CARACTERÍSTICAS BIOMÉTRICAS, MORFOLÓGICAS E ESPECTRAIS DE SEMENTES ASSOCIADAS À FORMAÇÃO E À QUALIDADE DE MUDAS DE Hymenaea courbaril L.

Abstract

Seed quality is a determining factor for the successful production of forest seedlings, directly influencing seedling emergence, vigor, and early plant establishment. In this context, the integration of biometric characteristics, morphological descriptors obtained through digital image analysis, and spectral variables can improve the understanding of the relationships between seeds and seedling performance. This study aimed to evaluate the associations between biometric characteristics, morphological descriptors obtained through digital image analysis, and spectral variables of jatobá (Hymenaea courbaril L.) seeds and seedling formation and quality. A total of 237 seeds were evaluated, in which biometric variables, morphological descriptors obtained using the ImageJ® software, and spectral signatures within the 350–2200 nm wavelength range were measured. After dormancy-breaking treatment, the seeds were sown in tubes, and the seedlings were evaluated for emergence, growth, and Dickson Quality Index. Data were analyzed using descriptive statistics, Spearman's correlation, logistic regression, linear regression, and generalized additive models. Seed mass and width showed positive associations with seedling formation, whereas morphological descriptors, particularly area, minimum Feret diameter, perimeter, and aspect ratio, showed greater ability to explain the variation in the Dickson Quality Index. Spectral variables showed associations at specific wavelengths with seedling characteristics, providing complementary information to biometric and morphological analyses. It was concluded that the integration of these approaches enhances seed characterization and contributes to a better understanding of jatobá seedling formation and quality.