Fenotipagem digital, caracterização biométrica e avaliação fisiologica de disporos de Spondias tuberosa Arruda

Abstract

The umbu tree (Spondias tuberosa Arruda) is an endemic species of the Caatinga biome with high socioeconomic and ecological value. However, intense extraction and fragmentation of its natural habitat require effective strategies for conservation and selection of elite genotypes. This study aimed to evaluate and select umbu tree matrices using classical biometrics, physiological tests, and digital morphometric analysis of diaspores from two populations in the state of Paraíba (Algodão de Jandaíra and Campina Grande). Diaspores were subjected to physical characterization (weight, length, and width), physiological evaluation (percentage of emergence, mean emergence time, and emergence speed index), and digital morphometric analysis using ImageJ® software (area, perimeter, circularity, roundness, and solidity). Data were evaluated using Analysis of Variance (ANOVA), genetic parameter estimation, and multivariate approaches (UPGMA, PCA, and heat map). The results indicated that there is no mandatory direct correlation between physical dimension attributes and germination physiological quality. Physical dimension descriptors, notably area, perimeter, length, and width, stood out as the most important, presenting high estimates of average heritability (h2 > 98%) and favorable variation indices (IV > 1.0), indicating strong genetic control and high potential for gains through mass selection. Multivariate clustering formed three distinct phenotypic groups, without a strictly geographic pattern, with the first two principal components explaining 77.5% of the total variation. It is concluded that classical biometrics reveals larger matrices promising for industrial use; physiological tests select genotypes with higher emergence speed indices for ecological restoration projects and seedling production; and high-performance digital morphometrics consolidates itself as an efficient tool to discriminate matrices and guide multidimensional selection. Keywords: Caatinga; genetic diversity; forest resources; seeds.