Fatorial fracionário na seleção de corretivos e bioestimulantes para bananeira 'Prata anã' clone Gorutuba: eficiência agronômica e rentabilidade

Abstract

Currently, agricultural experimentation has adopted smaller experimental designs and the use of new production technologies, derived from it, may allow brazilian banana culture to become more productive and profitable to the producer. In this work, we aim (i) to carry out the screening of 14 factors through the combination of a Definitive Screening Design (DSD and a 34 series complete factorial, previously selected for application in the 1st cycle of the 'Prata Gorutuba' banana as fertilizer supplements; (ii) estimate the effect of the main factors previously screened in the prior trial and; (iii) compare conventional fertilization management (CM) and alternative fertilization management (proposed in objective ii), regarding production, post-harvest and economic viability. We conducted the three experiments on commercial farm in the central region of Minas Gerais, Brazil. In the first experiment, we made a combination of a DSD and a factorial 34, a factorial scheme composed of 84 experimental plots, each with three 'Prata Gorutuba' plants. We divided the experimental area into four randomized blocks (RBD). We selected the 14 factors: seven soil conditioners and seven algae extract biostimulants. We carried out phytotechnical, production, cycle duration and post-harvest evaluations to obtain the best effects of factors and associations of these in the response variables. We submitted all variables to ANOVA and subsequently, we estimated the significant factors by the "t test" to obtain the direction of the causal effect. We obtained experimental savings, adopting only 38.88 % of the number of possible total plots. The factors that stood out the most were CC, CT, LT and ACA. We corroborated the main hypothesis of the study that some factors would be more important in the experimental design. In the second experiment, we elaborated a combination of a DSD and a factorial 33, a factorial scheme composed of 30 experimental plots, each with three 'Prata Gorutuba' plants. We divided the experimental area into six randomized blocks. We selected four factors: three soil conditioners and one algae extract biostimulant. The same evaluations were performed to estimate the effect of the main products and the same statistic as the previous experiment was carried out, both made by software R. The estimation of the factors CC, CT, LT and ACA, provided satisfactory results in the performance of the culture. We recommended using fractional factorial strategies, including DSD designs in agronomic assays, especially for estimating factor effects. In the third experiment, we adopted a CRD, with two fertilization treatments (CM and CM+3SC+1B) and ten replications, with 20 experimental plots, each with three 'Prata Gorutuba' plants. Were performed the same statistics and evaluations as in previous experiments. We verified the economic viability of the two treatments, based on productivity data and total operational costs of production, based on the main economic indicators. The CM+3SC+1B treatment increased by 4,032 tons ha-1 and was more profitable for the banana grower, with higher values of Gross Revenue, IRR and NPV. We suggested carrying out future research in this region with this alternative treatment, during subsequent cycles of the 'Prata Gorutuba' variety.