Parâmetros genéticos e índice de seleção MGIDI na identificação de genótipos superiores de mandioca

Abstract

The MGIDI selection index (multi-trait genotype/ideotype distance index) aims to obtain the distance from genotype to ideotype, helping to select superior genotypes in breeding programs. This work was carried out with the objective of obtaining estimates of genetic parameters through the mixed modeling methodology (REML/BLUP) and to identify superior cassava genotypes through the MGIDI selection index, which allows the selection of superior genotypes for the related traits. Diseases and agronomic characteristics. The research was carried out at the Experimental Farm of the Federal University of Grande Dourados – UFGD. The experiment was conducted for four consecutive years (2016/2017; 2017/2018; 2018/2019 and 2019/2020), with genotypes from the active germplasm bank of Embrapa Cassava e Fruticultura. The evaluated characteristics were symptoms caused by the diseases Anthracnose (CAD), Bacteriosis (CBB) and Superelongation (SED) and agronomic characters Size, Total root yield (PTR), Plant height (AP), Shoot yield (PPA), Dry mass (DM) and Survival (Survival). By deviance analysis, the likelihood ratio test (LRT) found significant effects (p<0.05) for the genotypes in relation to all evaluated characteristics. The heritability coefficient ranged from 7.15 to 74%, with four traits (AP, MS, PTR and PPA) showing values above 50%. For the same traits, the values of genetic variance were higher in relation to environmental variance, indicating that the phenotypic expression is necessarily due to the genetic character. The MGIDI index enabled the selection of 49 genotypes that were close to the ideotype. The nine characteristics evaluated were divided into four factors, in which each factor presented its strengths (with greater contribution, that is, the variables of this factor are farther from the ideotype) and weak (with less contribution, that is, closer to the ideotype are the variables of this factor). The variables related to anthracnose and bacteriosis diseases were the ones that most contributed to the distance between genotypes and ideotypes.