A total of 199 germplasm accessions collected from 10 administrative regions of Ethiopia and four released cultivars, which were used for estimating of error variance, of barley in Ethiopia were field evaluated for nine agronomic traits at Holetta and Bekoji Agricultural Research Centers of Ethiopia during the 2006 main cropping season using non-replicated augmented design plots consisting of four incomplete blocks. The objectives were to assess the extent and pattern of morphological variation in the barley accessions with respect to regions and altitude of collection, to classify the genotypes tested into relatively homogenous groups and to identify the major traits contributing to the overall observed diversity in the germplasm. Genotype variance estimate of regions and altitudes indicated wide variation among accessions depending on the traits involved. The presence of high morphological variation within regions and altitudes particularly above 2000 m a.s.l. indicated the potential of each region and high altitude zones for barley improvement and conservation in the country. The clustering of accessions did not show grouping on the basis of regions of origin. Traits like thousand kernel weight, plant height, days to head and days to maturity accounted for most of the gross variance among the barley accessions and played role in differentiating accessions collected from different regions and altitude classes into principal components. In general because of environmental factors on the observed morphological variation future germplasm collection should consider to explore wide geographical and climatic differences within the country.

Effective selection of parental material is an essential requirement for breeding success. Best linear unbiased prediction (BLUP) allows integration of all available information including pedigree information. However, for breeding self-pollinating crops the widely used coefficient of coancestry has disadvantages especially when pedigree Abbreviations: BLUP, best linear unbiased prediction; BLUP(E), best linear unbiased prediction considering environmental effects; BLUP(E1A), best linear unbiased prediction considering environmental effects and pedigree information; BLUP(E1GS), best linear unbiased prediction considering environmental effects and genetic similarities; MME, mixed model equations; QTL, quantitative trait locus.

Accurate and fast estimation of genetic parameters that underlie quantitative traits using mixed linear models with additive and dominance effects is of great importance in both natural and breeding populations. Here, we propose a new fast adaptive Markov chain Monte Carlo (MCMC) sampling algorithm for the estimation of genetic parameters in the linear mixed model with several random effects. In the learning phase of our algorithm, we use the hybrid Gibbs sampler to learn the covariance structure of the variance components. In the second phase of the algorithm, we use this covariance structure to formulate an effective proposal distribution for a Metropolis-Hastings algorithm, which uses a likelihood function in which the random effects have been integrated out. Compared with the hybrid Gibbs sampler, the new algorithm had better mixing properties and was approximately twice as fast to run. Our new algorithm was able to detect different modes in the posterior distribution. In addition, the posterior mode estimates from the adaptive MCMC method were close to the REML (residual maximum likelihood) estimates. Moreover, our exponential prior for inverse variance components was vague and enabled the estimated mode of the posterior variance to be practically zero, which was in agreement with the support from the likelihood (in the case of no dominance). The method performance is illustrated using simulated data sets with replicates and field data in barley.

Using breeding values in parental selection of self-pollinating crops seems to be superior to conventional selection strategies, where selection is often based on several traits which are correlated among each other. However, analysing each trait separately can bias estimates of breeding values. This study examined responses to selection for total merit indices based on breeding values resulting from single- and multiple-trait best linear unbiased prediction (BLUP). We generated data for a multi-environment trial of a "virtual" parental population in which the phenotypic value of inbred lines was influenced by additive, additive-by-additive epistatic, year, location, block and genotype-by-environment interaction effects. Two traits with heritabilities of 0.7 and 0.3 and nine different correlation scenarios between traits (estimated phenotypic correlation ranging from -0.39 to +0.36) were simulated. Gain in selection response was greater for multiple-trait than for single-trait breeding values, especially if traits were negatively correlated. For all correlation scenarios, the overall standard errors of difference of multiple-trait predictors were lower than those of single-trait analysis.

A common difficulty in mapping quantitative trait loci (QTLs) is that QTL effects may show environment specificity and thus differ across environments. Furthermore, quantitative traits are likely to be influenced by multiple QTLs or genes having different effect sizes. There is currently a need for efficient mapping strategies to account for both multiple QTLs and marker-by-environment interactions. Thus, the objective of our study was to develop a Bayesian multi-locus multi-environmental method of QTL analysis. This strategy is compared to (1) Bayesian multi-locus mapping, where each environment is analysed separately, (2) Restricted Maximum Likelihood (REML) single-locus method using a mixed hierarchical model, and (3) REML forward selection applying a mixed hierarchical model. For this study, we used data on multi-environmental field trials of 301 BC 2 DH lines derived from a cross between the spring barley elite cultivar Scarlett and the wild donor ISR42-8 from Israel. The lines were genotyped by 98 SSR markers and measured for the agronomic traits ''ears per m 2