Colorimetric variation and correlation of biometric parameters between umbu matrices

Abstract

The umbu tree is an endemic species and is found in the Brazilian semi-arid. Umbu is highly appreciated for its socioeconomic importance and significant diversity of quality attributes. This study aims to explore the color-related variations between umbu matrices, and to evaluate the correlation between the main biometric attributes of quality. With this, 14 matrices were explored at the commercial ripening stage and submitted to colorimetric analysis (a*, b*, C* and L*) and determination of biometric parameters (fruit mass, length, width, shape, firmness, percentages of peel, seed, pulp, and yield). The colorimetric parameters were evaluated by the Scott-Knott mean test and dendrogram, and biometrics were submitted to linear correlation analysis. It was observed that in the peel green predominates, compared to the pulp, and the lowest luminosity indexes (L*) consist of the fruits with hairy peel. The saturation (C*) has a direct relationship with parameter b*; in addition, the dendrogram indicates the formation of four groups between the matrices. The correlation analysis of biometric parameters showed significant tendencies. Therefore, the results reinforce that the parameters evaluated can provide an important tool in selecting and quality control these fruits, since the required characteristics change according to their commercial purpose.

Proposal of classes for umbu commercialization

“…However, the absence of classification standards underestimates the price paid for the fruit, as it disregards size, fresh mass and color. Differences occur depending on genotype or maturation stage (SARAIVA et al, 2022). Menezes et al (2017) For the stratification of size classes, the mean and standard deviation of 3,717 fruits within the mean confidence interval were considered using the equation: FS = x ± kSx, where x = mean fresh mass; Sx = standard deviation of the mean; and k = correction factor for class adjustment, avoiding very wide intervals, whose adopted value was k = 1.0, since CV was below 20%.…”