Growth models for morphological traits of sunn hemp

• Autor(a) principal: Bem, Cláudia Marques de
• Data de Publicação: 2017
• Outros Autores: Cargnelutti Filho, Alberto, Facco, Giovani, Schabarum, Denison Esequiel, Silveira, Daniela Lixinski, Simões, Fernanda Martins, Uliana, Daniela Barbieri
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Semina. Ciências Agrárias (Online)
• DOI: 10.5433/1679-0359.2017v38n5p2933
• Texto Completo: https://ojs.uel.br/revistas/uel/index.php/semagrarias/article/view/27740
• topic: Cover crop; Experimental planning; Modeling.; Modelagem; Planejamento experimental; Planta de cobertura.

The objective of the present study was to fit Gompertz and Logistic nonlinear to descriptions of morphological traits of sunn hemp. Two uniformity trials were conducted and the crops received identical treatment in all experimental area. Sunn hemp seeds were sown in rows 0.5 m apart with a plant density of 20 plants per row meter in a usable area of 52 m × 50 m. The following morphological traits were evaluated: plant height (PH), number of leaves (NL), stem diameter (SD), and root length (RL). These traits were assessed daily during two sowing periods—seeds were sown on October 22, 2014 (first period) and December 3, 2014 (second period). Four plants were randomly collected daily, beginning 7 days after first period and 13 days after for second period, totaling 94 and 76 evaluation days, respectively. For Gompertz models the equation was used y=a*e^((?-e?^((b-c*xi))and Logistic models the equation was used yi= a/(1+e^((-b-c*xi)). The inflection points of the Gompertz and Logistic models were calculated and the goodness of fit was quantified using the adjusted coefficient of determination, Akaike information criterion, standard deviation of residuals, mean absolute deviation, mean absolute percentage error, and mean prediction error. Differences were observed between the Gompertz and Logistic models and between the experimental periods in the parameter estimate for all morphological traits measured. Satisfactory growth curve fittings were achieved for plant height, number of leaves, and stem diameter in both models using the evaluation criteria: coefficient of determination (R²), Akaike information criterion (AIC), standard deviation of residuals (SDR), mean absolute deviation (MAD), mean absolute percentage error (MAPE), and mean prediction error (MPE).