Análise dialélica de populações S0 para precocidade e produtividade em arroz de terras altas

Abstract

The upland rice breeding programs in Brazil aims to develop early-maturing and high-yielding cultivars to integrate the crop into the production system. In this context, recurrent selection stands out as an effective method to obtain such genotypes, as it gradually increases the frequency of favorable alleles through successive selection cycles while maintaining the genetic variability of the population. Thus, this study aimed to identify the most promising populations for developing lines that combine high grain yield and early maturity, aligned with the main traits of interest in rice cultivation, based on the analysis of segregating populations from the first recurrent selection cycle of the Upland Rice Breeding Program – MelhorArroz UFLA. For this purpose, 21 S0 populations from the first cycle (CI) of the recurrent selection program were evaluated, along with the checks BRSMG Caçula and BRS Esmeralda. The experiment was conducted during the 2023/24 growing season in a randomized complete block design (RCBD) with three replications, using plots of 15 rows of 4 meters spaced 0.3 meters apart. The evaluated traits included days to flowering, grain yield, plant height, milled rice yield, head rice yield, grain dimensions, and chalkiness. The observations for the evaluated traits were standardized, and the sum of Zijk per plot was obtained, forming an index that simultaneously considered all seven traits. The collected data was subjected to basic assumption verification and analysis of variance. Dunnett’s and Scott-Knott’s tests were performed, along with Pearson’s correlation analysis. For each variable, diallel effect decomposition was performed to estimate the general combining ability (GCA) and specific combining ability (SCA) of the recombined parents. All S0 populations from the CI recurrent selection program exhibited a short cycle and high grain yield. Among them, the populations S0CIP-11, S0CIP-7, S0CIP-17, and S0CIP-19 stood out for displaying favorable estimates across multiple traits, showing potential for generating lines that combine high yield, early maturity, and grain quality. The parents CNA20658-B12 and CNA20666-B10 contributed with a high frequency of favorable alleles for multiple traits. The combinations CNA20658-B12 x BRSMG Caçula and CNA20666-B10 x CNA20666-B16 were the most complementary, suggesting that the progenies of these populations will have more probability of high mean performance and greater variability than the others.