Fitting nonlinear models to data from the biquinho pepper plant

Abstract

Pepper has several bioactive compounds with antioxidant, anti-inflammatory, and antibacterial properties. Due to its potential, it has several applications in food, industrial and pharmaceutical sectors. The main aim of the research was to fit non-linear growth models to biquinho pepper plant height data. Nonlinear models are important for obtaining information about plant growth that can help at different stages of cultivation. The data were adapted from an experiment carried out in a greenhouse at the were used. The sowing recommendations were suggested by the company: depth of 0.5 cm and composition of the soil substrate classified as Eutrophic Gray Clay. The plants were irrigated with 100% of their water requirement with wastewater treated by the UASB+ WETLAND sludge blanket up-flow anaerobic reactor. Assessments were carried out every seven days, totaling 21 assessments, 163 days after sowing (DAS), with two plants per replication. The models used in the adjustment were the logistic, Gompertz and von Bertalanffy. The assumptions under residuals of the adjusted models were verified using the Shapiro-Wilk (normality) and Bartlett (homogeneity of variance) tests. Furthermore, models with different variance structures and autoregressive terms were tested. All models evaluated proved to be adequate to adjust to data on the height of the biquinho pepper plant. However, according to the Akaike criterion (AIC), the von Bertalanffy model considering homogeneity of variance without the structure of autoregressive terms presented a better fit. The adjustment estimated an asymptotic plant height of 26.32cm and an inflection point of the curve when it reached approximately 52 days after sowing.