Predição de ganhos genéticos via REML/BLUP em progênies s2 de mamoeiro

• Ano de defesa: 2015
• Autor(a) principal: Silva, Tamiris Pereira da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Melhoramento vegetal; Carica papaya L.; Parâmetros genéticos; Modelos mistos; Acurácia seletiva; Caracteres agronômicos; Genetic parameters; Mixed models; Accuracy selective; Agronomic traits
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/19144

The reduced number of registered papaya cultivars has made breeders to access the genetic variability of the species in order to gather in one genotype the maximum phenotypes required by producers and consumers. Among the commercial hybrids,Tainung 1 and Calimosa are the most used, which the first has its seeds imported from Taiwan to high costs. In order to increase the accuracy and to obtain selective genetic gain, more efficient selection methodologies have been explored. For this reason, the adoption of more refined genetic statistical procedures is a trend in plant breeding, such as the REML / BLUP methodology. Based on the above, the aims of the present study are: (i) estimate the main genetic parameters in papaya progenies, (ii) predict genetic gain with selection of the best progenies and (iii) the selection of the best individuals within the progenies. The experiment was conducted at the Experimental Field of Curu, which belongs to Embrapa Agroindústria Tropical in the municipality of Paraipaba. The genetic material consisted of 36 S2 progenies and two witnesses, 17 derived from hybrid Calimosa and 19 Tainung1, all evaluated in a randomized block design with four replicates. The analyzed characters were height of first flower, fruit and plant at 6, 12 and 18 months; stem diameter at 6, 12 and 18 months; number of days after planting to emergence of first flower and fruit; number, mass and average mass of marketable fruits and number and mass of carpelloid, pentandric and bananiformes fruits. For data analysis, the methodology via linear mixed models REML / BLUP (Restricted Maximum Likehood / Best Linear Unbiased Prediction) was used, by which genetic parameters were estimated and the prediction of breeding values was made. The magnitudes of the genetic parameters indicated that the variability present in most of the characters enables genetic gain if the selection is practiced at the progenies level and not at individual plants, since the greater heritability and accuracies values are relative to the average of the progenies. For the selection of progenies, PROC-323, PROC-130, PROT-13 and PROC-217 were the most promising, bearing greater genetic gains among most of the analyzed characters. Regarding the selection among and within progenies, individuals 1421 and 1422 are the most potential for continuity of selection. Finally, the increase of the character expression was higher in number of marketable fruits and the reduction of expression was higher for pentandric number of fruits.