Multivariate analysis of quantitative characters variability in Ethiopian barley (Hordeum vulgare L.) landrace: Based on regions and altitude

Abstract

Barley (Hordeum vulgare L.) is Ethiopia’s most important highland cereal crop and widely growing in most part of the country. Based on high levels of genetic and phenotypic diversity, the country is considered as center of diversity for barley. Assessments of the amount of genetic variation within and among populations are crucial for effective and efficient genetic improvement of the crop. Hence, this study was conducted to assess the diversity of barley landraces collected from various altitudes and regions of Ethiopia. A total of 585 barley landraces and 10 checks were evaluated using augmented randomized complete block design consisting of six blocks. All the 585 landraces were planted in un-replicated plots and the 10 checks were replicated six times (ones in each block) to estimate an error variance. Data on 13 quantitative characters were subjected to calculation of descriptive statistics, ANOVA and multivariate analysis (Unweighted Pair Group Method Analysis (UPGMA) cluster analysis and principal component analysis). There were significant differences (ANOVA, P<0.01) among landraces for plant height, 1000-seed weight, number of seeds per spike, days to heading and days to maturity. All the genotypes were grouped into five clusters where 74.02% of the accessions (433) fall in cluster I, IV and V. Early matured accessions were grouped in cluster I, while late matured, high yielding and tall accessions were clustered in cluster IV. The highest intra-cluster distance was 23.12 for cluster III whereas the highest inter-cluster distance was 57.37 between cluster IV and V. The first three principal components contributed 51.76% of the total variations observed among the genotypes. Principal component one (PC1) alone had contributed 22.56% of the total variations mainly due to plant height, 1000-seed weight, grain yield and peduncle length in their respective order. Principal component two (PC2) contributed 18.94% of the total variations mainly through spike density, number of kernels per spike, spike weight and days to maturity in their descending order. Principal component three (PC3) had contributed 10.94% of the total variations through total number of tillers per plant, number of seed-bearing tillers, days to 90% maturity and days to 50% heading. Altitude of the original landrace collection sites also significantly impacted the various quantitative characteristics studied. Regional differentiations were also evident among the landrace collections. These results reveal the existence of significant agro-morphological variations among the landraces included in this study. Based on the characters considered and populations evaluated, the marked diversity observed among the barley landraces in Ethiopia could be utilized in future crop improvement for various agronomically important traits. The information generated complements the robust barley breeding program of competitive, stable and climate-resilient varieties of end users’ preferences in different agro-ecologies of Ethiopia.