The objective of the present study was to evaluate the agronomic performance of tri- and tetraploid bananas in two production cycles aiming to select the most promising for recommendation for planting in the Reconcavo Region of Bahia. The experimental design was in randomized blocks with 11 genotypes and three replicates with four plants per plot in 3 m x 2 m spacing. The following agronomic characteristics were evaluated: plant height (m); pseudostem diameter (cm); number of live leaves at flowering; number of live leaves at harvest; stem length (cm); stem diameter (cm); bunch weight (kg); hand weight (kg); single hand weight (g); number of hands; number of fruits per bunch; pedicel length (cm); pedicel diameter (mm); finger drop susceptibility (Lb); firmness of pulp with peel; fruit weight (g); fruit length(cm); fruit diameter (mm); elongation index; pulp weight (g); pulp/peel ratio; pulp yield (%); pulp diameter (mm), peel thickness (mm); pulp thickness (Lb), tritable acidity-TA (%), soluble solids-SS (ºBrix), SS/TA ratio and pH. Of the thirty variables evaluated, 12 were significant for the source of variation 'cycles´. As to the 'genotype x cycles' interaction it is possible to assert that there was no differentiated behavior in the genotypes from the first to the second production cycle, except for 'plant height', 'number of live leaves at flowering', 'fruit diameter' and 'peel thickness´. Results for the effect of 'genotypes' were significant for all characteristics, except for fruit 'finger drop susceptibility'. Considering the agronomic data, genotypes from the YB series and the cultivars BRS Princesa and BRS Garantida showed to be promising for cultivation in the Reconcavo region of Bahia for presenting good agronomic performance.
Musa sp.; variability; breeding; genotypes; banana crop
