Uso de covariáveis edafoclimáticas para caracterização de mega-ambientes e predição genômica de híbridos simples de milho
Nome: JOSÉ HENRIQUE SOLER GUILHEN
Tipo: Tese de doutorado
Data de publicação: 03/03/2020
Orientador:
Nome |
Papel |
|---|---|
| ADÉSIO FERREIRA | Orientador |
Banca:
| Nome |
Papel |
|---|---|
| ADÉSIO FERREIRA | Orientador |
| CLAUDIA TEIXEIRA GUIMARÃES | Examinador Externo |
| JURANDIR VIEIRA DE MAGALHAES | Examinador Externo |
| LUCIANO DA COSTA E SILVA | Examinador Externo |
| MARCIA FLORES DA SILVA FERREIRA | Examinador Interno |
Resumo: Environmental variability highly affect maize productivity, causing relevant
annual yield losses. Mega-environments (ME) are usually defined in breeding programs, to
identify appropriate selection strategies and cultivars adapted to specific ME. The present
study aimed to use environmental information to define MEs for the Embrapas maize
breeding program in Brazil. Geographic, soil and climatic covariates were collected for the
historical data of hybrid testing along 11 years, including 72 locations distributed in the
Brazilian territory, representing the first and the second maize cropping seasons. Twenty-nine
environmental covariates were collected for the whole experimental period in each locationyear combination (trial). Statistical analyses were performed to: detect multicollinearity;
verify the existence of environmental differences between locations; and identify
environment-similar groups of locations. In addition to the environmental covariates, the
phenotypic data was also used to investigate the similarities of maize hybrid performance for
grain yield between MEs within years. The combined analysis of both cropping seasons,
including 123 locations (i.e.72 and 51, for the first and the second cropping seasons,
respectively), divided the locations in two distinct groups, based on the cropping seasons. A
detailed study within each cropping season resulted in five groups of locations for the first
season, and four groups of locations for the second season. The groups for the first season
were: i) locations in the Cerrado; ii) locations in the Amazon; iii) low-latitude locations; iv)
high-latitude locations; and v) a mixed group of locations. Two major group in the second
season locations were: i) low-latitude in the South of Brazil; and ii) high-latitude in the
Central and North regions of Brazil. The use of environmental covariates efficiently identified
MEs, which were corroboratedby the phenotypic hybrid performance for grain yield within
ME.
