Selection of non-linear models and the study of conilon coffee fruit growth

• Autor(a) principal: Senra, João Felipe de Brites
• Data de Publicação: 2022
• Outros Autores: Silva, Josimar Aleixo da, Ferreira, Adésio, Esposti, Marlon Dutra Degli, Silva, Uliana Ribeiro, Milheiros, Idalina Sturião, Zacarias, Alex Justino
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/27093
• topic: Modelado; Biometría; Coffea canephora; Clones.; Modelagem; Biometria; Coffea canephora; Clones.; Modeling; Biometrics; Coffea canephora; Clones.

The objective was to develop growth curves for fruit dry mass in Coffea canephora clones, select the best non-linear regression model, estimate the rate of mass gain, analyze differences in fruit development located in the lower, middle and upper thirds of the coffee canopy and generate an equation that describes the process. Eleven data collections were carried out, starting in the pellet phase of nine clones with 30 plants, with 50 fruits being collected in each position of the coffee tree canopies. To obtain the dry matter mass, the fruits were dried in an oven with forced air circulation at 65 °C until constant weight. The mathematical models Brody, Gompertz, Logístico, Mitscherlich and von Bertalanffy were applied. The quality of the equations was evaluated using eight statistical parameters and the confidence intervals of β1, β2 and β3 of the regressions estimated based on the likelihood profile. After selecting the best model, the fruit growth curves were estimated considering the three positions in the coffee canopy. All statistical analyzes were performed in the R software. The Logistic model presents greater reliability to describe the accumulation of dry matter mass in fruits. There were no differences between positions in the coffee canopy. The β3 parameter can be used as an early indicator for Coffea canephora and guide breeding programs. Clones 204, 407 and P1 provided curves with higher quality in relation to the parameters evaluated.