Multivariate analysis for quantitative traits in finger millet (Eleusine coracana subsp. coracana) population collected from eastern and southeastern Africa: detection for patterns of genetic diversity.

摘要

The gradual change in climatic conditions, particularly rainfall distribution in tropical and sub tropical regions of the world necessitate looking for productivity enhancement of stress tolerant crops such as finger millet as one option. Assessing genetic variation is a crucial for varietal development and genetic resource conservation. To this regard, a study was conducted at Gute and Arsi Negele (Ethiopia) during 2011 cropping season on one hundred and forty four finger millet landraces collected from different regions of Ethiopia, some introduced from Kenya, Eritrea, Zambia and Zimbabwe to evaluate the genetic diversity for quantitative traits at population level and eco-geographical regions of origin. The trend of quantitative trait diversity revealed that the highest genetic diversity were observed at the lowest level (among landrace populations) followed among the regions or countries of origin and least among altitude classes. This leads to suggest, taking more samples within a locality or population would be a better approach to capture the range of variation in finger millet population. Cluster analysis indicated that finger millet populations from neighboring regions of Ethiopia, neighboring African countries and proximity in altitude classes shared strong similarity. The similarity could be either due to fact that farmer's selection criteria for a given traits might be similar particularly based on the adaptive role of traits for the environment, the primary seed source can be the same, or high tendency of seed exchange. Principal component analysis at populations level, geographical locations and agro-ecologies of origin indicated that grain yield per plant, thousand grain weight, days to heading, days to maturity, lodging index and biomass weight per plant were the most important traits contributing for the overall variability implying that breeding effort on those traits can meet the targeted objective.