Genetic Diversity of Shallot (Allium cepa var. aggregatum) Segregating Populations from Ethiopia Using Multivariate Analysis

Abstract

Shallot is an important traditional crop used for seasoning of various national cuisines. However, productivity of shallot is low partly due to lack of improved varieties that are adapted to diverse agro-ecologies of Ethiopia. It has been difficult to improve the genetic base of local shallot germplasm due to its vegetative propagation nature. However, some plants within the germplasm were found bolting and producing seeds providing the opportunity for broader genetic base. Therefore, the present study was initiated to characterize and classify some segregating genotypes so as to use them for future breeding program. The study was undertaken at Debre Zeit Agricultural Research Center. It comprised of sixty genotypes generated through natural out-crossing and three released varieties (Minjar, Huruta and DZSHT-005/-02/90 DZSHT-005/02) used as controls. The experiment was laid-out in augmented design with three blocks. Twenty bulbs of each genotype were planted on a ridge comprising two rows. The three control varieties were also planted in the same way but replicated at in each block. Data on yield and yield components, percent bolting and number of flowerstalks/plant were collected. Analysis of variance, cluster and principal component analyses were also undertaken on data recorded. The results of the study showed that the genotypes significantly differed in yield/plant, number of bolting plants and number of flowerstalks/plant. However, they did not differ in bulb diameter, bulb height and downy mildew severity. Eight genoytypes had better yield/plant than all the three controls. Cluster analysis grouped the genotypes into seven clusters. Clusters I through VII comprised of 1(1.6%), 2(3.2%), 14 (22.2%), 10(15.9%), 5(7.9%), 22(34.9%) and 9(14.3%) genotypes, repectively. The genotypes within Clusters I through VI had atleast 87.5%, 85.2%, 85.0%, 85.8%, 82.9% and 84.1% similarity, respectively. Cluster III had the second highest mean for yield/plant, bulb diameter and number of bulb splits/plant. On the other hand, Cluster VII had the highest mean for yield/plant, bulb height and doweny mildew severity. It had also high inter cluster distances with other clusters. The principal componenet analysis identified seven components, five of which contributed to 83.1% of the variation. Generally, the eight genotyes with better yield were recommended for further variety trials under different environments while maintaining the other genotypes as a source of for variation future for breeding activities.