Fenotipagem por imagem, estimativas de parâmetros genéticos e índices de seleção em alface biofortificada

Abstract

Developing biofortified foods is a constant objective of breeding programs. However, the number of genotypes to be evaluated and the time and monetary cost in determining leaf constituents represents an obstacle. An alternative is the use of image phenotyping and statistical methods to achieve a greater amount of information. Image phenotyping allows a quick, objective analysis of the entire experiment without destructing samples. The main statistical methods to obtain relevant knowledge about the germplasm analyzed include the estimation of genetic parameters, such as heritability, the coefficient of genetic and environmental variation, correlation between the characteristics analyzed, and trail analysis. Another method is the application of non-parametric indices. This enables the simultaneous analysis of various information and reveals the selection gain and the screening of promising genotypes for genetic improvement. In this context, two experiments were carried out: one with 31 genotypes of red lettuce (30 lines + cultivar Belíssima) and another one with 25 genotypes of green lettuce (22 lines + three cultivars – Grand Rapids, UFU-Biofort and Uberlândia 10000); all lettuce types were looseleaf. The experiment was carried out in 2018, at the Experimental Vegetable Station at the Federal University of Uberlândia, Monte Carmelo campus. The lettuce lines in both experiments came from the hybridization of the cultivars Belíssima versus Uberlândia 10000 (rich in carotenoids). Hybridization was followed by six successive self-foundations carried out between 2013 and 2017. This study aimed to prove the genetic diversity of lettuce germplasm, to validate the use of image phenotyping in the identification of carotenoid-rich genotypes, and to evaluate the efficiency of selection indices to chose biofortified looseleaf lettuce strains with favorable agronomic characteristics and good nutritional properties. The results indicated that the germplasm bank analyzed had a wide genetic variability. Highperformance phenotyping showed a high correlation with the traditional methodology to determine carotenoids; it is an alternative to identify different genetic backgrounds in a germplasm bank. These results can also help predict the nutritional values of the content of carotenoids in lettuce leaves before commercialization. The Mulamba and Mock index allowed the greatest selection gains for the characteristics analyzed in both experiments; it selected the superior genotypes in carotenoid levels and agronomic performance.