Divergence among guava genotypes for fruit quality estimated using REML/BLUP and UPGMA

Divergence among guava (Psidium guajavaL.) genotypes for quality traits can support the selection of superior fruit. This study aimed to estimate genetic parameters, predict genotypic values, and group genotypes according to fruit quality. Nineteen genotypes were evaluated over three growing seasons, with three replicates. Peel and pulp color, total soluble solids (SS), total titratable acidity (TA), firmness, and the SS/TA ratio were determined. Data were analyzed using a mixed model. Variance components were estimated by restricted maximum likelihood (REML), and genotypic effects were predicted using best linear unbiased prediction (BLUP). UPGMA clustering was performed using Euclidean distances calculated from standardized BLUPs for seven nonredundant traits with estimable genotypic variance. Genotype × growing season interaction variance exceeded genotypic variance for eight traits. The highest mean heritabilities were observed for pulp hue angle (0.795) and TA (0.675), with accuracies of 0.892 and 0.821, respectively. Peel hue angle and pulp lightness showed null genotypic variance. PSR 120 (0.292) and PR66 (0.269) combined positive BLUPs for SS and the SS/TA ratio, whereas PSP 38 (-0.046), PSP 110 (4.469), and PR 23 (4.328) stood out for firmness. The dendrogram showed a cophenetic correlation of 0.847, and cluster analysis identified two groups: one comprising 18 genotypes and another consisting exclusively of PSP 110, indicating a contrasting pulp color profile. The integration of REML/BLUP and UPGMA enabled the identification of superior and divergent genotypes, although the genotype × growing season interaction reinforces the need for evaluations across multiple production cycles.