Multivariate analysis for yield and yield-related traits of oat (Avena sativa L.) genotypes in Ethiopia

The study aimed to generate information on agro-morphological variability among the 120 oat genotypes in the central highlands of Ethiopia. The result showed that the hierarchical clustering grouped the genotypes into four clusters. The highest number of genotypes grouped into the first cluster (38) followed by the second cluster (35) and the fourth cluster (33) while the lowest genotypes (14) were recorded in the third cluster. High forage yield-producing genotypes were grouped into the second cluster while high grain-producing genotypes were grouped into the third cluster. The intra-cluster genetic distance values ranged from 4.3 (C3) to 2.3 (C1). The minimum inter-cluster genetic distance was recorded at 11.7 (between C1 and C2), while the maximum inter-cluster distance was 63.2 (between C3 and C4). Principal component analysis showed that the first four principal components with eigenvalues greater than one contributed 78.3% of the variability among genotypes. The highest contributing traits in the first component were forage yield production rate, biological yield production rate, panicle length at grain harvest, forage dry matter yield, and number of spikelets per panicle. The traits with the greatest weight on the second component were harvest index, grain yield, number of leaves at forage harvest, and plant height at grain harvest. Proper selection of best yield performing genotypes coupled with high yield contributing traits is very important to yield improvement and future oat breeding program. Therefore, selection based on the traits that had a high contribution to the first two principal components would be effective for yield improvement in oat genotypes.