Detalhes bibliográficos
| Autor(a) principal: |
Marcondes, Mariana Martins |
| Data de Publicação: |
2018 |
| Tipo de documento: |
Tese
|
| Idioma: |
por |
| Título da fonte: |
Biblioteca Digital de Teses e Dissertações do UNICENTRO |
| Texto Completo: |
http://tede.unicentro.br:8080/jspui/handle/jspui/1190
|
Resumo: |
To obtaining doubled haploid lines (DH) in maize involves the steps of haploid induction, haploid seeds identification and chromosome duplication of selected haploids. A phase that happens the haploid induction is the haploids selection, which had usually based on anthocyanin pigmentation in seeds, by the expression of the dominant R-Navajo marker (R1-nj). However, this marker has incomplete penetrance and variable expressivity depending on the genotypes used in the induction crossings and environmental factors. Consequently, a large number of diploid seeds, or false-positive seeds, passes to the doubling phase, which it justifies the adoption of an auxiliary tool to classify ploidies. With the purpose of discarding diploids into putative haploid medium in an analysis by means of logistic regressions, guard cell length (GCL) and stomatal density (SD) can be adopted, since GCL and SD measurements from different induction crosses (hybrids and haploid inducers) differing greatly between haploids and diploids. The efficiency to rank diploids and haploids by CGL and SD, had linked to both the hybrids and inducers and to the plants phenotype, being that the ideal cut-off point of CGL to separate haploids from diploids was variable in relation to the crossing from which the plants descended, and the ideal cut-off point of SD was more related to the source genotype. Similarly CGL and SD, the seedlings traits: coleoptile (CL) and radicle (RL) length, coleoptile (CD) and radicle (RD) diameter and number of lateral seminal roots (NLSR) have been shown to be an efficient device, jointly, to separating haploid plants from diploids. These traits were also related with hybrids, haploid inducers and phenotype, except for CD and RD, which were shown to be dependent only on the phenotype (haploid/diploid), and the ideal cut-off point for these traits was established according to each induction crossing. Low rates of FDR (False discovery rate) and FNR (False negative rate) were observed for the ploidy in classification by CGL, SD and seedling traits. In addition, the accuracy values of all the methods studied, and their correlation with the real ploidy of the plants (gold standard) was considered satisfactory, inferring the possibility of incorporation these traits in maize breeding programs from doubled haploid lines. |