Seleção fenotípica de progênies de arroz de terras altas auxiliada por marcadores moleculares

Abstract

Rice (Oryza sativa L.) is the dietary staple for approximately 2.5 billion people. In Brazil, there is a concerted effort to intensify upland rice cultivation as a sustainable alternative, given its reduced water requirements and lower greenhouse gas emissions. Within this context, breeding programs aim to develop early-maturing, high-yielding genotypes for integration into agricultural production systems, utilizing genetic diversity as a primary source for selection. Thus, this study aimed to evaluate the agronomic performance and genetic diversity of progenies from the Family Yield Trial (FYT) during the 2023/24 growing season, verified through molecular analysis using SSR (Simple Sequence Repeats) markers. A total of 169 progenies were evaluated using a 13x13 partially balanced lattice design with two replicates across three locations: Lavras (MG), Santo Antônio de Goiás (GO), and Sinop (MT). The traits evaluated included days to flowering (DTF), plant height (cm), and grain yield (kg ha⁻¹). Data were subjected to a joint analysis of variance via mixed models (REML/BLUP) using R software. Molecular characterization was performed with 24 microsatellite (SSR) markers analyzed by capillary electrophoresis. Of the 24 markers, 21 were successfully amplified, detecting 133 alleles (ranging from 2 to 13 alleles per locus). Cluster analyses identified 21 clusters for both phenotypic and molecular data. Twenty phenotypically superior progenies were identified; when contrasted with molecular clusters, these were distributed across nine distinct genetic groups. It is concluded that there is considerable genetic diversity to be exploited and that simultaneous selection based on agronomic performance and molecular divergence allows for the identification of promising progenies, thereby mitigating the narrowing of the program's genetic base.