Banca de DEFESA: MAYARA DE MELLO MARQUES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : MAYARA DE MELLO MARQUES
DATE: 29/04/2021
TIME: 14:00
LOCAL: Universidade Federal Rural do Rio de Janeiro
TITLE:

Estimation of Genetic Parameters via REML / Blup with the objective of Selecting Superior Genotypes of bitter gourd (Momordica charantia L.) for Fruit Production

KEY WORDS:

Mixed models, genetic parameters, medicinal plants, plant breeding, Blup-PCA.

PAGES: 65
BIG AREA: Ciências Agrárias
AREA: Agronomia
SUBÁREA: Fitotecnia
SPECIALTY: Melhoramento Vegetal
SUMMARY:

The bitter gourd (Momordica charantia L.) is a cucurbit, monoecious, diploid, with 2n = 2x = 22 chromosomes, known, popularly and scientifically, for its hypoglycemic action. Despite its importance, the improvement of bitter gourd is still in its infancy. Therefore, studies related to the inheritance of traits linked to production are extremely important to know the available germplasm, as well as to seek the best strategies for the selection process of superior genotypes. In this sense, the estimation of genetic parameters using the Maximum Likelihood Restritra (REML) method, and the estimation of the phenotypic values of the genotypes via Best Non-biased Linear Predictor (Blup) are important tools for the selection of bitter gourd genotypes with superior agronomic performance. Thus, the objectives of this study were to learn about the inheritance of 17 variables linked to fruit production in bitter gourd, as well as to select genotypes with high performance for fruit production in the crop. In this work, it was also proposed to associate the Blup with the PCA, whose objective was to allow a simultaneous interpretation of the phenotypic values of the genotypes considering all variables in the same two-dimensional plane. Therefore, 22 genotypes were planted in pots in a greenhouse and analyzed, during a complete production cycle, 17 variables relating to the number, total weight and average weight of fruits and seeds, as well as the total and average weight of the aryl of the seeds. Descriptive parameters and correlation coefficients between the variables were estimated, as well as the variances, repeatability and accuracy via REML for each variable. The phenotypic values and the ranking of the genotypes were obtained via Blup. These values were submitted to PCA analysis. Regression models were adjusted for the best genotypes, considering the number of fruits per plant (NFP). The estimated repeatability coefficients were classified as low for most of the variables analyzed. The NFP variable was the most promising for genetic gains, since it presented one of the greatest correlations with total production (fruits, seeds and aryl), combined with its individual repeatability estimate (r = 0.47). This estimate (r) is reasonable for cultures not yet improved and for highly quantitative variables. It was observed that there is a high and negative correlation between fruit production and fruit size. However, it is possible to select productive genotypes for larger fruit sizes. This was the case for the UFRRJ MSC028 and UFRRJ MSC087 genotypes. The most productive genotype, produced 146 fruits, and the least productive, only four fruits. According to Blup-PCA, the best ranked genotypes for fruit production were: UFRRJ MSC028, UFRRJ MSC042, UFRRJ MSC044, UFRRJ MSC053, UFRRJ MSC072 and UFRRJ MSC087. These genotypes were listed as female parents for the generation of superior progenies in the species improvement program at UFRRJ. The behavior of fruit production along the production cycle in these genotypes was expressed by a cubic regression model.

BANKING MEMBERS:
Presidente - 1712769 - PEDRO CORREA DAMASCENO JUNIOR
Interna - 2271445 - BRUNA RAFAELA DA SILVA MENEZES
Externo à Instituição - ROBERTO DOS SANTOS TRINDADE - EMBRAPA