brazil. For this, we used 10 sample plots, distributed in the municipalities of Lavras, Varginha, Carmo de Minas, Ijaci and Santo Antonio do Amparo. The criteria for choosing the plots was based on the existence of a history of productivity superior to 40 sc ha-1 . The necessary meteorological information were collected from the Main Climatological Station belonging to INMET and installed at the campus of the Universidade Federal de Lavras (UFLA, considering as representative of the region. We conducted a hydric balance for the period in which the study was conducted, encompassing the harvests of 2012/2013 and 2013/2014. The collection of phenological data was conducted during the months of September/October, December/January and March/April. The phenological data observed were number of flowers and fruits in the fourth and fifth productive knot of plagiotropic branches of the middle third of the chosen plants. The choosing of the plants was done randomly from walking in zigzag, every ten meters of distance in four between coffee lines, alternately collecting information between branches on the right and on the left. We sampled 40 plants in each plot, of which heights were measured to comprise the Productivity Phenological Indexes 1 and 2 (IFP1 and IFP2). For comparison and validation of the models, we obtained information regarding the production observed. The prediction results were submitted to statistical tests of coefficient of determination, coefficient of correlation, concordance index, C index, mean error, mean absolute deviation, mean absolute error percentage and mean quadratic error. The results indicated that the analyzed models are applicable to coffee crops of the southern region of Minas Gerais. IFP1 and IFP2 presented good correlation with the productivity observed. The adverse meteorological occurrences of the period influenced the performance of the model.