Ação gênica, parâmetros genéticos e ganho por seleção de atributos físicos relacionados a qualidade de grãos em arroz

Abstract

To assist researchers in the development of superior genotypes, it is necessary to understand the genetic processes of the desired agronomic traits. In the literature there are few studies regarding the genetics of quality attributes of rice grains using complete genetic design. This thesis aimed to determine genetic parameters such as heritability, genetic inheritance and type of gene action, analyze variance components, estimate selection gain, evaluate different selection pressures, reveal agronomic performance, trends and inter-genetic relationships of physical attributes of grain quality in irrigated rice genotypes. Three experiments were carried out at the Experimental Station Terras Baixas of Embrapa Clima Temperado in the seasons from 2014 to 2018, using the S21 digital scanner to measure the characters of rice grain quality. Gene action was determined by the method of analysis of segregating generations, with the complete genetic design of six generations, based on the complete model and the additive-dominant model, using the crosses between BRIRGA 409 x BRS Bojuru and their reciprocal, thus obtaining the F1, F2 and backcross generations (RC1 and RC2), conducted together in the same harvest. It has been determined that genetic inheritance is quantitative, and the additive genetic effect is the main contributor. For selection genetic gain, selections were performed in segregating generations F3, F4 and F5, in eight segregating populations (families) from different crosses. The analysis of genetic parameters, response to selection, was carried out using different selection pressures. Genetic gains for the studied characters were determined from selections in early generations and the response to selection with a pressure of 10% becomes very promising for the grain quality attributes of irrigated rice. Finally, grain quality trends and genetic interrelationships based on multivariate models were applied to elite irrigated rice lines. The experiment was carried out in a completely randomized block design, with three replications, with fifteen F6 lines and four control cultivars. The results showed that the lines PH18502 and PH18701 presented better agronomic performance for the studied characters, by univariate analysis. The linear correlations and canonical correlations presented, show potential in the direction of selection of multiple characters and point to the possibility of indirect selection among the relevant agronomic characters for the irrigated rice production chain.